Proceedings
                     of
    Non-point Pollution
Abatement  Symposium
                  Sponsors
            U S Environmental Protection Agency-
            Great Lakes National Program Office
                Wisconsin Department of
                  Natural Resources
                 Marquette University
                Water Resources Center
             University of Wisconsin, Madison
              Center for Great Lakes Studies
            University of Wisconsin. Milwaukee
             Metropolitan Milwaukee Sewerage
                   Commission
                  City of Milwaukee
                Southeastern Wisconsin
              Regional Planning Commission
              German Marshall Fund of the
                   United States
            Milwaukee, Wisconsin
              April 23-25, 1985

-------
                 NONPOINT  POLLUTION ABATEMENT
                            SYMPOSIUM
           TECHNICAL,  MANAGERIAL AND INSTITUTIONAL
                    PROBLEMS AND SOLUTIONS


                       April  23-25,  1985
                     Milwaukee, Wisconsin
Sponsors:  U.S. Environmental Protection Agency
            (Great Lakes  Program Office)
          Wisconsin Department of Natural  Resources
          Marquette University
          Water Resources Center, University  of Wisconsin-Madison
          Center for Great Lakes Studies,  University of Wisconsin-Milwaukee
          Metropolitan Milwaukee Sewerage  Commission
          City of Milwaukee
          Southeastern Wisconsin Regional  Planning Commission
          German Marshall  Fund of the United  States
Symposium Organizers:   Division of Continuing  Education
                      Marquette University

-------
             Symposium Organizing Committee
                  Dr. Gordon Chesters
    Professor and Director, Uater Resources Center
            University of Wisconsin-Madison

                  Dr. David Edgington
Professor and Director, Center for Great Lake Studies
           University of Wisconsin-Milwaukee

                 Mr.  C. Michael Farmer
 Assistant Director,  Division of Continuing Education
                 Marquette University

                    Mr. Kent Fuller
            Chief of Environmental Planning
          U.S. Environmental Protection Agency
                      Chicago, IL

                   Mr. David Gruber
       Milwaukee Metropolitan Sewerage District

                   Dr. William Katz
            Professor of Civil Engineering
                 Marquette University

                    Dr. John Konrad
          Chief, Non-point Pollution Section
       Wisconsin Department of Natural Resources
                     Madison, WI

                     Mr. Gary Korb
  Southeastern Wisconsin Regional Planning Commission
                     Waukesha, WI

                  Mr. Edwin Laszewski
                     City Engineer
                   City of Milwaukee

      Dr. Vladimir Novotny (Conference Organizer)
            Professor of Civil Engineering
                 Marquette University

                   Dr. Stuart Walesh
              Donohue 6 Associates, Inc.
                     Waukesha, WI

-------
                                FOREWORD

Each year billions of dollars are spent in the U.S.  and worldwide on sewers
and municipal point source abatement.  Yet, without limiting pollution in-
puts from both rural and urban non-point sources, the effect of the vast
expenditures on water quality of streams, lakes and estuaries receiving
point and non-point pollution loads may be only marginal.

Attacking the non-point pollution problem may not be an easy task.   Sources
are not clearly identifiable and the organization and institutional  struc-
ture necessary to resolve the problem may not be yet in place.

The Nonpoint Pollution Abatement Symposium was held in Milwaukee, Wisconsin
with the sponsorship of U.S. Environmental Protection Agency, Wisconsin
Department of Natural Resources, Metropolitan Milwaukee Sewerage  District,
Water Resources Center of the University of Wisconsin and  several other
agencies.  It attracted 145 participants from six countries (U.S.A., Belgium,
Canada, Germany, Great Britian, and Italy) and provided a  forum for the
exchange of ideas among the scientists, engineers and policy makers on the
problem of non-point pollution and its abatement.

The Symposium was divided into three programatic blocks:   (1) State-of-the-
Art presentations by invited leading scientists on the most important issues
dealing with the identification of the non-point pollution problem and its
solutions; (2) Panel discussions on technical, managerial, economical, and
legal issues; and (3) Technical presentations and case studies.  The pro-
ceedings contain all materials presented at the Symposium.

-------
                           Table of Contents
A-I  1-6         Address by C.D.  Besandy,  Secretary
                 Wisconsin Department of Natural  Resources


Keynote Presentation:

K-I 1-6          Nonpoint Pollution Problems from a Perspective   of
                 Municipal Governments
                 Henry W. Maier,  Mayor,  City of Milwaukee

K-II 1-10        Magnitude and Variety of Problems of Nonpoint Pollution
                 from  Agricultural  Practices
                 John  Harkin, University of Wisconsin

K-III 1-19       Magnitude and Problems  of Non-point Pollution from Urban
                 and Urbanizing Areas
                 Gary Oberts, Metropolitan Council of the Twin Cities Area

K-IV 1-13        Impacts of Nonpoint Pollution on Receiving  Waters
                 Peter Krenkel, University of Nevada

K-V 1-18         Use of Computer  Models  in Managing Nonpoint Pollution
                 from  Agriculture
                 Walter G. Knisel,  U.S.  Dept.  of  Agriculture, ARS

K-VI 1-24        Role  of Mathematical Models in Design and Selection of Best
                 Management Practices for Control of Pollution from Urban
                 and Urbanizing Areas
                 Vladimir Novotny,  Marquette University

K-VII 1-29       Management and Control  of Pollution by Urban Runoff
                 Richard Field, U.S. Environmental Protection Agency

K-VIII 1-6       Nonpoint Pollution Abatement Programs:  The Manager's
                 Dilemma
                 Frederick W. Madison, University of Wisconsin

K-IX 1-9         Institutional Arrangements for Nonpoint Pollution Control
                 Programs
                 Floyd Heft, Soil Conservation Society of America

K-X 1-14         Who Should Pay for Nonpoint Pollution Abatement?
                 Lawrence W. Libby, Michigan State University

-------
Panel Reports and Contributions

Panel I .- Planning Process for Nonpoint Pollution Abatement  Programs

P-I 1-4          Panel  Summary

P-I-A  1-7       Improvement of Environmental  Impact Assessment of Nonpoint
                 Sources of Water Pollution
                 Richard Homer, University of Washington

P-I-B 1-7        Quantification of Nonpoint Source Pollution
                 Jy S.  Wu, University of North Carolina


Panel II - Legislative  Means and Financing  of  NPS Control  Programs

P-II 1-3         Panel  Summary

P-II-A 1-6       Legal/Regulatory Approaches to Control  of Non-point  Sources
                 of Groundwater Contaminations
                 Robert Ehrhardtj Applied Management Sciences, Inc.

P-II-B 1-4       Urban  Wisconsin Should Join the Fight Against Nonpoint
                 Pollution
                 Kathleen M. Falk, Wisconsin Department  of Justice

P-II-C 1-10      Economic Incentives for Control of Nonpoint Pollution
                 Kathleen Segerson, University of Wisconsin

P-II-D 1-10      Legislation, not Litigation is the Solution to Non Point
                 Source Pollution
                 James  H.  Petersen, Milwaukee  Metropolitan Sewerage District
Panel III - Efficiency and Feasibility of Best Management Practices  for
NPS Control

P-III 1-4        Panel Summary

P-III-A 1-11     Evaluating BMP's in Pennsylvania's Conestoga Headwaters
                 Rural Clean Water Program
                 Bradley M. Crowder and C. Edwin Young,  The Pennsylvania
                 State University

P-III-B 1-12     Best Management Practices for Animal  Production
                 Jackie W.D. Robbins, Louisiana Tech.  University

P-III-C 1-11     Comparative Pollutant Removal Capability, Economics and
                 Physical Suitability for Urban Best Management Practices
                 in the Washington, DC Metropolitan Area
                 Thomas Schueler,  Robert Magill, Michael P. Sullivan,  and
                 Cameron Wiegand, Metropolitan Washington Council  of
                 Governments

P-III-D 1-5      Implementation of Urban Nonpoint Source Water Pollution
                 Abatement Programs
                 Robert P. Biebel and David B. Kendziorski, Southeastern
                 Wisconsin Regional Planning Commission

-------
P-III-E 1-17     NPS Pollution Control:   A  Consultant's  Perspective
                 Stuart G.  Walesh,  Donahue  & Associates


Panel IV- Organization and  Institutional  Arrangement  of  NPS Control Programs

P-IV 1-3         Panel Summary

P-IV-A 1-6       Organization and Institutional Arrangement of NPS Control
                 Programs
                 William J.  HorVath,  National  Association  of Conservation
                 Districts

P-IV-B 1-10      North Central States Conservation  Districts:  Effectiveness
                 of Regulatory Powers
                 Dean T. Massey,  University of Wisconsin

P-IV-C 1-2       Emerging Institutions for  NPS Control
                 Glenn E. Stout,-University of Illinois


Technical Reports and Case  Studies

T-I-A 1-10       Application of Geographical Information Systems  and Hydro-
                 logic Modeling to  an Agricultural  Watershed in Illinois
                 Ming T. Lee and  RodolfoCamacho, Illinois  State Water  Survey

T-I-B 1-19       Delineating Sources  of Agricultural  Nonpoint Source Pollution
                 W.T. Dickinson,  R.P.Rudra  and G.J. Wall,  University of
                 Guelph, Ontario, Canada

T-I-C 1-12       Coupling Nonpoint Pollution and  Water Quality Models:  An
                 Example for the  Green Bay-Fox River  Watershed
                 Thomas M.  Heidtke, Martin  T.  Auer, Raymond P. Canale, and
                 Theodore A.D. Slaweski, Michigan Tech.  University

T-I-D 1-12       Modeling Vertical  Flux of  Pesticides with CREAMS
                 R.A. Leonard, W.G. Knisel, and A.W.  Johnson, U.S. Department
                 of Agriculture

T-I-E 1-13       Nonpoint Pollution Abatement in  Tampa Bay
                 Bernard E. Ross  and  Mark Ross, University of South Florida

T-I-F 1-12       Pollution Abatement  in the Ruhr River Basin, West Germany
                 D.R. Albrecht and K.R. Imhoff, Ruhr  River Reservoir Manage-
                 ment Association, West Germany

T-I-G 1-12       Using Linear Optimization  for Urban  Nonpoint Source Poll-
                 ution Water Quality  Management
                 J.G. Garland III, F.S. Tirsch, C.A.  Markowski, and C.C.
                 Churn, Old Dominion  University

T-I-H 1-8        Design of Urban  Detention Basins for Water Quality Control
                 J. Bryan Ellis,  Middlesex Polytechnic,  England

-------
T-II-A 1-9       Using the Undiluted Event Mean Concentrations  to  Determine
                 Runoff Loads
                 Lei and L. Harms and Marsha Smith,  South  Dakota School of
                 Mines and Technology

T-II-B 1-12      Wet Weather In-Stream Water Quality Evaluation
                 John J.  Warwick, The University of Dallas  at Texas

T-II-C 1-18      Potential and Actual  Loading in a  Small  Mixed-Use Belgium
                 Basin
                 Yves Deleu, Belgium Institute for  Hygiene  & Epidemiology

T-II-D 1-11      Modeling Sediment Delivery from Field  to Channels: A Case
                 for Field Scale Data and Hierarchial   Data Structures
                 Kenneth Baun, Wisconsin Department of  Natural  Resources

T-II-E 1-13      Eutrophication Problems in North Carolina  and  Management
                 Approach for Nonpoint Sources
                 William A.  Kreutzberger, George T.  Everett and Alan Klimek
                 North Carolina Department of Natural Resources and Commu-
                 nity Development

T-II-F 1-9       Reducing Bacterial  Nonpoint Pollution  in Tillamook Bay,
                 Tillamook,  Oregon
                 James A. Moore and J.R.  Miner, Oregon State University

T-II-fi 1-14      Abatement of Nonpoint Pollution of Semi-Arid Streams Using
                 Livestock Grasing Management, Vegetation,  Instream Flow
                 Structures, and Beaver
                 Quentin D.  Skinner, Jerrold L.  Dodd, J.  Daniel Rodgers and
                 Michael  A.  Smith, University of Wyoming

T-II-H 1-17      Management Alternatives for Urban  Stormwater
                 Robert Pitt and Roger Bannerman, Wisconsin Department of
                 Natural  Resources

T-II-I 1-20      Nonpoint Sources Pollution of the  Venice Lagoon:  Perspec-
                 tives of Long-Term Abatement
                 F. Zingales, A. Marani, G. Bendoricchio, A. Rinaldo,
                 Universities of Padova and Venice, Italy

-------
                      REMARKS BY C. D. BESANDY, SECRETARY
                   WISCONSIN DEPARTMENT OF NATURAL RESOURCES
                TO THE NON-POINT POLLUTION ABATEMENT SYMPOSIUM*
     Welcome to Wisconsin on behalf of Governor  Earl.

     I apologize for the Governor's absence  and  convey  his  best  wishes  for  a
successful conference.

     Governor Earl was my predecessor at  the Wisconsin  Department  of  Natural
Resources (DNR) and as Secretary, he oversaw the  beginning  of  Wisconin's  non-
point pollution abatement program.  I commend his foresight and  thank him for
his continued budget support.

     Wisconsin is a leader in non-point source pollution  control,  not only
because of the financial and institutional commitments  we have assigned to  the
task, but because we have tackled the problem head-on from  an  interdisci-
plinary standpoint.

     The DNR is an integrated natural resources  agency  and  brings  to  bear a
variety of resources that are necessary to view  the non-point  issue in  a
total, holistic sense:  fish, wildlife, water quality,  toxics  management,
groundwater, forestry, water regulation,  etc.

     As important as non-point source control is,  it is more important  to view
the issue from an entire environmental systems standpoint.   The  variety of
disciplines represented at the conference and on  the conference  program is  a
good sign that the program's sponsors see the need for  such integration.

     It is also important to involve ecologists  and others  with  "a broader
view" in the non-point source decisionmaking process.   This has  not happened
much in the past, with engineers, agriculturalists, and bureaucrats dominating
program development and decisionmaking; but  it is  starting  to  happen.   I  want
to encourage that.

     It is appropriate that this significant non-point  source  pollution
abatement conference is held in Milwaukee.   Not only has Mayor Maier  placed
the issue high on his personal agenda, but the state is beginning  a major
rural-urban non-point source abatement program within the Milwaukee River
Basin.  Further, Milwaukee also has underway a major point  source  control
program that, when meshed with the non-point  program, will  make  the area  an
even more impressive showcase.  Finally,  numerous  individuals  have pointed  to
the important relationship between clean  water and the  region's  and state's
economic future.
*Text of a speech delivered at the Non-Point Pollution Abatement  Symposium.
                                   A-I-1

-------
      This conference is taking place within sight of the Milwaukee River,
 which we hope will be a national model for non-point source pollution control
 within a rural-urban watershed.  Why are the river and its clean-up important
 to  Milwaukee, to Wisconsin and to the conference?  The river should remind all
 within southeast Wisconsin that their economic, environmental and recreational
 fate  is tied together.  No one municipality or community can stand alone.  The
 830 square miles within the watershed reminds us all of the vastness and
 interconnectedness of our communities.

      The river represents a means for people to come together, recognizing
 that  divisive and unnecessarily bitter disputes in the past have cost the
 region time, money and lost opportunities.

      The river—for that matter,  water in general, as the Great Lakes
 Governors and premiers recently underscored with the signing of the Great
 Lakes Charter—represents significant, as yet not fully tapped, economic
 assets that can be enhanced and developed into more than a recreational
 asset.   They can be an important  selling point for the kinds of employees and
 employers who demand a quality working and living environment.  They remind us
 of  not  only Lake Michigan,  which  we are protecting through our point source
 pollution abatement program;  they remind us of the 84 smaller, inland lakes
 within the Milwaukee basin itself;  all recreational assets.

      For this conference, the River is important because it represents the
 complexity of issues facing us all  in non-point source clean-up.  Here are
 some  important points that show how the River's clean-up so adequately
 represents the challenges we all  face, no matter where we come from.

      1.   The benefits of the clean-up are difficult to quantify, yet we know
          they exist—in terms of  better quality of life,  increased economic
          opportunities and sustained agricultural productivity.

      2.   The technological challenges involved in the river clean-up are
          similar to challenges elsewhere.  Rural, agricultural sources have
          been identified and  must be addressed.   Suburban sources exist and
          must be addressed.   Urban  sources have been identified and must be
          cleaned up.

      3.   The institutional challenges involved in the river's  clean-up are as
          tough as  you will  find anywhere.   They are rooted in  years of
          disagreement over  numerous non-environmental issues and recent
          disagreements  over  the sharing of  the point source clean-up burden.

      The  vastness  of the river basin and the multiplicity of the problems
within  the basin have  pulled  us apart as  much  as brought  us together in the
past.    Examples  of this  are:

      1.  Non-point pollution  blame  has been heaped on upstream residents;
         while in  fact all in the basin are responsible for the problem and
         all  must  be involved in  the solution.

     2.  Many  diverse  communities live within  the basin.   There is suspicion
         and  distrust  among  these communities,  especially where there is a
         difference  in economic standing.


                                    A-I-2

-------
     3.  Past political and  economic wars—some  involving  litigation and  harsh
         words—are carried  over to the non-point  pollution  issues.

     4.  The misunderstanding  and  even mistrust  that  has often divided  rural
         and urban, urban and  suburban areas generally hangs  over  the non-
         point pollution issue, here in southeastern  Wisconsin and  elsewhere,
         too.

     As difficult as the problem may now seem  to Milwaukee and southeast
Wisconsin there are ample signs to indicate that the  job can  be  done.

     In the point source pollution area, seemingly impossible tasks have  been
achieved.  Some examples are:

     1.  The Wisconsin River in the central part of the state has  experienced
         a greater than 90%  reduction in biochemical  oxygen  demand  in dis-
         charges from pulp and paper mills and public wastewater treatment
         facilities during the last 10 years.

     2.  Wisconsin has highlighted more than 50  water quality success stories
         in a special report to Congress on progress  in the municipal con-
         struction grants program.

     3.  Wisconsin's most difficult point source pollution abatement program—
         right here in Milwaukee—is making good progress.

     There has also been progress in the non-point pollution  abatement  program
elsewhere in Wisconin.

     1.  In 1978, we became  the first state to enact  a statewide non-point
         pollution abatement program.  A 25-year strategy  was set to protect
         the 130 watersheds  that need attention.

     2.  Since initiation, 26  watersheds have  been identified as priority,  and
         planning and protection funding is underway.  In  some watersheds,
         like the Hay River in northwest Wisconsin, we already are  seeing
         progress in water quality improvement.

     3.  Complex intergovernmental and interpersonal  hurdles  have been  cleared
         in not only identifying priority watersheds, but  in  implementing a
         complex watershed management strategy.  Local government citizen
         advisors, federal agencies and other  state agencies  are involved.

     4.  While the Wisconsin program recognizes  the individual responsibility
         of the landowner to be a respectful steward  of the  land, it also
         provides some financial assistance to private property  owners  so the
         rest of society can share in protecting a common  public resource.

     For Wisconsin,  the nation, and the world, the need to wisely manage  and
protect our soil and water resources is a paramount issue.

     There is no question that the non-point pollution abatement program  is
"water quality driven."  It exists because of  water quality  laws and its
                                   A-I-3

-------
 benefits are calculated In terms of cleaner water for fishing, swimming,
 general recreation and public use.

      Another benefit of the non-point pollution abatement program is reduced
 soil erosion.  Much has been said recently about the farmer's plight.  If the
 farmers of our country lose their soil they won't be the only ones in
 trouble:  we will be, too.

      In Wisconsin total tons of soil lost from our fields has increased by 50%
 in the last  10 years.  In 1982, some 67 million tons of Wisconsin soil was
 lost to erosion.

      A third of Wisconsin's farmland is causing 85% of the soil loss.  The
 places where soil conservation experts say the loss is the worst is in the
 southern part of the state:  especially the hilly southwest, but also the
 gently rolling to flat southeast,  including the Milwaukee River watershed.

      The problem of agricultural soil loss is one that not only threatens
 Wisconsin, but the entire world and its ability to feed itself.  Whether it be
 in Wisconsin, or  on the Great Plains,  or in drought-stricken Africa,  soil
 erosion disaster  is being courted  by humankind.

      Farmers are  forced by the marketplace to make too many short-term
 economic decisions at a long-term  cost to the land.   Larger equipment and
 changes in farming practices have  resulted in less over-winter cover, and a
 decrease in  windbreaks and water retention practices.  Row crops such as corn
 and soybeans are  much more popular today than 10 to 15 years ago.   Land
 planted to row crops is much more  prone to soil erosion.   Outside  the United
 States,  people in need of firewood comb the countryside for anything  that will
 burn.   The land is stripped of trees that hold soil  and moisture in place,
 allowing deserts  to advance.

      Here at home, soil conservation at all levels has been starved for
 funding and  regulations.   The last  great  test of voluntary compliance is in
 the Chesapeake Bay region and 1 predict failure.  (Not enough attention has
 been paid to the  loss—in terms of  food production—of our vital coastal
 wetlands and estuaries.  This is another  example of  the potentially tragic,
 global  food-production consequences  if  non-point source pollution  is  not
 brought  under control.)

      Each of us needs to speak out  on  society's obligation to curb soil
 erosion  and  the failure of local,  state and national governments to engage the
 task simply  because they are  too short  sighted economically or meek politi-
 cally to  do  what  is right.

     The non-point source  pollution  abatement issue  and soil erosion  have some
 common problems and can draw  upon  some  similar solutions  in both rural  and
 urban areas.

     Politicians  need to  be the first  to  understand  the economic as well as
 esthetic value  of  natural  resource protection.   In Wisconsin this  is
especially important  because  not only was our state  built on its natural
resources, but  because  our  future may depend  upon how well  we steward and
allocate  them.
                                  A-I-4

-------
     Politicians  and  all  others  must  recognize  that in natural resources,  as
in business,  short-term gains  or expendiency  may  lead  to  long-term disaster.
American  business has gotten into trouble  by  adopting  a management strategy
based on  next quarter profit statements.   The American people  and especially
Wisconsin people  can  get  into  a  much  more  serious kind of trouble if they
squander  their land and water.

     The  watershed needs  to be recognized  and respected for what it is:  a
marvelous, interrelated environmental system  that demands comprehensive
management attention.  Failure to understand  the  linkage  among vegetation,
soil, water,  surface  activity  and human population pressure will delay or  deny
management success.   The  watershed must be viewed in a holistic fashion, with
humans as a part of that  whole.

     You  won't get something for nothing.   And  you can't  build Rome—or  clean
up the Milwaukee River watershed—in  a day.   The  degradation of America's  land
and water resources didn't happen overnight and those  who have profited  either
directly  or indirectly from the  degradation have  profited for  many years.   It
will take time and money—your money  and my money—to  clean up the mess  we
have made for ourselves.

     There are some of us who  won* t like others of us  for getting the job
done.  In addition to  spending funds  to save  our  soil  and clean up our water,
we need regulations—standards of social behavior that—if violated—will
result in unhappy consequences for the offender.   The  person—rural or urban,
private individual or  public official—who indulges in anti-social behavior
should be called on it and pay the public  price.

     We need to be proud  of our  successes  and brag to  others about our
accomplishments.  Our DNR recently published  a  report  on  Wisconsin's natural
resources, 15 years after the  first Earth  Day.  It felt good to list our
successes as an agency and as  a  people.  The  same should  be true for all who
take part in the effort to retain our soil and  clean-up our non-point source
water pollution.  Why:

     1.  The taxpayers need to know that government works;  in  Wisconsin  and
         elsehwere.

     2.  We need to praise and give recognition to those  private property
         owners and other who  pitched in and  made personal and financial
          sacrifices.

     3.  We need to market the fact that Wisconsin has something that other
         states fall short on:   good  water, a quality  living environment,  and
         a farsightedness that tells  investors, visitors  and others that we're
         looking beyond the next  quarter,  we're looking to the next century.

     I have several hopes for Milwaukee, for  Wisconsin, and for the entire
non-point pollution abatement effort,  here in the United  States and globally
as well:
                                    A-1-5

-------
      1.   That through this conference and the actions of the State of
          Wisconsin we will be recognized nationally and internationally as a
          leader in non-point source pollution control and long-term wise
          natural resource management.

      2.   That each of you at this conference gains knowledge and renewed
          conviction that the challenge of non-point source pollution control
          is a challenge that goes far beyond improving the esthetics of our
          natural environment.  It is a challenge that addresses the very
          ability of humankind to sustain itself through the wise protection
          and management of its life-giving soil and water.

      3.   That the federal government in Washington will take the issue of non-
          point source control as seriously as it took point source pollution
          control; and that all levels of government, as well as private
          landowners,  immediately engage the serious issue of soil loss,
          especially on our prime agricultural lands in areas affecting coastal
          estuaries.

      4.   That the land ethic of Aldo Leopold, who last Saturday was inducted
          into the Wisconsin Conservation Hall of Fame, will be sought out and
          embraced by  peoples of all lands at all levels.

      In conclusion, non-point source pollution control—the caring for our
soil  and  water and the respecting of the total interrelationship of all
elements  within our natural environment—is important to  Milwaukee, Wisconsin,
the United  States and the world.

     The  world's attention has recently been focused on the human disaster in
Africa where starving children have captured the hearts and moved millions of
global citizens to assist in relief efforts.  Yet the tragedy of Africa is as
much an environmental tragedy as a human one and in that  sense its enduring
consequences will be  felt long after the popular song—"We are the World,  We
are the Children"—has  passed from the  record charts.

     The  song goes like this:   "There comes a time, when  we heed a certain
call  ...  when the world must  come together as one.   We can't go on pretending,
day by day,  that someone,  somewhere,  will make a change."

     From the soil and  water  of  the world comes  food for  the people of  the
world.  To  the extent that  our soil and water are not wisely managed and are
lost to humankind, our  food and  our survival are threatened.

     History suggests that  there are  regions of  the world,  now deserts  that
once were gardens  of  plenty.   For whatever reasons—natural  and human-induced
—the deserts  are advancing,  not only on other continents but here in North
America.  We  must  ask ourselves  how long will we have  abundant  energy supplies
to mask the  consequences.

     As the  song goes:   "There's a choice we're  making, we're saving our own
lives."
                                   A-I-6

-------
Keynote Reports

-------
                NONPOINT POLLUTION  PROBLEMS  FROM  A PERSPECTIVE
                          OF A MUNICIPAL GOVERNMENT*

                   Henry W. Maier,  Mayor,  City  of Milwaukee
     I am most pleased to welcome all of you  to Milwaukee  and  this  most
excellent, high-level symposium  on  non-point  pollution.

     I want to thank Marquette University  for sponsoring this  event.  And  Dr.
Novotny, thank you for your role and your  efforts  in  planning  and organizing
this very important three-day symposium.

     Last May another seminar on the same  subject  was  held here  in  Milwaukee.
It was co-sponsored by Dr. Edgington's University  of Wisconsin-Milwaukee
Center for Great Lakes Studies,  the Milwaukee Metropolitan Sewerage District
(MMSD) and the Mayor's Task Force on Sources  of Pollution  from Outside the
MMSD.

     I organized that symposium  because I  felt there was a pressing need  to
launch a network that would act  as  a springboard to get action on the problem.
I was concerned that if significant steps  were not taken to deal with non-
point pollution coming into Milwaukee from up-river, that  our  own costly
point-source pollution abatement program would amount  to the civic  folly  of
our t imes.

     The MMSD has been operating under rigid  mandates  and  timetables set  down
by the state, the federal government and the  courts to clean up  local water
pollution.  But, little or nothing  was being  done  about up-river non-point
pollution.

     My concern was that after spending an estimated  $2.5  billion on
Milwaukee's point-source program that our  waters would still be  polluted.

     Part of the problem is that non-point pollution  is just not a  very
glamorous subject, particularly  for the layman.  I became  doubly acquainted
with that fact during the last presidential campaign.  I appeared in New  York
with a group of mayors to interrogate the  democratic candidates  for president
on their stands on various issues.

     I posed to them a question  on  non-point  pollution.  Although the question
had been stated in simple terms  and the candidates and audience  attending  the
forum had followed my line of questioning, reporters  for the New York Times
and The Milwaukee Journal later  wrote that my question was complicated.
*Text of a speech delivered at the Non-point Pollution Abatement Symposium.


                                   K-I-1

-------
      Following this incident I decided that it was time to redefine non-point
 pollution.   I came up with my own definition—unattended, up-river, up-stream
 pollution.   Because, in terms of our own situation, that is what it has been—
 unattended.

      It  is unattended because significantly, it had not been attended to by
 anyone in the past.   There were some faint efforts and little echoes related
 to  the problem,  but  no prior significant efforts.  It was unattended by the
 state, by the federal government, and the courts.

      Not so  for  point-source pollution.  Our own MMSB has had strict mandates
 and  strict timetables to meet.   It has amounted to a forced march on the
 problem  at tremendous expense.

      It  is estimated that by 1996, the pollution abatement program of the MMSD
 will  cost a  staggering $2.5 billion, conservatively.   City of Milwaukee
 taxpayers—in a  city containing 30% of the state's poverty—will pay about
 $500  million of  that cost.   That is enough to fund 7  years of our entire
 capital  budget.   That includes  schools, bridges, streets and alleys, sewer and
 water mains,  etc.

      So  we asked:  After all that money is spent, what will we have?

      Will we have  a  sparkling,  clean,  unpolluted Milwaukee River?  The answer
 is no.

      Will the river  be fishable and swimmable under standards called for by
 the Congress  and  the EPA?   The  answer  is  no.

      Will we have  finally ended any dangers of accompanying pollution related
 to Lake  Michigan?  The answer is no—we would not have done that sufficiently.

      What we will  have is  pollution flowing down from the north, continuing to
 pollute  the  Milwaukee River; continuing to pollute the very waters  we were
 spending $2,5 billion to clean  up.

      I saw the time  had come to get attention to the  problem.  I had outlined
 to the Common Council in my Charter Message! the previous year the extent of
 the problem.   I warned that action had to be  forthcoming from those with the
 power to  mandate endorsement and set timetables against up-river polluters.
 My warning got little attention.

      So  I decided  that  a major  Milwaukee  symposium on unattended, up-river,
 up-stream pollution  was  necessary if we were  going to get  everyone  into  the
act.  I might  add  it appears that all  you have to do  to get hasty action out
of the State  of Wisconsin is to  have the  Mayor of Milwaukee call for a
 symposium on  the subject.   As quickly  as  you  could say non-point pollution,
 the legislature drafted  a bill  designating  the Milwaukee River Basin as  a
priority  study area  and  ordering the development of a plan for action.
                                   K-I-2

-------
      I had  called on the state to develop a plan as far back as 1974,  but that
 plea  fell on  deaf ears.   So  has  my call  for mandated action by the state on
 non-point pollution.

      The prospectus  which has  been developed as a result of the recent state
 bill  could  provide a  solid foundation for action on non-point pollution.  It
 could,  if it  mandated clean-up.   It could, if the state guaranteed funding.
 But,  there  are no mandates,  no guarantees of funding.   While the prospectus
 does  set timetables  for  action paralleling those of the MMSD's point-source
 pollution abatement  program  the  strategies for achieving non-point clean-up
 are strictly  voluntary.

      The voluntary approach  simply is not good enough.   Previous efforts to
 get voluntary compliance have  had limited success.   I will say more about that
 a  little later.   It  is my contention that the problem will not be solved
 unless  mandates and  sanctions  are imposed similar to those imposed upon
 Milwaukee's point-source program.

      I  do not suggest that those up-stream simply be penalized.  Funds should
 be made available by  the state to get the job done.   We in Milwaukee are on a
 forced  march.  Why shouldn't those responsible for  non-pont pollution be
 forced  to clean up as well?

      During our symposium last May,  I suggested a way for the state to provide
 the needed  funding.   I said  that  the state should set up a special fund that
 would guarantee availability of  the money to do the job.  It would be a
 segregated  fund like  the Wisconsin Highway Fund.   It would alleviate the
 problem of  going  back to the Legislature every bienneium and convincing
 legislators who might have more  politically expedient uses for the money to
 provide the funds.

      We are going to  be  working  toward that end—to get legislation that will
mandate action and provide sufficient funds to get  the  job done.  I have no
 doubt  that  it's going to be  a  tough,  uphill fight.

      While my criticism  of the prospectus is aimed  at the voluntary clean-up
approach recommended,  the document is, as I said,  sound in other ways.

      And, I commend Kurt  Bauer and his staff of the  Southeastern Wisconsin
Regional Planning Commission (SEWRPC), the Department of Natural Resources and
its staff, and the members of  the Advisory Committee which assisted them,  for
 their efforts.

      I  might  point out,  to complete  the  study, they will need an additional
$350,000 from the  state  between  now  and  1987.   The  State Legislature and the
Governor should approve  those  funds.   I  urge them to do so.

      In the meantime,  our  own  city engineer Ed Laszewski,  will continue to
provide assistance as a  member of the Advisory Committee to SEWRPC and the
DNR.

      I  mentioned  the  symposium we held last year.   That effort resulted in
several very  positive  actions.  A resolution was  approved by the symposium
participants  that  has  led to establishment of a position for a District
                                   K-I-3

-------
 Director for Non-Point Pollution by the MMSD.  The new Director, Mr. Chris
 Potos,  is attending this conference.  He Will be on board next month (June
 1985)  and will begin working to get action on the problem.

      Secondly, at a meeting of the United iStates Conference of Mayors in
 Philadelphia last June, I sponsored a resolution calling for federal action
 that  would mandate states to submit plans ifor non-point pollution control
 before  the Senate and House which would achieve that.  Further, the bills
 would provide $150 million a year for actljon programs.  While that sum is
 nowhere near what is needed to get the job done, it would signify that the
 federal government recognizes the need,  tjnfo'fetunately, the fate of those
 bills may be determined by the White House1.  'It is my understanding that the
 Reagan  Administration opposes funds for the program and will be working
 against the legislation.  I would urge all of you attending this symposium to
 voice your support for the legislation and urge that the bills be strength-
 ened.

     The Conference of Mayors and a coalitibn'of national organizations has
 taken a strong stand in support of federal legislation requiring state
 mandated clean-up.  I have provided Dr. Novotny with copies of a February 21,
 1985  letter from the coalition to Sen. John Chafee,  chairman of the Senate
 Environmental Pollution Subcommittee.   It :outlines the coalition's position.
 I urge  you to write to your senators and representatives in Washington
 supporting that position.

     Let me discuss for a moment the state; bill I mentioned before.  That was
 the bill which led to the  recent prospectus report on the Milwaukee River
 Basin.   When the bill was  passed,  there was nothing in it addressing the need
 for timetables for unattended up-river pollution in areas where point-source
 pollution programs are underway or scheduled.  I was pleased,  however,  to find
 that the prospectus does call for implementing non-point source control within
 five priority watersheds in the basin on a schedule matching that of the
 MMSD's  point-source control program.

     The state's designation of the Milwaukee River Basin as a priority area
 for the study leading to the prospectus was a direct outgrowth of a task force
 I appointed in 1982.   The  Task Force was Charged with looking into the whole
 question of  pollution entering the Milwaukee River from up-stream.   It  was to
 determine whether implementation of the missive point source water pollution
 abatement  program by  the MMSD,  together wijth point source abatement programs,
 would produce streams and  water courses in the greater Milwaukee area that
 were "fishable and swimmable."  The Task Force concluded that  completion of
 the MMSD's  program would not by itself fully meet the objectives of producing
 "fishable/swimmable"  surface waters.

     As  a result,  the Task Force recommended the the MMSD petition the  state
DNR to  designate  the  Milwaukee  and Menomoriee River watersheds  as priority
watersheds.   It  also  urged the MMSD to seek to have  the Governor and the
Legislature  provide  the  necessary  funding for critical non-point pollution
abatement  projects, within a time  frame matching the Milwaukee District's
point-source  program.

     On  May  8,  1984,  just  3 days before the Milwaukee symposium, the Governor
signed  the bill  into  law designating  the  Milwaukee River Basin as  a priority


                                   K-I-4

-------
watershed.  As  I  said before,  my primary concern with the bill and the study
is  that  the action program contemplated is  one based on voluntary, rather than
mandatory compliance.

     The state  started a program of voluntary compliance in 1978 and its own
audit of that program,  released  in January  1984, indicated that the program
had only partial  success in getting polluters to participate.   It pointed out
that in  some watersheds participation levels  of upstream polluters were very
low, making it  extremely difficult to achieve water quality improvements.
And, even in those areas where participation  levels were higher, the audit
report noted that nonparticipation of just  a  few landowners in critical areas
can negate the  program which the majority is  carrying out.

     That is why  I have been against the voluntary approach.   It simply will
not work.  And  again  I  repeat:   It will be  a  tragedy to spend  all the money
that is  being spent on the District's point-source pollution abatement efforts
if  up-river pollution is still flowing down to us after the District's program
is  completed.

     Let me quote from the February 1985 issue of the Journal of the Water
Pollution Control Federation:

     "It is critical  that the  costs of not  addressing non-point source pollu-
tion control be considered.  Available data clearly show that  the effects of
non-point source  pollution are diluting the return of the billions of dollars
invested to date  to control  point sources ...  non-point source pollution
control  is a substantial barrier to achievement of the goals  of the Clean
Water Act."

     So,  while we and others are paying through the nose to build, operate and
maintain point-source systems, non-point pollution is negating our efforts.
In  our own case,  up-river polluters have the  nice option of "voluntarily"
installing pollution  abatement measures while we operate under rigid time-
tables and mandates.

     And,  if they volunteer, they will be doing it at an estimated 3% or less
than the  cost of  our  program,  even though they're responsible  for about 50% of
the problem.  Those estimates  are based upon  SEWRPC data contained in the
Mayor's  Task Force Report.   The  mismatch is as obvious as the  inequity.  There
is  no doubt that  this kind of  thing is being  repeated in a number of metro-
politan  areas throughout the country.   That is particularly true where cities
under clean-up orders have pollution coming in from rivers and streams that
cross jurisdlctional  boundaries.

     What is needed is  strong  federal legislation requiring the states to
force compliance  with Clean  Water Act  provisions by up-river/up-stream
polluters.  Without it  non-point pollution  will not be adequately controlled
or  abated.  In the meantime, those of  us under orders,  timetables and mandates
will continue on  the  forced  march to point-source clean-up at  a great cost.

     As  for non-point controls,  the big question is whether the legislature
would be  willing  to commit  the funds necessary to do the job,  even on a
voluntary basis.   It  has been  roughly  estimated that it could  cost between $15
and $20  million.   Frankly,  I don't know if  those are realistic figures.


                                  K-I-5

-------
Further,  there  are  no funding guarantees.  There is a draft plan.  We will
provide whatever  support  we  can to those responsible for seeing it through.

      In the meantime,  the symposium should bring added attention to the
problem and those dealing with it.  And, I urge all of those attending to read
Dr. Novotny's and Dr.  Gordon Chesters'  definitive work on the subject,
Handbook  of Non-Point Pollution.   It is an excellent book.   In it, they cite
the magnitude of  the  problem,  pointing  out that nearly half the polluting
sediments-j-nearly 2 billion  tons—result from runoff from agricultural lands.

      I might add  that  the head of  the Conservation Foundation, a national
organization that has  been trying  to persuade Congress to take action on non-
point pollution, has  stated  that the economic costs of rural soil erosion
totals between  $2.5 and $3 billion a year.

     And, the latest  federal Soil  Conservation Service report shows that in
Wisconsin about one-third of the land in the  state contributes 85% of the
erosion.  Our own studies have shown that tons of mud laced with pollutants
flow down from up-stream  and pollute the Milwaukee River and harbor estuary.
Professors Novotny's and  Chesters1  book also  contains some  very thoughtful
recommendations and strategies  to  deal  with the problem.   I am sure you will
be discussing some, if not all, of them during this symposium.

     Again, I want to  commend  Dr.  Novotny and Marquette  University for
sponsoring this significant  forum,  and  I wish you every  success in your
endeavors.
                                   K-I-6

-------
                Magnitude  and  Variety of  Problems  of  Non-Point
                     Pollution from Agricultural Practices

         John M. Harkin, Professor (Soil Science and Water Resources),
             Associate Director (Environmental Toxicology Center),
                        University of Wisconsin-Madison
     After about 20 years of exposure  to  the  ideals  of  environmental!sm,
modern  society now demands  that  soils  be  fertile  and that  waterbodies  and
waterways be oligotrophic,  that  is relatively infertile.   In  practice,  these
goals are incompatible.  The thermodynamic  forces of entropy  in nature dictate
that nutrient species present in or applied to  land,  being at a higher poten-
tial energy than lower-lying receiving waters,  must  gravitate toward the  lower
energy  regimes of streams,  rivers, lakes  and  groundwater.

     When a chemist looks at chemical  species and chemical reactions,  she/he
applies known physical constants to determine the amount of "free energy"—the
potential chemical energy—locked up in a particular chemical compound.
Whether this energy can and will be released  is determined by the availability
of chemical reactions to lead from this compound  to  products  of lower .free
energy.  Sometimes there is no way for such transitions to happen; sometimes
there is only one route, sometimes there  are  many competing routes, sometimes
there are many intermediate steps of progressively lower energies along any
one reaction pathway.  Often compounds of extremely  high energy content do not
lose their energy for lack  of a  convenient  pathway to follow.   In such cases,
the chemist applies catalysts to achieve  the  desired conversion, either
because she/he wants the reaction products  or desires to use  the chemical
energy.  The catalysts produce a reaction where none would otherwise occur or
speed up otherwise slow reactions.  An example of  this is  the oxidation of the
high-energy aromatic hydrocarbon solvents benzene, toluene or xylene.  Without
catalysis, these compounds  are stable and remain  unchanged.   In soils  or
water,  they are slowly degraded by microorganisms using oxidizing enzymes as
catalysts.  Here the release of  their energy  is so slow and gradual that  it  is
not particularly beneficial to the organisms  which accomplish the reactions.
They are concerned more with the destruction  of materials  which are toxic to
them.  However, when these  compounds, as  constituents of petroleum fuels, are
burned in furnaces or internal combustion engines, the release  of their energy
is fast and furious and can be applied to useful  purposes:  heating of homes,
generation of electricity or driving of engines.

     Conversely some chemical reactions occur spontaneously to  convert high-
energy to low energy products because of  the  ready availability of convenient
reaction pathways.  Rusting of iron is an example.   If undesired, such
reactions can be retarded by use of anticatalysts.   Good examples of this
phenomenon are the use of nitrite to retard spoilage of meat  and of antioxi-
dants to retard perishing of rubber and spoilage  of  food.  Car  tires last much
                                  K-II-1

-------
 longer nowadays since chemists recognized that ozone was the cause of cracking
 of  rubber  and  manufacturers  started putting antiozonants into rubber.

      Pollution of water by contaminants from nonpoint sources follows the same
 laws  of  thermodynamics as  chemical reactions.   Of  necessity, contaminants on
 upland sites have the potential to move to lower energy states in surface and
 groundwaters.   The only aspect of  this  process with which man can interfere is
 in  control of  the kinetics—i.e.,  the rate of  transport—by selecting agricul-
 tural practices which in effect act as  anticatalysts to contaminant  movement.

      With  this general physicochemical  concept: in mind, it is important for
 the soil conservationist/water quality  expert  to recognize this  natural law
 and to combine forces to determine what are the "reactions" which are of
 greatest qualitative  and quantitative concern  for  the preservation of highly
 productive fertile soils and clear, pristine surface waters or safe, uncon-
 taminated  groundwaters,  and  to decide what measures will function as good
 "anticatalysts" to retard  spontaneous pollution reactions.  In these delibera-
 tions,  it  Is important to  divorce  scientific thought from the emotional, knee-
 jerk  or  gut-feeling approaches to  water quality of the past.  We must consider
 not only the unaesthetic appearance of  eutrophication but less readily visible
 toxicological  concerns.  Some lakes—though pristine in appearance—may harbor
 greater  threats to human health and the environment in terms of  excess acidity
 or difficultly detectible  but biomagnifiable concentrations of pernicious
 mlcrocontaminants  such as  pathogenic bacteria  and  viruses;  toxaphene and other
 pesticides; volatile  organic compounds; polychlorinated biphenyls and dibenzo-
 furans;  or toxic metals, than lakes which look dirty and polluted.   Abundance
 of algae blooms may be a sign of a healthy lake!   Only by alloting appropriate
 emphasis to these  more important emerging concerns can scientists expect to
 generate the social pressures that will lead to fresh infusions  of funding to
 achieve  desirable  water  quality goals.

      Past  efforts  have largely curbed pollutant inputs from point sources and
 done  much  to reduce sediment and nutrient loadings from nonpoint sources to
 surface  water  and produced readily visible water quality improvements.  We
 have  made  great  progress toward ensuring that  the  nation's  waters are becoming
 "fishable  and  swimmable."  Concern must now be refocussed on whether it is
 safe  to  swim in  these  less murky waters or to  consume the fish we harvest from
 them.  The alliances  that  led to the first round of water quality improve-
ments—those with  groups or  agencies that sponsored primarily control of
 sediment and phosphate loadings—must be broadened to encompass  those that
espouse  techniques  to  control more insidious pollutants.   Only in this way can
we expect  to effectively establish pollutant control priorities  and  set
pollutant  standards and  criteria in a timely manner.   When  directed  toward
agriculture, nonpoint  source pollutant  control programs must couple  cost-
effective  means  of  minimizing water pollution  with least  disruption  to
efficient  farming  practices  and to production  of high-yield,  high-quality
agricultural crops.   This  approach will necessitate close interaction among
chemical industries,  federal and land-grant  university research,  and state and
county extension services.
                                  K-II-2

-------
     Fortunately most  of  the  infrastructure  for  this  collaboration  is  already
in place and is operating.  Just as necessary and  perhaps more  difficult  to
achieve will be an open-minded  receptiveness on  the part of  corporate  and
private farming enterprises to  take steps  to achieve  improved water quality
through modification of their farming  practices, especially  during  this period
of financial stress in the realm of agricultural production.   It  is up to the
ingenuity of the agricultural research and agribusiness communities to develop
ways to make both goals compatible.  Great strides have already been taken in
this direction, fortuitously  dictated  by other circumstances, such  as  the
energy crisis of the early seventies:   the increase in costs of tractor fuel
and fertilizers resulting from  the escalation of the  prices  of  petroleum
products provided a greater incentive  to the implementation  of  conservation
tillage than the loftier but  more esoteric goal  of water quality  enhancement
through sediment and nutrient control.  Who  cares?  The fringe  benefit of
efforts to reduce energy inputs to agriculture was tillage practices that do
such a good job of soil conservation and receiving water protection that  they
have become the cornerstone of  "best management  practices" for  nonpoint
pollution control.

     Thus, efforts to  protect water from nonpoint  pollution  from  agriculture
must adopt a forward looking  attitude  based  not on the past, but  on the
current state of agricultural practices.   To continue progress  in water
quality protection from nonpoint sources,  we must try to assess where  we  stand
at the moment and anticipate  where we  go to  from here.

     Agricultural pollutants  from nonpoint sources enter the environment  in a
diffuse but pervasive  manner.  The sources can be  land-based or airborne  and
the pollutants can be  transported in particulate, dissolved or  volatile
forms.  They can be naturally-produced or  man-made components.  Because of
their diffuse distribution they are difficult to isolate and often  appear in
low concentration, making their control a  complex task.  The fluxes of
contamination from land to water vary  widely in space and time, depending
primarily upon individual or  sequences  of  meteorological events,  so that
pollutant loadings associated with isolated  farming activities, derived from
whole or individual portions  of watersheds,  or introduced by long-range
transport from remote  geographic areas  become extremely difficult or
impossible to quantify.  Vast data bases of  information collected from
laboratory, field plot and field experiments and watershed monitoring  studies
have been assembled and a variety of clever  mathematical models generated to
evaluate such data to  aid in describing pollutant delivery systems  and
predicting benefits to be derived from their manipulation.  Yet the reli-
ability of these efforts is still too  questionable for them to  convince the
public and their elected representatives to  subsidize measures  which would
convert scientists'  fears, concern,  and suspicions into action  to support
curtailment of continued water quality  degradation.

     In these days of massive federal  government deficits, the  movement to
support water quality  protection is running  out of steam.  Section  208 of the
1972 amendments of the Federal Water Pollution Control Act provided funding
for the development of mananagement plans  to curtail water pollution from
nonpoint sources.   While this planning  effort was successfully  completed  the
follow-up implementation languished for lack of federal funding.  In only a


                                  K-II-3

-------
 few cases,  e.g.,  Wisconsin^ ' and Maryland' ' did state government jump in to
 fill the void by  creating state-supported nonpoint pollution control programs.
 Maryland understood better than most states through effective public relations
 and lobbying efforts to parlay their concerns into a $27 million federally
 funded program of research on the Chesapeake Bay estuary, which it followed by
 a  $36 million state program to curb pollution from both point and nonpoint
 sources.  In constrast, little has been done to utilize the findings of the
 PLUARG (Pollution from Land Use Activities References Group) studies and plans
 to implement measures to protect the Great Lakes.  The time is ripe for the
 Great Lakes  States to move forward in unison from the concerns of the past
 with regard  to agricultural nonpoint pollution — mainly BOD, bacterial, sedi-
 ment and phosphorus loadings — to those of the present and future, mainly toxic
 metals,  acidic deposition, pesticides, and organic supertoxicants such as
 PCBs,  PCDDs  (dioxins),  and PCDFs.

      The literature is replete with quantitative and qualitative information
 on pollutant losses from agricultural activities on a variety of scales — from
 field plot experiments through subwatersheds and watersheds to regions and the
 nation.   It  would be futile to try to summarize this information here.  A
 compact  summary is already available in the literature^ ' .  The nature and
 quality  of the data are highly variable „  but in summary the state-of-the-art
 is convincing enough to specify certain problems and to reveal that some
 measures being implemented to mitigate nonpoint pollution are indeed working
 effectively.  For example, conservation tillage has obviously reduced sediment
 loadings from agriculture.  Rather than recite a litany of  recognized
 pollutants,  and the successes achieved in managing them,  I feel it is more
 advantageous to examine what problems lie ahead.  Most important questions are
 whether  we want more of the same,  or whether we should redirect our attention
 to new or emerging problems, and in either case, who is going to foot the bill
 for research,  development, planning and mitigation programs.  We must start to
 examine  agricultural operations related to water quality  from new perspectives
 in the light of more recent scientific findings.  A good  way to do this is be
 using  a  few  of the prominent agricultural pollutants as examples.

      Concern with soil loss preceded concern with water quality deterioration
 by sediment  and any nutrients or pesticides associated with eroded particles .
 Measurement  of erosion rates on small plots of cropland,  although perhaps not
 a  good model of overall erosion potential because they neglect the long-
 distance delivery ratio,  do provide a means of pinpointing  "hydrologically
 active areas," i.e.  the areas most sensitive to constant  rill and sheet
 erosion.  It has  been calculated that half of the soil loss in 1977 occurred
 by water-induced  rill  and  sheet erosion from only 10% of  the cropland.   Such
 land was located  mainly in Mississippi Delta states (AK,  LA, MS), the South-
 east  (AL, GA,  SC,  FL),  Appalachia  (KY,  TN,  WV,  NC)  and the  Corn Belt  (IA,  MO,
IL,  IN, OH)^»  USDA National  Resource  inventories  for  1977,  as  revised in
1980 and 1984^ ' indicated that  the national  average  erosion rate  for  row crop
and small grain cropland was  5.4 tons /A/year.   Wind  erosion increases  this
average to 7 tons/A/year.  The  average rate of  soil  formation  is  placed  at a
mere 0.5 ton/A/year.  The same  data sources indicates that the "T value" — a
widely accepted criterion of  5  tons/A/year considered as a desirable objective
by the Soil Conservation Service and EPA — is  exceeded on over  112  x 10  acres
of cropland, including 33% of that in corn, 35% of that in cotton,  39% of that


                                   K-II-4

-------
in  sorghum and  44% of  that  in soybeans.   Clearly soil  erosion is  a national
problem with emphasis  in particular  regions.

     Using values for  the bulk density of soils, it  can be  calculated that a
soil loss of 5  tons/A/year  leads  to  loss  of  1  inch of  topsoil in  30-40 years,
depending mainly on  topsoil texture.  This may not seem like a significant
rate to a farmer in  Iowa growing  corn on  20-60 feet  of loess,  but it  is to
farmers in the  driftless area of  Wisconsin or  in Appalachia trying to eke out
an  existence on low  fertility thin soils  over  high bedrock.   These latter
areas are those that would  profit most from  erosion  control measures, but
present funding to curb erosion is not allocated by  prioritization.   It is
distributed equably  among states  and regions,  rather than by need or  the
prospect of deriving the maximum  benefit—i.e.  the biggest  bang for the
buck.  In any case,  there is  insufficient funding for  adequate programs of
soil conservation programs  anywhere!

     It will take imaginative new programs to  provide  appropriate funding for
soil conservation research  and  implementation  of conservation measures in a
prioritized plan.  One good way to achieve this would  be to follow the model
of  the Comprehensive Environmental Response, Compensation and Liability Act of
1980 and create a sort of Soil  Conservation  Superfund.   Although  nominally it
is  present-day  chemical manufacturers who pay  for cleanup of orphaned chemical
dumps through donations to  Superfund, regardless of  who the original  dump
operators were  and which companies deposited chemical  wastes in them,  in fact
it  is the general public which  is footing the  bill indirectly through higher
prices for the  chemical, textile, petroleum and other  products we use in
everyday life.  It seems appropriate that a  Soil Conservation Superfund should
be underwritten directly by the consumer,  in the form  of  a  small  checkoff tax--
say 1%—on all  food  products.   Such  a tax would be shared equably among the
beneficiaries of agricultural production—everyone according to his/her means.
Whoever buys choicer cuts of  meat and better quality fruit  and produce pays
higher prices and consequently more  Soils Superfund  tax.

     Regrettably there are  barriers  to implementation  of  such a scheme.  While
it would provide technical means  to reduce soil loss and  water pollution by
land erosion, its social and  political acceptance creates a "thermodynamic"
barrier.  Only  interaction between physical and social  scientists will develop
the catalysts needed to provide political decision-makers with the means to
provide the "activation energy" to make this reaction work.

     Fortunately we  are at a  point in the recognition  of  serious  water quality
problems associated  with materials other  than  sediment  or phosphorus  that may
help this reaction to go in a series of steps.

     Agriculturalists have some tendency  to try to justify  the erosion
loadings from their  land—or  at least to  place  them  into  perspective—by
comparing them with  erosion from  construction  sites, estimated at about 50
tons/acre or 85 million tons  annually from areas disturbed  for home construc-
tion (80,000 A) combined with highway construction and  other civil engineering
activities (1.6 million A).   This is actually  small  compared with 413 million
acres of cropland delivering an average of 5 tons/A/year  for an annual total
of  2.65 billion tons annually.  A bigger  concern than  the quantitative


                                  K-II-5

-------
 comparison is the fact that the eroding cropland soils are more heavily laden
 with nutrients (mainly nitrogen and phosphorus), pesticides (mainly herbi-
 cides) and, if manure and sewage sludges have been applied to the soils, toxic
 metals and bacteria.  Even in areas of relatively flat topography where little
 erosion normally occurs, or where wetlands act as sinks or sediment traps for
 these materials,  flooding or violent storms can catalyze a violent reaction
 and transmit pent-up pollutants with a scouring velocity to receiving bodies.

      The impact of sediment, or even of eutrophication-controlling phosphate
 may dim in importance compared to the effects of toxic contaminants from
 nonpoint sources.  Sport fisheries in many inland lakes and the Great Lakes
 are now suffering because prize catches are unfit to eat.  The EPA recently
 reduced the tolerance for PCBs, i.e. the concentration considered to be
 acceptable from a health safety standpoint, from 5 to 2 parts per million,
 defining that most Lake Michigan game fish are risky to consume.  Mercury
 levels from unknown sources in some Wisconsin lakes are so high that anglers
 have been warned  off.  PCBs in lake trout from Lake Superior are above the
 standard, fish from the Saginaw River, Saginaw Bay, and parts of South Lake
 Huron are polluted with PCBs,  PBBs and dioxin.  A fish consumption advisory is
 in effect for Lake Ontario because of mirex content—although this came from
 industrial rather than nonpoint sources.

      Being at the bottom of the thermodynamic progression, bottom sediments in
 the Great Lakes are apparently becoming the ultimate sink for many toxic
 metals and llpophilic ("fat loving") contaminants such as PCBs,  TCDD,  PCDFs,
 DDT and toxaphene,  which tend  to bioconcentrate in aquatic food  chains.
 Although these compounds are extremely insoluble in water, their vapor
 pressures are high enough for  them to enter lakes mainly through the
 atmosphere at a slow insidious rate.  In lakes they become associated with the
 organic fraction  of suspended  particulate matter and microbiota  (phytoplankton
 and zooplankton)  consumed by higher life forms and ultimately reach fish and
 man.   The best safeguard against long-term in--lake recycling of  these
 compounds is their burial in deep bottom sediments.  For this reason it would
 be  unwise to strive to completely eliminate sediment loadings to lakes,
 curtailment is a  cheaper and more prudent option from the standpoint of
 toxicant  immobilization.   Inactivatlon by sediment burial is the best  strategy
 for in-lake control of toxic metals as well as of supertoxic organics.

      The  guiding  principle  must be selectivity in the sediments  admitted to
 lakes—they should be those which will settle in the deepest sinks and which
 do  not  introduce  higher  loadings of priority  pollutants  than they  scavenge.
 Refinement  of these concepts will take further fundamental research.   Luckily
we  now  have good  models  for  the transport  of  many pollutants.  Acceptance  by
 the public,  regulatory agencies and decision-makers of the concept of  a
 tolerable  "beneficial"  sediment loading will  require a good  selling  job that
 physical  scientists  seldom  perform.

      Finally some deliberations about  groundwater pollution.   Nitrate
pollution of  groundwater  is  a  natural  phenomenon  that  is  greatly exacerbated
by  modern use of  synthetic  fertilizers,  especially in areas  of high  ground-
water tables  underlying  coarse-textured soils  or  fragmented  or creviced
bedrock.   Irrigation  in  such areas  can compound the problem.   Here use of  a


                                   K-II-6

-------
tolerance—EPA'a  primary  drinking  water  standard of  10 mg/L of  nitrate
nitrogen—has served to eliminate  consequences,  even if the risk per se
remains undiminished.  Where  the standard  is  exceeded,  parents  are advised to
provide water from nitrate-free sources  to  infants—the only population at
risk.  This strategy has  been effective:   cases  of nitrate toxicity are so
rare that they warrant publication in  public  health  journals^ '.   We have  just
conducted a survey of all pediatricians  and general  practitioners in
Wisconsin.  Only  20 of them reporated  ever  having seen a case of  infant
methemoglobinemia, none of them lately.

     In this case, fundamental research  coupled  with toxicological data and
regulatory guidelines have effectively eliminated the consequences of the
health risk.  The same approach can be used to cope  with emerging problems of
pesticides and volatile organic compounds  in  groundwater.   Some basic
principles are worth considering in this connection.   The  only  pesticides  that
will leach through soils  are  those that  have  high water solubilities and low
affinities for sorption onto  soil  particles.  Most pesticides do  not exhibit
these properties.  That is why there was formerly no requirement  for leaching
studies as part of the environmental fate  information required  by the EPA  for
pesticide registration.   Pesticides were not  expected to wash through the  soil
with percolating  water; indeed the earlier  generations  of  lipophilic pesticide
compounds did not.

     Only after the episodes  of incursion  of  aldicarb residues  into ground-
water on Long Island and  in Florida, Wisconsin and a few other  states did
information on pesticide  leaching  become part of the standard registration
protocol.  Actual leaching experiments need not  be conducted nowadays;  on  the
basis of chemical and physical properties  of  the active ingredients, accurate
predictions can be made as to whether  pesticides will leach or  not under
conditions of normal agricultural  use.  Advances in  instrumentation for
chemical analysis of pesticides in water have made it possible  to detect
extremely minute  concentrations of pesticides in groundwater.  Depending on
the compounds concerned and the sophistication of the method and  detector
system used, levels as low as  parts per  billion  or parts per trillion can  be
accurately quantitated.   The  EPA created short-lists  of compounds they
considered likeliest to leach.  Recent surveys and monitoring of  groundwater
using reliable methods to look for these compounds have not revealed evidence
of extensive contamination of  groundwater  or  field drainage with  significant
levels of pesticides, other than aldicarb  residues and  minor traces of  other
compounds.  To appreciate the  significance  of these  findings, toxicological
data must be considered.

     Aldicarb residues in groundwater  generated  concern mainly  because
aldicarb has such a low LD5Q  (the  dose lethal to 50%  of laboratory animals in
acute toxicity tests):  about  1 mg/kg  of body weight  in rats.  Fortunately
aldicarb residue do not exhibit any mutagenic, teratogenic,  carcinogenic or
other chronic/subchronic  toxic effects.  A "tolerance"  of  10 parts per billion
has therefore been set for groundwater; this  is  a nonbinding "health advisory"
suggested by EPA's Office of  Drinking  Water.  New York adopted  a  level of  7
ppb based on a value suggested by  a National  Academy of Sciences^ ' panel
before the pesticide was  actually  found  in  groundwater.
                                   K-II-7

-------
      Aldicarb is in effect one kind of worst-case situation:   it has a high
 water solubility,  is poorly sorbed by mineral soils and has a low LD^Q.
 Fortunately,  most  of the pesticides used in agriculture do not share these
 features:   they are less likely to leach,  and their acute toxicities are
 weaker.   Whether a pesticide is toxic to man or wildlife depends largely on
 its  mode  of biochemical action.  Chemists  try to design pesticides which are
 highly toxic  to pests but not to desirable plant and animal species.  They
 therefore  exploit  the differences in anatomy and physiology of the species of
 concern.   Insects  are different enough from vertebrates that  many pesticides
 specifically  toxic to insects but relatively innocuous to man and animal,  bird
 and  fish  species can be created.  The differences in physiology between higher
 animal species  and plants or fungi and molds are even greater.  Consequently
 in general herbicides and fungicides tend  to be much less harmful toward man,
 animals and fish than insecticides.

      Herbicides are the most commonly used pesticides:  nationwide,  farm use
 in 1980 was 445 million pounds versus 306  million pounds of insecticides,  and
 95 million pounds  of fungicides, rodenticides,  and other pesticides.  Farm use
 was  about  72% of total use.   Expanded use  of conservation tillage practices
 tends to  increase  use of herbicides  and fungicides—the less  hazardous
 pesticides.   Expanded use of integrated pest management is tending to reduce
 all  pesticide use,  but especially that of  insecticides, the most hazardous
 poisons.   Banning  of most of the persistent  organochlorine insecticides,  DDT,
 aldrin, dieldrin,  hexachlorobenzene, lindane,  etc. and some soil fumigants has
 stimulated industrial research for new,  more selective, superior pesticides.
 Some herbicides are presently under  development which promise to be  effective
 at application  rates which are only  a fraction  of those used  with the current
 products.  Thus, water quality problems with pesticides are expected to
 decrease.  Only those with the organochlorine  "antiques" remain a major
 continuing threat.   Ironically,  although banned in most western industrialized
 nations, the organochlorine  pesticides,  with DDT at the forefront, are still
 the  most widely used insecticides worldwide.  They are cheap  and easy to
 manufacture and are therefore still  popular  in  developing countries.   Long-
 range atmospheric  transport  is therefore likely to continue to add organo-
 chlorine pesticide  loadings  to U.S.  water  bodies but  at much  lower rates  than
 occurred during and following their  use—and overuse—in the  USA.

      DDT was banned,  not  because of  toxicity toward humans, but  because of its
 chronic toxic effects on widlife,  particularly  bird reproduction. A strong
 stimulus for the reappraisal of  the  environmental effects  of  organochlorine
 pesticides such as  DDT came  from important books such as Rachel Carson's
 "Silent Spring."  Reproductive problems  in mink fed on mink farms with Great
 Lakes  fish, once wrongfully  ascribed to  DDT  contamination in  the fish,  led to
 recognition of  the  chronic toxicity  of  PCBs, hitherto considered harmless
 because of their low acute toxicity.   High incidences of crossbill malforma-
 tions in herring gulls,  cormorants and  terns around Green  Bay  appear  to be
 associated with PCBs  or  other organic supertoxins (e.g.  TCDD  or  TCDF).

      Episodes like  these  have  led  to  close scrutiny of  the  chronic toxic
effects of pesticides.   Continuing tests of  chronic toxicity  may lead to  some
surprises.  One  is  already happening.   It has been known since  the middle  70's
that traces of  pesticides  occurred in cropland  runoff.   For example,


                                   K-II-8

-------
measurements  in Western  Iowa'  ',  revealed average  losses  for  the years  1973-75
of 0.21, and  0.15% of the organophosphorus  insecticide  fonofos  applied  at  a
rate of  2  Ib/A in the sediment  and water  fractions of runoff  under  conven-
tional,  till-plant and ridge-plant tillage  systems used for continuous  corn.
Corresponding losses of  the  triazine  herbicides  cyanazine and atrazine  were
0.28 and 1.81%.  Losses  of alachlor carboxyanilide herbicide  popular  for use
in corn  was 0.19 and 0.77% in the sediment  and water.   In terms of  the  acute
toxlcitles of these compounds,  one need not be concerned  with these minor
amounts.   Greater concern should  be with  the massive loss of  topsoil  (a silt
loam from  12-18% slope)  observed, equivalent to  about 23  and  10 tons/A  for the
first two  systems.  But  now  continuing tests have  shown for the first time
that alachlor causes cancer  in  some strains of mice.  Though  alachlor has  a
solubility in water of only  33  mg per liter versus 33 g per liter for
aldlcarb, meantime alachlor  has also  been detected in groundwater at  two sites
in Wisconsin  (probably not as a result of field  use) and  in samples of  tile
drainage in Ohio.  Canada is presently taking action to ban alachlor  use.

     Obviously, in the light of this  new  information, the use of this pesti-
cide must be  reexamined.  EPA is  subjecting it to  a "special  review."  Is  it
likely to  go  the way of  DDT? Not necessarily.   Better  understanding  of the
mechanisms of carcinogenicity are leading to cooler appraisals  of cancer risk,
so that  banning of useful products is not an automatic  consequence  of positive
results  in animal tests.  The Delaney Clause, which once  banned use of  any
food additive which produced any  cancer in  any test species did not hold up in
the case of saccharin as a result of  public pressure for  its  continued  use.
The case was made that the benefits of the  compound far outweighed  its  minute
cancer risk.  A popular  analysis  of national policies in  cancer research and
prevention' ' forms an antipode to "Silent  Spring" and  documents justifica-
tions for  the new regulatory approaches of risk  assessment and  risk manage-
ment.  The Office of Science and  Technology Policy recently published its
report on  the guiding principles  to be used by all federal government
regulatory agencies in dealing with carcinogenic risk *•  '.

     USA policy is therefore developing in a more  objective scientifically
considered manner independent of  Canadian policy and political  pressure.   In
the past, Canada has strongly influenced  USA policy with  regard to  phosphate
loadings to the Great Lakes.  It  is lobbying strongly to  influence  decisions
concerning sulfur dioxide emissions in the USA.  No doubt it  would  like to see
its actions on alachlor  duplicated in the USA.   This may  not  happen.  However,
the alachlor situation does  seem  to justify more intense  efforts to control
nonpoint pollution from  agriculture.

     If  nonpoint pollution from agricultural sources is to be further reduced,
there has to be funding  for  the effort and this  must be clearly justified. It
will be  necessary to reappraise the goals, and to  return  to my  thermodynamic
model to determine what  compounds, what reactions  and what reaction products
are of interest, which processes  need catalysts, and which ones need  inhibi-
tors.  Priorities must be decided.  Which loading  is more important:  5  tons of
sediment, versus 20 pounds of nitrogen, versus 1 pound  of phosphorus, versus
0.1 pounds of herbicide, versus 0.05  pounds of insecticide, versus  0.01 pounds
of cadmium, versus 0.001 pounds of PCB, versus 0.0001 pounds  of dioxin, versus
10° bacteria, or 105 PFU (plaque-forming  units—a  measure of  viruses) per


                                   K-II-9

-------
 acre?   Use of sediment and phosphate loadings as a management tool in policy-
 making  was useful in the first round of water quality protection from nonpoint
 source  contaminants.  Now is the time to review and realign our efforts and
 substitute the uncertainties of the present by a new "chemistry" of the
 future—a coalition of physical, biological and social sciences that will make
 reactions happen.
                                   References
 1*  J. G. Konrad,  J.  S.  Baumann and S.  E.  Bergquist.   Nonpoint pollution
     control:   The  Wisconsin experience.  J.  Soil Water Conserv. 40(1):55-61
     (1985).

 2.  W. L. Magette,  R.  A.  Weismiller and K. C,  Gugulis.  Saving the
     Chesapeake:  Maryland's agricultural education program.   J. Soil Water
     Conserv.  40(1):79-81  (1985).

 3.  R. C. Loehr.   Characteristics  and comparative magnitude  of nonpoint
     sources.   J. Water Poll.  Contr.  Fed.  4£:1849-1974 (1974).

 4.  S. S. Batie and R. G.  Healy (eds.).  The  Future of American Agriculture
     as a Strategic  Resources.   The Conservation Foundation,  Washington,  D.C.,
     1980.  90 pp.

 5.  U.S. Environmental Protection  Agency.  Nonpoint Source Pollution in the
     U.S.  Office of Water  Program  Operations,,  Water Planning Division,  EPA,
     Washington, B.C.,  4 chapters + 3 appendices,  January 1984.

 6.  D. Busch  and M. Meyer.  A case of infantile methemoglobinemia  in South
     Dakota.   J. Environ. Health 44(6);310-311  (1982).

 7.  National  Academy of Sciences.   Drinking Water  and Health.   Vol.  1,  pp.
     635-643,  Washington, D.C. (1977).

 8.  J. L. Baker and H. P.  Johnson.   The effect  of  tillage systems  on
     pesticides in runoff from small  watersheds.   Trans.  Amer.  Soc. Agric.
     Engrs. 22:554-559  (1979).

 9.  E. Efron.  The Apocalyptics.   Cancer and the  Big  Lie.  Simon & Schuster,
     New York, 1984.  589 pp.

10.  Office of Science  and  Technology Policy.  Cancer  risk from chemicals:
     Guidelines.  Fed.  Registers  50(50)-.10372-10442  (1985).
                                  K-II-10

-------
                  MAGNITUDE AND PROBLEMS OF NONPOINT POLLUTION
                        FROM URBAN AND URBANIZING AREAS

                                 Gary L. Oberts
                          Senior Environmental Planner
                  Metropolitan Council of the Twin Cities Area
                               St. Paul, MN 55101


                                    ABSTRACT

Pollution generated from urban and urbanizing area surfaces can be a signifi-
cant portion of the total pollution entering a region's receiving waters.
These nonpoint sources of pollution typically are extremely variable in char-
acter, as they are driven by rainfall and snowmelt events of variable magni-
tude.  Pollutant deposition and erosion rates are similarly variable.

Sources of contaminants contributing to urban nonpoint pollution include street
surface dust and dirt, atmospheric fallout, construction activities, vegetative
matter, chemicals from various uses, and miscellaneous sources such as rooftops
and channel erosion.  These contaminants result in pollutant concentrations and
loads that often exceed wastewater treatment plant outputs and recommended
guidelines for good quality water.  Receiving water biota can be severely
impacted by urban runoff, and groundwater can be impacted by the introduction
of chemicals at above-background levels.  Another impact often overlooked is
the physical destruction that results from increased amounts of urban runoff;
such destruction includes increased channel scour and erosion, flooding,
devegetation, sedimentation and debris deposition.

Efforts to properly address urban runoff must strive to adequately identify the
problem and its causes before any management can be effective.  Management
efforts should focus first on developing areas where maximum pollution controls
can be most easily implemented on sites where they are most needed.

Keywords: Urban runoff pollution, nonpoint pollution, stormwater runoff
quality, construction erosion.
                                   K-III-l

-------
INTRODUCTION

Defining the problems resulting from urban runoff is an elusive task, requiring
that an "impact" be seen.   In general terms, an  impact is seen if there is a
loss or curtailment in the  beneficial use of a water body.  A beneficial use,
in turn, is a  local perception and could include such things as body contact,
aesthetics, recreation, fishing, domestic or industrial consumption, or
boating.  The  task is elusive because it requires describing a phenomenon that
is a function  of several extremely variable inputs (precipitation, surface area
pollutant deposition, soil  erosion, complex overland and channel hydrology,
pollutant transport) that might have a measurable impact on a receiving water.

This paper is  a cursory review of literature that has in some way defined the
pollution character of urban and urbanizing area runoff.  The effort relies
quite heavily  on the results of the recently completed U.S. EPA Nationwide
Urban Runoff Program (NURP), on various NURP site studies and on water quality
monitoring results from Metropolitan Council efforts in the Twin Cities region
of Minnesota.  The actual documentation of "problems" from urban runoff is
scarce because it has only  recently been recognized as a problem, meaning
little good data exist; the effects are not well understood and are often
masked by point source pollution; the term "impact" is not universally agreed
upon; and receiving waters  are not well defined  (Heaney and Huber, 1984).

The best approach for addressing the topic of nonpoint pollution from urban
runoff is to adopt a planning strategy that consists of defining the problem,
determining the source of the problem, exploring methods for managing the
problem, and evaluating the results of the management program to assure
progress and make adjustments.  The information  that follows will assist water
managers with  the first two of these steps.

FACTORS CONTROLLING URBAN NONPOINT SOURCE POLLUTION

Runoff quantity

The mechanism  for mobilization of pollutants in  an urban environment involves
the generation of runoff from rainfall or snowmelt at a level with sufficient
energy to suspend or solubilize pollutants.  The total load or mass of a
pollutant in runoff is the  product of the pollutant's concentration (mass per
unit volume) times the flow volume carrying the  pollutant.  Research into the
dynamics of urban nonpoint  pollution loading has shown that typically the flow
factor plays a dominant role in loading behavior because it is the driving
force in the mobilization of pollutants (Bannerman et al., 1983; Colston, 1974;
Konrad et al., 1978; Oberts, 1982; U.S. EPA, 1983).  Concentration values might
be as variable as flow, but loading seems to mimic the pattern established by
flow, as shown in the load-hydrograph in Figure  1.  This behavior points very
clearly to the need to undertake comprehensive surface water management as a
first step in  urban/urbanizing area nonpoint pollution control.

The approach toward control of urban runoff has  undergone a large change in the
past 30 years.  Formerly the method used to handle surface water runoff was to
capture it and remove it as quickly as possible.  This approach, however, led
to many related problems downstream.  A graphic prepared by Tourbier and
Westmacott (1974) (Figure 2) shows how runoff becomes increasingly significant
as the amount of urban development (that is, paved surfaces) increases.  Some
of the washoff from impervious, paved surfaces drains to pervious areas or to

                                    K-III-2

-------
areas of  closed drainage that do  not  contribute to any receiving water.  The
most important  factor in loading, then,  has been found to be  the total
"hydraulically  connected" or "effective  impervious" area, rather than the total
impervious  area (Bannerman et al., 1983;  Miller and Mattraw,  1982;  Pitt and
Bozeman,  1982;  Sonzogni  et al., 1980).
               LL
                                         Time
                                                                   s
              FIGURE 1. FLOW, CONCENTRATION AND LOAD BEHAVIOR

                               DURING A RUNOFF EVENT
           40% Evapotranspiratlon
       10% Runoff
NATURAL
GROUND
 COVER
  25% Shallow
   Infiltration
             25% Deep Infiltration
                                                    38% Evapotranspiration
                                                20%
10-20% IMPERVIOUS
    SURFACES
                  21% Shallow
                   Infiltration
                           21% Deep Infiltration
            35% Evapotranspiration
   20% Shallow
    Infiltration
             15% Deep Infiltration
                                                   30% Evapotranspiration
                                              55% Runoff
                                                              75-100% IMPERVIOUS

                                                                  SURFACES
                   10% Shallow
                    Infiltration
                            5% Deep Infiltration
           FIGURE 2. EFFECT OF GROUND COVER ON URBAN RUNOFF
                       (from Tourbier and Westmacott, 1974).
                                     K-TII-3

-------
The volume and rate of urban runoff have impacts other than transporting
pollutants.  Porcella and Sorensen (1980) report that flood stages can be
increased by a factor of two to five as development occurs.  This increased
volume and the accelerated rate at which it can occur result in increased
channel erosion and bottom scour, lower baseflow because of decreased
groundwater recharge, and runoff exposure in areas previously unaffected.

The extreme variability in precipitation and snowmelt behavior add to the
complexity of runoff characterization.  Some generalization can, however, be
made based on the fact that on a long-term basis, precipitation character can
be statistically defined.  Rainfall-runoff coefficients can be generalized
based on actual and modeled runoff behavior.  Konrad et al. (1978), Oberts
(1982), and U.S. EPA (1983) have shown that the percentage of runoff increases
with the percentage of effective impervious area, ranging from 80 to 90 percent
for moderately large events in watersheds with over 90 percent effective
impervious area, to less than 10 percent for small events in low effective
impervious areas.  Typically for urban areas, most runoff comes from the
effective impervious areas with a minor amount contributed from pervious
areas.  In urbanizing areas and in developed areas during the period of frozen
ground, runoff from pervious areas becomes a far more significant contributor
to watershed runoff.

Runoff quality

Runoff picks up material on urban/urbanizing surfaces and moves it to receiving
water.  U.S. EPA (1984) in a report to Congress stated that urban nonpoint
source water quality problems affect 20 percent of the nation's river miles and
occur at some level in one-half of the nation's river basins.  Six of 10 U.S.
EPA regions and 36 of the 50 states have similarly identified urban/urbanizing
area runoff as a major problem.  The remainder of this paper focuses on the
pollutants that cause the largest receiving water problems in these areas.
Essentially everything that comprises the urban/urbanizing surface, or is used
within these areas, can be expected to occur in some form in runoff.  A list of
those pollutants and the problems they cause follows:

o   Solids—Inorganic (sediment) and organic debris can be moved by runoff
    either in dissolved or particulate form.  The particulate, suspended solids
    are the most significant problem commonly associated with urban/urbanizing
    runoff.  The problems caused by suspended solids in transit and after
    settling include turbidity, aquatic biota habitat destruction, transport of
    adsorbed pollutants, benthic accumulation, and direct adverse impact on
    aquatic organisms (Lager et al., 1977; Thronson, 1976).

o   Nutrients—Of particular importance are phosphorus and nitrogen.  These
    nutrients accelerate eutrophication in quiescent waters, exert toxic
    effects on aquatic organisms (ammonia), produce highly organic sediments,
    affect dissolved oxygen concentration, aid in the shift toward less
    desirable aquatic organisms and impact groundwater (nitrate) (Lager et al.,
    1977).

o   Toxicants—This category includes heavy metals, pesticides and organic
    chemicals.  U.S. EPA (1983) has listed 129 "priority pollutants", 77 of
    which were detected in EPA's recently completed NURP studies.  The effects
    of these chemicals are related to ill-health of humans and aquatic life.
    The toxicity of the various chemicals depends on hardness of the water,

                                    K-III-4

-------
    biological species being affected, degree of risk accepted and pollutant
    concentration.  Accumulation of toxicity in the environment, for example,
    heavy metals in sediment, means that they remain available for extended
    periods of time.

o   Oxygen-demanding substances—Organic and other oxidizeable material (ex.,
    iron) exert a demand for oxygen in order to degrade.  The measure of oxygen
    demand is termed biochemical oxygen demand (BOD) or chemical oxygen demand
    (COD): the latter is best suited to description of urban runoff impact.
    The effect of these pollutants is the depletion of dissolved oxygen,
    resulting in stress or death for aquatic organisms and generally degraded
    water quality (Abernathy, 1981; Keefer et al., 1979).

o   Bacteria and virus—Numerous bacteria and virus strains occur in high con
    centrations in urban runoff, but most studies measure bacterial pollution
    through the fecal coliform count.  Fecal coliforms are excreted from the
    intestinal tract of warm-blooded animals, and serve as a less-than-perfect
    indicator of human pathogens.  The direct impacts of bacteria and virus are
    related to human health and disease (U.S. EPA, 1983).

o   Chloride—The last major urban runoff pollutant of concern is chloride
    which results from the application of road salt.  High chloride levels in
    groundwater present health risks.  Elevated chloride levels can also
    inhibit the circulation of water in lakes (Lager et al., 1977).

Some of the above pollutants present a water quality problem in the short-term
(toxicity, turbidity, oxygen depletion), while others take longer to become a
threat (accumulation of metals, benthic oxygen depletion and nutrient
enrichment).  Specifics of pollutant levels will be covered in a later section.

SOURCES

The sources of the pollutants just itemized are numerous.  The categories that
follow are those best representing urban and urbanizing area activities that
contribute nonpoint pollutants.

Street surfaces

Dust, dirt and debris accumulated on impervious surfaces contribute substantial
amounts of pollution when runoff events wash these surfaces.  Material
deposited on street surfaces typically includes combustion emissions from motor
vehicles and industries, wind blown dust and dirt, traffic deposits (tires,
clutches, brakes, fluid, leaks, etc.), vegetation, material wash-in, roadway
material, anti-skid grit and deicing chemicals, and litter (APWA, 1969; Novotny
and Chesters, 1981; Pitt and Amy, 1973; Sartor and Boyd, 1972; Shaheen, 1975).
Table 1 summarizes the data presented in the above referenced efforts relative
to solids accumulation rates and associated pollutant concentations (Manning et
al., 1977).  This table reflects the material available for washoff from street
surfaces.  Bannerman et al. (1983) point out that a portion of this material is
very likely to be washed-off during melt and rainfall events.  They report in
their Milwaukee NURP study that street loads of solids and toxics are at their
peak in the snowmelt and early spring rainfall runoff.  Pitt and Bozeman
(1982) report after studying various sources of urban pollution that parking
lots and street gutters contribute the greatest loads of pollutants.  They also
indicate that 75 percent of street surface pollutants and 50 percent of parking

                                    K-III-5

-------
              TABLE  1.  SUMMARY  OF  MEAN  SOLIDS ACCUMULATION RATES AND
                    MEAN ASSOCIATED POLLUTANT CONCENTRATIONS
                           (From Manning  et  al.,  1977)
     Pollutant

Solids  (kg/curb  km/
        day)
BOD5
COD0
Total nitrogen (TN)
Dissolved phosphorus
(DP)
Cadmium (Cd)
Chromium (Cr)
Copper  (Cu)
Nickel  (Ni)
Lead (Pb)
Zinc (Zn)
Fecal coliform
(geometric mean)
   Single
   Family
Residential

     182

   5,260
  39,250
     460
      49

     3.3
     200
      91
      38
   1,570
     310
  82,500
 Multiple
  Family
Residential

     157

   3,370
  41,970
     550
      58

     2.7
     180
      73
      18
   1,980
     280
     925
Transpor-
 tation*
  2,300
 54,000
    235
     80
    120
    190
 12,000
  1,500
 38,800
Commercial Industrial
      45

   7,190
  61,730
     420
      60

     2.9
     140
      95
      94
   2,330
     690
  36,900
   288

 2,920
25,080
   430
    26

    3.6
    240
     87
     44
  1,590
    280
 30,700
*Data from one study only (Shaheen, 1975).
lot pollutants are eventually delivered to catchment outfalls.  As a result of
this high delivery rate, they found that the two areas contribute 50 percent of
the total solids, 85 percent of the TP, 80-100 percent of toxic metals, and
close to 100 percent of the fecal coliforms delivered to urban catchment
outfalls.

Most of the polluting material on street surfaces is associated with fine-
grained deposits rather than with the coarse-grained sand and grit.  Table 2
shows the results of a study by Sartor and Boyd (1972) on the percent of street
surface pollutants by particle size.  Similar work on particle size distribu-
tion from other areas has verified the general pollutant distribution, that is,
concentration of pollutants in this finer-grained fraction (less than 246
microns).  For example, 92.2 percent of the phosphates and 73.0 percent of all
toxic metals are associated with 43.4 percent of the total solids (less than
246 microns).  Bannerman et al. (1983) documented the phenomenon of fines-
associated pollutant wash-off, showing that most pollutants of concern peak in
concentrations in the .125-.250 micron range (fine sand, silt).  These results
are important to keep in mind during the development of management strategies
for pollution reduction, such as street sweeping and leaf/litter removal.
                                    K-nr-6

-------
         TABLE 2. PERCENT OF STREET SURFACE POLLUTANTS BY PARTICLE SIZE
                          (From Sartor and Boyd, 1972)


Particle Size
Less than
Pollutant
Total solids
Volatile solids
COD
BOD5
TKN°
Nitrates
Phosphates
All toxic metals
All pesticides
PCBs
2,000
24.4
11.0
2.4
7.4
9.9
8.6
0
16.3
. - _
- - -
246-2,000
32.2
29.4
17.5
35.8
31.6
14.4
7.8
32.4
- 27.0 	
- 66.0 	
(microns)

Less than
43-245
37.5
34.0
57.4
32.5
39.8
45.1
36.0
23.5
	 73.0
	 34.0
43
5.9
25.6
22.7
24.3
18.7
31.9
56.2
27.8
_ _ _
- - _
The exact behavior of pollutant buildup is a much debated phenomenon.  The best
data have been presented by Sartor and Boyd (1972), who define an exponential
buildup of pollutants, with the highest rate occurring two to four days after
washoff.  Lager et al. (1977) and Roesner (1982) present additional information
in support of those exponential functions.  Bannerman et al. (1983) studied the
total pollutant loads (rather than time) versus daily accumulation rates and
found no data to support diminishing accumulation rates with increasing street
loads; that is, buildup was not affected by existing load.  In short, accumula-
tion of street surface pollutants is a well documented occurrence that occurs
at an uncertain rate.  Management strategies that hope to reduce urban area
loadings must address the fine-grained materials occurring on impervious sur-
faces and do it in a way that reduces accumulation capacities.

Atmospheric

One contributing factor to street surface runoff pollution is atmospheric
deposition.  Such deposition, however, contributes also to all other urban area
surfaces to some degree.  Sources of pollution to atmospheric wetfall and dry-
fall are vehicular emissions, industrial emissions, soil erosion, improperly
applied chemicals, wind-blown vegetative debris, and fuel combustion (ex., wood
stoves).  Pollutants resulting from these sources include nutrients, inorganic
and organic solids, toxic metals and organics, pesticides and sulfur.

Dry deposition appears to be reflective of surrounding land use activities and
tends to be seasonal in behavior (Bannerman et al., 1983; Oberts, 1982).  The
driving mechanisms for dry deposition are particulate and aerosol suspension,
sorption, and particulate settling, all of which occur on a continual basis.
Data on dryfall deposition rates are shown in Table 3.  Ranges are given in
Table 3 because of the extreme variability of dry deposition.  It is importannt
to establish some local calibration data to determine location within the range.
                                    K-III-7

-------
                TABLE 3.  RANGE OF CONTAMINANT DRY DEPOSITION  RATES
     (From Bannerman et al.,  1983;  Novotny and Chesters,  1981;  Oberts,  1982)
                                                 Deposition  Rate
                     Pollutant                     (mq/m /day)"
Total solids
Total P
Total Pb
COD
Sulfate
TKN
11
0.007
0.001
1.2
0.52
0.52
- 735
- 0.421
- 0.390
- 86.7
- 8.67
- 1.58
Wet atmospheric deposition  occurs  as  rain  and  snow  scavenge pollutants from the
atmosphere.  Wet  deposition seems  to  be more uniform  in  quality  and does not
reflect  land use  activities because of the larger geographic mobility of pre-
cipitation.  Data on wetfall  are given in  Table 4.  As with dryfall, concentra-
tion ranges are given because of the  great variability in the data.
Seasonality of wetfall  loading  is  more a function or  precipitation behavior
than land use activities.

               TABLE 4. RANGE OF CONTAMINANT WET DEPOSITION RATES
               (From Bannerman  et  al., 1983; Konrad et al., 1978;
                      Manning et al., 1977; Oberts, 1982)


                  Pollutant                   Concentration (mg/1)
Total solids
Suspended solids
Total P
Total Pb
cop
Sulfate
TKN
pH
PCB and pesticides (ng/1)
2
0.5
0.001
0
2.5
0.50
0.20
3.0
0.5
- 100
- 58
- 0.162
- 0.110
- 34.0
- 13.0
- 2.20
- 6.4
- 32.0
Together wetfall and dryfall can contribute  a substantial portion of the total
load from a drainage catchment.  Contributed amounts of total nitrogen always
seem to be quite high, often contributing over 75 percent of the total load,
while total phosphorus is somewhat less, often ranging from 25 to 50 percent
(Bannerman et al., 1983; Lager et al.,  1977; Miller and Mattraw, 1982; Oberts,
1982; Reckhow et al., 1980).  Pitt and  Bozeman (1982) attribute generally less
than 15 percent of any pollutant to precipitation in San Jose, Calif.  This
finding is consistent with Novotny and  Chesters (1981) who point out that
loading from precipitation is expectedly high in the more humid midwest and
southeast portions of the U.S.  They also note the increased pollutant
scavenging efficiency of snow over rain because of the larger surface area of
snowflakes.

The outlook for improved atmospheric loading is mixed.  Much attention is being
paid to reduction of industrial and automotive emissions, but reserves of low

                                    K-III-8

-------
cost, low sulfur fuel are being depleted and many homes are increasingly chang-
ing to wood burning furnaces and stoves, to name just two forces working
against reduced pollution.

Construction

The highest per unit solids loading in urban/urbanizing areas comes from con-
struction activities that disturb the landscape (Thronson, 1976).  Construction
of residential, commercial/industrial, transportation and utility projects
presents challenges for those trying to minimize urban area nonpoint pollution.

The most obvious problem associated with construction activities is soil loss.
Sonzogni et al. (1980) report that soil texture is the most important land form
relating to erosion, but they also note significance for other soil character-
istics (water holding capacity, credibility, etc.), slope, topography, geology
and drainage patterns.  Thronson (1976) adds other nonland related factors
including climatic conditions, activity at site, equipment use, length of soil
exposure and management practices used to reduce soil loss.  U.S. EPA (1984)
reports that annually 1.6 million acres of land are developed.  This construc-
tion activity accounts for about five percent of the nationwide sediment load
coming from a very small, concentrated source area.

Erosion problems are caused by on-site operations such as clearing and grub-
bing, rough and final grading, foundation excavation, and site restoration/
landscaping.  Some reported soil losses from construction sites are (Heaney et
al., 1975; Konrad et al., 1978; U.S. EPA, 1973; Yorke and Herb, 1976):

    o Maryland residential - 354-42,350 MT/km2/yr
    o Washington D.C. residential - 16,800
    o Maryland residential, commercial - 1,000-100,000
    o Fairfax Co., Virginia highway - 12,600
    o Georgia highway - 17,500-52,500
    o Montgomery Co., Maryland residential - 8,770-42,000
    o Menomonee Basin, Wisconsin residential - 4,370

The on-site erosion noted above does translate into receiving water problems.
Oberts (1982) notes that the loads at the mouths of two watersheds in which
residential construction was occurring were 79 and 203 MT/km /yr, a substan-
tial reduction from the on-site values but dramatically above similar water-
sheds with no construction.  Konrad et al. (1978) report for the Menomonee
River watershed that 37 percent of the TSS load and 48 percent of the TP load
during the study period originated from 2.6 percent of the watershed under
development, thus indicating primary management attention focused on developing
areas.

Other pollutants associated with construction sites include nutrients, construc-
tion-related chemicals and biological contaminants.  The level of nutrients
associated with construction is a function of organic soil content, fertilizer
use for vegetative establishment and vegetative debris washoff.  As previously
noted, a substantial TP load (8MT/knr/yr) can result from construction activ-
ity (Konrad et al., 1978).  Oberts (1982) reports that a high level of on-site
nutrient washoff resulted in watershed outflow TP levels of about 0.2
MT/knr/yr.
                                    K-III-9

-------
Construction-related  chemical  problems  have  been noted, but documentation of
the problems  is essentially nonexistent.  U.S. EPA  (1984) and Thronson (1976)
report problems with  pesticides for weed  and pest control, petroleum products
from vehicles, construction materials (paint, curing compounds, solvents, salt,
asphalt,  acids, mortar,  dye,  insulation)  and general litter.  Additionally,
both sources  note problems with human wastes from improper sanitation practices.

Construction  activities  clearly have the  potential  to cause major degradation
of urbanizing area waters.  Even a small  amount of  construction in a watershed
(2.6 percent  in the Menomonee  River case) can cause substantial pollution prob-
lems.  This fact points  to the need for special effort to control construction
activities.

Vegetation

Vegetative material contributes to the  solids, organic and nutrient load from
urban areas.  Sources of this  vegetative  litter are leaves and twigs, grass
clippings, seeds, and weeds.  Problems  are associated with both particulates
and with  dissolved leachates.  Heaney et  al.  (1975) report that evergreen and
deciduous leaves contain 0.04-0.28 percent (dry weight) TP, and 0.58-1.25 per-
cent TN.  They also report that burning leaves concentrates the nutrients.  The
loading values from these percentages translate into 0.6-4.5 kg/ha/yr for ,TP
and 13.6-64.7 kg/ha/yr for TN.  Grasses can  contribute another 9-13 kg/ha of TP
and 48-361 kg/ha of TN.  Overall, Heaney  et  al. report that 3,000 kg/ha/yr of
vegetative litter is typical of urban areas.

Pitt and Bozeman (1982)  in their San Jose study identify vegetation as a major
source of loading from lawn and landscaped areas, vacant lots, sidewalks, park-
ing lots  and  street surfaces.  The authors follow this source analysis with an
indication that vegetative litter falling on impervious surfaces is more than
likely to be washed off  and contributed to watershed outfall loading.  They
report that vegetation is responsible for approximately 70 percent of the BOD
loading,  30 percent of total solids and 10 percent  of TKN.

Chemical use

As pointed out previously, every contaminant  used in an urban environment will
eventually be seen in urban runoff.  In northern climates, the need to deice
requires the use of salt, which can lead  to  chloride pollution problems.  The
snow and  ice  accumulate  all pollutants  for periods  of time between melts.
Oberts (1982) addresses  the pollutants  associated with snowmelt runoff and
indicates that essentially all of the salt added to a roadway can be accounted
for in runoff to drainage conduits and eventually to receiving waters.  This
poses a problem of groundwater and lake contamination.  Similar problems have
been noted in Wisconsin, Massachusetts, New  York, Michigan and Canada (Manning
et al., 1977).  Total salt use seems to be decreasing, but when safety is
concerned it can be assured that salt will always be used as the answer.

Pesticides and fertilizers are also chemicals commonly seen in runoff.  Again,
anything applied in residential areas,  in public parks, and on hobby farms
close to urbanized areas will eventually  be  seen in runoff.  In Minnesota, for
example, approximately one-half of all pesticides applied annually are applied
by inexperienced urban household dwellers, who may  or may not apply and dispose
of the chemicals properly.  Sonzogni et al.  (1980)  report that home fertilizer
use can also be a major  cause of nutrient impact to receiving waters and that

                                    K-III-10

-------
such nutrients are usually readily available to noxious organisms.  They also
indicated a wide spectrum of pesticides in runoff.  Novotny and Chesters
(1981) also point out that improperly applied chemicals can volatilize and
migrate as aerosols, or can migrate directly as adsorbed particles.

Similar migration problems can be seen for PCBs, which have been detected in
both Arctic and Antarctic ice (Novotny and Chesters, 1981).  Such dangerous
chemicals can be introduced into the environment by very routine, but ill-
advised practices such as hosing down spill areas, dumping chemicals down house-
hold or street drains, or failing to quickly repair leaking chemical con-
tainers.  A discussion in a later section shows the NURP data on the detection
of priority pollutants in urban runoff.

Other sources

There are some other source areas of urban runoff pollutants that are not as
obvious as those previously noted.  Bannerman et al. (1983) found that galva-
nized steel eaves troughs and downspouts are major sources of Zn in rooftop
runoff.  They found, however, that precipitation was the major source of most
contaminants to rooftop runoff.  They conclude that the relative importance of
rooftop runoff decreases with increased urban development because other sources
become more important.  Pitt and Bozeman (1982) report that residential, compo-
sition shingle roofs can be a potential significant source of oil and grease
and other toxic materials that might compose the shingles or be blown onto
them.  They found, for example, that 1 to 20 percent of some metals (As, Zn,
Cu) could come from rooftops.

Urban street catch basins can also be a little-suspected source of contamina-
tion.  A Chicago study by APWA (1969) found that runoff water in catch basins
becomes septic and could easily reach a BOD concentration of 60 mg/1.  COD
levels in the settled solids were 38,300 mg/kg before a wash-through and 24,500
mg/kg afterward.  Sartor and Boyd (1972) studied catch basins in San Francisco
and found COD levels as high as 143,000 mg/1; BOD up to 1,500 mg/1; TN up to
33.2 mg/1; and TP up to 0.3 mg/1.  They reference similarly high values from
Milwaukee and Baltimore.

Fecal coliform can get into urban drainage conduits from septic tanks, infil-
tration from sanitary sewers to storm sewers, leakage at manholes, interconnec-
tions and surcharges that cause overflows (U.S. EPA, 1983).  Extremely high
levels of fecal coliform in urban runoff are discussed in a subsequent section.

The final two sources of urban contamination, for which little documentation
exists, are litter and collection system erosion/deterioration.  Litter is a
very visible, yet little documented contributor.  Urban stream channel and bank
erosion, and conduit scour contribute pollutants directly to the runoff that
has already been captured and is in transit.

EFFECTS OF URBAN RUNOFF

The actual water quality effects of urban/urbanizing area runoff can be mea-
sured by pollutant concentration and loading, by biological impact, and by
groundwater impact.  The following sections address these effects.
                                    K-III-ll

-------
Concentration levels

The most  complete  set of data on  the  concentration of pollutants from urban
runoff  was collected in conjunction  with EPA's Nationwide Urban Runoff Program
(NURP)  (U.S.  EPA, 1983).  The NURP effort involved a great deal of  attention to
the technical methodology for data collection and analysis at 81 sites in 28
cities  throughout the country.   It truly represents state-of-the-art  urban
runoff  documentation.  As such,  the  majority of the conclusions relative to
urban runoff  presented in this paper are somehow derived from NURP.

Perhaps the most significant finding of NURP is that variability in flow and
pollutant concentration are so extreme from site to site and from event to
event that relying on specific land  uses and geographic locations to   base
pollution projections upon is not  statistically valid.  Rather, NURP  recommends
at a minimum  using NURP summary  data for concentration and applying it to local
runoff  data to determine loading.  This, of course, is in lieu of locally
collected data.

NURP concentration results are reported as  "event mean concentration"  or EMC,
which is  defined as the total mass load for an event divided by the total vol-
ume of  flow.   This could also be termed the flow-weighted mean.  The  distribu-
tion of concentrations was found to  be lognormal, that is, a logarithmic (base
e) transformation is required to describe the data in linear statistics.

Table 5 shows the various land use EMC data and the summary data for  all sites.
The all sites entry is recommended by  EPA "for planning level purposes as the
best description of the characteristics of  urban runoff."  The land use EMC
data are  sufficiently different that they can be used to "fine tune"  local
pollutant concentration input to best  fit local conditions.

                TABLE 5. NURP SUMMARY WATER  QUALITY CONCENTRATIONS
                         FOR URBAN  RUNOFF (US. EPA, 1983)
                                 Site Median EMC (Coeff.  of Variation)*
   Pollutant

TSS (mg/1)
BODe (mg/1)
COD5(rag/l)
Total P (ug/1)
Soluble P (ug/1)
TKN (ug/1) ;
NO?-NO,-N(ug/l)
Total Cu (ug/1)
Total Pb  ug/1)
Total Zn (ug/1)
Fecal Coliforra
 (warm weather,
 No./lOO ml)
Fecal Coliform
 (cold weather,
 No./lOO ml)
All Sites

Median
100(1-2)
9
65
0.5-1.0)
0.5-1.0
330(0.5-1.0)
120
1500
680
34
0.5-1.0)
0.5-1.0)
0.5-1.0)
0.5-1.0)
144(0.5-1.0)
160(0.5-1.0)
21,000(0.8)
90th
Percent! le
300
15
140
700
210
3300
1750
93
350
500
...
Resi- Mixed Com./ Open/
dential Urban Ind. Nonurban
101(0.96) 67(1.14) 69(0.85) 70(2.92)
10(0.41
73(0.55
383(0.69
7.8 0.52
65 0.58
263 0.75
143(0.46) 56(0.75
1900(0.73) 1288(0.50
736(0.83
33 0.99
144 0.75
9.3(0.31
57(0.39
201(0.67
80(0.71
1179(0.43
558(0.67) 572(0.48
27(1.32
114 1.35
135 0.84) 154(0.78
29(0.81
104(0.68
226(1.07
...
40(0.78)
121(1.66}
26(2.11)
965(1.0)
543(.091)
...
30(1.52)
195(0.66)
—
                   1,000(0.7)
    Coefficient of variation
     variable data).
                      standard deviation divided by mean (higher value indicates more
                                     K-III-12

-------
The data in Table 5 are analyzed by U.S. EPA (1983) relative to importance in
urban runoff.  The TSS concentration in urban runoff is typically much greater
than that from a well run secondary sanitary treatment plant (100 mg/1 vs 25
mg/1).  The urban runoff solids differ from treatment plant solids in that the
urban runoff solids are more mineral (less organic) and have more pollutants
adsorbed to them.  EPA suggests that control of solids be pursued where water
quality problems exist with solids and where sediment build up is a problem.

Oxygen-demanding substances (BOD and COD) are present in urban runoff at about
the same concentration as from a secondary treatment plant.  No NURP project
area defined low dissolved oxygen (DO) as a water quality problem resulting
from urban runoff.  Other studies, however, have shown impacts.  Keefer et al.
(1979) looked at 104 sites nationwide and found a likely probability that a
higher than average DO deficiency occurred at time of high flow or during
periods of precipitation.  They report that the diurnal cycle typical of DO
behavior disappears during high flow, and that DO levels are depressed 1.0-1.5
mg/1 below low flow levels for periods of 1 to 5 days.  Similar DO behavior has
been noted by Abernathy (1981), Manning et al. (1977), Mikalsen (1984) and
Rimer et al. (1978).  For these reasons it is possible that the "critical"
period for DO in streams is not simply during dry periods of low flow, but
rather during such periods when precipitation generates urban washoff.

Nutrients (N and P) are always present at elevated levels in urban runoff, but
generally their contribution is less than point source discharges.  For exam-
ple, TP from an advanced treatment plant is still 1.0 mg/1 and TN from a secon-
dary plant can be as high as 20 mg/1.  Oberts (1982), however, reports that
even though nutrients from urban nonpoint sources might be less than point
sources, quite often the receiving waters are urban lakes or quiescent streams
where eutrophication or toxic ammonia impacts can occur.

Toxic contamination can also occur from heavy metals and organic chemicals.
NURP sampled priority pollutant concentrations in urban runoff at 61 sites in
20 of the 28 study locations.  A limited sample was taken for 120 priority
pollutants, and supplemented by a special metals study; both of these were in
addition to routine metals analyses.  Of the 120 pollutants examined a total of
77 were detected at some level.  The detected group included 13 metals, cyan-
ide, and 63 organic chemicals.  Table 6 lists the most frequently detected pol-
lutants and shows the percentage of samples in which the pollutant was detected.

Of particular note is the 91-94 percent detection for Pb, Zn and Cu.  Levels of
these metals, other metals and many of the organics are above what can gener-
ally be called good quality water, violating U.S. EPA recommended or estab-
lished criteria.  The organics are seen at a lower frequency, but some type of
a criteria violation (aquatic life, drinking water, carcinogen and/or mutagen)
occurs for PCP, chlordane, bis (2-ethylhexyl) phthalate, gamma-BHC, alpha-endo-
sulfan, phenanthrane, pyrene and chrysene (U.S. EPA, 1983).  It must be noted
that the NURP data are the best available even though they are limited in
number with only two samples taken at each site.  The results do clearly point
to the fact that toxics are moving in urban runoff.

Coliform bacteria are present in very high levels in urban runoff, ranging in
the NURP studies from 10's to 100,000's/100 ml.  Every sample taken following a
runoff event showed very high levels, far in excess of the criteria established
by EPA (200/100 ml).  Table 5 shows the differences in concentrations between
warm and cold weather, a difference due to the dynamics of coliform and to

                                    K-III-13

-------
                TABLE 6.  NURP  PRIORITY POLLUTANT STUDY RESULTS
                                 (U.S. EPA,  1983)
       Inorganics                                  Organlcs

Pollutant     % Detected                    Pollutant                % Detected
Lead
Zinc
Copper
Chromium
Arsenic
Cadmium
Nickel
Cyanides
Antimony
Beryl 1 i urn
Selenium

94
94
91
58
52
48
42
23
13
12
11

Bis (2-ethylhexyl) phthalate
alpha-Hexachlorocyclohexane (BHC)
alpha-Endosulfan
Pentachlorophenol (PCP)
Fluoranthene
gamma-Hexach 1 orocyc 1 ohexane ( L i ndane )
Pyrene
Phenol
Phenanthrene
Dichloromethane
4-Nitrophenol
Chrysene
22
20
19
19
16
15
15
14
12
11
10
10
activity of contributors  in the watershed.  The health implications of high
coliform levels are hard  to define because of the fact coliform is a poor indi-
cator organism and people are generally not in contact with urban runoff water
during and shortly after  events.

The major NURP findings,  other than expanding the data base, are that EMC's are
not a function of geography, land use or watershed physiography, and that tradi-
tional point source oriented water quality criteria are not at all suited to
describe the true impact  of urban nonpoint pollution, particularly when EPA has
not recommended criteria  for many of the pollutants of concern.  It should be
noted that the factors  listed as not influencing EMC do influence runoff volume
and will, therefore, impact total load.

Much other data exist describing the nature of urban runoff pollutant concen-
trations.  Table 7 is a summary of some other data.  It must be noted that the
methods of collection in  these studies is in most cases not consistent with
NURP.  They are presented for comparative purposes for those who feel they
could adjust their local  situation.  Where single values are presented, only
one study was found that  collected that data.

Some brief comments can be made about runoff concentration.  Receiving water
studies have shown that the high pollutant concentrations noted in Table 7 do
translate into impacts on receiving water (Keefer et al., 1979; Konrad et al.,
1978; Oberts, 1982; Pitt  and Bozeman, 1982; U.S. EPA, 1983).  Examples of
impacts seen are DO depression, toxic contamination of the water column and
sediments, channel degradation and eutrophication.

The "first flush" phenomenon of pollutant concentration likely does occur, but
its importance in total loading is minimal for noncombined sewer areas (Konrad
et al., 1978; Manning et  al., 1977; Porcella and Sorensen, 1980; Rimer et al.,
1978).  The reason for this is that total load is more affected by flow char-
acter than by pollutant level.  Quite often peak concentration and peak runoff
rate are reached at the same time, yielding a maximum loading rate.

                                    K-IIL-14

-------
         TABLE 7.  URBAN RUNOFF CONCENTRATION RANGES OF REPORTED MEANS
                             FROM NON-NURP STUDIES
                                             Land Use Type
                                  Residential*
   Pollutant

TSS (rag/1)
BOD. (mg/1)
COD°(mg/l)
TP (ug/1)
Sol. P (ug/1)
TKN (ug/1)
N02-N03-N (ug/1
Total Cu (ug/1)
Total Pb (ug/1)
Total Zn (ug/1)
Fecal Coliform
 (1,000/100 ml)
Total Cd (ug/1)
Total Cr (ug/1)
Total Ni (ug/1)
  HDR

 88-439
 11-38
 68-289
400-2100
100-2100
1630-4200
 420-820
   10
 190-780
 122-270
  14-130

    4
   25
  MDR

 44-768
  6-37
 32-152
160-1390
100-460
570-5000
 70-790
  53
 20-400
 55-235
  5-31
  5-255
  16
   LDR

  6-578
  2-63
 22-278
 60-410
 60-320
500-1000
  400

   80

   22
  Comm./
   Ind.

 42-2052
  8-76
 45-597
100-1100
 20-1140
1000-2400
 530-700
   4-64
 146-2200
 130-490
 0.1-18

   2-4
    31
    33
 Mixed

 13-5000
  3-160
 20-600
200-1100
190-1100
960-2500
 50-800
  1-150
  1-460
 18-360
  4-126

  5-10
  2-251
 17-150
*HDR-High-density residential; MDR-Medium-density residential; LDR-Low-density
 residential.
Seasonal differences in runoff quality occur and are especially pronounced in
locales experiencing long periods of snow cover (Bannerman et al., 1983;
Novotny and Chesters, 1981; Oberts, 1982).  The highest unit loadings occur
during snowmelt and early spring rainfall events when accumulated winter debris
is washed off.  Lowest values are of course, during periods of little rainfall,
typically late summer and fall.

Loading levels

Another method to arrive at watershed loading is by application of normalized
loading factors.  NURP developed such factors, as have numerous other studies
of urban runoff.  The recommended reporting unit for loading is kg/ha/yr
(Ib/acre/yr).  Table 8 reports the results of NURP, normalized for a 30-inch
annual average precipitation.  In addition to NURP values, Table 8 lists ranges
of loading values, where they exist, from studies with conditions typical of
urban runoff from an annual 30-inch precipitation.

The NURP data in Table 8 show high relative loadings for commercial/industrial
areas, and near-equal loading for residential and mixed areas.  It should be
noted that the NURP data base contains only light industry.  The non-NURP data
show a very wide variability and do not necessarily reflect highest loadings
from commercial/industrial areas.  Again, no data base is nearly as complete as
NURP, and use should be judged accordingly.  Also, construction areas are not
directly reflected in the data, but might be contributing to the higher loads
noted for TSS and the nutrients.
                                    K-III-15

-------
             TABLE  8.  NURP  (U.S.  EPA,  1983)  ANNUAL  URBAN  RUNOFF  LOADS
                          IN  KG/HA/YR  (Non-NURP  Values*)

Pollutant
TSS
BODc
COD5
TP
Sol. P
TKN
NOo-NOo-N
Total Cu
Total Pb
Total Zn

Residential
412 (11-3650)
27 (0.64-9.7)
188 (60-178)
0.98 (0.04-2.8)
0.38 (0.20)
4.4 (2.1-8.0)
2.0 (0.30-2.8)
0.10 (0.03-0.80)
0.41 (0.01-5.6)
0.46 (0.02-1.7)
Commercial/
Industrial
1095 (50-5100)
74 (2-97)
500 (297)
2.6 (0.10-4.5)
0.90 (0.30)
11.6 (12.7)
5.2 (1.8)
0.26 (0.29-8.6)
1.11 (0.17-13.2)
1.23 (2.9-12)

Mixed
480 (16-8030)
32 (2.6-560)
219 (41-320)
1.1 (0.08-10)
0.38 (0.05-1.3)
5.0 (1.2-7.3)
2.7 (0.96-6.1)
0.11 (0.02-1.9)
0.49 (0.14-6.5)
0.54 (0.10-2.8)
     *References: Bannerman  et  al.,  1983; Konrad et  al., 1978; Lager et all.,
      1977; Novotny and Chesters, 1981; Oberts9 1982; Pitt and Bozeman, 1982;
      Porcella  and Sorensen,  1980; Sonzogni et al.,  1980; Whipple et al., 1976.

The primary factor in  load determination is runoff.  Control of loading, there-
fore, should focus on  curtailing the movement of water from  impervious areas.
Novotny and Chesters (1981)  recommend focusing on critical areas where large
volumes of runoff are  generated.  The dominance of point source vs nonpoint
source loading  is usually variable by location, indicating the need to define
this relationship before moving ahead with expensive corrective measures.

Biological impact

The biological  impact  of urban  runoff is an area only recently receiving much
attention.   The best  documented biological impact is that of solids, which was
referenced in a previous section.  Heavy metals also exert a strain on aquatic
organisms.  Morrison et al.  (1984) report that toxic forms of metals account
 for 63, 66 and 73 percent of Zn, Pb and Cd, respectively, in urban runoff.
Toxic forms of  Cu account for a relatively lower 32  percent.  The Pb in runoff
is mostly associated with the particulates whereas the other metals cited are
mostly dissolved or evenly divided.  What this means is that Pb readily settles
and accumulates in areas receiving urban runoff, and that several of the other
metals do the same or  become  available in dissolved  form in the water column.
Disruption to the benthic community  as streams flow  through urban areas has
been documented by Mikalsen  (1984) and Pitt and Bozeman (1982), and Whipple et
al. (1976).  The runoff impacts are  lowered organism diversity and total num-
ber, less desirable (pollutant  tolerant) species, and decreased growth rates.
NURP (U.S. EPA, 1983)  results from two studies showed that many toxicants are
moving in urban runoff, but  impairment from these pollutants is not evident;
that is, no acute or chronic effects can be documented.  EPA does, however,
recognize that  impulse loads  of toxicants can have drastic results, and recom-
mends that some criteria be developed to address diffuse source loading.  The
NURP studies did point out the  impact that physical  channel degradation can
have on aquatic life,  and suggest that in certain instances this can be more
severe than chemical impacts.
                                    K-III-16

-------
Groundwater impact

A limited number of studies have shown that urban runoff can impact ground-
water, but usually not to a large extent.  Konrad et al. (1978) report that
even though some chloride problems exist with groundwater in the Menomonee
River Basin, the total groundwater pollution load to the river is insignifi-
cant.  Similar high chloride levels in urban/traffic areas are noted by
Sonzogni et al. (1980).  U.S. EPA (1983) showed in two NURP studies that delib-
erate infiltration of urban runoff does not cause toxic introduction problems
to groundwater in the short-term.  However, long-term impacts are unknown.  The
same studies showed that chloride is not attenuated by infiltration, that atten-
uation of essentially all toxicants occurs in just a few meters of soil, and
that vegetation helps increase infiltration rates but has no effect on pollu-
tants.  Lager et al. (1977) report that some attention should be paid to nutri-
ent inflow to lakes from groundwater, since it could represent substantial
loading.

SUMMARY

Research efforts have shown that washoff of pollutants accumulated on urban and
urbanizing surfaces can adversely affect water quality.  In fact, some
researchers suggest that the critical period for water quality in a stream is
not a dry weather, low flow period, but rather a low flow period in which an
urban runoff event occurs.

Data have been presented to show that urban/urbanizing area runoff can easily
equal and exceed pollution loading from wastewater treatment plants, especially
for solids, toxics and oxygen-demanding substances.  Because of this nature of
urban runoff loading, it tends to accumulate and exert its negative influence
over long time spans, resulting in such problems as benthic accumulation of
toxics and eutrophication of lakes and streams.

Characterization of runoff quality is a difficult task since the variability in
quality seems to be independent of geographic locale, land use and watershed
character.  The best method for characterization, if local data are not avail-
able, is to apply NURP concentration data to locally developed runoff volume
data.  The impact of this runoff depends on several things, including climate,
local receiving water physical and biological conditions, local perceptions of
water quality, desired use of receiving waters, and regulations in effect.
Once the relative impact of urban runoff is determined, cost-beneficial imple-
mentation measures can proceed addressing both point and nonpoint sources.

It is apparent that short-term efforts for management should be focused on high
runoff, high pollutant concentration areas, particularly commercial/industrial
and construction areas.  Programs that grossly generalize management techniques
should be avoided in favor of site by site or watershed by watershed evalua-
tion.  Urban runoff controls can be very expensive, so their use should be
reserved for occasions when they are definitely needed.

REFERENCES

Abernathy, A.R. (1981). Oxygen-Consuming Organics in Non-Point Source Runoff.
     U.S. EPA Project Summary, EPA 600/3-81-033.
American Public Works Association (APWA) (1969). Water Pollution Aspects of
     Urban Runoff. U.S. EPA Report 11030DN501/69.

                                    K-III-17

-------
Banner-man, R., K. Baun, M. Bohn, P.E. Hughes and D.A. Graczyk (1983). Evalua-
     tion of Urban Nonpoint Source Pollution Management in Milwaukee County.
     Wisconsin. U.S. EPA Project Summary. P005432-01-5.
Colston, N.V.  (1974). Characterization and Treatment of Urban Land Runoff.
     U.S. EPA  Report, EPA-670/2-74-096.
Heaney, J.P. and W.C. Huber (1984). Nationwide assessment of urban runoff
     impact on receiving water quality.  Wtr. Res. Bull., 20(1).35-42.
Heaney, J.P.,  W.C. Huber, H. Sheikh, M.A. Medina, J.R. Doyle, W.A. Peltz and
     J.E. Darling (1975). Urban Stormwater Modeling and Decision Making.
     U.S. EPA  Report, EPA 670/2-75-022.
Keefer, T.N.,  R.K. Simons, and R.S. McQuivey (1979). Dissolved Oxygen Impact
     from Urban Storm Runoff. U.S. EPA Report, EPA 600/2-79-156.
Konrad, J.6.,  G. Chesters, and K.W. Bauer (1978).  Menomonee River Pilot
     Watershed Study; Summary Pilot Watershed Report. International Joint
     Commission, Windsor, Ontario.
Lager, J.A., W.G. Smith, W.6. Lynard, R.M. Finn and E.J. Finnemore (1977).
     Urban Stormwater Management and Technology: Update and Users' Guide.
     U.S. EPA  Report, EPA 600/8-77-014.
Manning, M.J., R.H. Sullivan, and T.M. Kipp (1977). Nationwide Evalautipn of
     Combined  Sewer Overflows and Urban Stormwater Discharges (Vol. III).
     U.S. EPA  Report, EPA 600/2-77-064c.
Mikalsen, T. (1984). Assessment of water quality changes resulting from
     urbanization, agriculture and commercial forestry in the state of Georgia,
     U.S.A. In Proceedings. Third International Conference on Urban Storm
     Drainage. P. Balmer, P. Malmqvist, and A. Sjoberg (eds.), Chalmers-
     University of Technology, Sweden, pp. 801-810.
Miller, R.A. And H.C. Mattraw, Jr. (1982). Storm water runoff quality from
     three land-use areas in south Florida. Wtr. Res. Bull.. 18(3). 513-520.
Morrison, G.M.P., B.J. Ellis, D.M. Revitt, P. Balmer, and G. Svensson (1984).
     The physio-chemical speciation of zinc, cadmium, lead and copper in
     urban Stormwater. In Proceedings. Third International Conference on Urban
     Storm Drainage. P. Balmer, P. Malmqvist, and A. Sjoberg (eds.), Chalmers
     University of Technology, Sweden, 989-1000.
Novotny, V. and G. Chesters (1981). Handbook of Nonpoint Pollution: Sources and
     Management. Van Nostrand Reinhold Co., New York.
Oberts, G.L. (1982). Water Resources Management; Nonpoint Source Pollution Tech-
     nical Report. Metropolitan Council Publication No. 10-82-016, St. Paul, Mn,
Pitt, R.E. and G. Amy (1973). Toxic Materials Analysis of Street Surface
     Contaminants. U.S. EPA Report, EPA R2-73-283.
Pitt, R.E. and M. Bozeman (1982). Sources of Urban Runoff Pollution and Its
     Effects on an Urban Creek. U.S. EPA Report, EPA 600/2-82-090.
Porcella, D.B. and D.L. Sorensen (1980). Characterizations of Nonpoint Source
     Urban Runoff and Its Effects on Stream Ecosystems. U.S. EPA Report, EPA
     600/3-80-032.
Reckhow, K.H., M.N. Beaulac, and J.T. Simpson (1980). Modeling Phosphorus Load-
     ing and Lake Response Under Uncertainty; A Manua1~and Compilation of
     Export Coefficients. U.S. EPA Report, EPA 440/5-80-011.
Rimer, A.E., J.A. Nissen and D.E. Reynolds (1978). Characterization and impact
     of Stormwater runoff from various land cover types. Jour. WPCF.50(2), 252-
     264.
Roesner, L.B.  (1982). Quality of urban runoff. In, Urban Stormwater Hydrology.
     D.F. Kibler (ed.), AGU, Water Resources Monograph 7.
Sartor, J.D. and G.B. Boyd (1972). Water Pollution Aspects of Street Surface
     Contaminants. U.S EPA Report, EPA R2-72-081.

                                    K-III-18

-------
Shaheen, D.G. (1975). Contributions of Urban Roadway Usage to Water Pollution.
     U.S. EPA Report, EPA 600/2-75-004.
Sonzogni, W.C. , and G. Chesters, D.R. Coote, D.N. Jeffs, J.C. Konrad, R.C.
     Ostry, and J.B. Robinson (1980). Pollution from land runoff. Envir. Sci.
     and Tech.. 14(2).148-153.
Thronson, R.E. (1976). Nonpoint Source Pollution Control Guidance: Construction
     Activities. U.S. EPA Report, Technical Guidance Memorandum: TECH-27.
Tourbier, J. and R. Westmacott (1974). Water Resources Protection Measures in
     Land Development-A Handbook. Univ. Delaware, Water Resources Center.
U.S. Environmental Protection Agency (U.S. EPA) (1984). Nonpoint Pollution in
     the U.S.: Report to Congress. U.S. EPA Summary Report.
     (1983). Results of the Nationwide Urban Runoff Program; Vol. I-Final
     Report U.S. EPA Report, Water Planning Division.
     (1979). A Statistical Method for the Assessment of Urban Stormwater.
     U.S. EPA Report, EPA 440/3-79-023.
	 (1973). Methods for Identifying and Evaluating the Nature and Extent of
     Non-Point Sources of Pollutants. U.S. EPA Report, EPA 430/9-73-014.
Whipple, W., Jr., B.B. Berger, C.D. Gates, R.M. Ragan and C.W. Randall (1976).
     Characterization of Urban Runoff. Rutgers Univ., Water Resources Research
     Institute.
Yorke, T.H. and W..J. Herb (1976). Urban-area sediment yield-effects of con-
     struction-site conditions and sediment-control methods. In Proceedings.
     Third Federal Interagency Sediment Conference, U.S. Water Resources
     Council.
                                    K-III-19

-------
IMPACT OF NONPOINT POLLUTION ON RECEIVING WATERS
                      Peter A.  Krenkel*  Dean
                      College of  Engineering
                      University  of  Nevada-Reno
                      Reno,  Nevada    89557
                  K-IV-1

-------
 INTRODUCTION

     There  Is no doubt that the qua! ity of our waterways has  improved con-
 siderably since the first permanent  Water Pollution Control act was passed In
 1956.  However*  the major thrust of water pollution control efforts has been
 towards the easily controllable pollutants emanating  from  point sources.   In
 many Instances*  the major contributions to a  water body are diffuse In origin
 and the dltnunltlon of point sources has not solved  water quality problems.
 In many cases* more  stringent controls on the  point  sources Is not rational
 Inasmuch as spending large amounts  of  money  on  what  might be  a fraction of
 the total  pollutant  load  reaching a  waterway  Is not in the public  Interest.

     It Is now time to put our water quality  problems In perspective*  define
 the real  effects  of various sources  of  pollutants on water quality and direct
 our control efforts towards those  measures  that  w I I I resuIt  In defIn ItIve
 Improvement In the quality of the  receiving water.   An  engineer Is taught
 that 90%  of solving a problem  Is  defining  the problem.  This concept Is
 particularly applicable to the management of  multi-origin pollutants.

     Many of the questions concerned  with source-receptor relationships  and
 their effects on  water quality can  only be answered  by adequate data.   While
 much data Is collected on a nationwide  basis*  they are not  collected to solve
 problems  and for the most part* their only value  Is to indicate "trends" in
 water quality.   In fact*  the  nation-wide  improvement In water quality  noted
 previously  Is mostly based on observations  made  In rivers and not in  lakes
 where major nonpoint problems exist.  Other major problems  with the data
 Include a  lack of adequate quality control,  a  lack of  comparable metho-
 dologies  in collection*  storage  and analysis and  inadequate  Interpretation.
 It Is obvious to  any  user of these data that  the entire col lection* storage
 and retrieval  system  needs  scrutiny.

     Inasmuch as  the  impact of pollutants on  receiving waters  Is essentially
 the same* whether they originate from point or nonpoint  sources  (Ignoring
 distribution)* it  is  important  to  determine their real significance.  For
 example*  one might find 3 mg/l of lead  in urban  runoff* however* If the flow
 only occurs for  a few moments*  the effects on  the  receiving water would
 probably  be Insignificant.  Thus* determinations  of  mass   loadings  are  para-
 mount  to determining impacts.  The  bottom  line  is whether or not  the
 resulting stream concentrations are sufficiently  high  to cause adverse
effects.   Of course,  possible long term sorption  by  sediments should be
considered.

 SOURCES

     The  EPA report to Congress  on  Nonpoint  Source Pollution  (1)  Identified
 five  source  categories of nonpoint pollution  as  follows:

                              Agriculture
                              Urban Runoff
                              SIIviculture
                              Mining
                              Construction

     One  should probably add atmosphere sources  to this  I I st although  the

                                K-I.V-2

-------
relative contributions are probably not significant in many cases.

     Agriculture  contributes pollutants from cropland*  grasslands  and live-
stock operations and  is probably  the  most pervasive diffuse source.  Pol-
lutants originating from  agricultural practices Include sediment*  micro-
organisms*  pesticides*  nutrients*  organic  materials.   It has been estimated
that agriculture  contributes more  than 50% of the sediment delivered to the
nations  waterways.   The  role  of agriculture  practices In contributing
nitrogen* phosophorus and pesticides  to our waterways  is  also significant.
Figure 1  demonstrates the relative concentrations of nutrients contributed by
agriculture to other land  uses  (2).

     Urban  runoff also may play a significant  role in contributing to water
quality problems.   Urban  rainwater may yield heavy  metals*   microorganisms*
sediment nutrients and organic materials.   While  urban runoff appears  to have
the potential  for resulting  in serious impacts on receiving  waters as shown
by  Figures 2  (2)  and 3  (3),  the  International Joint  Commission (IJC)
concluded  that pollution from  urban  runoff does  not  have  a  major impact on
water  quality In the Great Lakes (4).  It should be noted*  however* that
urban runoff's contributions of heavy metals. PCB's* etc. over  a  long term
basis must be considered.  The potential  for these and other priority pol-
lutants to accumulate to significant  levels over long periods  of time Is a
viable possibility.   In addition*  significant adverse water  quality  impacts
do occur on a  localized basis.

     In silviculture*  similar  to agriculture* the major pollutant by volume
is sediment.  Other pollutants that might be contr ibuted by siIvlcultural
practice  include nutrients* organics and pesticides.  In addition*  concern
has been expressed  over the resulting increase in  river temperature when
trees are  cut  along river  banks.

     Mining activities* while not  as widespread as other diffuse  sources* can
result in  serious degradation of not only  waterways,  but  groundwater.  Con-
taminants  emanating from  mines may consist of  numerous  substances  Including
sediment*  acid* salts* metals* radioactive material  and  organics.  The prob-
lems from  abandoned mines  are local  in nature, however, their multitude and
seriousness makes them a nationwide problem.  The problems  are exacerbated in
the West because  of  the paucity of water.

     Construction  activities, while localized in nature*  may  result  in severe
water quality  impacts.   The major construction site pollutant  is sediment*
however*  its  contribution  is only  4  to 5% of  the total  sediment  load to
receiving  waters (I).  Other  pollutants  contributed by construction may
include nutrients,  pesticides*  petroleum derivatives and debris.

     As pointed out  by Krenkel and Novotny (6), the atmospheric  contribution
•to nonpoint pollution may  include  nutrients* organics* acid  deposition*
^articulates and  any materials  that  may be carried  into the atmosphere and
subsequently  washed out  by precipitation.  While the concentrations appear
relatively  small,  the area I distribution  may  be significant in some areas.

EFFECTS

     The effects  of nonpoint pollution are similar to  those  of  identifiable

                                K-1V-3

-------
                         MEAN TOTAL PHOSPHORUS CONCENTRATIONS
RO, 09-BUM
 59 FOREST
170 MOSTLY FOREST

 30 10XKD

 11 MOSTLY URBAN
 06 MOSTLY AGRIC

 91 AGRICULTURE

                LAND USE
 DATA ON 473 SUBDRAINAGK AXKAS IN EASTERN U.S.
0.014
    -- 0.035
         0.040
                  0.066
, — .
!
L 	 »
n.oo

1 1 . i. . 1 1 i
.05

it i i
.10
0.135
i 4 i- J
.15
                                  MILLIGRAMS PER LITER
HO. w sow
 59 FOREST
170 MOSTLY FOREST
 58 UDffiD
 11 MOSTLY URBAN
 08 UOmYAGRXC
 81 AQRICIILTUKK
                 L  i
               a. ao
  MEAN TOTAL NITROGEN CONCENTRATIONS
                   T5P
               LAND USE
 DATA ON 473 SUBBRAINAG2  AREAS IN EASTERN U.S.
   0.650
    0.665
       1.886
              1.812
           i   i    i   i    i   i
         1.67             3.33
          MILLIGRAMS PER LITER
4.170
t    i   .J
     5.QD
   FIGURE 1 RELATIONSHIPS BETWEEN GENERAL LAND  USE AND TOTAL PHOSPHORUS
                 AND TOTAL NITROGEN CONCENTRATIONS IN  STREAMS
                                      K-IV-4

-------
                             0  K-NITROGEN


                             A  LIAD


                             O  TOTAL SUSPENDED SOLIDS,


                                TDJESE3
   600
800
1000
                                         r 80.0
1200     1400
                                             0.0
           TIME (HRS)
73GURE 8 POLLtTTlNT VARIATIONS ¥/Q AM) TDJE
       (BKDRAim AFTER COLSTON (B))
              K-IV-5

-------
             SCIOTO R. AT CHILUCOTHE.OH.
  15
     I-
\
o
a
H
HH
O
fc
w
Q 5
O
p
-
-

-

-
-
-
/
T7TTTT
                     TTTT ft tT T ]T1 Tl"
ft T t rjTTTT-
         SATURA110N AT 20C  ~9.07
                                  STORM Fl/)TT
                                NORMAL FLOW
    ^                          ~-	———.	
Q ktxj-i.ii 11 ill ii t J11 ii.J null i-ij.J.i i i.i 11i i.iJj.u.1.In.

  0   10  20  30  40   50  60  70   80  90  100

    COLUMBUS                          CHILUCOTHE

                                        MONITOR
                  DISTANCE (MI)
flOUKE a STKUl'Ktt-PHKLPS ANALYSIS R88UILT8 FOB 8C30TO RIVER AT CHIXIICOTH8.0H,
                        K-IV-6

-------
point sources.  However* their distribution with time and the manner In which
they are Introduced  Into a water body may differ.  Therefore* the traditional
analysis of diffuse sources of pollutants requires a somewhat  different
analytical  treatment.   Figure (4)  Illustrates the way  in  which the typical
nonpoint pollutant enters a  receiving water (7).

     A discussion of the impacts of these pollutants  Is therefore In order.

                      Oxygen Demand i nfl  Substances

     The contribution of biodegradable organ Ics to a receiving  water causes a
depletion in  oxygen  resources.  Should the concentration of oxygen be dimin-
ished below acceptable water qua! Ity standards* the lower dissolve oxygen
levels  will  be  Inimical to aquatic  11fe.  As shown on Figure 3,  nonpotnt
pol lutlon may cause a  dlmunltlon of disolved oxygen subsequent to a storm
event that may  result  in  an oxygen deficiency.  This can be  particularly
serious In the case of a small river or  Impoundment.

                            M i croorganIsms

     The Introduction of microorganisms  Into a receiving water Is  of concern
because of  the  potential for  the presence of pathogens.   This  may result in
impaired recreational use of the water*  an increase  in water treatment costs
and a potential human  health hazard.  The prime nonpoint source of these
materials Is  from agriculture and urban  runoff.

                                Metals

     Metals  in  the  aquatic environment may  have  a multitude  of effects
because of  their toxlclty.   If  the  concentration  is  sufficiently high* they
may cause death of  aquatic life and  lower levels  may cause chronic toxlclty
resulting In  adverse effects on propagation and  life  expectancy.

     The metals may also be sorbed  In  the sediments and become toxic to
benthlc organisms*  be biologically methylated and released to  the water for
fish accumulation and enter the food chain for subsequent biomagnlfIcatI on.

     Metals have also  caused economic problems by affecting commercial
fishing* sport fishing and Increased  water treatments costs.  The synerglstlc
effects of metals with other metals and contaminants  should also be con-
sidered.

                               Nutrients

     Probably the contaminants  from  nonpoint  sources causing the most prob-
lems  are nitrogen and phosphorus.   If  one  of these elements  Is  growth
limiting for either phytoplankton  or macrophytes* their  Introduction  Into a
lake may Induce prolific growths  of algae and/or attached aquatic plants.
This hastens  or  accelerates the problems  associated  with eutrophlcatlon.  It
should be noted that most North American lakes  are phosphorus  limiting* thus
making the  control  of phosphorus of significance.

     Excessive algal growths and aquatic weeds  may diminish the use of  lakes
for recreation,  increase the costs of water treatment,  increase hypollmnetic

                               K-IV-7

-------
            RAINFALL INTENSITY
                 •CONCENTRATION OF POLLUTANT
                 4IASS OF POLLUTANT
                 .FLOW (HYDROGRAPH)
                         TIME
FIGURI 4 FLOV AND QUALITY HISTOGRAM FROM NONPOINT SOURCES
                       K-IV-3

-------
oxygen depletion and result  In toxic effects on fish and humans.  The ap-
pearance of the water may become aesthetically unacceptable thus causing
major adverse public reaction. The major nonpolnt source of nutrients  In the
U.S.  Is  from agriculture.

                                 Salts

     The salinity of a  water  may have  drastic effects* In particular  In the
arid parts  of  the country where nonpolnt  sources are the major culprit.  For
example* the sodium content of parts of the Colorado River are near 120 mg/l
and the water Is used  for drinking  purposes.   Sodium also has an adverse
effect on soils and salinity  may  affect crop yields and ultimately render the
sol I toxic to crops.  The salt water Intrusion along our coasts  Is a welI
known problem and could be considered as  nonpolnt pollution.

                       Herbicides and pesticides

     Herbicides and pesticides  are obviously toxic materials and their
presence In the aquatic  environment can cause contamination of fish resulting
In  a public  health hazard*   sub-lethal  effects on the entire  food chain*
blomagnlfIcatlon and adverse effects  on reproduction.  The major contribution
of these materials  Is from agriculture.

                              Sediment

     Sediment Is the largest nonpolnt problem* not only because of Its magni-
tude* but because It may resuIt  In all of the previous!y described I mpacts
due to the  materials associated with  the  sediments.

     In addition* sediment may  cause decreased  water  transparency* thus
Interfering with photosynthesis* destroy spawning areas* clog fish gills*
cause density  problems  and Interfere with the benthlc  population.  Turbidity
also results  In aesthetic problems and  Increases water treatment costs.

BIOLOGICAL  AVAILABILITY

     A major question with regards to some of the pollutants contributed from
diffuse sources  Is  their  availability to be utilized  by  the microorganisms*
for example*  much of the phosphorus contributed  via runoff Is In the form  of
apatite*  which for  all practical  purposes*  Is biologically unavailable (8).

     This  Is an  Important concept Inasmuch as It has  been Impl led that one
reason for placing emphasis on  the control of point sources rather than
nonpolnt sources of phosphorus  In the Great Lakes Is because of  the high
bioavailablIIty of phosphorus in wastewater and the large quantities  of
biologically  unavailable phosphorus in  many sediments  (9).

     The uncertainties caused by the lack of definitive resolution of the
bloavallablIIty  question needs Immediate attention.  Every effort  should  be
made to resolve  this issue at an  early  date.

EXAMPLES OF THE EFFECTS OF NONPOINT POLLUTION

     The effects of nonpoint pollution can  best be Illustrated by some

                                K-IV-9

-------
 examples that Mill serve to show the magnitude of the problem.

                            The Great
    The latest International Joint Commission report stated  that "even If all
the commitments with respect to phosphorus control are met and detergent
phosphorus limitations are continued,  the full  extent of the phosphorus
problem will still not have been addressed"  (10).   In  addition,  the GAO
performed a study on the water quality  improvement efforts on the Great Lakes
and concluded that unless  nonpoint sources are addressed* the Great Lakes
water quality objective may not be achieved even  if all other sources of
pollution are completely controlled or  eliminated  (9).

     Table I.  taken from that report, demonstrates the problem.  According to
the GAO report,  while 7455  of the  phosphorus  load originates from diffuse
sources, the major emphasis on control  has  been on point sources.
                                TABLE  I

               197$ Phosphorus Loadings liy. Source (Percent)

                    Point                              Nonpotnt

           Municipal    Industrial    Total    Air  Upstream   Land

Superior        7            3         10       37      0       54      91
Michigan       39            5        44       26      0       30      56
Huron          11            2         13       23     14       50      87
Erie           39            3        42        4      6       49      59
Ontario        24            2        26        4     41       31      76
     Assuming that the relative contributions shown in Table I  are correct.
it is obvious that if  the water qua! ity in the Great Lakes  is to be further
improved.  Immediate attention must be given to the control of diffuse sources
of phosphorus and  other contaminants.

     The situation  existing In the Great Lakes  is not unique.   Duda and
Johnson report that  "the damage being caused  by nonpoint sources across the
country  is  staggering" (11).   This conclusion is  based on EPA and  state
studies.  As  pointed  out  by Carter, "virtually all  of  the 53  states and
territories acknowledge that they have a severe nonpoint source pol lution
problem" (12).

                           M1nIng ProbI em

     As previously noted,  nonpoint sources from mining activities can cause
serious local  problems.   An example of  the  Impact of  mining  wastes to a
waterway is shown on Figure 5 (13).   As can be seen,  the  water upstream from
the tailing ponds was relatively low In suspended solids.  As the  water
passed the  areas  where the tailing ponds were being discharged   Into the
river, it picked up  considerable  amounts of  material  resulting In concen-


                               K-IV-10

-------
1000
"T
     -   CARTERSVILLE ¥ATER INTAKE
 100
  10  r
~r
                                      L	
   1
     0        4       8       12       16       20      24.
   1B/N                       18N                       lOIK
                        HOUR OF DAY
      FIGURE 5 VARIATION OF SUSPENDED SOLIDS WITH TIME ON
   20 AU6 64 AT ALLATOONA TAHJUCE AND CARTERSVILLE INTAKE
                         K-IV-11

-------
 tratlons varying from 4 to 500 mg/I  s.s. within a few  hours.  The resulting
 Impact on the water treatment plant Is obvious.  In addition, significant
 concentrations of  Iron and manganese  were added with the solids causing
 problems In the water treatment plant, the distribution system and at the
 tap.   Barium was also being added to  the river via the  sediments.

      The concentration of sediment reached  levels  such that the regulatory
 agency stated  that  the receiving  waters were unfit for reasonable uses  In-
 cluding  fish  and  wildlife development and  recreation.   The  river  discharges
 Into a reservoir, thus compounding the sediment problems.

                              Lake Lahonton

      An  Interesting study of  the relative  Importance  of nonpolnt and point
 contributions has been the subject of much  controversy In Northern Nevada.
 Severe phosphorus  restrictions have been placed on the effluent from the
 Reno/Sparks Wastewater Plant  for several reasons,  one  of them  being because
 of the al leged Impact on Lake Lahonton; which  is east of Carson City.  Ob-
 viously,  the  degree of phosphorus removal  required will have  a significant
 Impact on the cost of wastewater treatment.

      A comprehensIve analysis of the phosphorus loading revealed that the
 contribution from the Reno Plant was a small percentage of the  total load to
 the lake.  A significant part  of the  phosphorus being utilized  by the phyto-
 plankton  was  found  to be that released  from the sediments.  Other nonpolnt
 sources were also found to be  of significance.  The water quality problems on
 this lake have become so bad  that the  lake was closed for  recreation  for a
 period of time.

      The  impacts of closing the lake  are exacerbated by the paucity of
 recreational lakes  in Nevada.  Thus> nonpoint pollutlon control  Is of the
 utmost Importance to the area.

SUMMARY

      The problems of nonpoint pollution  have been discussed  as  well as  their
 sources and Impacts  on receiving waters.   It was  pointed out that the lack of
 adequate water quality data  is a major  problem.  In addition,  it was noted
 that until diffuse pollution  is ameliorated,  the attainment of  the goals of
 PL 92-500 will  not be achieved in many waterbodies.
                                K-IV-12

-------
1.    U.S.  EPA,  "Nonpolnt Source Pollution  In the U.S.", Report to Congress,
           Jan 1984.

2.    U.S.  EPA,  "The Influence of Land Use on Stream Nutrient Levels",  ERL,
           ORD,  EPA-600/3-76-014, Jan 1976,  Corvallls, Oregon.

3.    U.S.  EPA,  "Dissolved Oxygen Impact from Urban Storm Runoff", MERL,  ORD,
           EPS-600/2-79-156, Nov 1979, Cinclnattl, Ohio.

4.    IJC., "Nonpolnt Source  Pollution  Abatement  in the Great Lakes Basin",
           Report to  the Great  lakes Water Quality  Board,  August 1983,
           Windsor, Ontario.

5.    Colston,  N.V., Jr., "Characterization and   Treatment  of  Urban  Land
           Runoff, EPA 670/2-74-096., U.S.  EPA,  Dec  1974, Cincinnati, Ohio.

6.    Krenkel, P.A.,  and V. Novotny: "Water Quality  Management",  Academic
           Press,  1980.

7.    Novotny V.  and G. Chesters,  "Handbook  of Nonpolnt  Pollution", Van
           Norstrand Relnhold, 1981.

8.    Expert Committee on Engineering and Technological  Aspects of Great Lakes
           Water Quality",  "Biological Availability of Phosphorus",  Report  to
           the Great Lakes Science Advisory  Board,  1979.

9.    U.S.  General Accounting Office, "A More Comprehensive Approach Is needed
           to Clean  up the Great Lakes",  CED-82-63,  21 May 1982.

10.   IJC,  2nd Biennial  Report under  the Great  Lakes  Water  Quality  agreement
           of 1978 to  the Governments of the United States and Canada and the
           States and Provinces of the Great Lakes  Basin", Dec  1984.

11.   Duda, A.M.  and R.  J.  Johnson,  "Lakes are Losing  the  Battle In Clean
           Water Programs",  Jour.,  Water  Pollution Control  Federation,  July,
           1984.

12.   Carter, K.B.,  "Protecting our Investment In Clean  Water",  Monitor,
           Jour., Water  Pollution Control  Federation,  Feb, 1985.

13.   Krenkei, P.A., "Invest igat ion  of Manganese Concentration In AIlaroona
           Reservoir", Georgia, U.S.  Army  Corps  of Englners,  Mobile,  Alabama,
           August  1965.
                                 K-IV-13

-------
    USE OF COMPUTER MODELS  IN MANAGING  NONPOINT  POLLUTION  FROM  AGRICULTURE

                                 W. 6.  Knisel
                         Research Hydraulic  Engineer
                       U. S. Department  of Agriculture
                        Agricultural Research  Service
                   Southeast Watershed  Research  Laboratory
                               Tifton,  Georgia


                                   ABSTRACT

Concepts  and  philosophies  of modeling  are  presented to  provide a  background
for some of the more  widely used models for assessing nonpoint  pollution from
agriculture.   A few  models are briefly reviewed,  and  suggestions  given  for
model selection and application.  Application  of the CREAMS model  is  presented
to demonstrate  the use of  computer models  in  evaluating management  practices
for  control  of nonpoint  source pollution.    Runoff, percolation  through  the
root  zone,  sediment  yield,  pesticide runoff,  and pesticide leaching are ana-
lyzed to estimate  the effects of agricultural  management practices on edge-of-
field pollutant loads.


Keywords;  Computer models, mathematical models,  nonpoint pollution,  runoff,
           erosion,   percolation,  groundwater,   water  quality,   agricultural
           practices, management practices,  pesticides, plant nutrients.
                                   K-v-l

-------
 INTRODUCTION

 The development  and  improvement of  the  digital computer  during the  past  30
 years  has made  it  possible  for engineers  and  scientists  to  perform  complex
 repetitive  calculations  at  extremely  high  speeds.    Also,   improved  storage
 capabilities now permit operation of complex  computer  programs,  or mathemati-
 cal  models  programmed for computer operation.  With these developments, scien-
 tists  have  been  formulating  models  to  represent physical  processes  and their
 interactions in  nature  that  relate to  soil and water  resources conservation
 and  nonpoint source pollution.   Numerous models have been developed during the
 past 20 years for various  uses  with  later combining of  some  models  as compo-
 nents  to form  more complex models  for  evaluating nonpoint  source pollution.
 This paper  describes  the application of  computer  models  for  evaluating  non-
 point  pollution  and  developing  management  practices  for  its  control.   The
 specific  model  applied  is a mathematical  description of physical  processes.

 COMPUTER  MODELS AND MODELING  PHILOSOPHIES

 Backgrounds  of  the  symposium  participants  are  quite  varied,  and  perhaps it  is
 desirable to  describe and discuss computer models before reviewing some of the
 more common  ones.  What  is a  computer model?  This  question is not intended  to
 insult  anyone's intelligence; rather, it  is intended to get  all  the partici-
 pants thinking  along  the same lines.  A  computer model  is the formulation  into
 a computer  language of the modeler's concepts  of the physical  system and  pro-
 cesses.   Models are simplifications  of  the real system,  and the degree  of  sim-
 plification  may be  the result of the modeler's  understanding  of  the  processes
 or  the  desired  degree of  accuracy,  or  the  purpose  of  the  model   itself.   For
 example,  a  scientist  conducting research  on  infiltration of  rainfall  into the
 soil might very precisely describe mathematically the infiltration process and
 generation of runoff  volume,  but might  grossly simplify the movement  of  sur-
 face runoff  rates  and the erosion  processes.   Likewise,  engineers familiar
 with runoff   and  erosion  processes  might  disregard  or  oversimplify chemical
 adsorption to soil  particles  and the processes  of chemical transport.   If the
 end  product  from  the  model is  estimation  of chemical  movement,  a  scientist may
 formulate the  adsorption/desorption processes  in  considerable  detail,  but
 drastically  simplify the  runoff and erosion processes.   Therefore,  one can
 readily  see  that a  computer  model  represents the modeler's  concepts   of the
 processes involved  and the model  is  developed  for a  specific  purpose.

 There are--two important  factors  in  any model:   space and time.   How  large (or
 how  small)  an area  can be represented?   What  is the time  step (increment)  in
 the  model?    Before  discussing  these questions,  some consideration  should  be
 given to  the  overall  problem.   To talk  about  nonpoint pollution  generally im-
 plies water  quality,  but  where  is  the  interest  or  concern in water quality?
 In  a lake?   In  a stream?   In a  pond?  At the edge of a field?   Water  quality
 is  measured  at  some  receiving body  of  water,  but not  at the edge of  a field
 unless  that  field drains into a pond.   When management practices  are  applied
to reduce nonpoint  source  pollution,  where are those practices applied?   Is  a
given practice  recommended for  every field  on  every  farm in  the  drainage ba-
sin?  Management  practices are  applied  on  individual fields within the basin,
but  it  is necessary to know the  impact  that practice will  have on  the  quality
of  water  in  the  receiving body.  The same practice on  every field  will not
necessarily  have the  same impact on  water  quality  downstream.    Then water
                                    K-V-2

-------
quality specialists really  need  two models:   one  for  field-size  areas  to  exam-
ine alternate management  strategies,  and one  for watershed or basin  scale  to
estimate  the  aggregate affects  of field  treatment on in-stream processes  and
downstream  loads.   Thus, we  will discuss  models of two  scales:    relatively
small field-size areas and  basin  scale.

Within a computer model,  some  time step or  time  interval of calculation is  in-
volved.  The time step varies  with  component and with modeler as well as with
scale.  For example, a modeler conceptualizing the  infiltration  process may be
concerned with  time steps of  1  minute  or 5 minutes  on  field size areas.   An
ecologist often thinks of a 1-year increment when concerned  with nitrogen  and
phosphorus cycling  in  basin-size areas.   A  chemist may  think of one  day time
intervals for pesticides.   A  one-year time  increment would  be meaningless  for
pesticides which have  a half-life of 1  to 4 weeks.  Another element of time
that more nearly reflects scale  is  the  routing interval  within or  through  the
drainage areas.  Routing, if  done on  field-size  areas, may require increments
less than an  hour,  possibly 15-minute intervals.   For  basin-size  areas, such
as  a 200-square kilometer drainage in  the  Coastal Plain  or Coastal  Flatwood
physiographic areas, a 6- or  12-hour  routing interval might  be sufficient.   A
routing interval of one day may  be adequate  for  an  area  of  1000  square kilome-
ters.  Differences  in  infiltration  rates  from one soil  type  to  another proba-
bly  has  no significance  when simulating  for 1000-square  kilometer  drainage
area.   The impact  of  management  practices  become  less  significant  in such
large basins  unless the  treated  areas  are  near  the  stream or drainage  basin
outlet.  For  example,  a  terrace  system or  a no-till  practice on a 20-hectare
field will have  negligible  effect on the  pollutants  leaving the basin.unless
the field is  a  critical  source area adjacent to  the  outlet.  This  emphasizes
the need for  both field-  and  basin-scale models.   The  field-scale models  are
needed to develop management  practices,  and basin-scale models  are needed  to
simulate  the  aggregate affects  of  treating  some,  or all,  of  the  individual
fields.

Another important factor  to consider  in computer  models  for nonpoint  pollution
is  the time period  for model  application.   In the  design  of road  culverts  or
other drainage structures,  a "design" type  storm  is generally used  whereby  the
expected  rainfall  amount during some predetermined  time  span  is  estimated.
The  resulting  runoff  rate  and  volume  generated with the model are  used   to
determine the  size  of structure to be built  or the sediment yield from  the
storm.  These are practical  applications of models.  However,  in the case  of
water quality, the  design-type storm  generally is not sufficient.   If  sediment
is the pollutant of concern,  then  a model  that simulates for a  single design-
type storm is satisfactory.    If pesticides  are  suspected   pollutants, a  daily
or  continuous  simulation model  must  be used  to consider  possible ranges   in
pollutant  loads or concentrations  of  toxic  compounds.     Since   pesticides
degrade in time, storms that  occur  within  a few days or weeks after  pesticide
application are potentially the  most  dangerous to  the  environment.   If  rain-
fall does  not occur within  two  or three weeks  after  pesticide application,
generally very little losses will occur even with  large  rainfall events.

Storms that produce small amounts of  runoff  on  the  day  of  pesticide  applica-
tion,  or  one day  after,  may  result  in  the most  hazardous  toxic  condition.
Large runoff  producing events  on that same  day  may result in dilution (lower
concentration), and a  toxic condition may not exist.   Therefore, all  possible


                                    K-v-3

-------
 conditions must  be  examined in  order to  assess  potential  pollution.   Since
 weather conditions do  not  repeat exactly  each  year,  it is not  sufficient to
 make model application for  a single  year,  or even two or three  years.   It is
 recommended that a representative climatic  record  be  used,  20 years for exam-
 ple,  in order that a range  of  conditions can be  considered  and  risk analysis
 made to determine the probability of exceeding some potentially hazardous con-
 centration.

 Pollution may occur from nitrate  leaching  into  groundwater  supplies or leach-
 ing  below  the  plant  root zone  arid  moving  laterally  into  water bodies.   Al-
 though  potential problems  may  be greatest shortly after fertilizer applica-
 tion,   such  potentials  may  exist most  any  time  due  to continuing  nitrogen
 cycling in the  natural  system.   For example, in a climatic region where signi-
 ficant  deep percolation occurs  in the winter,  high  losses  of  nitrate leaching
 may  occur  even  though  several  months elapsed  since  fertilizer  application.
 Combinations of low nitrogen uptake by the previous crop, return of crop resi-
 due  to  the  soil, nitrogen  mineralization,  and  possible  animal  waste applica-
 tion  can result in  high soil nitrate  levels  and leaching below the root zone.
 Here  again,  long-term  continuous  simulation  is  needed  to  assess  pollution
 potential.

 It is impossible to formulate a computer  model  to  exactly represent a physical
 process or processes.   Some model  developers  think  this  inexactness  can  be
 overcome by using observed data to calibrate a  model  before making production
 or simulation runs.  The  modeler recommends setting  some  initial  value  of  a
 few  (five or less) parameters,  run  the model,  and  compare the  model  results
 with  observed or measured data.   Then parameter values  would  be changed,  one
 at a  time,  -until the best comparison  is  obtained  between model  simulation and
 observed data.   If  observed  data are  not  available for a given site or problem
 area, then data from a similar site can be  used  (similar soil, similar  cli-
 mate, similar pesticide,  etc.).   If  the  model  is  physically based,  that  is,
 represents  physical  processes,  and parameters represent  physically measurable
 quantities,  the calibration amounts to "fine tuning"  those  sensitive paramet-
 ers.    Calibration  using  optimization techniques  with  observed data  is  not
 required.   A model  that is  not  entirely  physically based,  and  parameters  are
 not physically  measurable,  require more detailed calibration.   This results in
 additional  burden,  expense,  and uncertainties for  the model user.    Data  must
 be  obtained  and computer  simulations made  for calibration.   Probably  more
 important  is  the fact that data for calibration represents  some  given manage-
ment practice.   What  happens to those  calibrated parameter  values when differ-
 ent management  practices  are represented  in simulation?   The user  would  not
 have  data for  alternate  practices.    Can  the user be  sure  that the  original
 calibration  is  valid?   Probably so if  the alternate  practices  merely represent
changes  in  fertilizer or pesticide  application  rates  and dates.   If,  on  the
other hand,  the  alternate practices  involve different  types  of tillage such  as
chisel  plow  or  no-till compared  with moldboarding,   these  practices  affect
 infiltration  rates, crop residue  incorporation,  etc.   Are the  originally cali-
brated  parameter values valid for  the  alternate  practices?   That  is a decision
the user must make.   Every  model  should be validated with observed  data where
possible.   Fine tuning  of parameter  values is  desirable to improve the  pre-
dictability  of  a model, but  is  not essential  when  considering  relative differ-
ences between management practices.
                                   K-V-4

-------
A  question  that potential  model  users  always  ask is:   "How accurate  is  the
model?".  Model  developers may claim  their model  "predicts" within  75%,  80%,
or or 90% of measured  values  of runoff,  sediment,  or  chemical  load.   What  they
do not say  is that the level  of  accuracy is only  for  one  location  where cali-
bration was performed, the same  management  practice  was used,  and only a few
storms were simulated.   The problem of model accuracy touches  on  two aspects:
the proposed use  of  the model and  model  validation.    Before discussing these
two points, it would be  well  to discuss  measured or observed  data.

How accurate are  measured data?   A model  must  contain  some  relationships  for
estimating  runoff from  rainfall  and  it generally will  include a  soil-water
accounting procedure.  Measured data  really does not  include  measured rainfall
and measured  soil-water  content.   Rather,  these  are estimated from samples.
Rainfall is measured by,  at  best,  a few raingages.  For example, there  may be
two or three raingages in a 10-hectare  field.  A  standard raingage  is  200 mm
in diameter and approximately 31,400 square mm in  area.   Thus,  rainfall  is
estimated from  samples.   Soil water  content in that  10-hectare field  may be
"measured"  at 10 "sample"  sites using  a 25 mm diameter tube.  The  soil  is  not
absolutely  uniform  over the field, and  there is  a system of ridges  and  fur-
rows.   A  sample depth of  150 mm  on the  ridge  has a  different  elevation,  and
possibly soil texture  and  water content, than a sample depth of 150  mm  in  the
furrow.  Tractor wheels  run  in some furrows and the soil  is  more  compact  than
the soil on the ridge.   Then  soil water  content  for the field is "measured" by
weighting or  averaging  the content  determined  for  the  same  depth  from  all
locations.  Measurement  of rainfall  depth  and soil  water content in  samples
may be quite accurate, but  how well the  samples  represent  the field  is another
matter altogether.   Therefore, the true rainfall  and soil  water content  are
not known.  Conversely,  runoff from the  field  passes through the gaging  sta-
tion equipped with  a time-continuous  chart  and hydraulic  engineers  say  runoff
can be measured +_ 10 to  15% of the  actual  value.

Now consider sediment  yield from the  10-hectare   field.   All  of the sediment
cannot be trapped  and  measured.   The only  feasible method is to collect  sam-
ples of  sediment-laden  flow  periodically during  a storm event and  precisely
determine the  concentration  of  sediment.    Just  as   the  soil  in   the  field
varies, eroded  soil  varies and the sediment concentration in the flow  varies
both across the flow and  vertically through the  flow.   The sediment  concentra-
tion may  somewhat pulse  in  the  flow  just  as we  sometimes  see runoff  itself
pulse past  a gaging  station.  Measurement  of sediment concentrations in  sam-
ples may be very accurate.  These measured  sediment concentrations  are applied
over some  time  period  to calculate the  sediment  load.   Then  sediment  yield
must be sample estimates  rather than measured data.

The same water and sediment samples may  be  further processed  for dissolved and
adsorbed chemicals.  The  same degrees  of uncertainty  exist relative  to samples
versus measurements.   When chemicals  (fertilizers  and  pesticides)  are  applied
on the 10-hectare field,  concentrations of the chemicals  can be "measured" in
the soil periodically  just as soil water content  was measured,  i.e. by  sam-
pling.  It  was  stated  above that  a ridge  and furrow  system may be applied on
the 10-hectare field and  the  point  was made about elevation  of  the  respective
150-mm sample depths.   If a chemical  was spread uniformly over  the field,  in-
corporated  uniformly by  tillage,  and  then the ridge-furrow system  was formed,
                                  K-V-5

-------
 where is the chemical?  All in the ridge?  Partially in the ridge?  Where ver-
 tically?  How is "average" concentration in a 150-mm layer "measured" from the
 samples?  How many samples are needed to obtain the average concentration?  If
 the soil surface  is  rough and cloddy,  where  is  the reference  level  for sam-
 ples?  Are ihe clods part of the sample?   In  stony soils  such as in glaciated
 areas,  are the stones taken  as part of the soil  sample for water and chemical
 content?  When samples are taken for chemical  determination, does the sampling
 device  scrape soil from the sides of the hole each time it is entered or with-
 drawn,  thus contaminating the  sample?   Every  one  of these  questions  are very
 valid and have been  asked by  good  researchers many times over.   There  is not
 an  exact answer,  but in  defense  of the  research profession,  techniques are
 continually improved and refined to the degree that research data are general-
 ly  quite good if quality control  is maintained

 Reconsider  the runoff samples which  are used  to  "measure" sediment and  chemi-
 cals.   The  samples are generally collected by electrically operated mechanical
 or  electronic samplers.   These samplers are  not  fail-safe,  i.e. malfunctions
 will  occur:   power  (battery)  failure,  intake plugged  by sediment  or  debris,
 timer malfunction, or switching mechanism  failure.  In many  instances  one  or
 two storms  account for 90 to 95% of the annual runoff,  sediment yield or chem-
 ical  yield.  When "measured" data are closely examined,  it  is not uncommon  to
 find  that equipment failure resulted in only two  samples during the storm when
 as  many as  20 to 30 samples  should  have been collected.

 Now that everyone has been thoroughly dismayed by these discussions,  how does
 a modeler define  the  accuracy of a model  or  calibrate model parameters?   It
 has been stated  by authoritative researchers  that  sometimes  a model  is  better
 than  the data being  used for  validation.    The foregoing discussion does not
 imply that  a modeler should  take the attitude  that all  data are bad and  he has
 a perfect model  that should not be  subjected  to  validation.   It simply means
 that  caution must be exercised in  the way  a model  is tested  and used in  appli-
 cation.   Rather  than attempt  to  use  models   for  absolute  predictions,  they
 should  be used to examine relative  response between alternate management prac-
 tices.   Obviously model  developers  and  model  users alike hope  to  achieve the
 most  accurate results possible, but it  may be  much more desirable to determine
 if  one  practice  results  in  lower pollutant loads than another  practice  with
 the same long-term climatic  record.

 REVIEW  OF MODELS

 Now that everyone  thoroughly  understands  what computer  models are and  what
 they  can and cannot  do,  and  no one  trusts  models  or modelers,  we will  review a
 few of  the  more  commonly used  models before demonstrating  an  application.  The
 review  will  be brief  since we  will  not consider details of models  or  compari-
 sons  of  models.

 The Federal Water Pollution  Control Act  Amendments  of 1972,  Public Law No.
 92-500,  was the  legislation that  attracted concern about pollution from  dif-
 fuse  agricultural  and forestry sources, and  provided  the stimulus for  model
 development.   It was not  considered initially that mathematical models  would
 be  developed and  used;  in other words,  PL  92-500  did  not state  "Let there  be
models".  There  was considerable support for  monitoring fields,  watersheds  or
 basins  in Order to  evaluate  the pollution  and  any  improvement  in water quality


                                   K-V-6

-------
resulting from  improved management  practices.   But  in  practicality,  monitoring
is expensive  and data  from  a  specific  practice  on  a particular soil,  in a par-
ticular  climatic  region  could not  be  extrapolated  to  another practice  at  the
same  location or the same practice  on  another  soil  type,  or  in another climat-
ic region,  without the use of  models.  The use of models is really  the  only
way that  research  data can  be transferred  from the location  of  collection  to
another  location.

Engineers have  used models  for  many years to estimate storm  runoff  from rain-
fall.  Kuichling  (1889) developed the rational  formula, Q =  CiA,  to  calculate
urban  storm  sewer  discharge rate,  0,  using  a  runoff coefficient, C,  rainfall
intensity,  i, and  drainage  area, A.   The rational  formula is a model,  and  it
can be programmed  for  a computer very easily.   This model is  being  used today
to make  quick estimates  of  peak discharge rate for  selected  recurrence inter-
vals.   In  fact,  it is  an  accepted  engineering  practice supported  by  legal
precedent.

In the 1950's,  the U.  S.  Department of Agriculture, Soil  Conservation Service
(1972) developed  a procedure, referred to as the  SCS  curve  number method,  to
estimate  direct surface  runoff  volume.   Although  the commonly  used  procedure
is in  the graphical form, mathematical relationships are  available for use  in
computer models.

Wischmeier  and  Smith   (1965)  published the first  comprehensive erosion  equa-
tion.   The  Universal  Soil  Loss Eauation  (USLE),  which  estimates  long-term
average-annual  gross erosion, has become  the basis  of many erosion models.

One of the  earliest published  hydrologic computer  simulation models  was  the
Stanford Watershed Model (Crawford  and Linsley, 1962).  Although the  model  was
developed for large watersheds  or basins, it was later adapted  for  field  size
areas.

Negev  (1967)  developed the first  erosion/sediment  yield counterpart of  the
Stanford Watershed Model.   Negev utilized some  of  the  basic  concepts  of Wisch-
meier  and Smith (1965) on rill  and  interrill (sheet) erosion, but also inclu-
ded a component for channel erosion.

Glymph and  Holtan  (1969)  developed  an infiltration-based  hydrologic model  to
determine the effects of  management  practices  on  streamflow.    The  computer
model, called USDAHL (U.  S. Department of Agriculture Hydrograph  Laboratory),
later became the hydrologic component  for a  nonpoint pollution model.

Other  hydrology and erosion/sediment  yield models  have  been developed,  and
have  been used  successfully for  specific purposes.   However, these  aforemen-
tioned models are the  more  important ones that  later were  used in  the  develop-
ment of nonpoint source pollutant models  that include  plant  nutrients  and  pes-
ticides as well  as sediment

Under  the terms of PL 92-500,  the Environmental  Protection  Agency (EPA)  had
responsibility  for  agricultural  chemical  pollution  and its reduction.   There-
fore, the EPA had  a major role in research  and  development,  both  "inhouse"  and
contracting,  relative  to  nonpoint pollutant models.   The  U.  S.  Department  of
Agriculture,  Soil  Conservation  Service,  was responsible for  reducing  sediment
                                   K-V-7

-------
 as  a pollutant  and  as  a chemical  pollutant  carrier.   These  respective roles
 were later changed,  but  for the purposes  of  model  development  and  review in
 this paper,  the dual  agency roles were important.

 The  first  EPA-contracted model  for nonpoint pollution was the Pesticide Trans-
 port and  Runoff (PTR)  model  developed  by Crawford  and  Donigian  (1973)  for
 field-size areas.   The PTR model  used  modifications  of  the Stanford Watershed
 Model  (Crawford and  Linsley,  1966) and Negev's  (1967)  erosion/sediment yield
 model.   Although  the hydrology  and  sediment  yield components  were developed
 for  basins,  they were applied on field-size areas with pesticide transport in-
 corporated.   A real  danger may exist when  scaling a  component  up or down, and
 that is  over sensitivity  of  some parameters,  or distortion  of  the parameters
 out  of the range for  which they were originally developed and tested.  Another
 common pitfall  is  that  a component  previously  developed  may  not be consistent
 or provide the exact  information needed in a developing  component.  It is more
 desirable  and efficient  to develop all  components simultaneously.

 Constraints  of various  kinds can result in not being able to  develop a compre-
 hensive  model on the  first trial, and this was no different for PTR.  In test-
 ing  and  further development of PTR,  Donigian  and Crawford  (1976a) incorporat-
 ed,  among  other items,  plant nutrient transport and  produced  the ARM (Agricul-
 tural Runoff  Management)  model.

 During  the development  of ARM,  it was  realized that a  simpler  version  was
 needed to  operate in connection  with  urban models.   Thus, Donigian  and Craw-
 ford (1976b)  developed  the NPS (NonPoint  Source) model.  The  hydrologic algo-
 rithms of  NPS were used with the Hydrocomp Simulation Program  (HSP)  to simu-
 late nutrient  loadings  in surface  runoff  from urban and  agricultural  water-
 sheds.

 The  need for  basin-scale models  resulted  in the  natural  evolvement under con-
 tract from EPA, and  the  development of the HSPF: model  (Johanson et al.,  1980).
 HSPF is  the principle basin-scale model  in use today.

 At the same time that some of the  EPA contracting occurred,  efforts  were made
 in the U.  S.  Department  of Agriculture to develop models  that  reflected  exper-
 tise in  hydrology  and erosion/sedimentation.   Frere,  Onstad,  and Holtan  (1975)
 used the USDAHL model  (Glymph  and Holtan, 1969) and  a modification of the USLE
 (Wischmeier  and Smith,  1965)  to develop  the  Agricultural  Chemical  Transport
 Model (ACTMQ).   ACTMO was  a basin-scale model, but has been  applied  on  field-
 size areas as well.   Somewhat  simultaneous with the ACTMO development,  Bruce
 et al. (1975)  developed  a  parametric model  for water-sediment-chemical  (WASCH)
 runoff for single  storm  events  on field-size areas.  The WASCH  model  required
 data for  parameter  calibration,  and  its chemical  component  only  considered
 pesticides.

 Lack  of  documentation and  validation of  ACTMO and the  data requirement  for
 calibration of  the WASCH model  resulted  in little use  of  these  two models.   In
 order to fulfill their  role in  research to the Soil Conservation  Service,  the
Agricultural  Research Service established  a  national  project that  culminated
 in development  of the CREAMS model  (Knisel,  1980).    CREAMS  is  a physically-
 based model  that can be used,  without  calibration,  to  compare  edge-of-field
 pollutant  load  response  for alternate  management  practices.   The  SCS  curve


                                    K-V-8

-------
number method  (I).  S.  Department  of  Agriculture,  1972)  is  used  as one option in
CREAMS to  simulate daily runoff from daily  rainfall.   A second  option  (Smith
and Williams,  1980)  uses a modified  Green  and Ampt (1911)  infiltration  equa-
tion to simulate runoff  from  time-intensity  rainfall data.   The erosion  compo-
nent  is  the Foster  et  al.    (1977) modification of the  USLE  (Wischmeier  and
Smith, 1965) on a  storm  basis.   CREAMS  also  includes plant  nutrients (Frere et
al., 1980)  and pesticides (Leonard  and  Wauchope, 1980).    A paper to be  pre-
sented later in this symposium (Leonard et al.,  1985) will  describe modifica-
tions of CREAMS to consider  pesticide leaching.   CREAMS  has been used  widely,
both nationally and  internationally,  to  develop  or  select management practices
to reduce nonpoint pollutants  delivered  to the edge of the  field.

Steenhuis and  Walter (1980) used the  SCS curve number method (U.  S.  Department
of  Agriculture,  1972)   and  the USLE  (Wischmeier  and  Smith,  1978)  for  the
hydrology  and  the  erosion components of their CPM (Cornell Pesticide  Model).
CPM actually was developed to make  use  of most any hydrology  and erosion  com-
ponents that generated volumes of water  and  loads of sediment.

The above referenced  models do not  include groundwater components.   Enfield et
al. (1983) developed  the  PESTANS model to provide a one-dimensional  projection
of vertical chemical  movement through the unsaturated zone.   Although  PESTANS
treated very  simplified  conditions,  it   led  to  the later  development  of  the
Pesticide Root Zone  Model (PRZM) by Carsel  et  al.  (1984).   PRZM contains  the
SCS curve number method  (USDA-SCS,  1972)  to  partition rainfall  into  runoff and
infiltration,  and  contains Williams  and Berndt's (1977)  modification  of  the
USLE  (Wischmeier   and  Smith,   1965)  to   calculate sediment  yield.   PRZM  was
developed to evaluate movement of pesticides  within the root zone and the  low-
er unsaturated/saturated  zones.

Leonard et  al. (1985) modified  the CREAMS  model  (Knisel,  1980)  to  calculate
pesticide movement within the  active  root zone.   Although the  modifications do
not constitute a groundwater  model,  percolation  of pesticides  below  the  active
root zone provides estimates  of  loadings  to  the  vadose zone.

A number  of  models have  been  developed  for  erosion and  sediment yield  only,
and do  not  include  plant  nutrients or pesticides.    Beasley  et al.   (1977)
developed the  Areal  Nonpoint  Source Watershed Environment  Response  Simulation
(ANSWERS} model to consider erosion/sedimentation in a basin for  a design-type
storm.  ANSWERS has  been used to identify critical  areas of erosion for  pro-
tective practices.   Also,  ANSWERS  has  been  applied  on  field-size areas  to
examine variability  of erosion and  deposition within heterogeneous fields.

Simons et al.  (1977)  modified  the USLE (Wischmeier and Smith,  1965)  to  develop
a procedure  for  estimating source  erosion,   and  developed  a  sediment  routing
component  for  calculating basin sediment yield.   The  model  operates  with  a
design-type storm.

Williams and Hann  (1978) developed a computer model  to   analyze  runoff,  sedi-
ment,  and plant nutrient  losses  from  basin-size  areas.  Although  the model  was
developed  to  consider economic  factors, it  provided  information on sediment
and nutrient losses  from  a basin.
                                   K-V-9

-------
 The models herein reviewed  obviously do not represent an exhaustive  list,  but
 they do represent  those models most commonly used by  action agencies  today.
 Furthermore, it is easy to  see what  basic hydrology and  erosion  components  are
 included.   Due to  this  author's involvement  and  experience with CREAMS, this
 model  was selected  for  demonstration of how  a computer model can  be used  to
 develop management practices.

 MODEL  SELECTION AND APPLICATION

 Model  Selection

 Model  developers  never  envisioned  that models,  either  field scale  or basin
 scale,  would be used to  identify water  quality problems.  That is, it was  not
 anticipated  that a  model  such as, for  example,  HSPF (Johanson  et al., 1980)
 would  be applied on some basin  for  existing management  practices, and be able
 to say "there is a  problem  of toxicity of  pesticide  'X1  in this water  body".
 On the  contrary,  it  was  envisioned  that  after  a  water quality  problem  was
 identified,  such as  concentrations  of  pesticide  "X"  at  toxic levels,  models
 would  be useful in  analyzing alternative practices for  alleviating the prob-
 lem.   Intuitively,  a person  can generally  be  more specific about the  problem.
 For example, pesticide  "X"  is highly adsorbed to the soil, and erosion/sedi-
 ment transport is  essentially the  only method of transport.   Then management
 practices  can be evaluated on the basis of  sediment and  pesticide delivered  to
 the edge of  the field or to the outlet of  the basin.  This sounds simple  and
 straightforward,  but caution  must  be exercised.  If  management  practices  are
 developed  to reduce erosion and sediment yield,  include conservation .tillage
 or minimum tillage, a new  spectrum  of  herbicides,  insecticides, nematicides,
 or fungicides may be required.   If  any of the new  pesticides  are very  mobile
 (in water),  the improved management practices may actually  result in more pol-
 lutant  runoff.   Another  possibility may be that little of the mobile pesticide
 is lost in  surface  runoff,  but  significant loads  may  be  transmitted  tothe
 aquifers.  Therefore,  the full range of management effects  should be evaluated
 with models  over some  extended climatic record.  These kinds of situations  are
 common  occurrences  that  must be considered totally.

 A  large number of models  has been developed,  and  this  is  both  good  and bad.
 It is  good in the sense that  the  model  user has  several  potential  tools from
 which to choose  for  their job.  It is bad in that the burden  is  upon  the user
 to decide  which model  is appropriate for  the  problem  at hand.   The  user must
 learn the  model  concepts, assumptions,  and  limitations   and whether or not   it
 will adequately treat the problem  of concern.,  The user also must  carefully
 consider the input  requirements, if required data  are  available,  how much time
 is  required  to  develop parameter sets, what is involved  in changing parameters
 to reflect management  changes, and  what the computer requirements are  relative
 to storage and  run  time.   In essence, the model user  must  know as  much as the
 developers of  several  models.   A person really cannot  learn a model until they
 actually run it,  get familiar with  it,  and  feel  comfortable  in  what  they are
 doing.   In other words,  a model user must  become  an  expert before  really un-
 derstanding  a model.   This  does not  mean that  no one should use a model before
 they become  an  expert.  A potential  model  user oftentimes  is  in  a  dilemma  as
 to  which model  or  models should be  used.   Some thought  has been  given to es-
 tablishing an  independent unbiased  group  (council)  to test and  run  available
models  to  be able to  recommend model A, model B,  or model C for  a  specific


                                   K-V-10

-------
user's problem.  This  is
decisions are not  always

Model Application
not really  feasible,
easy to make.
but it points out  the  fact  that
Details of the  CREAMS model  will  not be given  in  this  paper.   Model concepts,
assumptions,  limitations,  or specific physical  processes  are  not important in
demonstrating  the  model applicability.   This  information  is  fully documented
by Knisel  (1980),  and  readers are referred  to  that  publication if  interested.
However, a schematic representation  of the physical  system addressed by CREAMS
is shown in  figure 1.   The  field  (watershed)  is depicted  as  being  made up of
some geology,  soil, and topography which is acted upon by natural climatic in-
put (precipitation, radiation,  and temperature).  The  farmer  superimposes his
management  input  (crops, tillage  practices, and pesticide  and  fertilizer ap-
plications),  and the  field responds  to  produce such  output as surface runoff,
erosion/sediment yield  and associated dissolved  and  adsorbed  chemicals.  Some
fraction of  precipitation  infiltrates into  the  soil,  and  a portion of the in-
filtration may go  to  percolation  which  also may  contain  dissolved chemicals.
Thus, figure 1  is  a flow chart of the  system that was used as  a guide in the
development  of  CREAMS.

                             MANAGEMENT INPUT
LAND
USE
                 NATURAL
                                                        OUTPUT
                               FIELD SYSTEM
                              (GEOLOGY, SOIL
                                TOPOGRAPHY)
Figure 1.  Flow chart  of  system for evaluating nonpoint source pollution.

A very simple  scenario is used to  demonstrate  application  of a computer model
to select management  practices for reduction of  nonpoint  pollution.  In order
to not compound the system  response among crop rotation, year-to-year climatic
changes,  and  fertilizer-pesticide-tillage  interactions,  a  simple continuous
crop was  used.   Continuous corn  was  simulated in a  conventional  tillage sys-
tem.   A 20-year  climatic record  at  Tifton, Georgia  was  used  for simulating
runoff,  percolation,  sediment yield, pesticide runoff,  and  pesticide leaching
on a loamy sand soil  in  the Coastal Plain.  Rainfall  at Tifton averages about
                                    K-V-ll

-------
 1200 mm per year.  Several of  the  Coastal  Plain soils are subject to develop-
 ment of plow pans  which  result from continued  tillage at  the  same depth year
 after year.  These dense  comoaction layers are similar to  the  clay pan soils
 that occur in Wisconsin.   The  layer restricts  root  growth  and  water movement
 at a depth of approximately 20 cm.  This is really the effective rooting depth
 for a crop unless tillage practices are  used  to break the pan  condition.  The
 main difference between the Wisconsin  clay pan  and  the Coastal  Plain plow pan
 is that the  clay  pan  is more  restrictive to  water  movement.    Also,  chisel
 tillage is more effective in breaking the plow pan.

 For demonstration purposes, three management practices were simulated:

      (1)   continuous  corn, conventional moldboard tillage (no surface
           residue), 20 cm root depth;

      (2)   continuous  corn, chisel 30 cm deep,  residue management, 45 cm
           root  depth;

      (3)   continuous  corn, chisel 30 cm deep,  residue management, 45 cm
           root  depth, with a rye winter-cover  crop.

 In all  three management  practices,  two generic pesticides  were  considered to
 be broadcast on  the  surface at  planting time.   The  two pesticides  have  the
 same solubility (10 ppm)  and same half-life (approximately three weeks).  Pes-
 ticide  A  is  transported  mainly  in  the water.,  whereas   pesticide  B  is  more
 strongly  bound to  the soil particles  and  organic matter  and  thus  move  with
 sediment.   These  management practices are not  recommendations,  and the results
 would differ  for  other soils,  other climatic  regions,  and other  cropping sys-
 tems.   They are useful,  however,  to demonstrate the  use of a model  in managing
 nonpoint pollution  from  agriculture.

 Simulation results  of direct surface runoff for the  three management practices
 are presented  in  figure  2.  The  largest  accumulated  runoff  volume  was simula-
 ted for management  practice 1  representing a  20  cm  root  zone.    The  second
 highest accumulated runoff was  simulated for management practice  2.   The chis-
 el  tillage in practice 2  affected three parameters in the model:   (a) increas-
 ed  the  rooting depth,  (b)  increased  the  hydraulic  conductivity of  the  root
 zone,  and  (c)  reduced the evaporation potential due  to  crop residue  left  on
 the surface.   These changes resulted  in  a  significant reduction in  runoff  by
 about one  third compared  with  practice 1.  The  rye winter  cover  of  practice  3
 resulted  in some plant transpiration,  but  only about  7%  additional  reduction
 in  runoff  occurred  as depicted  by the bottom line in  figure 2.   These results
 represent  the runoff response for  the three  management  practices  which  has
 further implications  relative  to  erosion  and pesticide losses.

 Results of simulation of  percolation below  the root zone, i.e. 20 cm for  man-
 agement practice  1  and 45  cm  for management practices 2  and  3,  are  shown  in
 figure  3.  The  same order  differences  as  for runoff  are apparent:   more  perco-
 lation  for practice 1,  intermediate for practice 2, and least for  practice  3.
 This requires  a little further interpretation.   If  runoff was less  for  prac-
 tice 2  than practice  1, how can percolation be less,  also?  The explanation  is
 that for practice 2,  there is  an  increased soil reservoir  (storage)  and  also
more plant use  of water.   It should be  noted  that the difference  in percola-


                                   K-V-12

-------
         225


         200


         175


         150
       E
       3 125
         100
          75
          50
          25
                                      PRACTICE I-
                                                         PRACTICE 3
           1957
                      1961
                                  1965         1969
                                       YEARS
                                                         1973
                                                                     1977
Figure 2.  Accumulated  runoff  simulated  with  the CREAMS  model,  1958-77.
         750
         625
       _ 500
       E
       u
       O  375
       8
         250
          125
                                           PRACTICE I-
PRACTICE 2
            1957
                       1961
                                   1965        1969
                                       YEARS
                                                          1973
                                                                      1977
 Figure 3.   Accumulated percolation simulated with  the CREAMS model,  1958-77.
                                       K-V-13

-------
 tion between practice 1  and  practice 2 is  relatively small  compared with  the
 difference in runoff for the  same  practices (10% for percolation).  The  addi-
 tion of a winter cover crop in practice 3 results  in  water uptake,  thereby  re-
 ducing soil  water stored, during the time that  significant amounts  of rainfall
 and  percolation normally occur.  Thus, percolation for practice 3  {bottom line
 of figure 3) is reduced an additional 16%.  Another  important  point  about fig-
 ures 2 and 3 are the  relative magnitudes of runoff  and  percolation.  The  20-
 year accumulated  values  show  that percolation  is almost four  times greater
 than surface runoff.  This has significant  implications in nonpoint pollution
 as will  be shown later.

 Figure 4 shows the  cumulative 20-year simulated  sediment  yield  for the  three
 management practices.   Sediment yield for practice 1 is quite high  (figure 4)
 when considering the low volume of  runoff (figure 2).  The reduced  runoff  and
 residue management  from  practice 2  resulted  in  a 43%  reduction  in sediment
 yield.   An  additional  10% reduction resulted  from the  winter  cover crop  for
 practice 3 (figure 4).
           400
           350
           300
           250
         3 200
         u
           150
         8 100
         
-------
 (figure 5).  The 20-year  percolation loss was 43% of  the  amount  applied.   Al-
 though percolation  volume was  reduced  by about 10%,  for management practice 2,
 it  resulted in a 41%  reduction  in  percolation loss  of pesticide  A.   Since the
 pesticides  were  applied each year of  simulation  on  the date of  corn  planting
 and the half-life  was   about three  weeks, percolation reduction  by  the winter
 cover crop  was not  effective in  further reduction of pesticide  losses  (manage-
 ment practice 3).
                   1957
                           1961
                                  1965      1969
                                      YEARS
1973
        1977
Figure 5.  Accumulated leaching  loss  of  pesticide  A simulated  with  the CREAMS
           model, 1958-77.

The 20-year  sediment-associated  loss  of pesticide  B for management  practice 1
amounted to  less than 100 g/ha,  whereas  runoff  loss was  1640 g/ha and  percola-
tion loss was  16,900  g/ha.   Data from  simulation  of surface runoff losses  of
pesticide B  are shown in figure  6.  The  20-year runoff  loss  for practice 1  was
2.7% of  the amount applied.   Reduced  runoff from management  practice 2  re-
sulted in a  37% reduction  in the runoff  loss.   Although the  runoff reduction
for practice 3 compared  with practices 1 or  2  was very small   (figure  2),  the
pesticide runoff loss was significantly  reduced by practice 3,  a 53% reduction
from that of practice 1.

Figure 6 shows another  very important point  made  earlier  in  this paper.   The
flat portions of the  curves  between 1965 and 1972 differ drastically  from  the
1958-64  and  1974-77 periods.   If  a  5-year  period between 1965 and 1972  had
been selected, relative  difference  between practices would have  been  negligi-
ble.  If the 5-year period  (1960-64)  had  been selected  for simulation,  projec-
ted difference would have been exaggerated.   It is important that longer, more
representative climatic  records  are used.  In the 1965-72 period,  either run-
off was  low  each year,  or  runoff producing rains  did  not  occur  for some time
after pesticide application.
                                      K-V-15

-------
                 1800-
                 1500 •
                 1200 -
                 900 -
                                   1965      1969      1973      1977
                 600 •
                 300 •
                   1957    1961
Figure  6.   Accumulated surface runoff loss of pesticide B simulated with the
            CREAMS model,  1958-77.

The  results of the several CREAMS model  simulations  do not  show that there is
or is not  a water quality problem because CREAMS is not a water quality model.
The  simulations were  not  intended to be  used  to make such claims.   They are
merely  edge-of-field  and  bottom-of-root  zone losses  for  three  different man-
agement practices.   If there had been a  problem in  water  quality,  as deter-
mined at some  off-site location,  this kind of  model  application could be made
to estimate response from existing and alternative management practices.  Even
then, the  simulation results would not indicate an offsite water quality prob-
lem  and its "cure".   A basin-scale model  would  have  to be applied for routing
the  aggregate  field  response to the  receiving  waters  -- river,  lake, estuary,
or etc.  Thus, one  has  to be  aware  of  the problems,  needs,  tools,  benefits,
pitfalls,  and  other  obstacles  and opportunities  to  address  nonpoint  pollution
problems.

SUMMARY

Computer models are  valuable tools with  which  sound  professional  judgment can
be used in evaluating management  practices and  treatments for nonpoint pollu-
tant loads.  The  several  models currently being used  all have their strengths,
weaknesses,  assumptions,   and limitations.   None are predictive  in  terms  of
absolute quantities.   They can  be extremely  useful to  consider response from
different  management  practices over  a  long  representative  climatic  period.
Also, they can be used to extrapolate research results from  the  specific re-
search  location to  another  location  (soil,  climate,  management  practice)  of
interest.
                                   K-V-16

-------
There  are  no  water quality models.  There  is  no  model  that this  author  knows
about  that can be  used to  say  "these management practices  cause eutrophication
of  this  lake",  or that  "these practices caused pesticide  concentrations  that
are toxic to  fish  (or man)  in  this  stream".  What  the  computer models can  tell
the user is that certain pollutant  loads may occur over  some  period of time as
a result  of  some  practices.   Models,  used  with  sound professional  judgment,
can be valuable aids in making sound management decisions.

REFERENCES

Beasley, D. B., Monke,  E.  J.,  and Muggins,  L. F.   (1977).   ANSWERS:   A model
for watershed planning.   Purdue Agricultural Experiment Station Journal  Paper
No. 7038.
Bruce, R. R., Harper, L. A., Leonard,  R. A., Snyder, W.  M., and Thomas, A.  W.
(1975).  A model for runoff of pesticides from  small upland watersheds.  J.  of
Environmental Quality, 4(4):541-548.
Carsel,  R.  F.,  Smith,  C.  N., Mulkey,  L.  A.,  Dean,  J.  D.,  and Jowise,  P.
(1984).  Users manual  for  the  Pesticide  Root  Zone Model  (PRZM) Release 1.  U.
S. Environmental Protection Agency.Athens, Georgia.
Crawford, N.  H.,  and Donigian, A.  S.,  Jr.  (1973).   Pesticide Transport  and
Runoff Model  for  agricultural  lands.   U. S. Environmental Protection Agency,
Environmental Protection Technology Series,  EPA-660/2-74-013.
Crawford, N.  H.,   and  Linsley, R.  K.    (1962T!    The  synthesis of  continuous
streamflow hydrographs on  a digital computer.  Stanford University,  Department
of Civil Engineering, Technical Report No.  12.  Stanford,  California.
Crawford, N.  H., and Linsley,  R. K.(1966).   Digital  simulation  in hydrology:
Stanford Watershed Model  IV.    Stanford  University,  Department of  Civil  Engi-
neering, Technical Report  No.  39.
Donigian, A.  S., Jr.,  and  Crawford, N.  H.   (1976a).   Modeling pesticides  and
nutrients on  agricultural  lands.  U. S.  Environmental Protection Agency,  Envi-
ronmental Protection Technology Series,  EPA-600/2-76-043.
Donigian, A.  S., Jr., and  Crawford,  N.  H~!(1976b).Modeling  nonpoint pollu-
tion from the land surface.  U. S.  Environmental Protection Agency,  EPA-600/3-
76-083.
Enfield, 6.  G.,  Carsel, R.  F., Cohen,  S.  Z., Phan,  T.,   and  Walters, D.  M.
(1983).   Approximating  pollutant  transport  to   groundwater.    Ground Water
20:711-721.
Foster, G. R.,  Meyer,  D.  L.,  and  Onstad, C. A.   (1977).   A  runoff  erosivity
factor and variable  slope length  exponents for soil  loss  estimates.   Trans.
Am.  Soc. Agric.  Eng. 20(4):683-687.
Frere, M. H., Onstad, C. A., and Holtan,  H.  N.  (1975).  ACTMO,  an Agricultur-
al  Chemical  Transport  Model.   U.  S.  Department  of Agriculture,  Agricultural
Research Service,  Headquarters, ARS-H-3.
Frere, M. H., Ross,  J. D., and Lane,  L. J.   (1980).   The nutrient  submodel.
In:   W. G. Knisel  (Ed.).   CREAMS:   A field-scale model for Chemicals,  Runoff,
and Erosion from Agricultural  Management Systems.   U.  S.  Department  oF Agri-
culture, Conservation Research Report  No. "26.  pp. 65-87.
Glymph, L.  M.,  and  Holtan,  H. N.   (1969).   Land  treatment  in  agricultural
watershed hydrology  research.    In:   W.  L.  Moore  and  C. W.  Morgan  (Eds.),
Effects of Watershed Changes   on Streamflov, Water Resources  Symposium No.  2,
University of Texas, Austin, Texas, pp.  44-68.
Green, W. A., and  Ampt, G. A.   (1911).   Studies  on soil  physics,  I.   The  flow
of air and water through soils.  J. of Agricultural  Science, 4:1-24.


                                      K-V-17

-------
 Hydrocomp Incorporated (1969).  Hydrocpmp  Simulation  Programming:   Operations
 Manual.   2nd  edition.   Palo Alto,  California.
 Johanson, R.  C., Imhoff, J. C., and  Davis, H.  H.   (1980).   User's  manual  for
 the  Hydrologic Simulation Program  — FORTRAN (HSPF).   U. S. Environmental Pro-
 tection  Agency,  Environmental  Protection Technology Series, EPA-600/9-80-015.
 Knisel,  W. G.,  (Ed.).   (1980).   CREAMS:   A field-scale model  for  Chemicals,
 Runoff,  and Erosion from Agricultural IManagement Jystems.  U. S. Department of
 Agriculture,  Conservation Research  Report No. 26.
 Kiiichling, E.(1889).   Hie relation between the rainfall and the discharge of
 sewers  in populous districts.   Trans, of  the  Am.  Soc.  of Civil  Engrs., Vol.
 20,  pp.  1-56.
 Leonard,  R. A.,  Knisel,  W.  G., Still, D. A., and Johnson, A.  W.   (1985).  Mod-
 eling  vertical flux of  pesticides  with  CREAMS.   Proceedings of  the  Nonpoint
 Pollution Symposium, Milwaukee,  Wisconsin,  April 23-25,  1985.
 Leonard,  R. A.,  and Wauchope,  R. 0.   (1980).  The  pesticide submodel.   In:   W.
 G. Knisel  (Ed.).   CREAMS:   A field-scale model  for Chemicals, Runoff,  and Ero-
 sion from Agricultural IManagement  Systems.  U., S. "Department of Agriculture,
 Conservation  Research  Report No. 26T  pp.  88-112.
 Negev, M.  A^(1967).Sediment  model on a digital computer.   Stanford Univer-
 sity, Department of Civil  Engineering, Technical Report  No. 76,  Stanford, Cal-
 ifornia.   109  pp.
 Simons,  D.  B., Li, R.  M.,  and Ward,  T.  J.  (1977).   A simple  procedure  for
 estimating on-site soil erosion.   Proceedings  of  the International  Symposium
 on Urban  Hydrology, Hydraulics,  and  Sediment Control, University of Kentucky,
 Lexington, July  18-21.   pp.  95-102.
 Smith, R. E., and  Williams,  J.  R.    (1980).   Simulation of  the  surface water
 hydrology.  In:   W.  G.  Knisel  (Ed.).   CREAMS:   A  field-scale model  for Chemi-
 cals, JRunoff,  and  Erosion  from Agricultural Management jjystems.   U.  S. Depart-
 ment of Agriculture, Conservation Research  Report  No.  26.  pp. 13-35.
 Steenhuis, T.  S.,  and  Walter,  M. R.(1980).Closed  form  solution  for pesti-
 cide loss  in  runoff water.   Trans.  Am. Soc. Agric.  Eng.  23:615-620,  628.
 U.  S.  Department  of  Agriculture,   SoilConservation  Service.    (1972).   SCS
 National  Engineering Handbook, Sec.  4, Hydrology.   548 pp.
 Williams,  J. R.,  and Berndt, H.  D.(1977).Sediment yield  predictions based
 on watershed hydrology.   Trans.  Am.  Soc.  Agric. Eng.  20(6):1100-1104.
 Williams,  J. R,  and Hann,  R.  W., Jr.(1976).Optimal  operation  of large  ag-
 ricultural watersheds  with  water quality constraints.  Texas  A & M University,
 Texas Water Resources  Institute, Technical  Report  No.  96.
 Wischmeier, W. H.,  and  Smith,  D~U!    (1965)~Predicting  rainfall-erosion
 losses from  cropland east  of  the  Rocky  Mountains -- Guide  for selection  of
 practices  for  soil  and water  conservation.   U. S. Department of  Agriculture,
 Agr. Handbook  No. 282.  47  pp.
Wischmeier, W. H.,and  Smith,  D.  D.   (1978).   Predicting  rainfall  erosion
 losses -- a guide  to conservation  planning.  U. S. Department of  Agriculture,
 Agric. Handbook  No. 537.  58 pp.
                                   K-V-18

-------
              ROLE OF MATHEMATICAL MODELS IN DESIGN AND SELECTION  OF
               BEST MANAGEMENT PRACTICES FOR CONTROL OF POLLUTION
                        FROM URBAN AND URBANIZING  AREAS

                                      by

                               Vladimir Novotny
                  Professor, Department of Civil  Engineering
                             Marquette University
                         Milwaukee, Wisconsin  53233

                                   ABSTRACT
Mathematical models are an integral  and indispensable part of any nonpoint
pollution abatement project.   There are two types of the models  which  have
been developed:  screening overview models, and detailed hydrological  models.
These models were divided into five levels according to their complexity.

The paper discusses the types of models and their usefulness in  identifi-
cation and abatement of nonpoint pollution from urban and urbanizing areas.
The approach is demonstrated on a practical example of a use of  a medium
complexity hydrological model on a typical urban watershed.

Keywords:  Complex models, hydrological models, urban runoff models, water
           quality models, nonpoint pollution, nonpoint pollution models,
           runoff, pollutant build-up, pollution by urban runoff, urban
           stormwater modeling, mathematical models.
                                  K-VI-l

-------
 INTRODUCTION

 Hydrological mathematical models have become an integral part of urban drain-
 age  engineering.   Recent advancement of small computer hardware technology
 and  availability  of  relatively  inexpensive microcomputers have meant that
 almost  every city engineer's or consultant's office can now possess a
 computer  capable  of  an advanced application of urban hydro!ogical models.

 Therearenow a  large number of  models available from various agencies,
 universities and  software vendors that range from simple applications of
 basic hydro!ogical procedures to more advanced and more complex urban run-
 off  quantity and  quality models.  A user may have a problem of selecting a
 model that would  be  reliable, perform the tasks that are needed, would have
 adequate  documentation and support, its acquisition would not be costly,
 and, finally, select a model that has been tested and verified.

 Why Models Are  Needed

 One question model advocates must first answer is "why we need the models?"
 There are a number of reasons for and against the use of models, but the
 most obvious one  is  that models can go far beyond the results of any moni-
 toring  program.   Monitoring programs are always time limited.  A typical
 very expensive  monitoring program (such as those conducted during the Na-
 tional  Urban Runoff  Project) may cover two years of monitoring on a few
 experimental small watersheds.  Chances are that neither of the two years
 will reflect extreme or average conditions required for the design.  In
 this case the loading estimates from the monitoring program may be mis-
 leading,  considering the probabilistic hydrologic nature of the nonpoint
 pollution generation process.   In a dry year, the loading will be low, in
 a wet year too  high.  In this case the model is a tool that can help to
 organize  the data, and bring them in line with the average or extreme
 hydrologies! conditions.

 Another extreme  view is to rely solely on models and disregard the data.
 As it will be shown  in this article, monitoring field data are needed to
 calibrate and verify the model.  A model  that has not been calibrated
 (tuned) loses its practical usefulness because the answers given by the
 model can be inaccurate.  In an ideal  and most common situation both
 monitoring and modeling programs should be a part of a project.  Further-
 more, the monitoring program should be designed to provide the necessary
 calibration and verification data.

 TYPE AND  BASIC STRUCTURE OF THE MODELS

 Nonpoint  pollution models are part of a category of "loading models,"
which represent and simulate generation and movement of water and its
 pollution content from the point of origin to watercourse.    These models
may interface with water quality models for receiving waters and provide
 input concentrations and flow rates (hydrographs and quality loading
 histograms),
                                  K-VI-2

-------
Urban runoff and nonpoint pollution models fall generally into two categories
that can be further subdivided, (Novotny and Chesters  (1984); Novotny (1984);
Sutherland  (1980)). The first broad category includes more simple screening
models and the second category covers more complex computerized hydrological
assessment models.  A good review of hydrologic fundamentals of urban runoff
models was presented by Hall (1984) and by the APWA (1981).

Screening Models

Screening models rely on --or evolved from established traditional urban
drainage design procedures and/or statistical results of long term monitor-
ing programs.   They can be further subdivided into:  I.  Simple statistical
(desktop) procedures and unit loads with no interaction  among the processes,
and II. Simplified procedures with some interaction among the processes.

The hydrological fundamentals of these models are not complete or even may
not be considered and some components are missing (for example a procedure
may simulate only runoff from impervious areas and neglect subsurface flow).
These models do not commonly require calibration and verification with
measured data.

Screening models and computerized procedures often require a determination
of a design storm.  The magnitude of the design storm and the frequency are
usually selected by the designer according to local or state building codes
and flood protection insurance regulations.

Examples of the procedures included in category I include the Rational
Formula for estimating peak runoff rate and the British  Transport and Road
Research Laboratory (TRRL) Hydrograph Method (Watkins (1982), Terstriep and
Stall (1969)). These simple models estimate design flows for sewer systems.
Simple water quality estimation procedures for urban areas rely on estimated
unit loadings  of pollutants.  A unit loading is a simple value or function
expressing pollution generation per unit area and unit time  for each typical
urban land use.  The units are expressed in kg per hectare per year or season.
In some cases, the unit loads for developed urban areas  are  expressed as a
mass of pollutants per unit of curb length.  Then the unit load can be ob-
tained by multiplying the curb loads by curb density (curb length/area).
The magnitudes of unit loadings and simple procedures can be obtained from
a publication  by McElroy et al.  (1976).

A number of Simplified Hydrological Computertized procedures (modeling
category II) have been developed in the past fifteen years.   The most
known hydrological procedures include the Soil Conservation  Service - run-
off hydrograph method (SCS(1975))in the U.S. or the Wallingford Procedure
developed in England and described by Price (1984).  The SCS Method provides
a means of estimating surface runoff volume and hydrograph using a semi empir-
ical relationship between the rainfall and runoff based  on a runoff curve
number that depends on land use and soil  types, vegetation,  and antecedent
moisture conditions.  The runoff can then be routed overland by a simple
triangular unit hydrograph.
                                      K-VI-3

-------
Water  quality  components  in  the  category of screening models usually rely
on  the Universal  Soil Loss Equation  (Wischmeier and Smith (1965)) that
estimates  soil  loss from  pervious surfaces and use the pollutant build-up-
washoff concept for impervious urban areas (Novotny and Chesters (1981)),
The modeling methodologies have  been published by Johnson et al. (1978),
Sonzogni et al. (1980), and  Haith and Tubbs (1980).

In spite of their questionable accuracy and reliability the simple screening
procedures have found wide application in areawide pollution abatement
planning efforts.  One reason explains this popularity:  these procedures
provide a  simple  mechanism and quick answers to pollution problems of large
areas  where more  complicated efforts would fail because of the enormous
amount of  information required.  The screening concepts enable indentifi-
cation of  hydrologically  active  nonpoint pollution generation areas within
a large watershed that should be subjected to futher studies and subsequent
management.

The effect of various abatement measures on the nonpoint pollution loads
using  the  simplified screening procedures can be estimated only very
approximately by  using  simple rule-of-thumb factors and pollution removal
efficiencies.   For example,  it may be assumed that an effective street sweep-
ing program will  reduce pollution loads from a high density residential urban
area by 20 to 30  percent  or  that a properly designed wet retention pond may
reduce particulate pollution loadings by 50 percent.

Hydrologic Simulation Models

The more complex  hydro!ogic  simulation models in the second group represent
a description of  the hydro!ogic rainfall-runoff transformation process with
associated erosion, pollution build-up and washoff, and other quality com-
ponents.  These models estimate the  pollution generation from an area based
on the  accumulation during a dry period and their subsequent transport by
rainfall and convective water movement.

Similarly to the  hydrologic models used for modeling nonpoint pollution
from agriculture, the scope of such  a model can cover a subdivision or a
small  section of  a city.   Application of these models to entire large urban
watersheds is usually tedious requiring a break-down of the watershed and
a step-by-step additive simulation.   Most of the urban models are limited
in their scope and in a single application they are limited to relatively
small watersheds.

Urban Hydrological models have the following basic components (Figure 1):
1.  The surface runoff generation component that describes the transforma-
    tion of rainfall into excess rainfall  and surface runoff. Most of the
    nonpoint pollution in urban areas originates from so called hydrologi-
    cally active  areas, that is areas from which surface runoff is generated.
    The purpose of this component is to locate these areas and to determine
    the magnitude of the  surface runoff.
2.  The soil and  ground water movement component that is so important for
    agricultural models,  is  in most  urban models represented by simple
    relationships and empirical equations or is not included.
3.  Accumulation, removal and washoff of pollutants from impervious surfaces.
    This component balances  the particulate pollutants that accumulate near

                                  K-VI-4

-------
the curbs of paved roads during the dry days and estimates their washoff
into the runoff during storms.  The accumulation mass is a result of
deposition of particulate pollutants from various sources (atmospheric
dry deposition, litter deposition, traffic, salt application) and their
removal from the curb storage by wind, traffic induced turbulence and
street sweeping.
Soil erosion component estimates soil loss from pervious areas.  The
Universal Soil Loss Equation (USLE) is the most common model used to
represent the process.
Pollutant adsorption components that in agricultural models is represented
by complex adsorption-desorption relationships, is in most urban models
replaced by factors that relate the mass of a particular pollutant to
that of total suspended solids.
Runoff and pollution routing component transforms the excess rainfall
and its pollution content into a inlet or stream hydrograph or quality
                            RAIN ( SNOW MELT )
                          DRY AND WET ATM. DEPOSITION


                           111111

SURFACE
STORAGE
COMPONENT

-J
SURFACE
RUNOFF
EROSlOr


'ARTICLES

NFILTRATION
SOIL WATEf
COMPONEN1


GROUNDWA
AQUIFER
1
s
TER
>ERVIOUS A
i
MX
1
1
s
ADSORBED
POLLUTANTS
MOTION


AOSOHPTiON
DISSOLVED
POLLUTANTS

pq

WATER
REAS



URFACE
(UNOFF

INTERFLOW

-------
     histogram.   Common  techniques  for  excess  rainfall -- surface runoff
     routing (convolution)  rely  on  a  synthetic unit hydrograph.  In  the
     absence of  a detailed  sediment redeposition and pollution trapping
     description during  the overland  flow  rate,   delivery and enrichment
     factors are used  to   relate the sediment and pollutant yields in the
     receiving water body to those  at the  source.

 Following  Chow's (1972) definition of  hydrological models, most of the urban
 runoff quantity-quality models  can be  classified as deterministic lumped
 parameter  models.  Deterministic means that for one set of inputs there is
 only one set of outputs.   This  differentiates them from so called stochastic
 hydrologic models whereby  outputs  from the model may vary depending on the
 probability distribution of the inputs and model parameters.

 The  lumped parameter models treat  the watershed or a large portion of it as
 one  unit.   The  various characteristics of the modeled  areas are lumped
 together (for example slopes are represented  by an weighted average slope)
 often with the  use of weighting empirical equation.

 Models have been designed  to run on  an event  or continuous basis.   Discrete
 event modeling  simulates the response of a watershed to a major rainfall or
 snowfall.   Continuous models simulate flows and pollutant loadings over an
 extended period  of time that may range from a season to several  years.  The
 principle  advantage of event modeling over continuous simulation is that it
 requires relatively little  meteorological data and can be operated with a
 shorter computer run time.  A principal disadvantage of event runoff
 modeling is  that it requires specification of the design storm and the
 antecedent dry  period, determining subsequently the amount of accumulated
 pollutants  and  soil moisture conditions.  This a priori  implies  that the
 recurrence interval for the rainfall  and runoff and pollution content are
 the  same.

 The  principal advantage of  continuous modeling is that it provides long time
 series of water  and pollutant loadings that can be statistically analyzed as
 to their frequency.  A principal disadvantage of continuous modeling is that
 it requires  long simulation runs,  this imposes restrictions on the amount
 of alternatives  that can be investigated.  It also requires historical data
 on precipitation often in less  than  hourly intervals, which may  not always
 be available.
As to their complexity the hydrological urban runoff quantity-quality models
can be divided into the following subgroups (Sutherland (1980)):

III.  Simplified models - either event oriented or continuous.   In these
      models some hydrologic components are missing   some other are repre-
      sented by a simple relationship.   There  will  be some interactions
      among the processes but this is usually done in an approximate way.
      Water and pollutant routing may not be included.
IV.   Sophisticated single event simulation models.  Thesemodels provide for
      extensive interactions among the various processes that are important


                                  K-VI-6

-------
      in the simulation of stormwater quantity, including infiltration, over-
      land flow, gutter and pipe flow, and flow routing.  The quality compo-
      nents may be less sophisticated, relying on simple pollutant build-up-
      washoff concepts.  Street sweeping effects can be included only during
      a predetermined relatively short antecedent dry period, and, in general,
      such models may not be suitable for evaluation of typical  long term
      best management practices for controlling urban runoff pollution.
      Their primary use is in design and evaluation of urban drainage systems.
V.    Sophisticated continuous models.  These models provide for extensive
      interaction among the various processes that are important in the sim-
      ulation of both stormwater quantity and quality.  Routing  of flow and
      quality are included.  Modeling of intermittent dry periods enables to
      account for soil moisture, pollutant build-up and removal  and other
      important processes occurring between the rainfalls.

Examples of simplified models in category III include the Corps of Engineers'
(Anon. 1975) Storage-Treatment-Overflow-Runoff Model (STORM), Wisconsin Urban
Runoff Model developed by this author (Novotny (1983))  and a number of pro-
prietory models.  Most of these models simulate runoff either by Modified Ra-
tional Formula (STORM) or by the SCS Runoff Curve Model (Wisconsin Urban
Runoff Model).  Most of these models simulate street drainage and storage-
treatment systems.

The most known models in the category IV (event oriented models) are the
EPA's Stormwater Management Model  - SWMM (Huber et al. (1975)) and Illinois
Urban Drainage Area Simulator-QUAL-ILLUDAS (Terstriep and Stall  (1974)).

SWMM is a widely used stormwater runoff quantity-quality model which is
readily available from the U.S. EPA laboratory in Athens, Georgia.  It has
an excellent documentation and its continuously maintained and updated.
The model can analyze single storm events on multiple catchments.  The model
is primarily used for large urban projects where considerable data are
available for calibration, verification and sensitivity analyses.

The ILLUDAS model is available from the Illinois State Water Survey.  The
model has been widely distributed in the U.S.  and Canada as well as in many
other countries.  ILLUDAS uses a design storm hydrograph and physical  basin
characteristics to predict stormwater runoff hydrographs from both paved and
unpaved areas. Its use for water quality prediction is limited.

The Fortran version of the HSP-F model developed by Hydrocomp, Inc. (Hydrocomp
(1979)) is an example of the level  V models capabable  of simulating hydro-
logic time series of runoff quantity-quality events.  This version of the
model is again available from the U.S. EPA Environmental Laboratory in Athens,
Georgia.   The model had originally evolved from the Stanford Watershed Model.
It is a large model and similar to SWMM, it requires a considerable effort
when applied to a watershed.  The model can be applied to urban  areas in a
fashion similar to a previous Hydrocomp model  called Nonpoint Simulation
Model - NPS (Donigian and Crawford (1976)).
                                   K-VI-7

-------
 RELIABILITY AND  USEFULNESS  OF THE MODELS

 The mathematical models  are only a  rough approximation of the real world.
 All real world processes  aremore-or-less stochastic in nature, meaning that
 it is  not possible  to exactly predict the outcome of a series of events in
 the future.  This is especially true for hydro"!ogic processes such as flow
 and pollution generation.   Ideally, such processes should be modeled using
 statistical techniques and  models.  Yet, almost all urban runoff quantity-
 quality models that have  been developed recently are deterministic.

 A common error made by many planners and users of the results of modeling is
 that they accept simulation results as true and accurate.  In order to avoid
 disappointment and  court  challenges, users should be aware of the limited
 accuracy of the models.

 Figure 2 shows  schematically    reliability of various types of models and
 relative magnitudes of errors.  The most accurate models (± few percent) are
 hydrologic models simulating runoff from small, uniform impervious areas,
 the least reliable  (on order of magnitude of more) are water quality models
 for large watersheds simulating pesticides and similar complex chemicals.
 The errors can be reduced if the model  is properly calibrated and verified
 by actual measured  data.  It should be noted that in modeling pollutant
 transport, hydrologic (runoff) components must be calibrated first, followed
 by erosion components, and  finally pollutant generation-transport components.

 In spite of the errors involved in modeling complex systems such as nonpoint
 pollution from urban areas, the model as a planning tool  cannot be replaced
 by any "rule-of thumb" approach.  The use of models is beneficial and greatly
 enhances the planning process for the following reasons (Novotny and Chesters
 (1981)):

 1.  Models can provide a forecast of the impact (although only approximately)
    of planned actions on water quality and pollution loadings.
 2.  Models can provide an understanding of the processes  involved in pollution
    generation from nonpoint sources.
 3.  The data base necessary to construct and calibrate the model is useful
    for other planning activities.   Many problems will  be solved and questions
    answered or become clearer just by evaluating the data and compiling them
    into an appropriate input format.
4.  Critical  processes and areas of concern can be delinated and detected
    by modeling.
 5.  Models can generate numerous alternatives according to the specifications
    of the users.  Various strategies can be investigated, and the impact of
    remedial  measures can be evaluated.
 6.  Although the absolute accuracy of the outputs from the model are limited
    and sometimes even small, a comparison and ranking of outputs for  various
    alternative remedial  measures are reliable and in most cases more than
    adequate.
7.  Models can estimate and analyze trade-offs between planning  objectives.
    A system providing the lowest pollution may not be optimal  for other ob-
    jectives such as development.   If the environmental  objectives are known
    the alternatives to achieve them can be measured in terms of economic and
    welfare efficiency by considering the willingness of  those involved and
    affected by the nonpoint pollution  abatement measures to pay for the
    consequences and the benefits.

                                  K-VI--8

-------
31

8

-------
 SELECTION OF THE MODEL

 After the models and their capabilities  have  been  introduced in the preceding
 sections, one question managers  and  urban  planners may ask is which model
 to select and for what purpose.   The ideas of an universal model that could
 do everything that was advocated in  1970's by some developers of the mathema-
 tical  models, may not be  as popular  and  feasible today.  The progress and
 availability of small  and relatively cheap micro- and minicomputers have re-
 volutionized the model  applications.  A  highly powerful (500K RAM memory) fast
 microcomputer can now be  purchased for $1500  to $2500.  These computers are
 comparable in their  capacity and speed to  medium size computers of the 1970's.

 What  this means is that a menu of smaller  models may be more useful to a
 practicing urban engineer and manager than one large model.  Thus, one should
 look  at  the model  selection question from  the point of view of what purpose
 and tasks the model  should serve.  The model  itself should be looked upon
 as a  tool  to accomplish the task and objectives.

 As stated in the preceding section,  most of the hydrological  models are
 limited  in their areal  extent.   A typical  size of the watershed best suited
 for detailed hydrologic modeling would be  from a fraction to several km2.
 For example the applications  of  ILLUDAS model should be limited to basin
 areas  of less than 13  km2.   Thus, most of  the applications have been limited
 to small  experimental watersheds and  techniques on how to expand the modeling
 to larger watersheds  are  still in evolution.  Bearing in mind that commonly
 only  a portion  of a  typical  urban or  urbanizing watershed is responsible for
 most  of  nonpoint pollution,  overview  or screening modeling techniques should
 be made  compatible with the  hydrological detailed models.   In a typical  water-
 shed  nonpoint pollution abatement process  the modeling effort should proceed
 in the following  phases:
 (I)    Overview screening  models  should identify the problem areas.
 (II)   Detailed continuous models (level III  or V) could be used for obtaining
       more  accurate nonpoint pollution loadings figures and for screening
       the various management practices and structural  measures and their
       water quality  impact.  Such modeling activities  must consider both the
       source strength  and pollutant  delivery.
 (Ill)  A  detailed  event oriented model (level IV) could be used to finalize
       the  design  of some  technically complex structural  and  nonstructural
       measures  selected and evaluated in Phase II. Such modeling  may include
       design of  storage for multiple overflows, various treatment schemes,
       and  redesign of drainage  systems.
 A  schematic  diagram of  the process is shown on Figure 3.

 Calibration  and verification with field measured data may be  necessary for
 Phase II modeling  if accurate loadings of pollutants are desired.   Calibra-
 tion is  less  important  if  the effect  of various management practices on  the
 loadings  is  studied.

This approach leads to a hierarchical modeling process  whereby  the knowledge
and extensive testing associated with some large and complex  urban runoff-
pollution models can be extrapolated to the practical  load management models
and approaches  used for selecting of hazardous critical  areas within the

                                  K-VI-10

-------
Goals and
Objectives
Preliminary Data
Collection and Surveys
Delineation of Hazardous
and Problem Areas
                                                Screening Model
                                                Level  I or II
Selection of Critical Areas for
Detailed Modeling, Selection of
Alternatives and Management
Practices
Data Collection for
Calibration
Detailed Modeling and
Evaluation of Alternatives
Final Detailed Design
                                                Selection of a Detailed
                                                Hydrologic Model  -
                                                Level  III or V
                                                Model Calibration
                                                and Verification
                                                Event Oriented
                                                Model (Level IV)
                              Figure 3
                 Schematic of the Modeling Process
                              K-VI-ll

-------
watershed and for selection of appropriate management practices.  A detailed
hydrologic model calibrated and verified by small plot field studies can be
used to estimate unit loads for a screening overview model as shown by
Novotny and Bannerman (1980).

This concept has been used by the Wisconsin Department of Natural Resources
to identify the problem areas within so called "priority watersheds" that
is watersheds with a serious nonpoint pollution problem.

The Model Enhanced Unit Load concept (MEUL) enables selection of the priority
watersheds and the problem areas within them.  The units loadings were gen-
erated by a calibrated and verified hydrologic-nonpoint model.    In this
study loadings were generated for 1 km2 (100 ha) area! units located on four
hydrologically uniform soil types with a uniform slope of 6%.  A time series
summarizing several years of meteorological data was statistically analyzed
to provide long term average loadings under different land use conditions.
In a subsequent study that focused on urban areas only, a hydrological level
III model, calibrated and verified by 3 year monitoring program encompasing 12
small urban watersheds, was used to extrapolate the findings into more general
relationships (Novotny (1983, 1984), Novotny et al. (1985)). Figure 4 shows
such a relationship whereby pollutant loadings from hypothetical unit urban
areas (residential  and commercial land use) are related to the percent imper-
viousness of the area soil type, and street sweeping practices.

USE OF THE MODELS FOR SELECTION OF CONTROL OF URBAN RUNOFF

There are a number of measures that can be used to control pollution by urban
runoff.   These methods will be presented in detail in a subsequent paper.
From the modeling standpoint, urban runoff (quantity-quality) control  measures
can be categorized as follows:

(1)  On-site source control measures that include control of atmospheric de-
     position, litter control  programs, leaf and grass pick-up,  street sweep-
     ing.   These measures affect the input and accumulation of pollutants
     on urban impervious areas.
(2)  Control of urban pervious areas, including mostly control  of bare and
     unprotected soils within the watershed, especially those due to con-
     struction.   This is modeled by adjustment of a few hydrological  para-
     meters (depression storage, runoff curve number, infiltration rate) and
     by the cover factor, C, in the Universal  Soil Loss Equation.
(3)  Hydrologic modification of urban watersheds.   Many measures in this
     category of urban nonpoint pollution control  are similar to those
     used for control  of flooding.   But not all flood control practices can
     control nonpoint pollution.  Simple flow retardation, for example, in  a
     dry detention pond or surface ponding on impervious surfaces in areas
     served by storm sewers may have no effect on reduction of pollution
     loads.   Hydrologic modification for modeling purposes can be described
     as  follows:
     (a)  practices that increase permeability and enhance infiltration, such
          as the use of pervious pavements or vegetation infiltration  strips;
     (b)  practices that will  increase hydrologic storage, such  as rooftop
          storage on flat roofs, temporary surface ponding, and  restriction
          of stormwater inlets;

                                  K-VI-12

-------
    500
 O  400
 LJ
 CO
    300  -
  CD
  z  200
  o
     100   .
IMPERVIOUS /
  UNSWEPW
                     20        40        60 .
                         % IMPERVIOUSNESS
      80
100
Figure 4 -  A schematic representation of sources of urban particulate
          nonpoint pollution related to imperviousness, soil type and
          street sweeping.  The loadings are expressed for 1 km2 unit
          watersheds with 6% slope.
                            K-VI-13

-------
      (c)   practices that will  reduce directly connected  impervious  areas,
           that include disconnnecting roof drains,  letting  surface  runoff
           overflow on adjacent pervious  areas, use  of  dry wells, and use
           of pervious pavements.
 (4)   Reduction of delivery of  pollutants.   After  the particulate pollutants
      leave the source areas, they can settle  or be  removed  on  route from the
      sources to the receiving  water  body.   A  delivery  ratio factor  expresses,
      in the models, the ratio  of  the amount of pollutants delivered to a
      receiving water body (pollutant yield) to the  amount of pollutant
      generated at the source.

      Urban areas with a good storm drainage have  a  delivery ratio close to
      one,  meaning that almost  all  pollution,  after  it  leaves the source,
      is eventually delivered to the  receiving water body.

      Residential  areas with a  natural  (swale)  drainage have delivery ratios
      ranging from a few percent to about 50 percent.   As a  consequence, urban
      areas with a natural  drainage produce much less pollution than areas
      with  a good storm sewer system.

 (5)   Runoff detention-retention.   The terms "detention" and "retention" have
      been  used synonymously in most  literature interpretations and  in common
      usage.   However, with recent  advancements of the  detention-retention
      technology,  detention refers  to a stormwater storage facility  that is
      normally dry and is  designed  to temporarily  hold  stormwater during high-
      peak  runoff events.   Such basins have only a very limited efficiency
      in controlling urban  runoff pollution.   A retention facility always
      contains a substantial volume of water to serve other  purposes such as
      recreation or aesthectics.  The modeling  of  the effects of these facili-
      ties  to control  urban runoff  is  significant.   The modeling of  the effects
      of these facilities may vary  in  sophistication.   The simple techniques
      will  adjust the erosion control  practice  factor P of the Universal Soil
      Loss  Equation using  rule-of-thumb factor.   Such factors have  been
      developed by Driscoll  (1983)  and they can be used in combination
      with  the Level  II  or  III  models.

      A number of  more complex  models  of  retention facilities evolved in the
      last  ten years.   A summary of the models  was presented by Prasad (1984).
      One of  the most advanced  retention  facility models is  SEDIMOT  (Wilson
      et al.  (1983)).  Most of  the  Level  IV and V models mentioned herein
      have  a  retention storage  segment.

Example of the Model  Appli cati on

The Level  III  Urban  Runoff Model developed by  the author (Novotny (1983)
and Novotny  et al.  (1985)) was  used  to study  the  effect of  several  low cost
management practices.

In the first phase  of the  research the model was  calibrated and verified by
monitoring data of  eight small  urban  experimental watersheds gather, by the
Wisconsin  Department of Natural Resources  (Bannerman et al.  (1984)).
                                  K-VI-14

-------
For each watershed the monitoring program provided the following information:
     (a)  atmospheric deposition rates (wet and dry),
     (b)  pollutant accumulation in the curb storage,
     (c)  runoff quantity and quality,
     (d)  traffic count and land use characteristics.

The calibration process preceded as shown on Figure 5:

     (a)  The generated street pollutant accumulation was matched against the
          monitored data (Fig. 6).
     (b)  When adequate fit  of street accumulation data was obtained the cali-
          bration then preceded to matching runoff quantity for storm events
          with monitored data.
     (c)  Runoff quality calibration was the last step (Fig. 7).

Generally good to satisfactory fit  of measured and computed data  was  obtained
for most of the pertinent watersheds.

After the calibration precipitation data for 1981 and 1982 was used to generate
seasonal and annual (excluding winter) loadings, the output from the model
provided event runoff volume and quality for five constituents -- suspended
solids, volatile suspended solids, nitrogen, phosphorus, and lead.  These
simulated loads and breakdown of sources of pollution are shown in Table 1.

This type of study provides the manager with information on magnitudes of
nonpoint pollution and on the source strength.  For example, in the Milwaukee
study modeling indicated that atmospheric deposition accounted for less than
5% of the input, traffic became an important source of lead only in high
traffic watershed, and effect of street sweeping was not significant.

The effect of several management practices on water quality was then docu-
mented on a case of a "hypothetical" watershed that was set-up using the
surface, soil, and land use data similar to those established for the study
sites.  The watershed consisted of the following land uses:

     medium density residential zone      100    ha
     low density residential zone           7.5  ha
     open unused land (natural grass)      23.6  ha
     commercial (retail)                   12.7  ha
     transportation (freeway)              14.5  ha
     parks                                 19.50 ha

Figure 8 is a map of the hypothetical watershed.  The soils were in silt
and silty clay loam textural classes.  Slopes ranged from 0 to 6 percent.

The following alternatives were studied:
     Alternative 1 - No action.
     Alternative 2 - Effective street sweeping.
     Alternative 3 - Street sweeping and grass strips along the roads.
     Alternative 4 - Street sweeping and two retention ponds providing 50%
                     reduction of pollutants.
     Alternative 5 - Street sweeping, two retention basins and grass strips
                     along the roads.


                                  K-VI-15

-------
STREET POLLUTION

 ACCUMULATION
                            MONITORED DATA
RUNOFF QUANTITY
RUNOFF QUALITY
Litter Deposition
Rate
Pervious Area - CN
Pollutants Cone.
in Litter
Litter Particulate
Distribution
Impervious Area - DS
Breakdown Factor
    DUSTFALL
TRAFFIC
1

4 *
MEDIUM

i
LITTER
1
I
COARSE


                              BREAKDOWN
              Figure 5 - Calibration Process Schematics

                               K-VI-16

-------
  250
                              MEASURED  COMPUTED
                         FINES     A

                         MEDIUM   X


                         COARSE   O
                      100              140


                  DAY OF THE YEAR
     Figure 6 - Calibration Results of Street Solids
             Accumulation Process
  1000-^
0>
S 100 -d


O
CO

Ol
CO
z>
Q
LU


OC
UJ
CO
CD
O
        10
               -i	1	1—i—i—i  i i
                100                    1000


COMPUTED SUSPENDED SOLIDS, mg/l
       Figure 7 - Calibration Results of Runoff Quality
                       K-VI-17

-------
Figure 8 - The Experimental Watershed
               K-VI-18

-------
                 Table 1
Computed Sources and Outputs of Pollutants
         From an Urban Watershed
         Overall Mass Balance

Street Loads
Initial
Final
Daily Inputs
Dry Atm.
Litter
Street Dirt
Total
Event Inputs
Wet Atm.
Soil
Total
Outputs
Blown Out
Swept Out
In Runoff
Total
Susp. Solids
(ton)

9.225
6.641

8.052
93.614
271.149
372.814

1.496
14.577
16.073

328.965
.000
62.507
391.472
Organics
(kg)

897.443
529.347

797.155
24527.625
59919.688
85244.438

748.006
831.012
1579.018

73871.187
.000
13320.262
87191.437
Nitrogen
(kg)

14.761
8.715

12.883
149.813
362.217
524.913

299.202
346.410
645.613

465.394
.000
711.177
1176.571
Phosphorus
(kg)

21.139
12.645

8.052
257.288
631.968
897.308

3.740
20.118
23.858

801.693
.000
127.966
929.659
Lead
(kg)

13.838
24.615

8.052
140.444
960.308
1108.804

3.740
5.831
9.571

961.323
.000
146.275
1107.597
                 K-VI-19

-------
 Figure 9 shows the unit loads  of solids  for  the  no action alternative (Alt. 1).
 Such figures can aid managers  and designers  in their decision making to iden-
 tify the problem areas  and  sources of pollution.  In a typical mixed land use
watershed identification  of hazardousland is  necessary since not all lands and
land use segments require management.  For example, established low density
residential  areas with  natural drainage  have loadings comparable to undis-
turbed natural  prairies and woodlands.

Figure 10 shows  the cummulative probability plot of suspended solids concentra-
tions.   The  results  indicated that grass strips along the roads (2 meter grass
strips between  the impervious road surface and the drainage system - sewer
or  swale)  provided the  best water quality benefit both in reducing the con-
centrations  by  80% and  runoff volumes by 50% for medium and by 25% for longer
storms,  respectively.

Street sweeping was  not an  effective pollution control measure in this sensi-
tivity analysis,  although it did  reduce the pollutant outputs from highly
impervious segments.

The  suggested optimal best management practices for this hypothetical  case
study would  include  street  sweeping in the commercial  zone,  two meter  grass
strips along the  roads  and  streets, and retention ponds to intercept flows
from open  lands and  residential areas located on high slopes and poor  (silty
clay loam) soils.

In addition, grass clipping  control, leaf pick-up during the fall  season and
litter control ordinances should be implemented.

Conclusions

The urban  runoff  quantity-quality models are indispensable tools in the  eval-
uation of  sources of pollution from urban and urbanizing areas and  the  design
of the abatement measures.   Such models can:

(1)  Estimate loadings  of pollutants from nonpoint sources under different
     hydrological and meteorological  conditions.
(2)  Estimate source strengths of different contributing areas within  urban
     and urbanizing watersheds.
(3)  Evaluate relative  efficiency of various  management practices  and
     structural measures.
 4)  Enable design of nonpoint control  measures.
 5)  Aid in watershed planning.

Both screening and hydrological models should be used,  however, not on the
same scale and with the same detail.   Screening models  are useful  for  identi-
fication of problem areas within larger watersheds.   Hydrological  models are
then subsequently used for detailed modeling  of nonpoint pollution loadings
and evaluation of the impact of abatement.
                                   K-VI-20

-------
UNIT LOAD

kq/ha year
            250.300
            500-IOOO

            XOOO
     Figure 9 - Units Loadings of Solids from the Experimental
              Watershed for Alternative 1 (Ho Action)
                         K-VI-21

-------
 1000
 9oo-

 o
 V)

 o
 ui
 o
 z
 llj
 a.
 CO
 D
    10-
10     20
                                40     60
 i

80
90   95
                      PROBABILITY 4.



Figure 10 - Cumulative Probability Plot of Suspended Solids Concentrations
           1n  the Runoff for 5 Alternatives Measured by the Model

Alt.  1 - No actlpn,
Alt.  2 - Street sweeping only,

Alt.  3 - Street sweeping in coranericial  zones plus grass strips along the
        roads  in residential zones,
Alt.  4 - Street sweeping plus two retention ponds,

Alt.  5 - Same as Alt.  3 plus two retention ponds.
                           K-VI-22

-------
REFERENCES:


American Public Works Association (1981).   Urban Stormwater Management.
Spec. Report No. 49, Chicago, IL.

Anon. (1975).  Urban Storm Water Runoff -  STORM.  The Hydro!ogic
Engineering Center, U.S. Army Corps of Engineers, Davis,  CA.

Bannerman, R., et al. (1984).  Evaluation  of Urban Nonpoint Source Pollution
Management in Milwaukee County, Wisconsin, Vol.  I.  Urban Stormwater  Charac-
teristics, Pollutant Sources and Management by Street Sweeping.  A Report
for U.S. EPA, Region V., Chicago, IL.

Barfield, B.J., Kao, D.T.Y, and Toller, E.W.  (1975).   Analysis  of the Sedi-
ment Filtering Action of Grasses Media.  Res. Pap. No. 90, University of
Kentucky Water Resources Research Institute,  Lexington, KY.

Chow, V.T. (1972).   Hydrologic Modeling.   Journal of the  Boston Society  of
Civil Engineers, 60, pp. 1-27.

Donigian, A.S., and Crawford  N.H. (1976).   Nonpoint Pollution Model from
the Land Surface.  EPA 600/3-76/083. U.S.  EPA, Washington, DC.

Driscoll, E.D.  (1983).  Performance of Detention Basins for Control of Urban
Runoff Quality.  Proc. International Symposium on Urban Hydrology, Hydraulics,
and Sediment Control.  University of Kentucky, Lexington, KY.

Haith, D.S., and Tubbs, L.J.  (1981).  Watershed Loading Functions for Nonpoint
Sources.  Journal Env. Eng. Div., ASCE, 107,  pp. 121-137.

Hall, M.J. (1984).   Urban Hydrology.  Elsevier Applied Science  Publ., Ltd.,
Essex, England.

Hydrocomp, Inc. (1979).  User's Manual for the Hydrologic Simulation  Program/
Fortran.  Mountain View, CA.

Huber, W.C., Henry, J.P., Nix, S.J., Dickinson,  R.E., and Polinon, D.J.  (1984).
Stormwater Management Model User's Manual  - Version III,  EPA-600/S2-82-84-109
a & b, U.S. EPA, Cincinnati, OH.

Johnson, M.J.,  et al. (1978).  Management  Information Base and  Overview  Model-
ing.  Inter. Joint Commission, Windsor, Ontario, Canada.

McElroy, A.D.,  et al. (1976).  Loading Functions for Assessment of Water
Pollution from Nonpoint Sources.  EPA 600/2-76-151. U.S.  EPA, Washington, DC.

Novotny, V. (1983).  Model of Nonpoint Pollution of Eight Small  Watersheds in
Milwaukee County.   A Report by Novotny Associates to Wisconsin DNR,  Milwaukee,
WI.

Novotny, V. (1984).  State-of-the-Art of Hydrological and Water Quality  Models
Used for Simulation of Nonpoint Pollution  from Agriculture.  Proc. International
Conference on Agriculture and Environment  1984.   Venice,  Italy, June  11-15.


                                  K-VI-23

-------
Novotny, V.  (1984).  Efficiency of Low Cost Practices for Controlling Pollu-
tion by Urban  Runoff.  Proc. 3rd International Conference on Urban Storm
Drainage, G'dteborg, Sweden, June 4-8, pp. 1241-1250.

Novotny, V., and Banner-man, R. (1980).  Model Enhanced Unit Loadings of
Pollutants from Nonpoint Sources.  Proc. Hydraulic Transport Modeling
Symp., Publ. No. 4-80, ASAE, St. Joseph» MI.

Novotny, V., and Chesters, G. (1981).  Handbook of Nonpoint Pollution:
Sources and Management.  Van-Nostrand-Reinhold Publ., New Yrok, NY.

Novotny V., Sung, H.M., Bannerman, R., and Baum, K. (1985).  Estimating
Nonpoint Pollution from Small Urban Experimental Watersheds.  Paper to be
published by Journal of Water Pollution Control Federation.

Prasad, N.S. (1984).  Modeling Urban Stormwater Retention Basins.  M.S.C.
Thesis, Department of Civil Engineering, Marquette University, Milwaukee, WI.

Price, R.K. (1984).  Development and Implementation of the Wallingford, Pro-
cedure.  Proc. Third International Conference on Urban Storm Drainage.
GSteborg, Sweden, June 4-8, pp. 445-452.

Sonzogni, W.C. et al. (1980).  WATERSHED:  A Management Technique for Choosing
Among Point and Nonpoint Control Strategies.  Proc. of the Seminar on Water
Quality Management Trade-offs.  U.S. EPA, Chicago, IL.

Sutherland, R.C. (1980).  An Overview of Stormwater Quality Modeling.  Proc.
International Symposium on Urban Storm Runoff, University of Kentucky, Lexing-
ton, KY, July 29-31.

Terstriep, M.L., and Stall, J.B. (1969).  Urban Runoff by Road Research
Laboratory Method.  Journal of the Hydr. Div.-ASCE 95, pp. 1809-1834.

Terstriep, M.L., and Stall, J.B. (1974).  The Illinois Urban Drainage Area
Simulator.  Bulletin 58, Illinois State Water Survey, Urbana, IL.

U.S. Soil Conservation Service (1975).  Urban Hydrology for Small Watersheds.
Technical Release No. 55.

Watkins, L.H. (1962).  The Design of Urban Sewer Systems:  Research into the
Relation Between the Rate of Rainfall and the Rate of Flow in Sewers.   Road
Res. Tech. Paper No. 55 (HMSO, London)

Wilson, B.N., Barfield, B.J. and Moore, I.D. (1983).  Design Manual for the
SEDIMOT II - Hydrology and Sedimentology Model.   University of Kentucky,
Dept. of Agr. Engineering, Lexington, KY.

Wischmeier, W.H., and Smith, D.D.  (1965).  Predicting Rainfall-Erosion Losses
from Cropland East of the Rocky Mountains.   U.S.D.A. Agricultural Handbook
No. 282, Washington, DC.
                                  K-VI-24

-------
              MANAGEMENT AND CONTROL OF POLLUTION BY URBAN RUNOFF
                             Richard Field, Chief
                        Storm and Combined  Sewer Program
                      Office of Research and  Development
                     U.S. Environmental Protection Agency
                               Edison, NJ  08837
Abstract

     The paper reviews progress on urban  stormwater management  and  pollution
control with emphasis on non- and low-structurally intensive  techniques  along
with the total system approach encompassing  control-treatment.

     Many of the United States Environmental Protection Agency's demonstra-
tion-evaluation projects are presented to exemplify:   Land Management
Techniques, i.e., land use planning, best use of natural drainage,  dual  use of
retention and drainage facilities required for  flood  control  designed
simultaneously or retrofitted for pollution  control,  porous pavement,  surface
sanitation, and chemical use control; Collection Systems Control, i.e.,
catchbasin cleaning, flow regulators (including swirl and helical blend
devices), and the new concepts of elimination or reduction of unauthorized
cross-connections in-channeI/conduit storage and/or other forms of  storage for
bleed back to existing treatment plants;  In-Receiving Water Storage;
Treatment, i.e., physical/chemical, disinfection, and a treatment-control
planning and design guidebook; Sludge and Solids Residue from Treatment; and
Integrated Systems, i.e., storage/treatment  dual-use  wet-weather flow/dry-
weather flow facilities, and reuse of stormwater for  non-potable purposes.

     Recommendations for the future in the areas of:   control based on
receiving water impacts, toxics characterization and  their control, sewer
system cross connections, integrated stormwater management, and
institutional/sociological/economic conflicts are also presented.


Keywords:  Storm sewers, Overflows—sewers,  Drainage,  Surface water runoff,
           Runoff, Hydrology, Combined sewers,  Water  pollution, Water  quality,
           Wastewater, Sewage, Contaminants, Waste treatment, Storage  tanks,
           Disinfection, Mathematical models, Remote  control, Hydraulics
                                     K-VII-1

-------
 INTRODUCTION

      The mission of the Storm and Combined Sewer Program (SCSP) of the U.S.
 Environmental Protection Agency (EPA) is to develop methods for controlling
 pollution from urban stormwater discharges and combined sewer overflows
 (CSO).   This paper will emphasize research relevant to urban stormwater.  This
 program has  been involved for 20 years in the development of a diverse
 technology including CSO and stormwater control technology, instrumentation,
 problem assessments, best management practices (BMP) development and
 evaluation,  stormwater management models,  sludge handling and disposal
 methods,  infiltration inflow (I/I) control, erosion control, and many
 others.   Because the time allotted will not allow complete coverage, just the
 more  important products will be highlighted with emphasis on some recent
 developments.

 POLLUTION PROBLEM ASSESSMENT

 Characterization
      The  quality and quantity characterization of urban stormwater discharges
 is necessary for problem assessment,  planning,  and design.   Summaries of
                                                           /I _0\
 characterization from many research studies are available v   '.
      About 70  Ib/ac/yr of  biochemical oxygen demand (BOD) in urban stormwater
 runoff discharges contribute about 45% of  the annual BOD load if secondary
 treatment  is provided  for  the dry-weather  flow (DWF).
      Since 1974,  EPA supported an urban rainfall-runoff-quality  data base
 // 5\
 v ' ' for  two  important  data  requirements:   characterization of  urban storm
 runoff and calibration and verification of  urban stormwater management models.
      From an in-house  project,  preliminary  screening of  urban wet-weather
discharges from  24  samples  from nine  urban  areas  found that approximately one-
half of the  129  EPA's  priority pollutants  (Table  1).   The heavy  metals were
consistently found  in  all  samples.  Polynuclear aromatic hydrocarbons from
petroleum were the most  frequently  detected  organics.


                                   K-VII-2

-------
Table 1.  Distribution of Priority Pollutants  Frequently Detected
       Pollutant
                                                                 **
   Frequency*      ^	
LiquidSediment   Liquid (pg/1)Sediment (yg/kg)
                                                            Range
Phenols

Polynuclear Aromatic
Hydrocarbons (PAH's)
Fluoranthene
Pyrene
Anthracene
1,2-Benzanthracene
Benzo(a)pyrene
3,4-Benzofluoranthene +
  11,12-Benzofluoranthene
Chrysene
Fluorene
Phenanthrene
  5
  4
3
3
3
2
2
2

2
2
2
                       0.3-0.9
0-98
0-76
1,200-4,000
1,200-3,000
  375-7,000
1,100-4,100
  600-25,000
2,800-4,600

1,500-4,100
  210-300
  375-2,000
Esters
Bis(2-ethylhexyl)phthalate
Di-n-butyl phthalate
Butylbenzyl phthalate
Heavy Metals
Arsenic
Beryllium
Cadmium
Chromium
Copper
Lead
Nickel
Silver
Zinc
Antimony
Selenium
Thallium
Mercury

6
6
—

7
7
7
7
7
7
7
7
7
7
7
7
7

3
3
2

3
3
3
3
3
3
3
3
3
3
3
3
3

0-160
0.5-3.4
—

2-15
1-2
2-7
9-36
11-49
30-400
9-40
3-8
90-330
<20
3-10
2-10
0.2-0.56

130-59,000
300-25,000
450-1,200
Sediment (mg/kg)
2.8-4.8
0.1-0.2
0.06-1
8.10-28
7-140
24-370
3-16
0.5-2.1
37-380
3-3.3
0.06-1.4
<0.2
2.1-2.4
 *Reported when found four or more times out of seven  liquid  samples  and  two
  or more out of three sediment samples.
**These are mostly grab samples and therefore cannot be accurately  related  to
  pounds of pollutants discharged.  Also, the mass emission of pollutants will
  vary for each rainfall event.
                                  K-VII-3

-------
      A project in Syracuse, NY using the Ames Test to evaluate urban runoff
 mutagenicity' ' ' found detectable responses for 22% of the samples.  It is
 significant that some mutagenic substances are present with a potential for
 entering the food chain.
      The discharge of pathogens from CSO and storm sewers is a major public
 health concern.  Indicators such as fecal coliform have long been known to be
 present in stormwater discharges in densities sufficient to cause contraven-
 tion of standards.  A study in Baltimore, MD identified actual pathogens and
 enteroviruses in storm sewer discharges' '.  Cross-connections from sanitary
 sewers were strongly implicated as the major cause.  This problem is not
 confined to Baltimore.  The problem is likely to be widespread particularly
 where sewers are older and cross-connections are present.

 Receiving Water Impacts
      Knowledge of the receiving water impacts resulting from urban wet-weather
 discharges is a basis for determining the severity of problems and for
 justifying control.   Problem studies of receiving water impacts are described
 in a proceedings from a national conference^ '  and in a journal paper^  '.
      Under certain conditions,  storm runoff can govern the quality of
 receiving waters regardless of  the level of DWF treatment provided.  Based on
 national annual mass balance determinations, urban wet-weather oxygen demand
 loads  are greater  than the dry-weather  Loads from the same areas^  '  '•
 Hence,  control of  runoff pollution can be a. viable alternative for maintaining
 receiving water  quality standards.   However,  documented case studies of  urban
 runoff  impacts are scarce.
     The SCSP has  had only partial success  in finding urban storm flow
 generated receiving  water impacts  employing the conventional dissolved oxygen
 (DO)  concentration criteria'   ''.   The problem appears to be in the
 application  of conventional dry-weather monitoring techniques  to  the unsteady-
 state  flow regimes  caused by storms.  Further studies should be undertaken to
 resolve  the  anomaly  between actual  runoff  loadings  and  observed receiving
 water  impacts.   The  SCSP has  been  more  successful in  sediment  analysis than in
water column analysis  for DO  depletions.   Direct  evidence  has  been obtained
 (from the  Milwaukee  River Project'   ')  of  how a disturbed benthos depletes DO
 from the  overyling waters.   Studies have  also shown that  storm and CSO

                                   K-VII-4

-------
adversely affect sediment by  toxics  enrichment  and  resultant  biological
upsets^*   »"~'.   Since particulate matter in untreated  storm and CSO  is
larger, heavier, and  in significant  quantities  when compared  to treated
sanitary effluent, more needs to be  known about the fate and  transport of
settleable  and separable materials.  Hydrodynamic solids separation and
sediment transport routines must be  added to receiving water  models to take
care of the neglected or presently omitted  significant particulate and bed
flow fields.
     A simplified continuous  receiving water quality model  has been developed
for preliminary planning and  screening of areawide  wastewater treatment
alternatives in terms of frequency of water quality violations'  •'.  Also,
a general methodology has been developed for evaluating the impact of CSOs on
receiving waters and  for determining the abatement  costs for  achieving various
water quality goals'  '.  In  addition, a methodology for defining criteria for
wet-weather quality standards has been developed'  !""/.   in  recognition  of  an
important gap in the developed methodologies, the  duration of water quality
standards vs. species survival was taken into consideration.

USER'S ASSISTANCE TOOLS

     User's assistance tools include instrumentation, stormwater management
models, manuals of practice (MOP), methodologies,  compendiums, and state-of-
the-art (SOTA) reports.

Instrumentation
     Storm flow measurement is essential for process planning, design,
control, evaluation, and enforcement.  Sampling devices do not provide
representative aliquots.  Conventional flowmeters apply to steady-state flows
and not to the highly varying storm flows.
     Flowmeters have been developed to overcome these adverse storm condi-
tions^  '  '.  A prototype sampler for capturing representative solids in
storm flow has also been developed and a design manual is available^  '.  This
gave manufacturers the incentive to perfect samplers by increasing intake
velocities and in other ways.  SOTAs are available for flow measurement and
                                   K-VII-5

-------
 sampling'z'»z°).  Because storm flow conditions are extremely adverse, the
 manuals and instruments developed are useful for monitoring all types of flow.

 Simulation Models
      The SCSP has fostered the development of models for assessment, planning,
 design, and control of urban stormwater pollution.  The program's urban water
 management analysis approach involves four levels of evaluation ranging from
 simple to complex that can be worked together.
      The various levels of the Storm Water Management Model (SWMM) are the
 most  significant model products in terms of past SCSP resources,  user accep-
 tance, and overall popularity.  SWMM is one of the most widely used urban
 models and its benefits for planning and design have been demonstrated.  It
 has been employed by consulting engineers to design sewers and to analyze
 pollution control alternatives.
      There are significant recent enhancements of SWMM.  Probably the most
 significant is Version IIl'"»J1' which includes:   better documentation and
 improved stability of the dynamic flow routing model,  EXTRAN (Extended
 Transport Block);  documentation of continuous simulation and snowmelt
 capabilities;  new flexible physically-based storage and treatment routines
which  can provide estimates  of treatment (by settling)  in storage basins;  an
addition of the physically-based Green Ampt infiltration model; and flexible
water  quality  routines  and other routines.
      Documentation and user's manuals  are also available for three continuous
stormwater  planning models.   One is  a  detailed model,  the Quantity-Quality
                       (32 33}
 Simulation model  (QQS)V   '    .  The  other two are  macroscopic planning models,
ABMAC<34»35> and  EPAMAC<36'37>.
     Operational  models which have been implemented in Detroit^   ',
Minneapolis'   ',  Seattle'   ',  and  San  Francisco produce control decisions
during storm events.

SOLUTION METHODOLOGY

     A major emphasis  of  the  SCSP  was  solution methodology,  implemented
through development of SOTA reports, MOPs,  and  user manuals.


                                   K-VII-6

-------
     The  SOTA texts, user's  guide  and  the  assessment  of  urban  stormwater
technology are recommended documents^   '   '.   Separate engineering manuals are
available for storm flowrate determination^   »',  porous  pavement
design^5~^8\ cost estimating^9"51 \  storm  sewer  design^52"5^, planning and
design guidance'   ', and for conducting stormwater  studies'  '.   Seminar
proceedings with themes of "modeling,  design,  operation, and costs" have  been
published.
     The  SOTA document on particle size and settling  velocity^   ' offers
significant information for  solids  treatability and their  settlement in
receiving waters,  important  areas  overlooked  in planning and design.  An
excellent film is  being distributed by  the General  Services Administration
(GSA), National Audio Visual Center which  covers  the  EPA CSO Research,
Development, and Demonstration Program, and in particular  full-scale control
technologies'  '.
     A report entitled, "Urban Stormwater Management  and Technology:  Case
Histories,"'  ' presents 12  case histories which  represent the most promising
approaches to CSO  and stormwater control.  The case histories were developed
by evaluating operational facilties that have  significant  information for
future guidance.
     Three illustrative methodologies  for conducting  stormwater  facility
planning have been published'      '.

MANAGEMENT ALTERNATIVES

     The next major SCSP area is management alternatives.  First the choice of
where to attack the problem; at the source by  land management, in the
collection system, or off-line by  storage.  Pollutants can be removed by
treatment and by employing integrated systems  combining  control  and treatment.

Land Management
     Land management includes structural, semi-structural, and non-structural
measures for reducing urban  and construction  site stormwater runoff and
pollutants before  they enter  the downstream drainage  system.  Various concepts
have been fostered by the SCSP including:

                                   K-VII-7

-------
    .  Land use planning;
    .  Natural drainage which will reduce drainage costs and pollution, and
      enhance aesthetics,  groundwater supplies, and flood protection; and
    •  Retention and drainage facilities, and other management techniques
      required for flood and erosion control which can be simultaneously
      designed for pollution control.
      Retention on-site or upstream can provide for the multiple benefits of
 aesthetics,  recreation, recharge,  irrigation,  or other uses.
 Porous  Pavement
      Porous pavements provide storage, enhancing soil infiltration that can be
 used  to reduce  runoff and CSO.   Porous asphalt-concrete pavements can be
 underlain by a  gravel base course with whatever storage capacity is desired
 (Figure 1).
      Results from a study in Rochester, New York,  indicated that peak runoff
 rates were  reduced as much as 83%'  '  where porous pavement was used.  The
 structural  integrity of  the porous pavement was not impaired by heavy load
 vehicles.   Clogging did  result  from sediment from  adjacent  land areas during
 construction; however, it was relieved by cleaning (flushing).   The construe-
 tion  cost of a  porous pavement  parking lot is  about equal to that of  a
 conventional lot  with stormwater inlets and subsurface  piping.
      A  project  in Austin,  Texas developed design criteria for porous
 pavements^   ' and compared porous asphalt pavement to six other conventional
 and experimental  pavements'   »'.

 Surface  Sanitation
      Maintaining  and cleaning urban areas can  have a significant impact on the
 quantity of pollutants washed off  by stormwater.   Tests  under real work
 conditions  in San Jose,  CA showed that street  cleaning  can  remove up  to 50% of
 the total solids  and  heavy metal  yields  in urban stormwater  with once or  twice
 a day cleaning'   '.   Typical  street cleaning programs of once or twice a  month
proved ineffective.   Organics and  nutrients  could  not be effectively
 controlled even with  intensive  cleaning.
      In Bellevue, WA, street  cleaning  proved ineffective; however,  a  modified
regenerative air  Tymco*  street  cleaner showed  promise'   '   '.   Street cleaning

                                   K-VII-8

-------
                                   POROUS ASPHALT COURSE
                                   FILTER COURSE
                                   RESERVOIR COURSE
                                   VOLUME DESIGNED FOR RUNOFF
                                   DETENTION AND FROST PENETRATION-
                                   EXISTING SOIL
                                   MINIMAL COMPACTION TO RETAIN
                                €  POROSITY AND PERMEABILITY
Figure 1.  Porous Asphalt Paving Typical  Section
                       K-VII-9

-------
 is no panacea for stormwater pollution control (and is site specific dependent
 upon climatic conditions),  but if integrated with other methods, could reduce
 city-wide costs.

 Chemical Use Control
      Until assessment in 1971^  ', there had been limited research on highway
                                                  /£Q \
 deicing effects.  A search  to define alternatives*-  ' confirmed the need for
 an economic impact analysis and for identifying a hydrophobic substance to
 reduce  ice adhesion.  The economic analysis has been conducted^  ' and
 hydrophobic substances have been identified and investigated'  »  '.  Even
 though  their costs appear greater than salt, when considering an estimated $3
 billion annual damage to the environment, the costs are acceptable.
      The 1973 assessment'  ' identified the problems of sloppy salt storage
 practices and over application, resulting in manuals of practice'   *' for
 improvement in these areas.   These manuals were recognized as highly
 significant.  The Federal Highway Administration reprinted them, and over
 8,000 copies have been distributed.
      SCSP work has prompted several states to enact legislation controlling
 salt application and storage.   Also the SCSP's expert testimony helped
 Wisconsin to change its laws to contain controls for roadway deicing.   And
 lastly,  Program work has greatly assisted the Federal Highway Administration
 and  the Army's Cold Climate  Research Laboratory who are utilizing  SCSP results
 and doing further work with  hydrophobic substances.

 Collection System Controls
     The next overall SCSP  category,  collection system controls, pertains to
 management  alternatives for  stormwater interception and transport.   These
 include:   improved maintenance and design of catchbasins,  sewers,  in-pipe and
 in-channel  storage,  elimination of sanitary and industrial cross connections,
 and remote flow monitoring and control.

 Catchbasins
     In  a  project  conducted  in Boston,  MA,  catchbasins  were shown  to be
potentially  quite  effective  for solids  reduction  (60-97%)^75>76\   Removals  of
associated  pollutants  such as  chemical oxygen demand (COD)  and BOD, were  also

                                    K-VII-10

-------
significant  (10-56% and  54-88%,  respectively).   To  maintain the  effectiveness
of catchbasins for pollutant removal  requires cleaning  probably  twice  a  year
depending upon conditions.

Regulators and Hydrobrakes
     At present, there is a strong  need  to  develop  and  have a reserve  of
control hardware for urban runoff control and to effectively reduce  the
associated high cost implications for conventional  storage  tanks,  etc.  The
dual functioning swirl flow regulator/solids concentrator has shown
outstanding  potential for simultaneous quality  and  quantity control^   '.  A
helical type regulator/separator has  also been  developed based on  principles
similar to the swirl.  It is felt that the  swirl/helical type regulators,
previously applied only  to CSO,  can also be installed on separate  storm  drains
before discharge and the resultant  concentrate  flow can be  stored  in
relatively small tanks,  since concentrate flow  is only  a few percent of  the
total flow.  Stored concentrate  can later be directed to the sanitary  sewer
for subsequent treatment during  low-flow or dry-weather periods, or  if.
capacity is  available in the sanitary interceptor treatment system,  the
concentrate may be diverted to it without storage'   '.
     These methods of stormwater control may be more economical  than building
huge holding reservoirs  for untreated runoff, and offer a feasible approach to
the treatment of separately sewered urban stormwater.
     A completed demonstration project in Boston, MA evaluated the swirl and
helical bend for separate stormwater  treatment.  A  complete swirl/helical
design textbook has been published'   '.
     Hydrobrakes™ have been demonstrated by the SCSP in Rochester, NY.'*'3)  and
Cleveland, OH'  >  '.  Hydrobrakes have no moving parts, require no  external
energy source to operate and are not  susceptible to blockages.   They can be
used as upstream off-line attenuators, and in-line  flow back-up  devices.
Conceptually, they allow flow release without orifice movement or  closure  at a
constant predetermined rate regardless of upstream  head.

New Concepts for Stormwater Control
     Research efforts have shown that  sanitary  and  industrial contamination of
separate storm sewers is a nationwide problem.
                                  K-VII-11

-------
      The SCSP has fostered a new concept: in response to this problem which
 includes simple methods of checking for cross-connections.  If found, they can
 be eliminated or the storm sewer can be dealt with as though it were &
 combined sewer whichever is more feasible.  Other new concepts are using storm
 sewers and storm drainage systems for in-channel storage applying CSO in-sewer
 storage and routing technology and other storage facilities for ties into the
 existing sewage treatment system.  Thus making better use of existing
 facilities for overall pollution control.

 Storage
      Because of the high volume and variability associated with stormwater,
 storage is considered a necessary control  alternative.   It is  the SCSP's best
 documented abatement measure.   Project results and theory indicate storage
 must  be considered  at all times in system  planning,  because it allows for
 maximum use of existing dry-weather and other treatment plant  facilities and
 results in the lowest cost in  terms of pollutant removal.
      Storage facilities may have auxiliary functions, such as  sedimentation
 treatment,  flood  protection, sewer relief,  flow transmission,  and dry-weather
 flow  equalization.
      Storage concepts investigated include the conventional concrete holding
 tanks  and  earthen basins,  and  the minimum  land requirement concepts  of:
 tunnels, underground and underwater containers, underground "silos," natural
 and mined  under and  above  ground formations,  and the  use of abandoned
 facilities and existing sewer  lines^82'83^.
     The in-receiving water  flow balance method (Figure  2),  is  a  recently
 developed  storage alternative'   '.   In-receiving water  storage  facilities
 contain stormwater between floating plastic curtains.  After cessation of the
 overflow,  pumps start automatically and the surrounding waterbody will enter
 the compartments  and  push  the  stormwater back  towards the  first compartment
 where  it is  pumped to the  plant.   Thus,,  the waterbody is used  as  a flow
 balance medium.
     The storage  method is low cost due to the employment  of low  cost
materials  (plastic and  wood),  the  time  required to install  the  unit  (several
days to months vs. months  to a year for more  structurally  intensive
                                   K-VII-12

-------
Figure 2.  In-Receiving Water Flow — Balance Method
                            K-VII-13

-------
 alternatives), and the absence of land requirements.  Studies show that costs
 could be about 5 to 15% of conventional concrete tank costs.
      The facility which was tested at three locations in Sweden performed very
 satisfactorily, and was able to take ice and wind loads without adverse
 impact.  It is desirable to demonstrate a. facility in a harsh urban estuarlne
 or marine site, such as the SCSP project will be doing shortly in Fresh Creek
 Basin in New York City, NY.
      A storage/sedimentation planning/design manual'  ' is being finalized
 which includes the following:
      Major Issues
         Applicability,
         Integration into existing treatment systems,
         Costs,
         Effectiveness as treatment/control device.
      Secondary Issues
         Citing impacts,
         O&M requirements,
         Sludge production and  handling.

 Treatment
      Due to adverse and intense flow conditions  and unpredictable shock
 loading effects,  it has been difficult to  adapt  existing treatment methods  to
 storm-generated overflows,  especially the  microorganism dependent biological
 processes.   Physical/chemical  treatment  techniques have  shown more promise
 than  biological processes  in overcoming  storm shock loading effects.   To
 reduce  capital investments,  projects  have  been directed  toward high-rate
 operations  approaching maximum loading.
      Storm-flow treatment methods  demonstrated by  the  Program include
 physical, physical-chemical, wetlands, biological,  and disinfection'   '.
 These processes,  or combinations of  these  processes, can be adjuncts to the
 existing sanitary plant or  serve as  remote satellite facilities  at the
outfall.

Physical/Chemical Treatment
     Physical  processes/systems with  or without  chemicals,  such  as:  fine mesh
screening, swirl  degritting, fine mesh screening/high-rate  filtration  (HRP),

                                   K-VII-14

-------
sedimentation, and  fine mesh  screening/dissolved  air  flotation (DAF),  have
been successfully demonstrated.  Physical  processes have  shown importance  for
storm-flow treatment because  they  are  adaptable to automated operation,  rapid
startup and shutdown, high-rate operation,  and resistance to shock  loads.   The
high-rate processes/systems (DAF,  fine mesh screening,  and HRF) are felt to be
ready for municipal installation.
     The swirl has  also been  developed for grit removal.   The small size,  high
efficiency and absence of moving parts offer economical and operational
advantages over conventional  degritting  facilities.   A  full-scale demonstra-
tion of a (16 ft diameter/11  mgd design  flowrate) swirl degritter has  been
completed in Tamworth, Australia'   *'•   Removal efficiencies confirmed
laboratory results.  Compared with a conventional grit  chamber,  construction
costs are halved, and operation and maintenance costs are considerably lower.

Disinfection
     Because disinfectant and contact  demands are great for storm flows'   ',
research has centered on high-rate applications by static and mechanical
mixing, higher disinfectant concentrations'     ', and  more rapid oxidants,
i.e., chlorine dioxide^91'9^, ozone^90\  and ultraviolet (UV)  light;  and  on-
site generation'  »»'.  Demonstrations  in Rochester,  NY'   '; Syracuse,
      . Eagt Chicag0) iN; and Philadelphia, PA^93\ indicate that adequate
reductions of fecal coliform can be obtained with contact  times of  2 minutes
or less by induced mixing and dosing with chlorine and/or  chlorine  dioxide.  A
pilot scale UV demonstration with a contact time of  less than  10 seconds was
conducted in New York City, NY.
     The hypochlorite batching facility is still being used  in New  Orleans, LA
to protect swimming beaches in Lake Ponchartrain^  ' .  The SCSP supported  the
development of a brine hypochlorite generator now being used in industry'  ' .

Treatment/Control Design Guidebook
     A compilation of the SCSP's best research efforts in  CSO  treatment /con-
                                              (Qf.\
trol over the past 18 years has been published''  '.
Treatment Process Performance
     Treatment process costs and performance in terms of design influx rate
(gpm/ft2) and BOD5 and SS removal efficiency is provided in Table 2.  The

                                   K-VII-15

-------
 high-rate performance of the swirl, and the microstrainer, screening/HRF and
 screening/DAF systems, is apparent when compared to sedimentation.


 Table 2.   Wet Weather Treatment Plant Performance Data
   Device
Control Alternatives
  Design
Loading Rate
 (gpm/ft2)
                                                         Removal Efficiency (%)
BO DC
SS
Primary





Secondary


Swirl concentrator
Microstrainer
High-rate filtration
Dissolved air flotation
Sedimentation
Representative Performance
Contact stabilization
Physical-chemical
Representative Performance
60
20
24
2.5
0.5




25-60
40-60
60-80
50-60
25-40
40
75-88
85-95
85
50
70
90
80
55
60
90
95
95
 Sludge-Solids
     Another  SCSP  area  is  the sludge and solids associated with storm flow
 treatment.   Sludge handling  and  disposal must  be considered an integral part
 of  treatment because  it significantly affects  the efficiency and cost of the
 total waste treatment systems.   A study  quantified solids  residue and sludge
 that would result  from  the treatment of  separate urban stormwater on a
 nationwide basis(98»99).

 INTEGRATED SYSTEMS

     The most promising and  common approach  to urban storm flow management
 involves the integration of  control and  treatment.   Integrated systems are
 divided into storage/treatment,  dual-use wet-weather flow/dry-weather flow
 facilities, and control/treatment/reuse.

 Storage/Treatment
     When there is storage,  there is treatment by settling,  pump-back/bleed-
back to municipal works, and  sometimes disinfection.   In any case,  the break-

                                   K-VII-16

-------
even economics of  supplying  storage must  be  evaluated when  treatment  is
considered.  The SCSP has demonstrated all of  these  storage/treatment concepts
at full scale.

Dual-Use Wet Weather Flow/Dry Weather Flow Facilities
     The concept of dual use is — maximum  utilization of wet-weather facilities
during non-storm periods and maximum utilization of  dry-weather facilities
during storm flows.  The SCSP has demonstrated the full-scale  dual-use of
high-rate trickling f ilters^100), contact stabilization(101~103),
and equalization basins^     .  Various municipalities are employing  dual-use
microscreening.

Control/Treatment/Reuse
     "Control/Treatment/Reuse" is a  "catch-all" for all  integrated systems.  A
prime consideration should be the various nonstructural  and land-management
techniques.  In Mt. Clemens, MI, a series of  three "lakelets" have been
incorporated into a CSO treatment /park development'   '.  Treatment  is being
provided so that these lakes are aesthetically pleasing  and allow for
recreation and reuse for irrigation.

RECOMMENDATIONS FOR THE FUTURE

Receiving Water Impacts
     Ties between receiving water quality and storm flow discharges  must be
clearly established and delineated.  Quantification of the impairment of
beneficial uses and water quality by such discharges is  a major goal.  Project
results indicate the potential for significant impact of wet-weather flows on
receiving waters.  Control of runoff pollution can be a  viable alternative for
maintaining receiving water quality standards.  However, the problems found
seem to be site-specific in nature.  Therefore, site-specific surveys are
required that must consider the effects of larger materials and floatables
near the outfalls.  Based on results from these surveys, control may be
warranted.
                                   K-VII-17

-------
 Toxics Characterization and Control/Treatment
      Results from a limited in-house effort and the Nationwide Urban Runoff
 Program (NURF) indicate that urban stormwater runoff and CSO contains
 significant quantities of priority pollutants.  Additional investigation of
 the significance of concentrations and quantities of toxic pollutants with
 regard to their health effects oi: potential health effects and ecosystem
 effects is required.  A need exists to evaluate the removal capacity of
 alternative treatment technologies for these toxics and to compare their
 effectiveness with estimated removal needs to meet water quality goals.

 Sewer System Cross Connections
      Investigations have shown that sanitary and industrial contamination of
 separate storm sewers is an extensive nationwide problem.   In practice, a
 significant number of separate stormwater drainage systems actually function
 as  combined sewer systems.   Therefore,  a nationwide effort on both federal and
 local levels, to alleviate  the pollution impacts from discharges of these
 systems  is required.  It is better to classify such cross-connected drainage
 systems  as combined systems for pollution control priorities.

 Integrated Stormwater Management
      The most effective solution methodology for wet-weather pollution
 problems must consider:   (1)  control of wet-weather pollution impacts  in lieu
 of  continually and blindly  upgrading existing municipal plants,  (2) structural
 vs. non-structural techniques,  (3)  integrating dry- and wet-weather flow
 systems  to make maximum use of  the existing drainage during wet  conditions and
 maximum  use  of  wet-weather  control/treatment facilities during dry-weather,
 and (4)  the  segment  or blend  on the percent pollutant  control vs.  cost curve
 in which  incremental costs  exceed  incremental  control  efficiencies  subject  to
 load  discharge  or receiving water  requirements.
      Flood and  erosion control  technology  must  be  integrated  with  pollution
 control,  so  that  the retention  and  drainage facilities  required  for flood and
erosion control can  be simultaneously designed  or  retrofitted for  pollution
control.   If  land management  and non-structural techniques  are maximized and
integrated costs  for  the extraction  of  pollutants  from  storm  flows  in  the
potentially more  costly  downstream  plants  will  be  reduced.

                                   K-VII-18

-------
 Institutional and  Socio/Economic Conflicts
     Some of the most promising opportunities  for  cost-effective  environmental
control are multipurpose in nature.  However,  there are  institutional  problems
that hinder their  implementation.  First, the  autonomous Federal  and local
agencies and professions involved in flood and erosion control, pollution
control, and land  management and environmental planning  must  be integrated  at
both the planning  and operation levels.  Multi-agency grant coverage must be
adequate to stimulate such an approach.  For example, the EPA would have  to
join with the Corps of Engineers, Soil Conservation Service,  Department of
Transportation, and perhaps other Federal agencies as well as departments of
pollution control, sanitation, planning, and flood control at the  local level.
     Another problem is that construction grant  (and other) incentives are
geared towards structurally intensive projects which may counter  research
findings in the area of optimal solutions.  Optimized wet-weather  pollution
involves a city-wide approach including the integration  of structural  as  well
as low-structural  controls.  The low-structural  measures are  more  labor
intensive.  Construction grant funding does not  presently address  this expense
and accordingly, municipalities are discouraged  from using them.

CONCLUSIONS

     In general, on a mass basis, toxic compounds, oxygen demanding, and
suspended and visual matter in urban stormwater  are significant.   Ignoring  the
problem because it appears to be too costly or too difficult  to solve, will
not make the problem go away.  The integrated  approach to wet-weather
pollution control  is the only way which is going to be feasible,  economical
and, therefore, acceptable.  Potentially tremendous savings can be derived
from wet-weather pollution control research fostering integrated  solutions.
As you can see, the SCSP has investigated a problem, proven its significance,
and developed a gamut of design and control techniques.   Better advantage
needs to be taken of proven technology.
                                   K-VII-19

-------
                                   REFERENCES

  1.  Lager,  J.  A.,  and W.  G.  Smith,  Metcalf & Eddy Engineers, Inc., Palo Alto,
     CA -  "Urban Stormwater Management and Technology:  An Assessment," USEPA
     Report  No.  EPA-670/2-74-040,  NTIS No. PB 240 687.

  2.  Lager,  J.  A.,  et  al.,  Metcalf & Eddy Engineers, Inc., Palo Alto, CA -
     "Urban  Stormwater Management  and Technology:  Update and User's Guide,"
     USEPA Report EPA-600/8-77-014,  NTIS No. PB 275 654.

  3.  Sullivan,  R. H.,  et al.,  American Public Works Assoc.,  Chicago, IL -
     "Nationwide Evaluation of Combined Sewer Overflows and Urban Stormwater
     Discharges,  Vol.  Ill  - Characterization of Discharges," USEPA Report No.
     EPA-600/2-77-064c, NTIS  No. PB  272 107.

  4.  Huber,  W.  C.,  et  al.,  University of Florida, Gainesville,  FL - "Urban
     Rainfall-Runoff-Quality  Data  Base," USEPA Report No. EPA-600/2-81-238,
     NTIS  No. PB 82-221 094.

  5.  Huber,  W.  C.,  et  al.,  University of Florida, Gainesville,  FL - "Urban
     Rainfall-Runoff-Quality  Data  Base," USEPA Project Summary  No.  EPA-600/S2-
     81-238.

  6.  Spiegel, S.  J., et al.,  O'Brien & Gere Engineers, Inc., Syracuse,  NY, and
     Ott,  R., Department of Drainage and Sanitation,  Onondaga County, North
     Syracuse,  NY - "Evaluation of Urban Runoff and Combined Sewer Overflow
     Mutagenicity," USEPA  Report No. EPA-600/2-84-116,  NTIS  No.  PB 84-211168.

  7.  Spiegel, S.  J., et al.,  O'Brien & Gere Engineers,  Inc., Syracuse,  NY, and
     Ott,  R., Department of Drainage and Sanitation,  Onondaga County,  North
     Syracuse,  NY - "Evaluation of Urban Runoff and Combined Sewer Overflow
     Mutagenicity," USEPA  Project  Summary No.  EPA-600/S2-84-116.

  8.  Olivieri, V. P.,  et al.,  The  Johns Hopkins University,  Baltimore,  MD -
     "Microorganisms in Urban  Stormwater," USEPA Report No.  EPA-600/2-77-087,
     NTIS  No. PB  272 245.

  9.  Proceedings  of National  Conference, Orlando,  FL,  November  26-28, 1979,
     "Urban  Stormwater and  Combined  Sewer Overflow Impact on Receiving  Water
     Bodies," USEPA Report  No.  EPA-600/9-80-056,  NTIS No. PB 81-155 426.

10.  Field, R.,  and Turkeltaub, R.,  USEPA,  Edison,  NJ  - "Urban Runoff
     Receiving Water Impacts:   Program Overview,"  Journal of the  Environmental
     Engineering  Division,  ASCE, Vol.  107,  No.  EE1,  Feb.,  1981, pp.  83-100.

11.  Sullivan, R. H.,  et al.,  American Public  Works Assoc.,  Chicago,  IL.  -
     "Nationwide  Evaluation of  Combined Sewer  Overflows  and  Urban Stormwater
     Discharges, Volume I - Executive  Summary,"  USEPA Report No.  EPA-600/2-77-
     064a,  NTIS No.  PB 273  133.


                                   K-VII-20

-------
12.  Heaney, J. F., et al., University of Florida, Gainesville,  FL -
     "Nationwide Evaluation of Combined  Sewer Overflows and Urban  Stormwater
     Discharges, Volume  II - Cost Assessment and  Impacts," USEPA Report  No.
     EPA-600/2-77-064b, NTIS No. PB 266  005.

13.  Reefer, N., et al., The Sutron Corporation,  Arlington, VA - "Dissolved
     Oxygen Impact from Urban Storm Runoff," USEPA Report No. EPA-600/2-79-
     156, No NTIS number.

14.  Melnholz, T. L., et al., (Rexnord)  Metropolitan  Sewage District  of  County
     of Milwaukee, WI -  "Verification of the Water Quality Impacts of  Combined
     Sewer Overflows," USEPA Report No.  EPA-600/2-79-155.

15.  Pitt, R., Private Consultant, Blue  Mounds, WI; Bozeman, M., Woodward-
     Clyde Consultants,  San Francisco, CA - "Sources  of Urban Runoff  Pollution
     and Its Effects on an Urban Creek," USEPA Report No. EPA-600/2-82-090,
     NTIS No. PB 83-111 021.

16.  Pitt, R., Private Consultant, Blue  Mounds, WI; Bozeman, M., Woodward-
     Clyde Consultants, San Francisco, CA - "Sources  of Urban Runoff  Pollution
     and Its Effects on an Urban Creek," USEPA Project  Summary No. EPA-600/S2-
     82-090.

17.  Tomlinson, R. D., et al., Municipality of Metropolitan Seattle,  Seattle,
     WA - "Fate and Effects of Partlculates Discharges by Combined Sewers and
     Storm Drains," USEPA Report No. EPA-600/2-80-111, NTIS No.  PB 81-118 390.

18.  Medina, M., Duke University, Durham, NC - "Level III:  Receiving  Water
     Quality Modeling for Urban Stormwater Management," USEPA Report  No. EPA-
     600/2-79-100, NTIS No. PB 80-134 406.

19.  Medina, M. A., et al., Duke University, Durham,  NC - "River Quality Model
     for Urban Stormwater Impacts," USEPA Journal No. EPA-600/J-81-234,  NTIS
     No. PB 81-221 087.

20.  Moffa, P. E., et al., Stearns & Wheler, Civil and  Sanitary  Engineers,
     Cazenovia, NY "Methodology for Evaluating the Impact and Abatement  of
     Combined Sewer Overflows:  A Case Study of Onondaga Lake, New York,"
     USEPA Report No. EPA-600/8-80-048,  NTIS No.  PB 81-141 913.

21.  Mancini, J. L., Mancini & DiToro Assoc., Valley, NB - "Development  of
     Methods to Define Water Quality Effects of Urban Runoff," USEPA  Project
     Summary No. EPA-600/S2-83-125.

22.  Mancini, J. L., Mancini & DiToro Assoc., Valley, NB - "Development  of
     Methods to Define Water Quality Effects of Urban Runoff," USEPA  Report
     No. EPA-600/2-83-125, NTIS No. PB 84-122 928.

23.  Mancini, J. L., Mancini & DiToro Assoc., Valley, NB - "A Method  for
     Calculating Effects on Aquatic Organisms, of Time Varying Concentra-
     tions," Water Research, Vol. 17, No. 10, pp. 1355-1362, 1983.
                                   K-VII-21

-------
 24.   Foreman,  K.  M., Grumman Aerospace Corp., Bethpage, NY - "Field Testing of
      Prototype Accoustic Flowmeter," USEPA Report No. EPA-600/2-79-084, NTIS
      No.  PB 80-121 544.

 25.   Anderson, R. J., and Bell, S. S., City of Milwaukee, WI - "Wastewater
      Flow Measurement in Sewers Using Ultrasound," USEPA Report No. EPA-600/2-
      76-243, NTIS No. PB 262 902.

 26.   Shelley,  P.  E., EG&G Washington Analytical Services Center, Inc.,
      Rockville, MD - "Design and Testing of a Prototype Automatic Sewer
      Sampling  System," USEPA Report No. EPA-600/2-76-006, NTIS No. PB 252 613.

 27.   Shelley,  P.  E., and Kirkpatrick,  G. A.,  EG&G Washington Analytical
      Services  Center, Inc.,  Rockville, MD - "Sewer Flow Measurement - A State-
      of-the Art Assessment," USEPA Report No. EPA-600/2-75-027, NTIS No.  PB
      250  371.

 28.   Shelley,  P.  E., and Kirkpatrick,  G. A.,  EG&G Washington Analytical
      Services  Center, Inc.,  Rockville, MD - "An Assessment of Automatic Sewer
      Flow Samplers - 1975,"  USEPA Report No.  EPA-600/2-75-065,  NTIS No. PB 250
      987.

 29.   Huber, W.  C., et al., University  of Florida, Gainesville,  FL - "Storm
      Water  Management Model  User's Manual,  Version III," USEPA Report Nos.
      EPA-600/2-84-109a,  NTIS No.  PB 84-198-423 and EPA-600/2-84-109b, NTIS No.
      PR 84-198-431 (EPA-600/2-84-109b  is the  "Addendum EXTRAN").

 30.   Huber, W.  C., et al., University  of Florida, Gainesville,  FL - "Storm
      Water  Management Model  User's Manual,  Version III," USEPA Project Summary
      No.  EPA-600/S2-84-109a  & b.

 31.   Roesner,  L.  A.,  et  al.,  Camp  Dresser & McKee Inc.,  Annadale,  VA - "Storm
      Water  Management Model  User's Manual Version III,  Addendum I - EXTRAN,"
      USEPA  Report No.  Pending (Interim availability:   University of Florida).

 32.   Geiger, W. F.,  and  Dorsch, H.  R.,  Dorsch Consult  Ltd.,  Toronto,  Ontario,
      Canada -  "Quantity-Quality Simulation (QQS):   A  Detailed Continuous
      Planning Model  for  Urban Runoff Control  - Volume  1  Model Description,
      Testing,  and Applications," USEPA Report No.  EPA-600/2-80-011,  NTIS  No.
      PB 80-190 507.

 33.   Geiger, W. F.,  and  Dorsch, H.  R.,  Dorsch Consult  Ltd.,  Toronto,  Ontario,
      Canada -  "Quantity-Quality  Simulation  (QQS):   A  Detailed Continuous
      Planning Model  for  Urban Runoff Control  - Volume  2  User's  Manual," USEPA
      Report No. EPA-600/2-80-016,  NTIS No.  PB 80-221 872.

34.   Litwin, Y. J.,  et al.,  Ramlit  Assoc.,  Berkeley, CA  - "Areawide  Stormwater
     Pollution Analysis with  the Macroscopic  Planning  (ABMAC) Model,"  USEPA
     Project Summary  No.  EPA-600/S2-81-223.
                                 K-VII-22

-------
35.  Litwin, Y. J., et al., Ramlit Assoc.,  Berkeley,  CA -  "Areawide  Stormwater
     Pollution Analysis with the Macroscopic  Planning (ABMAC)  Model,"  USEPA
     Report No. EPA-600/2-81-223, NTIS No.  PB 82-107  947.

36.  Smith, W. G., and Strickfaden, M.  E.,  Metcalf  &  Eddy  Engineers,  Inc.,
     Palo Alto, CA -  "EPA Macroscopic  Planning Model  (EPAMAC)  for  Stormwater
     and Combined Sewer Overflow Control:   Application Guide and User's
     Manual," USEPA Report No.  EPA-600/2-83-086,  NTIS No.  PB 83-259  689.

37.  Smith, W. G., and Strickfaden, M.  E.,  Metcalf  &  Eddy  Engineers,  Inc.,
     Palo Alto, CA -  "EPA Macroscopic  Planning Model  (EPAMAC)  for  Stormwater
     and Combined Sewer Overflow Control:   Application Guide and User's
     Manual," USEPA Project Summary No.  EPA-600/S2-83-086.

38.  Watt, T. R., et al., Detroit Metro  Water Department,  Detroit, MI  -
     "Sewerage System Monitoring and Remote Control," USEPA Report No. EPA-
     670/2-75-020, NTIS No. PB  242 107.

39.  Metropolitan Sewer Board,  St. Paul, MN - "Dispatching Systems for Control
     of Combined Sewer Losses," USEPA  Report  No.  11020FAQ03/71, NTIS No.  PB
     203 678.

40.  Leiser, C. P., Municipality of Metropolitan  Seattle,  Seattle, WA  -
     "Computer Management of a  Combined  Sewer System,"  USEPA Report No. EPA-
     670/2-74-022, NTIS No. PB  235 717.

41.  Lager, J. A., et al., Metcalf & Eddy,  Inc.,  Palo Alto, CA - "Urban
     Stormwater Management and  Technology:  Update  and User's  Guide,"  USEPA
     Report No. EPA-600/8-77-014, NTIS No.  PB 275 654.

42.  Lager, J. A., and Smith, W. G., Metcalf  & Eddy Engineers,  Inc., Palo
     Alto, CA - "Urban Stormwater Management  and  Technology:   An Assessment,"
     USEPA Report No. EPA-670/2-74-040,  NTIS  No.  PB 240 687.

43.  Brater, E. P., and Sherrill, J. D., University of  Michigan, Ann Arbor,  MI
     - "Rainfall-Runoff Relations on Urban  and Rural  Areas," USEPA Report No.
     EPA-670/2-75-046, NTIS No. PB 242 830.

44.  Yen, B. C., and Chow, V. T., University  of Illinois,  Urbana,  IL - "Urban
     Stormwater and Runoff Determination of Volumes and Flowrates," USEPA
     Report No. EPA-600/2-76-116, NTIS No.  PB 253 410.

45.  Field, R., et al., USEPA,  Storm and Combined Sewer Program, Edison,  NJ  -
     "Porous Pavement:  Research; Development; and  Demonstration,"
     Tranportation Engineering  Journal of ASCE, Vol.  108,  No.  TE3, May 1982,
     pp. 244-258.

46.  Diniz, E., Espey, Huston & Assoc.,  Inc.  Albuquerque,  NM - "Porous
     Pavement: Phase I - Design and Operational Criteria," USEPA Report No.
     EPA-600/2-80-135, NTIS No. PB 81-104 796.
                                 K-VII-23

-------
 47.   Goforth,  G.  F.,  Espey,  Huston & Assoc., Inc., Austin, TX; Diniz, E. V.,
      Resource  Technology,  Inc.,  Albuquerque, NM;  and Rauhut, J. B., Austin, TX
      - "Stormwater Hydrological  Characteristics of Porous and Conventional
      Paving  Systems," USEPA  Report No.  EPA-600/2-83-106, NTIS No. PB 84-123
      728.

 48.   Goforth,  G.  F.,  Espey,  Huston & Assoc., Inc., Austin, TX; Diniz, E. V.,
      Resource  Technology,  Inc.,  Albequerque, NM;  and Rauhut, J. B., Austin, TX
      - "Stormwater Hydrological  Characteristics of Porous and Conventional
      Paving  Systems," USEPA  Project Summary No. EPA-600/S2-83-106.

 49.   Benjes, H. H., Gulp,  Wesner,  Gulp  Consulting Engineers, El Dorado, CA -
      "Cost Estimating Manual-Combined Sewer Overflow Storage Treatment," USEPA
      Report  No. EPA-600/2-76-286,  NTIS  No.  PB 266 359.

 50.   Benjes, H. H., Jr., Gulp, Wesner,  Gulp Consulting  Engineers, El Dorado,
      CA - Field,  R.,  Storm and Combined Sewer Program,  Edison, NJ - "Estimate
      Sewer Overflow Facility Costs," Water  & Wastes Engineering,  September
      1978, pp. 56-62.

 51.   Heaney, J. F., et al.,  University  of Florida, Gainesville, FL -
      "Nationwide  Evaluation  of Combined Sewer Overflows  and Urban Stormwater
      Discharges,"  Volume II  - Cost Assessment and Impacts," USEPA Report No.
      EPA-600/2-77-064b, NTIS No. PB 266 005.

 52.   Sonnen, M.,  Water Research  Engineers,  Walnut Creek, CA - "Abatement of
      Deposition and Scour  in Sewers," USEPA Report No. EPA-600/2-77-212, NTIS
      No. PB  276 585.

 53.   Yen, B. C.,  et al., University of  Illinois,  Urbana, IL - "Stormwater
      Runoff on Urban  Areas of Steep Slope,"  USEPA Report No.  EPA-600/2-77-168,
      NTIS No. PB  272  755.

 54.   Kaufman, H.  L.,  and Lai, F. H.,  Clinton Bogert Assoc., Fort  Lee,  NJ -
      "Conventional and Advanced  Sewer Design Concepts for Dual Purpose Flood
      and Pollution Control - A Preliminary  Case Study, Elizabeth, NJ," USEPA
      Report No. EPA-600/2-78-090,  NTIS  No.  PB 285 663.

 55.   Field, R., and Weisman,  D. A.,  USEPA Storm and Combined Sewer Program,
      Edison,  NJ -  "A  Planning and  Design Guidebook for Combined Sewer  Overflow
      Control," USEPA  Report  No. EPA-600/2-82-084,  NTIS No.  PB 82-259 235.

56.  Wuelschleger, R.   E., et al.,  Envirex Environmental  Sciences  Div.,
     Milwaukee, WI -  "Methdology for  the Study  of Urban  Storm Generated
     Pollution and Control," USEPA Report No.  EPA-600/2-76-145, NTIS No.'PB
      258 243.

57.   Dalrymple, R. J.,  et al., Beak Consultants for American Public Works
     Assoc.,  Chicago,   IL - "Physical and Settling Characteristics  of
     Particulates in  Storm and Sanitary Wastewater," USEPA  Report No.  EPA-
     670/2-75-011, NTIS No.  PB 242  001.
                                   K-VII-24

-------
58.  Field, R., Storm and Combined  Sewer Program, USEPA,  Edison,  NJ  -
     "Stormwater Pollution Control:  A New Technology," 28 Minute Film
     distributed by the General  Services Administration (GSA),  National Audio
     Visual Center, Washington,  B.C. 20409.

59.  Lynard, W., et al., Metcalf &  Eddy Engineers,  Inc.,  Palo Alto,  CA -
     "Urban Stormwater Management and Technology:   Case Histories,"  USEPA
     Report No. EPA-600/8-80-035, NTIS No. PB 81-107  153.

60.  Heaney, J. F., et al., University of Florida,  Gainesville, FL -
     "Stormwater Management Model:  Level  I—Preliminary  Screening
     Procedures," USEPA Report No.  EPA-600/2-76-275,  NTIS No. PB  259 916.

61.  Heaney, J. F., et al., University of Florida,  Gainesville, FL -
     "Stormwater Management Model:  Level  I—Comparative  Evaluation  of
     Storage/Treatment and Other Management Practices," USEPA Report No.  EPA-
     600/2-77-083, NTIS No. PB 265  671.

62.  Moffa, P., and Karanik, J., Onondaga County, NY  - "Methodology  for
     Evaluating the Impact and Abatement of Combined  Sewer Overflows - A  Case
     Study of Onondaga Lake, NY," USEPA Report No.  EPA-600/8-80-048, NTIS No.
     PB 81-141 913.

63.  Murphy, C. B., et al., O'Brien & Gere Engineers, Inc.,  Syracuse, NY;
     Quinn, T. J., and Steward, J.  E., Monroe County  Division of  Pure Waters,
     Rochester, NY - "Best Management Practices:  Implementation," USEPA
     Report No. EPA-905/9-81-002, NTIS No. PB 82-169  210.

64.  Pitt, R. E., Woodward Clyde Consultants, San Francisco, CA - "Demonstra-
     tion of Nonpoint Pollution Abatement Through Improved Street Cleaning
     Practices," USEPA Report No. EPA-600/2-79-161, NTIS  No. PB 80-108 988.

65.  Pitt, R. E., Consulting Environmental Engineer,  Blue Mounds,  WI -
     "Characterization, Sources, and Control of Urban Runoff by Street and
     Sewerage Cleaning," USEPA Draft Report for Project No. CR-805929,
     January, 1984.

66.  Pitt, R. E., Consulting Environmental Engineer,  Blue Mounds, WI -
     "Characterization, Sources, and Control of Urban Runoff by Street and
     Sewerage Cleaning," USEPA Draft Project Summary  for  Project  No.
     CR-805929.

67.  Struzeski, E., Storm and Combined Sewer Overflow Section,  USEPA, Edison,
     NJ - "Environmental Impact of  Highway Deicing," USEPA Report No.
     11040GKK06/71.

68.  Murray, D. M., and Eigerman, M. R., Abt Associates,  Inc.,  Cambridge,  MA -
     "A Search:  New Technology for Pavement Snow and Ice Control,"  USEPA
     Report No. EPA-R2-72-125.
                                  K-VII-25

-------
 69.   Murray, D. M., et al., Abt Associate's, Inc., Cambridge, MA - "An Economic
      Analysis of the Environmental Impact: of Highway Deicing," USEPA Report
      No. EPA-600/2-76-105.

 70.   Ahlborn, G. H., and Poehlmann, H. C,,, Jr., Ball Brothers Research Corp.,
      Boulder, CO - "Development of a Hydrophobic Substance to Mitigate
      Pavement Ice Adhesion," USEPA Report No. EPA-600/2-76-242.

 71.   Krukar, M., and Cook, J. C.,  Washington State University, Pullman, WA -
      "Optimization and Testing of  Highway Materials to Mitigate Ice Adhesion,"
      USEPA Report No.  EPA-600/2-78-035.

 72.   Field,  R., et al., Storm and  Combined Sewer Technology Branch, USEPA,
      Edison, NJ - "Water Pollution and Associated Effects from Street
      Salting," USEPA Report No. EPA-R2-73-257.

 73.   Richardson, D.  L., et al., Arthur D. Little, Inc., Cambridge, MA -
      "Manual for Deicing Chemicals:  Storage and Handling," USEPA Report No.
      EPA-600/2-74-033.

 74.   Richardson, D.  L., et al., Arthur D. Little, Inc., Cambridge, MA -
      "Manual for Deicing Chemicals:  Application Practices," USEPA Report No.
      EPA-670/2-74-045.

 75.   Freeman,  P. A.,  Peter Freeman Assoc.,  Inc., Berlin,  MD - "Evaluation of
      Fluidic Combined  Sewer Regulators Under Municipal  Service Conditions,"
      USEPA Report No.  EPA-600/2-77-071, NTIS No. PB 272 834.

 76.   Field,  R.,  USEPA,  Storm and Combined Sewer Program,  Edison,  NJ - "The
      Dual-Functioning  Swirl Combined  Sewer  Overflow Regulator/Concentrator,"
      USEPA Report No.  EPA-670/2-73-059, NTIS No. PB 227 182.

 77.   Field,  R.,  et  al., USEPA,  Storm  and Combined Sewer Program,  Edison,  NJ -
      "Treatability  Determinations  for a Prototype Swirl Combined  Sewer
      Overflow Regulator/Solids-Separator,"  Prog. Wat. Tech. Vol,  8,  No.  6, pp.
      81-91.   Printed in Great Britain.

 78.   Field,  R.,  and  Masters,  H. E., USEPA,  Edison,  NJ - "Swirl Device for
      Regulating  and  Treating  Combined Sewer  Overflow,"  USEPA Report  No.  EPA-
      625/2-77-012,  EPA  Technology  Transfer  Capsule  Report,  CERI (Cincinnati)
      2012.

 79.   Sullivan,  R.  H., et al., American Public Works  Assoc., Chicago,  IL -
      "Swirl  and  Helical Bend  Pollution Control  Devices:   Design Manual,"  USEPA
      Report  No.  EPA-600/8-82-013,  NTIS No.  PB 82-266 172.

80.  Matthews, T. M., et  al., Snell Environmental Group,  Akron, OH -
      "Hydrobrakes Regulated Storage System  for  Stormwater Management,"  USEPA
      Report  No.  EPA-600/2-83-097,  NTIS No. PB 84-110 378.
                                  K-VII-26

-------
81.  Matthews, T. M., et al.,  Snell  Environmental  Group,  Akron,  OH -
     "Hydrobrakes Regulated  Storage  System  for  Stormwater Management,"  USEPA
     Project Summary No. EPA-600/S2-83-097.

82.  Field, R., and Struzeski, E. J., USEPA,  Storm and  Combined  Sewer Program,
     Edison, NJ - "Management  and Control of  Combined  Sewer  Overflows,"
     Journal Water Pollution Control Federation, Washington,  D.C.,  Vol. 44,
     No. 7, pp. 1393-1415, July 1972.

83.  Field, R., Storm and Combined Sewer Program,  Edison,  NJ;  Lager, J. A.,
     Metcalf and Eddy,  Inc., Palo Alto, CA  -  "Urban Runoff Pollution Control -
     State-of-the-Art," Journal of the Environmental Engineering Division,
     American Society of Civil Engineers, Vol.  101, No. EE1,  pp.  107-125,
     February 1975.

84.  Soderland, H., Kjessler & Mannerstrale AB,  Swedish Council  for Building
     Research, Stockholm, Sweden - "Flow Balancing Method for Stormwater and
     Combined Sewer Overflow," ISBN  91-540-3765-4: 017:1982,  pp.  1-27,  1982.

85.  Smith, W. G., et al., Metcalf & Eddy Engineers, Inc., Palo  Alto, CA -
     "Storage/Sedimentation Facilities for  Control of  Storm  and  Combined Sewer
     Overflows Design Manual," USEPA Report No.  Pending,  (USEPA  Contract No.
     68-03-2877).

86.  Field, R, and Weisman, D. A., USEPA, Storm and Combined Sewer Program,
     Edison, NJ - "A Planning  and Design Guidebook for  Combined  Sewer Overflow
     Control and Treatment," USEPA Report No. EPA-600/2-82-084,  NTIS No. PB
     82-259 235.

87.  Shelley, G. J., et al., George  J. Shelley  Consulting Engineers, Tamworth,
     New South Wales, Australia - "Field Evaluation of  a  Swirl Degritter at
     Tamworth N.S.W., Australia," USEPA Report  No. EPA-600/2-81-063, NTIS No.
     PB 81-187 247.

88.  Shelley, G. J., et al., George  J. Shelley  Consulting Engineers, Tamworth,
     New South Wales, Australia - "Field Evaluation of  a  Swirl Degritter at
     Tamworth N.S.W., Australia," USEPA Project  Summary No.  EPA-600/S2-81-063.

89.  Field, R., et al., USEPA, Storm and Combined  Sewer Program,  Edison, NJ  -
     "Proceedings of Workshop  on Microorganisms  in Urban  Stormwater," USEPA
     Report No. EPA-600/2-76-244, NTIS No.  PB 263  030.

90.  Glover, G. E., and Herbert, G. R., Crane Company,  King  of Prussia, PA -
     "Microstraining and Disinfection of Combined  Sewer Overflows  - Phase  II,"
     USEPA Report No. EPA-R2-73-124, NTIS No. PB 219 879.

91.  Drehwing, F. J., et al.,  O'Brien & Gere  Engineers, Inc.,  Syracuse, NY -
     "Combined Sewer Overflow  Abatement Program, Rochester,  NY - Volume H
     Pilot Plant Evaluations," USEPA Report No.  EPA-600/2-79-03lb,  NTIS No.  PB
     80-159 262.
                                 K-VI1-27

-------
  92.   Moffa,  P. E.,  et al., O'Brien & Gere Engineers, Inc., Syracuse, NY -
       "Bench-Scale High-Rate Disinfection of Combined Sewer Overflows with
       Chlorine and Chlorine Dioxide," USEPA Report No. EPA-670/2-75-021, HTIS
       No.  PB  242 296.

  93.   Maher,  M. B.,  Crane Company,  King of Prussia, PA - "Microstraining and
       Disinfection of  Combined Sewer Overflows - Phase III," USEPA Report No,
       EPA-670/2-74-049,  NTIS No. PB 235 771.

  94.   Drehwing, F.,  et al., O'Brien & Gere Engineers, Inc., Syracuse, NY -
       "Disinfection/Treatment of Combined Sewer Overflows, Syracuse, New
       York,"  USEPA Report No.  EPA-600/2-79-134, NTIS No.  PB 80-113 459.

  95.   Leitzs  F. B.,  et al., Ionics, Inc., Watertown, MA - "Hypochlorite
       Generator for  Treatment of Combined Sewer Overflows," USEPA Report No.
       11023DAA03/72, NTIS No.  PB 211 243.

  96.   Pontius,  U.  R.,  et al.,  Pavia Byrne Engineering Corp., New Orleans, LA ~
       "Hypochlorination of  Polluted Stormwater Pumpage at New Orleans," USEPA
       Report  No.  EPA-670/2-73-067,  NTIS No.  PB 228 581.

  97.   Conrick,  D.  J.,  et al.,  Environmental  Design & Planning, Inc., Allston
       (Boston)  MA -  "Evaluation  of  a Treatment Lagoon for Combined Sewer
       Overflow,"  USEPA Report No. EPA-600/2-81-196,  NTIS No. PB 82-105 214,

  98.   Carr, D.  J., et  al.,  Marquette University,  Milwaukee,  WI - "Characteris-
       tics and  Treatability of Urban Runoff  Residues," USEPA Report No. EPA-
       600/2-82-094, NTIS No.  PB  83-133  561.

  99.   Carr, D.  J., et  al.,  Marquette University,  Milwaukee,  WI - "Characteris-
       tics and  Treatability of Urban Runoff  Residues," USEPA Project Summary
       No.  EPA-600/S2-82-094.

100.   Homack, P.,  et al., E.  T.  Killlam Assoc., Inc.,  Millburn,  NJ -
       "Utilization of  Trickling  Filters  for  Dual  Treatment  of  Dry  and Wet
       Weather Flows,"  USEPA Report  No.  EPA-670/2-73-071,  NTIS No.  PB 231 251.

101.   Benedict, A. H., and  Roelfs,  V. L.,  Whitely-Jacobsen  and Assoc.,
       Portland, OR - "Joint Dry-Wet Weather  Treatment  of  Municipal Wastewater
       at Clatskanie, Oregon,"  USEPA Report No.  EPA-600/2-81-061, NTIS No.  PB
       81-187 262.

102.   Benedict, A. H., and  Roelfs,  V. L.,  Whitely-Jacobsen  and Assoc.,
       Portland, OR - "Joint  Dry-Wet  Weather  Treatment  of  Municipal Wastewater
      at Clatskanie, Oregon,"  USEPA Project  Summary  No. EPA-600/S2-81-061.

103.  Agnew, R. W., et al., Envirex, Milwaukee, WI - "Biological Treatment of
      Combined  Sewer Overflow  at Kenosha,  WI,"  USEPA Report  No.  EPA-670/2-75-
      019, NTIS No. PB 242  107.
                                  K-VII-28

-------
104.  Innerfield, H., and Forndran, A., New York City Deparatment of Water
      Resources, New York, NY - "Dual Process High-Rate Filtration of Raw
      Sanitary Sewage and Combined Sewer Overflows," USEPA Report No. EPA-
      600/2-79-015, NTIS No. PB 296 626/AS.

105.  Welborn, H. L., Y-T-0 & Assoc., Walnut Creek, CA - "Surge Facility for
      Wet- and Dry-Weather Flow Control," USEPA Report No. EPA-670/2-74-075,
      NTIS No. PB 238 905.

106.  Mahida, V. U., and DeDecker, F. J., Spalding DeDecker Assoc., Madison
      Heights, MI - "Multi-Purpose Combined Sewer Overflow Treatment Facility,
      Mt. Clemens, MI," USEPA Report No. EPA-670/2-75-010, NTIS No. PB 242
      914.
                                  K-VII-29

-------
        NONPOINT POLLUTION ABATEMENT PROGRAMS:   THE  MANAGER'S  DILEMMA

                            Frederick W. Madison
              Associate Professor, Department of Soil Science,
              Wisconsin Geological and Natural  History Survey,
                       University of Wisconsin-Madison
                             Madison, WI  53706
                                  ABSTRACT

The passage of PL 92-500 in 1972 set governments at all levels—state, local
and federal—on a collision course.  The legislation changed national
pollution abatement policy in that it directed the pollution control efforts
be focused on treating pollution at its source rather than in receiving
waters.

The challenge of dealing with non-point sources of pollution is that it
interfaces land-use and water quality problems.  In the country,
historically, the federal government has dealt with water policy whereas
responsibility for land-use decisions has been delegated to local units of
government.  As a result, there exists a tremendous institutional complexity
which must be overcome if nonpoint problems are to be dealt with
effectively.  Implementation programs and mechanisms have evolved and
nonpoint programs can, and should be put in place.


Keywords:  Nonpoint source pollution, water policy, land-use policy,
           nonpoint source programs, PL-92-500, institutional arrangements,
           abatement program implementation
                               K-VIII-l

-------
 NonPoint Pollution Abatement Programs:   the Manager's  Dilemma

 The passage of PL 92-500 in the early 1970's set governments  at  all
 levels—state, local  and federal—on a  collision course.   This landmark
 legislation was the result of extensive deliberations  by  the  Congress which
 focused its attention on the continuing deterioration  of  the  nation's
 surface waters even after several  years of fairly significant federal
 investment in the upgrading of municipal  sewage  treatment plants.

 The legislation,  of course, represented,  in the  final  analysis,  a
 significant change in policy for the nation in its struggle to control the
 continuing degradation of its waters.   Pollution control  efforts were now
 going  to be focused on treating pollution at its source,  rather  than in
 receiving waters.  Permits were to be issued to  control any pollutants
 discharged from a point source and pollution arising from the wide array of
 man's  land-use activities was also to be  treated.

 The term nonpoint source pollution was  coined to cover basically all
 pollution arising from diffuse sources.   It remains interesting  that in the
 entire text of PL 92-500, the term nonpoint source pollution  was never
 defined;  it was basically taken to mean everything else.

 Section 208 of that same legislation launched the  nation  on a massive
 planning effort designed to devise strategies for  the  control of both point
 and nonpoint sources  of pollution.   Unfortunately,  the language  in that
 section was also  apparently vague  and it  took several  years and  many fits
 and starts before that program got  underway.

 The challenge of  dealing with nonpoint  sources of  pollution is that it
 interfaces land use and water quality problems;  the inevitable collision
 course that it puts units of government on  is deeply rooted in the
 institutional  and legal  history of  this country.   Since the early days of
 the nation,  the federal  government  has  played a  major  role in water
 resources development as well  as in water quality  and  policy.  In 1824, the
 U.S. Supreme Court concluded that  navigation  and related  water resource
 programs  were within  the purview of the Congress under its authority to
 regulate  interstate commerce.

 Over the  years, the federal  government  has  been  directly  involved in the
 construction of dams,  canals,  flood control  projects,  irrigation projects,
 and  the like.   A  minimal  responsibility for pollution  control was placed in
 the  U.S.  Public Health  Service in  1912  although  the first abatement
 authority was  probably  given  to  the Corps of  Engineers under  authorities
 granted  in  the  Refuse Act of  1899.   Pollution abatement efforts,  of course,
 have been  strengthened  enormously over the  years with  significant
 legislative  overhauls  in  1948,  1965, and  1972.

 The  states,  basically,  have  exercised authority  over water resources
 problems  only  in  those  areas  clearly without  federal jurisdiction.  Since
 the  1950s,  they have  been  involved  in implementing federal water quality
 programs,  all of  which  has  been  done with pretty strong federal  policy
 guidelines.   States do  have the  responsibility for the regulation of
municipal  and private use  of water, water supply, groundwater, and
 industrial waste.   It  is  noteworthy that  these regulatory responsibilities

                                 K-VIII-2

-------
have been carried out primarily by the states themselves and that very
little responsibility in these areas has been delegated to local units of
government.

Policy governing land-use management has followed a very different course.
The idea that a "man's home is his castle" is deeply rooted in our national
tradition.  Much of the impetus behind the westward movement of the pioneers
in this country was the desire to own a piece of land to which an individual
could do whatever he or she wanted, presumably without any interference from
anybody.  Although the states, primarily through their police powers, have
the authority to regulate land use, this authority has traditionally been
passed on to local units of government.  States will often define areas of
interest where land use activities affect more than local jurisdiction or
will set limits to local land use activities but above and beyond this, the
tendency has been to interfere as little as possible in land-use issues.
The result is a very diffuse approach to land management questions which is
in sharp contrast to the rather sharply focused, water-quality management
programs.

The federal role in land-use plicy is very limited.  They do, of course,
administer public lands which does assume significance in some western
states where those holdings make up a large percentage of the land area.
Federal policies on soil and water conservation, however, are carried out at
the local level by unique state-local institutions which were created at
federal insistence, the Soil and Water Conservation Districts (SWCDs).

But PL 92-500 set the stage.  It established the fact that nonpoint source
pollution was a significant contributor to the degradation of surface water
resources and that it was a problem that had to be dealt with.  Dealing with
it meant that land-use policies had to be reviewed, amended, updated, and in
some instances, established in order to solve water quality problems.

PL 92-500 also stimulated a number of research projects to provide the
technical backdrop for institutional, legal, and programmatic changes that
the implementation of nonpoint source control programs would require.  Many
of the efforts examined and attempted to quantify, loadings from land-use
activities and the impacts of those loadings on surface-water quality.
Funds for some of this work came from Section 208 although perhaps the
largest program was mounted by the Great Lakes National Program Office in
Region V of the U.S. Environmental Protection Agency (EPA) with funds
appropriated under Section 108 of PL 92-500.  That program was linked
closely to a massive effort by both U.S. and Canadian researchers under the
auspices of the International Joint Commission.

Rural and urban nonpoint problems were identified and assessed.  Best
management practices were evaluated.  Technical solutions to nonpoint
problems were tested and a fairly good understanding of how to deal with the
majority of the diffuse pollution problems emerged.  Of all these projects,
the Washington County Project in southeastern Wisconsin was unique because
it was designed to examine the legal and institutional complexities
associated with the implementation of nonpoint source control programs and
even had, as one of its specific program objectives, the development of a
sediment control ordinance.  It was a bold and innovative effort funded by
the Great Lakes National Program Office.

                                K-VIII-3

-------
 The project involved 11  cooperating agencies representing state,  local,
 federal  and regional units of government with the main actors being the
 Water Resources Center of the University of Wisconsin-Madison,  the State
 Board of Soil  and Water Conservation Districts, the Wisconsin Department of
 Natural  Resources (DNR), and the Washington County Soil  and Water
 Conservation District Supervisors.   Project staff included a smattering  of
 soil  scientists, agricultural engineers, and water chemists but was
 dominated by lawyers, planners,  sociologists, and economists.  The challenge
 was fairly straight-forward, namely, to work through the legal-institutional
 maze which had evolved as a result  of the differing approaches  taken to  land
 and water management in  this country and to determine if the very
 heterogeneous  mix of institutions could be woven into some kind of pattern
 that could deal  effectively with the control of nonpoint source pollution
 problems.

 At  the inception of the  Washingotn  County Project, EPA's interest was
 twofold,  i.e.,  they were committed  to focusing responsibilities for
 implementation on a local unit of government, preferably the Soil and Water
 Conservation District, and they  were committed to regulating sediment loss
 in  rural  areas.

 In  Wisconsin at  that time (the mid-1970s)  the set-up of  Soil  and  Water
 Conservation Districts was somewhat unique in that the districts  were
 coterminous with county  boundaries  and the SWCD supervisors were, by law,
 those members  of the county board who served on the Agriculture and
 Extension Education Committee.   A project  sponsored study of the  County
 Board in  Washington County concluded that  change—in this instance,  perhaps,
 even  radical change,  namely sediment control--could probably best be
 accomplished at  this  government.  The reasoning behind this was basically
 that  county board supervisors were  somewhat isolated from their constituents
 who generally  have very  little notion of what county government was  all
 about.  As  a result of this,  county supervisors tended to be  re-elected  time
 after time  without much  voter concern about what they were doing.   The
 thinking  was,  then,  that perhaps  major change could be accomplished  at this
 level  of  government if it were done carefully.

 County government in  Wisconsin is basically an  arm of state government and
 really has  only  those authorities that the state grants  it.   A  legal  review
 suggested that much  could be  accomplished  within the existing statutory
 framework but that the role of the  SWCD could be strengthened by  making  it  a
 standing  committee of the County  Board and by integrating its function with
 other  county land use functions.  This led to an extensive revision  of state
 statutes which abolished Soil  and Water Conservation Districts  in Wisconsin
 and replaced them with a committee  of the  County Board called the Land
 Conservation Committee.   Authorities  of the old SWCDs  were retained;  new
 responsibilities  for  planning for erosion  control  and  animal  waste
management  have  been  added.   Major  funding for  implementing these programs
may be forthcoming from  the current  session  of  the  Wisconsin  legislature.

A sediment  control  ordinance was  developed which  set  a performance standard
for soil  losses  from  an  individual  farm based on  the  Universal  Soil  Loss
Equation.  A farmer not  meeting the  standard  could  adjust  his management
practices in anyway he wished as  long  as the  predicted soil loss  rate  met
requirements established  by the county.  The  Washington county  SWCD

                                K-VIII-4

-------
supervisors were prepared to take the proposal to a referendum vote in a
part of the county when a certain amount of waffling on the question of
sediment regulation occurred both in the state legislature in Madison and at
EPA headquarters in Washington.  That was the end of that proposal; the
substance of the idea was adopted finally by the SWCD as a policy or
position paper.

Efforts to regulate sediment losses from subdivisions in urban and urbanized
areas proved to be more successful.  In a simple sense, urban residents are
much more accustomed to regulation than their rural counterparts and are
required to obtain permits and to meet all kinds of requirements for almost
everything they do.  One more requirement one way or another didn't seem to
make much difference.  The major problem encountered involved convincing
local officials that, in fact, a problem existed.

During this period, the Wisconsin Department of Natural Resources launched
its Nonpoint Source Pollution Abatement Program.  Five watersheds were
selected for the program initially involving multiple governmental
jurisdictions.  Local government agencies with implementation authority--
called Designated Management Agencies or DMAs—were identified to run the
programs.  One watershed, the Root River in southeastern Wisconsin, had 23
DMAs.  The program established the pattern for nonpoint source
implementation, namely, that local units of government with appropriate
authorities would implement programs guided by firm state policy
directives.  The institutional complexities, however, remain.

This past year the EPA convened a Nonpoint Source Task Force to develop a
National Nonpoint Source Policy to accelerate the implementation of nonpoint
source control programs.  The proposal policy urges continued federal
leadership and coordination but goes on to emphasize that the states will
have the lead in developing and implementing nonpoint source management
strategies.  Local governments are to be charged with developing nonpoint
source strategies in coordination with the states and ultimately, with
implementing those strategies.

All the pieces would seem to be in place, yet the question still remains:
Can nonpoint control programs be implemented?  The linkages between the
water and land agencies have been forged although a great deal of
institutional complexity remains.  Federal agencies, particularly those in
the U.S. Department of Agriculture, are falling prey to the budget cutter's
ax.

This year is the 50th Anniversary of the establishment of the U.S. Soil
Conservation Service.  The agency was formed in response to the Dust Bowl
crisis of the 1930s and under the leadership of the inspirational Hugh
Hammond Bennett represented a bold and innovative effort to solve nagging
soil erosion problems.

Now, 50 years later, we are again faced with many of the same problems and
perhaps now it is time once again to look for innovative answers to the
problems plaguing nonpoint control programs.

Authority for passing money through to states for technical and cost-sharing
assistance has existed for some years now.  Maybe the time is right for
trying that approach.

                                K-VIII-5

-------
Participation  of  landowners  in  rural  nonpoint  programs  today  is a problem
much  as  it was with  PL-566 Watershed  Projects  20  and 30 years  ago.   Is
regulation the answer?   The  jury  seems  to  still be out  on that question.
More  and more  states are looking  to regulation  and the  idea of establishing
some  realistic goal,  like T  by  2000,  seems to  have become very appealing.
The established goal  is  to be reached in steps  and most proposals call for a
regulatory mechanism to  kick in if interim goals  are not met.

It goes without saying that  as  point  sources of pollution are  increasingly
controlled, nonpoint sources assume even greater  significance  in the overall
effort to control the continuing  degradation of the nation's water.  Many
problems have  been solved, some still remain.   The where-with-all to mount a
major program  to  control  nonpoint source has been developed.   The time has
come to put that  knowledge into action.

SELECTED REFERENCES

Arts, J.L., and Church,  W.L. (1982).  Soil Erosion—the Next Crisis?
     Wisconsin Law Review, Vol. 4, No.  5, pp. 535-626.

Best Management Practices for Nonpoint  Source Pollution Control Seminar
     (1976).   EPA 905/9-76-005.

Final Report on the  Federal/State/Local Nonpoint Source Task Force and
     Recommended National Nonpoint Source Policy (1985).  EPA,
     Washington, D.C.

Madison, F.W., Arts,  J.L., Berkowitz, S.J., Salmon, E.E. and Hagman, B.B.
     (1979).   The Washington County Project:  Final Report.  EPA
     905/9-80-003.

Madison, F.W.  (1981).  Institutional   and Legal Complexities of Nonpoint
     Source Pollution Control.   EPA 905/9-81-005.

Voluntary and  Regulatory Approaches for Nonpoint Source Pollution Control
     (1978).   EPA 905/9-78-001.
                                K-VI1I-6

-------
     INSTITUTIONAL ARRANGEMENTS FOR NONPOINT POLLUTION CONTROL PROGRAMS

                                Floyd E. Heft
                                Past-president
                     Soil Conservation Society of America
                              4319 Brookie Court
                              Columbus, OH 43214
                                    U.S.A.
                                  ABSTRACT

Institutional arrangements for nonpoint pollution control must be structured
to provide maximum local acceptance, implementation, and enforcement.  The
more remote the institutional arrangements are from the local polluter and his
understanding and trust, the more suspect and difficult implementation
becomes.  The ultimate in institutional arrangements is about 180 degrees from
point-source pollution abatement arrangements.

Strong educational and technical assistance to the polluter, coupled with
realistic subsidies for pollution abatement practices that are not
cost-effective for the polluter, are the key to acceptance and adoption.
Local institutional arrangements must prevail in these efforts.  Enforcement
must be established with a low profile and used as a last resort.

State and federal institutional involvements must carry a low profile a's well,
with realistic assistance to the local institutions.  State and federal
dictates will only slow implementation.  Federal and state subsidies and
implementation funds can be administered most efficiently at the local level
and should be provided through grants.

The standard procedures for monitoring point-source discharges and
establishing tolerances becomes nearly impossible and indefensible for
nonpoint pollution, particularly from an economic point of view.  This dilemma
merits the alternative solution of preventing nonpoint pollution as the most
acceptable means of control.

Local institutional arrangements must and will vary from state to state and
community to community according to the forms of institutional authorities and
responsibilities and the severity of nonpoint pollution problems.  State and
federal efforts to stereotype local institutions and programs will usually
hinder local initiatives and implementation.

Coordination of local, state, and federal institutions for maximum effect and
efficiency is necessary for success and for gaining of the public's confidence
that the nonpoint pollution problem is being addressed in a realistic and
productive manner.  Rearranging authorities, responsibilities, and support
programs of existing institutions can and must prevent the formation of
additional institutions at any level of government.

Institutional arrangements adopted within a state must provide the capability
for implementation if responsibilities are not carried out by the primary
institution of authority, thus assuring the public of positive action.


                                    K-IX-l

-------
       INSTITUTIONAL  ARRANGEMENTS FOR NONPOINT POLLUTION CONTROL  PROGRAMS
Distinct  differences  exist  between  point-source  and  nonpoint-source  pollution.
These  differences  must  be taken  into account in  developing  acceptable
institutional  arrangements  for the  control  of nonpoint-source  pollution.
Following are  the  major differences:

*  Nonpoint-source pollutants are primarily diffuse  in  nature.   Most are
   associated  closely with  surface  runoff,  usually following land-disturbing
   activities,  in  both  rural and urban  settings.  In contrast,  point-source
   pollutants  are  primarily associated  with industrial  and  municipal
   discharges  that emanate  at a  singular  point.

*  Point-source pollutants  frequently are by-products of  commercial  activity
   with little or  no  use or value.   Nonpoint-source  pollutants,  on the other
   hand,  often are of economic value to the generating  unit and,  therefore,
   represent an economic loss.   Pollution abatement  efforts must  consider the
   fact that the motivation to pollute  or not to pollute  involves these forces
   of  opposing value.

*  Nonpoint pollutants  from agricultural  operations  mainly  include sediment,
   fertilizers, pesticides,  and  animal  wastes.  All  represent  purchased or
   natural elements required for the  production of food and fiber.   The
   positive value  of  these  elements  seemingly would  be  sufficient motivation
   for agriculture to attempt to retain such pollutants on  the  land.
   Unfortunately,  it  is  necessary to  expose these potential pollutants, which
   are necessary for  agricultural production,  to rainfall and  wind during
   land-disturbing activities.

*  An  individual's or firm's ability to recover point-source pollution
   abatement costs is possible—the  individual or firm  generally  controls a
   product's supply and  market price.   Nonpoint polluters,  particularly
   agricultural operations, have little,  if any, ability  to recover  pollution
   abatement costs through  supply control or price setting.  Therefore,
   effective pollution  abatement costs  cannot  be recovered  in normal
   production  and  marketing activities.

*  Monitoring  and  permitting point-source polluters  is  quite simple  and
   precise.  These activities usually generate the factual  data,  records, and
   analysis that are  crucial to  enforcement and prosecution if necessary.
   Because of  their diffuse nature, the identification  of sources of nonpoint
   pollution and the  volumes of  pollutants  they contribute makes meaningful
   monitoring,  control,  and enforcement nearly impossible.

*  Issuing permits to point-source polluters,  a relatively easy process,
   usually is  accomplished  by establishing  an  acceptable  level of pollutants
   to be discharged or  issuing an order to  the polluter requiring compliance
   with an established  standard.  The technical and  financial burden for
   compliance becomes a  production cost to  the point-source polluter; this
   cost is programmed into  the cost of producing a product  and ultimately is
   passed along to consumers.  The same procedure could be used to control and
   finance the elimination of municipal sewage pollution.  However, 75 percent


                                    K-IX-2

-------
   of the technical and financial burden for constructing such facilities is
   assumed by the federal government in the name of public benefits, which
   alleviates a large portion of the municipality's direct cost.  Increased
   federal taxation is necessary to pay for this federal subsidy.  The higher
   tax becomes an added cost to the nonpolluter having no direct obligation.

Permitting nonpoint sources of pollutants has little, if any, support from the
public or governmental institutions because of the impossibility and
prohibitive costs of monitoring the discharges.  Fluctuations in discharges,
both volume and concentration, by virtue of variations in both rainfall
intensity and soil characteristics create a nearly impossible condition for
reliable and defensible monitoring.  Enforcement under a permit system would
be questionable at best.

A strategy for nonpoint pollution control

The differences between point-source and nonpoint-source pollution are the
basis for development and execution of two drastically different abatement
strategies.

Success in reducing point-source pollution through the sequence of permitting,
monitoring, prosecution, and penalties has been effective because of
governments' extensive authorities.

Success in reducing nonpoint pollution cannot make use of this same sequence
of tactical actions by governments.  Success can only be attained through
education, demonstration, technical assistance, subsidization, and, as a last
resort, enforcement by an institution capable of assessing a penalty.

Technological deficiencies are substantial in nonpoint pollution abatement.
Identification of a pollutant's source, its concentration or volume, best
control techniques, economic impacts, and the practicality of solutions are
all matters that agricultural research has dealt with in only limited fashion.
Agriculture instead has pursued profit-motivated research and development
strategies generally acceptable within our free enterprise system.  Little
effort has been devoted to determining the environmental effects of new
products or technologies.

Extensive investigation gradually produced a prevention strategy as the most
acceptable and defensible approach to nonpoint-source pollution control.  As a
strategy, prevention represents a system of management practices designed to
reduce, alter, or eliminate landowner and land operator actions that generate
nonpoint pollutants.  Other than sediment, the agricultural pollutants of
primary concern are phosphates, nitrates, pesticides, and animal wastes.
Phosphates and pesticides affix to soil particles upon application to the
land.  Nitrates pollute water by becoming an ingredient of surface water
runoff or by percolating down through the soil into underground water
supplies; the degree of pollution depends upon rainfall intensities, nitrate
application rates, and timing of application.  Animal wastes pollute water if
applied to the land in excessive quantities or through the uncontrolled
concentration or confinement of animal units without adequate animal waste
management facilities.  The focus of the prevention strategy has been to keep
agricultural pollutants from entering streams and underground water supplies

                                   K-IX-3

-------
through the reduction or elimination of  soil movement  and  storm water runoff
from the  land  into  streams  and  lakes.

This prevention  strategy focuses directly on farmers and their actions to
produce food.  Owners and developers of  residential, industrial, and
transportation sites are examples of nonagricultural land-disturbing
activities causing  erosion  and  sediment  pollution of our waters.  Nonpoint
pollution by sediment itself contributes little chemical pollution, although
it does produce  extensive turbidity and  damage to the  biological environment.

The identification  of major nonpoint pollutants and why they enter water
supplies puts the prevention strategy squarely in front of millions of
landowners and land operators.  Neither  owners or operators have an economic
motivation to pollute, nor  do they understand to what  extent they may or may
not be polluting.

Nonpoint Abatement  Goal—A  First Priority

Identification of nonpoint  pollutants by volume and source provides the
foundation for developing institutional  arrangements for their control.
Acceptable and attainable nonpoint pollution control goals dictate what
institutions should be involved and the extent of that involvement.

Institutional structure must also recognize three facts:  that landowners and
land operators are  independent, private entrepreneurs  by choice; that most
dislike regulation; and that resistance to regulation  magnifies as higher
levels of government become involved.  Many landowners still harbor the
position that a  fee simple  title grants complete management authority,
regardless of the impact on the public's interest.  With few exceptions,
however, most will  accept change if they understand the need for pollution
control and if control efforts  are both justified and  fair.

For 10 years I was  involved in  the development and early implementation of
Ohio's nonpoint  pollution abatement program.  The merits of that state's
strategy deserve attention.  Other states have structured similar strategies
with minor differences due  to varying forms of local and state governments.

Four major determinations influenced Ohio's final institutional arrangement:

1.  All means were used to  avoid the creation of a new institutional structure
    to address nonpoint pollution control.

2.  It was determined which existing institutions might logically and
    willingly expand their responsibilities and authorities to implement a
    nonpoint pollution abatement program.

3.  Consensus was secured among agricultural, governmental, and environmental
    interests regarding their desired institutional arrangement.

4.  Legislation was written to  authorize the selected  institutions to produce
    the necessary standards, rules, and regulations and implementation
    procedures for an effective nonpoint pollution abatement program.
                                    K-IX-4

-------
Among the Important factors influencing the selection of a final institutional
structure and implementation strategy were the following:

1.  Existing federal and state agricultural agencies were unwilling to
    restructure their voluntary-service approach to include enforcement.

2.  Controlling the largest volume of agricultural pollutants was attainable
    through application of soil erosion control and other land management
    practices recommended by existing local conservation institutions.

3.  Good working relationships and a reasonably high level of trust and
    respect existed between local conservation institutions and farmers.

4.  Farmers seemed relatively comfortable with the identification of pollution
    problems by local agencies and recommendations from those agencies for
    abating the problems.

5.  Pollution abatement practices, enforcement procedures, penalties for
    noncompliance, and legal recourse for polluters were generally formulated
    and agreed to prior to legislative action granting state and local
    institutions the necessary authorities for implementation.

6.  Institutional arrangements acceptable to the U.S. Environmental Protection
    Agency and, presumably, the public for point-source pollution control
    proved both inadequate and inappropriate for nonpoint-source pollution
    control programs.

Local Responsibilities

Conservation districts in Ohio accepted the authority to investigate and
specify the necessary combination of land management practices required to
reduce soil loss to the state-adopted limits.  Districts provide technical
assistance to landowners and land operators when necessary and recommend
cost-sharing where applicable.  Should enforcement procedures become
necessary, a district's administrative board issues the first warning to a
landowner or land operator.  The board also establishes a time frame for
compliance and offers the farmer any technical or financial assistance
available.  If the landowner or land operator complies within the designated
time frame, the enforcement process ceases.  If the landowner or land operator
fails to comply, he or she is declared to be in violation of the rule, and the
district board informs the individual that the designated state enforcement
agency, the Division of Soil and Water Conservation within the Ohio Department
of Natural Resources, is to be notified of the violation.  The violator is
also informed that he or she has a designated period of time within which he
or she may appeal the local board's proposed action, if desired, prior to the
board's forwarding of the citation to the state agency.

Public complaints about pollution problems can be registered with districts or
the Division of Soil and Water Conservation.  Complaints must be in writing,
and they must be signed by the complainant.  Complaints registered with the
state agency are referred to the conservation district for investigation and
recommendations.  A district representative discusses the complaint with the
accused violator.  Technical and financial assistance are offered, if


                                    K-IX-5

-------
appropriate, to facilitate  installation of  any needed management practices.
This contact usually results  in a  cordial solution to the problem.  Both the
individual who filed the  complaint and the  accused are  informed of the
district's findings and disposition of the  complaint.

This strategy for controlling agricultural  pollution is premised on education
and information, technical  assistance, and  financial subsidy, where
appropriate.  Enforcement is  a last resort.  But to view the strategy as weak
and cumbersome is inappropriate because enforcement is administered, when
necessary, to attain compliance.

State Responsibilities

Every state has an agency with administrative responsibility for soil and
water conservation and conservation districts.  In Ohio, this unit is the
Division of Soil and Water  Conservation, which also has been assigned certain
responsibilities for nonpoint pollution abatement.

Specifically, the state legislature assigned the Division responsibility for
developing and implementing a program for abatement of agricultural and urban
sediment nonpoint pollution.  The  program consists of rules and regulations
for the control of pollution from  soil erosion, pesticides, fertilizer, animal
wastes, and urban sediment  generated by land-disturbing activities.  Criteria
were also established for subsidizing the installation of sediment control
practices and animal waste  control facilities that are not cost-effective to
the landowner but were required for public  benefits.

The Division disburses these subsidies and  administers the enforcement
procedure for conservation  districts when recommended.  It also grants the
accused landowner or land operator a hearing prior to its final determination
and issuance of a court order and  penalty schedule if that becomes necessary.

Federal Responsibilities

The U.S. Environmental Protection  Agency has responsibility for setting the
parameters within which a state formulates  its nonpoint pollution abatement
plan for meeting requirements of the Clean  Water Act.  These parameters
provide the needed flexibility to  states for implementing a nonpoint pollution
abatement program tailored  to a state's particular problems and institutional
arrangements.  These institutional arrangements at the state and local levels
are not and should not be a federal concern beyond the assurance that the
chosen institutional structure has the capability to get the pollution control
job done.  Periodic progress reports to EPA are justified as a measure of a
state's progress in meeting the nation's clean water goals.

EPA must assume leadership  responsibility for other federal institutions in
compliance with state-adopted program standards and local implementation
procedures.

Urban Sediment Control Strategy

Institutional arrangements  and strategies for controlling sediment pollution
from land-disturbing activities in urban areas are basically the same as in


                                    K-IX-6

-------
agricultural areas.  The major responsibility for action must rest at the
local level.  County and city governmental units should be given the authority
if they are not already authorized to enact and enforce sediment control
ordinances.  Both governmental units control development and building permits
within their jurisdiction, and such permits should contain sediment and storm
water management plans.  No special permit need be required.  In fact,
additional permits are economically and politically frustrating to the permit
seeker.

Conservation districts may encourage adoption of such sediment control
ordinances and offer technical assistance, or even enter into working
agreements for review, approval, and implementation services, as the city or
county may desire.

Should cities and counties fail to accept these local responsibilities, state
initiatives should be authorized to assure progress and compliance with
state-adopted rules and regulations.

Enforcement should automatically fall to the legal arm of the enforcing unit
of government, be it local or state.  Also, local units of government should
be given the authority and opportunity to address urban sediment pollution
problems before involving state and federal institutions, beyond encouragement
and guidance from those levels.

The state can play a vital role by developing and adopting urban sediment
control rules, regulations, and a model ordinance as guides for local
initiatives.

Logistical Support Strategy

Local institutions that assume the responsibilities for nonpoint-source
pollution abatement have many state and federal agencies available to assist
them in varying degrees and ways.  It is not only logical but imperative that
these agencies be used to the maximum extent.  The structuring of new
governmental units should be avoided.

The major areas of assistance needed and normally available include the
following:

Education and Information.  Education of the public and landowners
specifically about the chemical and physical aspects of nonpoint pollution is
essential to acceptance of an abatement program.  The Cooperative Extension
Service has a long-standing, highly respected reputation for public education
in agricultural technology.  This education is founded on research by state
and federal agencies.  Extension's participation in education should be
sought.  But information, of course, can be disseminated by any public agency
as well as commercial interests and the public media.  These educational and
informational forces are extremely important to landowner acceptance and
public support.

Research.  Research must address the causes of and cures for nonpoint
pollution.  The Agricultural Research Service and state agricultural research
                                   K-IX-7

-------
institutions must produce practical and economical pollution control
solutions.

EPA has essentially been ineffective in its strategy of allocating research
funds for nonpoint pollution control by requiring answers in three years.  No
defensible data or research finding can be produced in such a short time span
when incorporating such unpredictable variables as rainfall quantity and
intensity.  A more realistic time frame would attract new research urgently
needed and produce reputable findings.

Technical Assistance.  Technical solutions to preventing nonpoint pollution
are directly related to land and water management systems.  Both are directly
affected by soil and water conservation practices.  The Soil Conservation
Service, state conservation agencies, and conservation districts have
specialized in planning and applying these practices in cooperation with
landowners for about 50 years.  These agencies can continue to provide the
technical assistance needed to deal with nonpoint pollution.  These same
agencies must exercise caution in using the Universal Soil Loss Equation as a
pollution abatement equation.  The equation must be used only as a measurement
of soil loss from a given site.  Delivery of such sediment to a given stream
or body of water requires the development of a delivery formula or model.

Local institutions must provide technical and administrative expertise above
and beyond that received from state and federal institutions.  Acceptable
implementation rates for pollution abatement systems dictate that local
institutional capabilities be strengthened.

Financial Assistance.  Financial support is imperative for the successful
implementation of a nonpoint-source pollution abatement program.  All levels
of government and the polluter must share the financial burden for pollution
control.

Federal funding to help solve nonpoint pollution problems must be commensurate
with the point-source pollution abatement funding relative to public benefits.
To date, spending for nonpoint pollution has been insignificant from a federal
point of view.  Institutionally, federal funds for nonpoint pollution
abatement should represent a realistic portion of the federal funding of
pollution abatement efforts by the congressional public works committees to
the U.S. Environmental Protection Agency.  Such funds should not be provided
through the congressional agricultural committees to agricultural agencies.
It is logical and imperative that such funds be viewed by the public as water
quality funds benefiting the public and not just agriculture as another
subsidy.  Federal funds need to be allocated to state and local institutions
in the form of grants for implementation and subsidies.  Experience dictates
that this procedure will not only be more productive but more efficient per
dollar spent.  Any other system of federal financial support will be less
effective and produce financial waste.

State funding should supplement federal funding to assure that local
implementing institutions are adequately staffed and subsidy needs in a state
or area are sufficiently met.  State funds can be most effective if they are
used by local institutions responsible for implementation.
                                    K-IX-8

-------
Local funding should provide for basic facilities and support personnel.
Special needs and cooperative efforts can best be funded locally to assure
jurisdictional control and direction.

Subsidies to private landowners from any institutional or governmental level
should be restricted to those practices required of the landowner that are not
cost-effective to the individual but are required to achieve public benefits.
Any subsidized landowner practice must be accompanied by a mandatory
maintenance provision, and there must be a means for recovering public funds
if a practice is not maintained.  This mandatory maintenance provision is now
used by many state and local institutions, but it has never captured the
support of federal agricultural institutions administering subsidy programs
for erosion control.  The public has a right to require this protection of its
investment.

Other Assistance and Structural Issues.  Other assistance, such as legal
support and private-sector support, must be built into institutional
arrangements on the basis of various state and local jurisdictions and
capabilities.

The most successful institutional arrangements for nonpoint pollution
abatement must be acceptable to the local implementing institutions first,
then to state and federal institutions, in that order.  The nature and
complexity of the problem dictates these priorities of acceptance.

Institutions implementing nonpoint pollution programs must accept the fact
that nonpoint pollution cannot and will not be eliminated, but reduced by the
best of institutional arrangements.  Nature by itself produces significant
nonpoint pollutants through natural processes.

The number of polluters essentially equals the number of landowners
nationwide.  This requires a strong, persuasive effort prior to enforcement
action.  It also requires a graduation of compliance, dealing with flagrant
violators first, which will generate rapid public support.

Progress toward nonpoint-source program goals and clean water objectives must
and will move at a somewhat slower pace than has been the case in point-source
programs.

A benefit secondary to the immediate purpose of pollution abatement, but
possibly an even greater benefit in the long run, is that pollution abatement
practices will equally benefit the protection of the food-producing base of
our nation.  These co-benefits should generate public support politically and
financially if properly implemented through realistic institutional
arrangements.
                                    K-IX-9

-------
                             WHO SHOULD PAY FOR
                       NONPOINT POLLUTION ABATEMENT?

                                Lawrence W. Libby
                  Professor, Department of Agricultural Economics
                             Michigan State University
                               East Lansing, Michigan
                                    ABSTRACT

Nonpoint pollution  is an  expensive social problem,  but  abatement of  nonpoint  costs
society as well.  Nonpoint policies involve implicit comparison of these two categories of
cost and their distribution among policy participants.  This paper assembles  economic
evidence of costs of nonpoint and cost of abatement, discusses distribution, and suggests
policy direction.

Emerging policy in  nonpoint abatement  must place greater emphasis on national efforts
rather than state or local.  Because those who cause nonpoint  are usually separated by
both time  and space from those  damaged by pollution, institutions  that  bring  these
interests together are essential.  Further, a "polluter pays" philosophy will become more
prominent  in  nonpoint  policy.  Society will insist that the  polluter bear a  greater
obligation for the impacts. Economic incentives that redistribute cost (a tax, control, or
cross compliance)  or  benefit (benefit share)  will  improve  chances  for design of
acceptable, workable policy.

Key Words: Nonpoint pollution policy, economics, institutions, distribution of cost.
                                   K-X-1

-------
                                 Who Should Pay for
                            Nonpoint Pollution Abatement

                                         by
                                Lawrence W. Libby*
Introduction
      Nonpoint pollution is an expensive social problem.  It imposes enormous costs on
water users, a cost recently estimated to total  about $6 billion a year (Clark,  1985a).
Sediment and  pollutants ruin the  ecosystem habitat for  fish,  waterfowl  and other
organisms that are valued by  people for various  reasons.  Clogged harbors and channels
raise the cost  of water  transport.   Nutrient  enriched lakes and  streams are  less
attractive for recreation or as part of a living environment; damaged ground and surface
water must  be treated before use in municipal systems.  The costs and consequences of
nonpoint  have  been  discussed  and  documented throughout  this  conference.   But
abatement of nonpoint costs society also.  Nonpoint policies involve implicit comparison
of this cost  of abatement with the cost of not abating and the distribution of the two
categories of cost.  Nonpoint  is the unfortunate yet predictable side effect of activities
that are generally worthwhile — food  production, construction,  and others.  Abatement,
then, may mean less of that valued activity, or can be direct financial outlay to reduce
the water runoff.

      This paper addresses the  economics of  abatement  — both  the efficiency and
distributional implications of those actions taken to reduce nonpoint damage.Primary
emphasis is  on  agricultural  nonpoint pollution.  Effficiency  consideration basically
establish the technical parameters, the feasible set within an abatement strategy  may be
selected. Distributional impacts influence choice by discribing  who pays and who  gains
from the options available.  The  most  economically efficient abatement  technique,
whatever it  may be, has important distributional consequences. Thus it is just one among
many abatement options, evaluated by political actors in terms of what it gives them at
what price.

      Purpose of  this paper is to assemble the economic evidence being generated and
draw conclusions that may  be useful for this conference and for any subsequent policy
recommendations that might be  developed. A further purpose is to help set the stage for
more specific technical sessions to follow.

      First, a series of assertions about "the nature of things" in  the political economy of
nonpoint will establish context.

      1.   "Efficient" solution of  the  nonpoint  problem  implies the  greatest possible
abatement for the money spent, achievement of a given level of  abatement at  least
possible cost and/or most importantly, comparing the cost of an additional increment of
abatement to the benefits from achieving that increment.  The  assumption is  that  when
     Professor of Agricultural Economics,  Michigan State  University.   Paper prepared
for the Symposium on Nonpoint Pollution Abatement, Marquette University, Milwaukee,
WI, April 23-25, 1985.

                                     K-X-2

-------
an increment costs more than it  gains, whoever gains or loses, society has acquired
enough reduction of nonpoint pollution.  Economists love to repeat this basic rule, but it
has real  meaning in the  social process of  allocating public effort among many valued
services, of which clean water is just one.

     2.  Benefit, cost and therefore efficiency in  resource use are defined within the
structure of  property rights  that  expand  or  constrain  the rights of individual water
users.  These rights are established and reinforced by public authority.  Thus any solution
to pollution is valid or efficient only to the extent granted by publicly acknowledged and
protected property rights. There is nothing meta-physical about economic efficiency, in
abatement or any other production process. It is a fabrication, a result of  transactions
among actors granted the priviledge of commerce,  with  publicly protected  rights to
impose a price  for a  productive factor.  Those rights could  change,  thus  altering the
efficient solution.

     3.  As with most forms of pollution, nonpoint agricultural  pollution involves impacts
that are  separated in  time and space from those causing the problem.  Pollution is an
additional output of a production process generating a "good" of  some kind.   Property
rights protect the producer, limiting his liability for the cost of that undesireable output
in the production decisions that  determine quantity  and  form of output.   Thus,  the
benefits  of abating nonpoint  are distributed differently from  the costs  of abatement.
While those benefiting from abatement might  be willing the  "bribe" those causing the
problem  through  some sort  of  compensation program, the  costs of  organizing  the
transaction would be formidable.

     4.   Participants in the policy process  form  positions on  options  based on  a
comparison of separable  benefits to  separable  costs.  There is no inclination to pay or
bear inconvenience if that action produces no obvious result  of consequence to  the
actor.  Some people will bear personal cost to create benefits for others out of  a sense of
community or altruism.   But  it is  risky to construct  nonpoint policy on the assumption
that that will occur.

     5.  In nonpoint policy as in  other areas, good policy  is acceptable policy.  Policy
changes that  may be technically correct but require too great a sacrifice of other valued
services  are irrelevant.   "Non-degredation  by  1985" is  and  has been such an  irrelevant
policy  goal.   Changes occur  in  incremental adjustments  as  competitors compare  the
relevant  consequences of  options and bargain on that basis. Policy goals and objectives
may be  stated  and used  as targets  so long a participants do not expect to  actually
achieve them.  Goals and objectives  evolve in policy  bargaining.  They must be flexible
enough to accomodate  reality (Lindblom, 1979).

     In  nonpoint  abatement,  the   efficiency  questions  involve   comparing   the
consequences of pollution with the  consequences of abatement. Policy  participants must
understand the cost of failing to reduce erosion and compare to the cost of  cleaning up,
to arrive at the marginal benefit/marginal cost comparisons discussed above.


The Evidence on Cost;  A  Brief Review

     Various empirical studies have sought to establish  the magnitudes of the  two types
of cost — pollution and  abatement.   There is further  development of methodology to
improve  the  comprehensiveness  and reliability of  these estimates.   Shortage of  site
specific water quality data continues to be a problem in seeking economic estimates of
damage.

                                      K-X-3

-------
      Cost of Nonpoint. The cost of nonpoint pollution represents a reasonable starting
 point for estimating benefits of abatement.  Some non-user benefits might be omitted
 from the cost figures, thus cost of pollution may understate the actual benefits from
 abatement.  But certain instream costs to the biological ecosystem are also omitted.
 There is  an implicit property  rights  issue  imbedded  in  this cost/benefit template  of
 nonpoint pollution.  Those hurt by pollution are basically trying to defend something they
 already have ~ a clean water supply.  The polluter, however, implicitly owns the right to
 convert that clean water into the combination of dirty water and a product,both of which
 are  outputs of the production process in question.  The payment that  the downstream
 user would be willing to accept to  part with clean water would likely be greater than
 what he would  be willing to pay to acquire a new supply of  clean water (Bishop and
 Heberlein, 1980).  Initial allocation  of property rights, therefore, will influence benefit
 estimates  of  nonpoint abatement.   Magnitude  of  that difference would be a  useful
 empirical question to pursue.

      The  most definitive  national  estimates of nonpoint costs have  been  generated
 recently by  the Conservation Foundation(Clark, 1985b).   Authors assemble  the most
 useful empirical literature on  the  subject, impose a  few defensible assumptions, and
 develop some estimates.  While more  specific case  studies within particular watersheds
 could improve the quality  of the national figures,  the CF estimates  are a significant
 contribution to  policy development  in  the area.  Pollutants include sediment, nutrients
 and  chemicals.   Instream  damages are  defined to  include impacts  on the aquatic
 environment,  reduction in  recreation benefits  from  fishing and  other  water sports,
 reduced  water storage and holding  capacity of  reservoirs and  lakes, increased cost  of
 maintaining harbors and water ways, and loss of certain non-user values associated with
 clean water.  Their estimate of these instream damages is $4.3 billion per year, of which
 agriculture causes $1.6 billion.  Off-stream  damages result from increased flooding
 associated with suspended  sediment, less  efficient  water conveyance systems (canals),
 and more expensive water treatment systems for power generation. The total off-stream
 estimate  is $1.9  billion of which agriculture  accounts for $660  million.   These are
 acknowledged to be "order of magnitude" estimates (Clark, 1985a, p. 22).

      A recent case study of the Obion-Forked Deer River Basin in Tennessee estimated
 off site damagaes from  erosion totalling $74 million per year plus losses associated with
 flooding of productive hardwood forests (USDA, 1980).  Impact of declining water quality
 on value of waterside property along St. Albans Bay on Lake Champlian was estimated
 using an hedonic model.  The water quality variable constituted approximately 20%  of
 property value or $4,500 per property difference attributed to  quality  of  the adjacent
 waterbody. Total damage experienced by owners of 430 single family residences was $2
 million (Young and Teti, 1984).  Other  case studies are needed to get a more accurate fix
 on economic burdens associated  with nonpoint pollution.

      Benefits of Abatement.  Another perspective  on the cost  of pollution side of the
 nonpoint policy question may be gained by estimating benefits of cleaning up the water.
 Benefit is  more than just damage avoided.  It  also includes the liklihood  of  continued
 availability of clean water  and the  impact of that  liklihood  on willingness to invest  in
 waterside property or other  services  that require clean water.

      The  Economic  Research  Service of USDA  is  putting considerable  effort into
 estimating benefits  of abatement in  the various rural clean  water projects (RCWP)
around the country.  Work is still underway, but a few results are available.  Improving
the quality of water in St.  Albans Bay in the northeastern part of Lake Champlain  in
Vermont would  produce annual  benefits estimated at $230,300 through a contingent
valuation technique  and about twice that using a travel cost approach (Ribaudo,  1984).

                                      K-X-4

-------
The travel cost model involved estimating the cost that a recreator has been willing to
bear to  enjoy  the bay for various recreation experiences  and the amount of travel
expense that recreator would pay if the  bay  were cleaner.  Number and costs of  trips
from various origins in the region were estimated. Relationship between the willingness
to bear travel  cost and distance from  St. Albans, availability of alternative sites and
income levels was calculated for current water quality and for the hypothetical situation
of a cleaner bay.  Mean  difference in  willingess to  pay for recreation attributable to
improved  water  quality was $123  per  recreator,  for  a total recreation benefit of
$530,700.  The  contingent valuation model involved interviews of prospective recreators
who  were asked  to rate several  recreation opportunities  involving  St. Albans Bay and
then re-rate those options with the possibility of a cleaner bay. These indifference maps
were then converted to monetary benefits associated  with improved water quality in the
bay.

     Benefit estimates of several other clean water projects will be  published soon.
Each is unique, with  land use, soil, and  water characteristics that differ  significantly
from site to  site.   Generalization is impossible.  Yet these point estimates  can improve
chances for public agencies  and  various  political participants involved in  clean water
debates to draw conclusions that are economically sound.  With reasonable estimates of
abatement costs and their distribution, one may draw a conclusion about the efficacy of
a particular abatement program.

     Cost of Abatement.  The greatest amount of published work in  the economics of
agricultural nonpoint pollution deals with the costs that selected management techniques
could impose on the  farmer.   These studies generally assume  initial distribution of
property rights that gives the farmer the right to decide for or against a  reduction in
run-off.    In  a  1976  linear  programming  analysis,  Kasal  examined farm  income
consequences of  imposing  limits  on fertilizer use and soil loss as measures to  reduce
nonpoint pollution. For the sample farms selected and alternative measures imposed,
nonpoint abatement reduced farm incomes by from 10 to 36 percent (Kasal, 1976). A
national LP model  run by Wade and Heady made similar kinds of estimates  of increased
production costs  associated with nonpoint measures,  within the overall requirement of
meeting projected food demands. The minimum sediment solution  to the production cost
minimization model was 42% higher than the unrestricted solution.  In both cases there
was considerable  cropland adjustment within the model to retain sufficient production to
meet output requirements (Wade and Heady, 1977).

     Alt, e£ al  (1979) examined the farm  level costs implicit  in reducing sediment
delivery to a specific  reservoir in central Iowa from the Iowa River watershed. A 10 ton
acre limit on gross erosion from farms in that watershed  would increase production costs
17%  while reducing sediment delivery by 91%.  White and Partenheimer (1979) examined
income consequences  of  erosion reduction  plans for a  sample of  Pennsylvania dairy
farms.  Impacts varied among the specific farms studied, with two farms experiencing
increased  revenues from the  recommend  soil  conserving  plan,  four with  income
reductions of less than 5% and six with reduction from 7% to 30%. A  no-till option was
introduced with the result that all but one farm showed increased returns.  The net
on-farm/off-farm  economic  consequences of  erosion reduction were examined for  a
particular river basin  in northeast Texas.  The general conclusion reached in  this analysis
was  that the positive  on-farm impacts  of  soil conservation overshadowed whatever
revenue  reduction  might  be associated  with complete  attention  to reducing off-farm
damages.  Thus, authors conclude, it is in the farmer's interest to conserve soil  and no
regulation or subsidy  is  warrented  assuming  that  farmers are economically rational
(Reneau and Taylor, 1979).

                                      K-X-5

-------
      Results  of  these and  many  other case studies of the economic costs of various
erosion control techniques have provided important analytical back-up for recommending
best management practices in specific state and local water quality plans.  The whole
approach to nonpoint reduction has been  local solutions to local problems, with "best
practices" a function of the  physical, economic and institutional circumstances prevailing
in that area;   A systematic  approach for considering the various  factors involved in
selecting best management  practices for a  particular political/economic/hydrologic
sub-region has been suggested by Bailey and Waddell (1979).  The general policy goal has
been to establish an abatement program that avoids major cost impact on local farmers.
Procudures  for considering  the most important firm  level costs have been recommended
by USDA and EPA (1975) for the development  of those BMP's (1*).  Inter-regional  and
national  consequences  of  these  strategies  have  been estimated  also.   The  more
aggregated  the  analysis or modeling becomes,  the  more imprecise and unreliable  the
estimates because of obvious differences among areas.

      A significant and widely acknowledged gap in the evidence on alternative nonpoint
strategies  is  the   link  between  technique  and  water   quality   improvement
(Christensen,  1983).  We know that nonpoint costs money; we know that abatement costs
money; we even know that  the two costs are distributed differently.  But we need much
better evidence  on the  aggregate water  quality results of farm  level practices that
reduce run  off,  so that  the  estimates of  economic  benefit of those actions may be
improved.

      Reduced tillage alternatives show particular promise  since they can, under some
circumstances, lead to increased net farm revenue while reducing run-off. Pope, et a[
(1983) concluded from their study of Iowa agriculture that conservation  tillage systems
have less negative impact on  farm income than is true for conventional systems.  Study
of alternative tillage  systems on Ohio  farms found  that no-till options  that reduced
erosion to T  or  less  also produced higher net revenue than the base solution of  fall
plowing  on  soybeans  (USDA,  1983).  In his study of  cash  grain farms  in the Jackson
Purchase Area of Kentucky, Kugler found that  for nearly all farms there  was a tillage
conversion that  would increase net returns while reducing erosion (1984).  Black, et al
(1984) concluded that since yield is  not  affected by  tillage system for the sample of
Michigan farms  studied in  the Saginaw Bay watershed, the  net  revenue advantage for
conservation tillage comes from reduced machinery and labor costs. Determination as to
whether or  not reduced tillage makes sense as a best management practice for nonpoint
abatement  depends on  the net effect of  that  practice on the environment.   While
conservation  tillage often  reduces  soil movement  and runoff,  it  may also  result in
increased chemical concentration in  the run-off water that does leave the field.  The
plant residue  left in  the  field as  part of the reduced  tillage system tends to increase
insect problems  and  reduce the effectiveness of some herbicides (Baker, and Laflen,
1984).  Any calculation of the cost of nonpoint abatement using reduced tillage or  any
other erosion  reducing technique must consider offsite damage as well as income impacts
for the farmer. Conservation tillage may reduce erosion at little direct economic cost to
the farmer but if it also increases chemical contamination of rivers and streams because
of  increased  use of pesticides,  its  viability  as  a best  management  practice is
questionable.

     The real difficulty for  an  individual  or  government  in  reaching an  efficiency
decision  on nonpoint  investment  is that information comes in  fragments.   Cost  and
benefit are not felt at the same time or place.  At the  farm level, an action that reduces
nonpoint may have advantages in net  revenue.   On  the other  hand, a practice  that
reduces  erosion  and  increases the  farmer's revenue, may  worsen the water quality
problem.  As  analysts we try  to  isolate factors, examine them one at a time  for their

                                     K-X-6

-------
positive and negative consequences. In fact, however, the farmer makes decisions within
a complex of incentives,  risks, and options.  It may be that an incentive necessary to
attract socially rational investment in nonpoint abatement should be artifically linked to
the action that would reduce runoff.  That is the general idea behind cross-compliance in
soil conservation policy.  As  long  as the positve  incentive is  greater  than any  income
burden from  the resulting nonpoint technique, it can be a rational choice for the farmer
(Dinehart and Libby,  1981).   So long as  the public cost of the incentive,  including
whatever unintended side impacts it may  cause,  in less than the incremental  cost of
pollution it can be an improvement.

     Success of alternative  policies to reduce nonpoint pollution will depend primarily on
the distribution of both the  cost of pollution and the cost of abatement. In policy we are
more concerned with who pays than with the overall size of the  bill.


Who Pays?

     Distributing the burden for  reducing agricultural nonpoint pollution, assuming there
is general agreement that nonpoint should  be reduced, is the central policy question in
this area.  The prevailing theories  are 1) those who benefit from abatement should pay
for it, and 2)  those who cause the problem should pay for its solution.  The former implies
that the  right to permit run-off remains with the polluter; those who want less pollution
because  they currently experience a pollution cost, must pay  for abatement.   The
problem, then, is to decide who  benefits and levy a tax accordingly.  In large measure
this is the current  system.  Farmers and other  polluters retain the  right to pollute;
abatement is voluntary; cost sharing and  technical assistance are paid  for by taxpayers.
State  level cost sharing  programs exist  in  many  states, implying that benefits are
received by all state taxpayers.   As of  1983, sixteen states have their own cost share
programs (Braden and  Uchtmann, 1984).   Federal  cost sharing, through ASCS of the US
Department of Agriculture, implies that all US taxpayers would benefit from reduced
pollution  and should  therefore  pay for  it.   In  practice,  state  programs  actually
supplement this federal support.

     The latter approach to  distributing the cost of pollution implies  that the polluter
has full liability for his actions and that  any cost of abatement must be included in the
cost of production for  the commodity or service that creates to pollution.  Thus far, the
only method  for implementing this approach to  distributing abatement  cost  has been
through regulation.  County soil conservation districts in most  states have the authority
to  impose regulation  against   excessive  erosion,  though few  have exercised that
authority.  Five states have sediment control ordinances that include agriculture.  Iowa
has the most aggressive and comprehensive regulations against excess erosion.  Illinois
has a newly enacted law designed to "jawbone" farmers toward acceptable erosion levels
by the year 2000.  Pennsylvania has an innovative  permit system functioning through the
state's Clean  Stream Law (Holmes,  1979, p. 63-93).

     Other mandatory approaches being examined for soil conservation  programs include
various cross compliance measures that would require soil conservation as pre-condition
for  eligibililty for price supports and other income support programs.  While these have
not really been  designed as nonpoint abatement measures, they could  help meet water
quality goals.

     The matter of "who  should  pay" for  nonpoint is  obviously a  matter of opinion.
Conclusions presented here  are based not  on perception of right and wrong (though some
element  of that may creep into  the discussion) but on a judgement  of  what is likely to

                                      K-X-7

-------
happen.  The "benefits received" approach is well received in the agricultural community
including, incidentally, most of the Land Grant system.  The obvious problem is, however,
that it has not worked. Policy development for the next decade will shift more generally
toward the "polluter pays" approach reflecting concern for cost effectiveness in nonpoint
and other natural resource policy  as well as a growing impatience among the pollution
control professionals about progress toward  reducing  nonpoint.  The magnitude  of the
estimated national cost of nonpoint pollution and agriculture's portion of that total has
the attention of the Soil Conservation Service.  Estimated  annual costs imposed by
agricultural nonpoint pollution are several orders of magnitude  higher than annual cost of
soil productivity  lost  to  erosion.    SCS  must  give greater  attention  to off-site
consequences of erosion in these days of  national budget stringency  for agriculture
natural resources and for SCS in particular.

      Policy Conclusions. The following conclusions on the general trends in policy and
cost distribution are based on a review of the economic evidence on nonpoint summarized
briefly above,  review of policy literature concerned  both  with soil conservation and
water pollution, and a general "testing of the wind."  Following these brief conclusion, is
a list of suggested innovations for policy in this area.

      1.  Abatement of nonpoint must be a national program,  not just the sum of state
programs.  Despite the fact that 208 water  quality planning, the Model Implementation
Program and the Rural Clean Water Program emphasize  state action,  and the Reagan
administration  prefers decentralization, water pollution is just as is national as national
defense. The implicit assumption that beneficiaries of abatement are only the first line
pollutees who directly feel  the results of dirty water,  and that they reside in the same
state as those causing  the problem cannot  be sustained.  In  fact, benefits of a clean
water supply go well beyond  immediate users.  They include  various non-user benefits
from  continious availability of water when and if it might  be used. They include various
multiplier impacts on the overall  economic environment of a region where the  farms,
cities and countryside become attractive places to live and work.  Clean water is part of
a broad  sense  of  well  being or security  that will encourage investment and general
popular support for change.

      A recent  article in the 3ournal of Soil and Water Conservation was entitled "Saving
the Chesapeake:  Maryland's Agricultural Education Program," (Magette, 1985). Even  if
the title were  changed to refer  to "Maryland's Ironclad Regulation Against Erosion,"  it
would be wishful thinking.  There is just no way that the problems  of the Chesapeake Bay
can be handled separately by  Pennsylvania  and Maryland.   Even when the pollution
problem is reasonably well confined as in Saginaw Bay, Michigan or Green Bay, Wisconsin
benefits  of  abatement extend  well  beyond.  There  has been important institutional
innovation in states, and that must be supported. The Wisconsin nonpoint program is an
exceptional  example  of  a  balanced and integrated  program with all components from
problem identification to implementation (Konrad, et al,  1985).  RCWP's have become
testing grounds for coordination among federal and  state agencies.  They have been
strongly endorsed  by EPA and UDSA  (Groszyk,  1979; Unger,  1979).  But beyond these
intermediate purposes,  reducing nonpoint to  acceptable levels will  require federal
action.  It is unreasonable to  expect an  agricultural state to enact abatement measures
sufficient to produce  adequate  water  quality benefits for downstream users in  other
states.  To a large extent we have allowed the traditions of state and local prerogative
on land use management to run  the nonpoint program. We begin with the irrevocable
presumption that locals know best when it comes to land use. Perhaps that is true, but  it
will not  solve  the nonpoint problem.   If  land  use  controls  are the only solution to
nonpoint, and land  control must  be a  local government function, then nonpoint will be
with us for a long time.

                                      K-X-8

-------
     A national  program for nonpoint abatement means that the cost will be  shared
among payers of national income tax.  State programs may supplement but never replace
the national focus.  Taxpayers in agricultural states will  pay more than taxpayers
elsewhere since farmers in those states will likely have sufficient political clout  toshift
the abatement cost and get additional cost sharing programs installed.

     2.  Farmers and others causing nonpoint pollution will pay an increasing portion of
the abatement cost.  Most  of this cost will be indirect, in the form  of fewer  property
rights to use land in ways that cause nonpoint pollution and through any reduced revenue
associated with less erosive farming systems that might reduce output. There  is simply
no  possible  way  that  continued  reliance on  technical assistance,  cost sharing and
voluntary action can satisfactorily  cope with the nonpoint problem.  Reduced tillage has
really been  the only hope for this softer  approach and recent evidence on the  pollutant
concentration in  run-off raises serious question about the viability  of these techniques.
This should not suggest that the voluntary approach with or without the other elements is
inadequate in every case.   There have  been important successes.  But in a national
nonpoint program, these measures are simply inadequate.

     There  is nothing  earthshaking about this conclusion.  Resource policy professionals
have known  for  somtime  that only by  changes in farmer  behavior will agricultural
nonpoint be reduced, and current incentives do not  favor the necessary changes.  The
RCWP era  in water quality  policy has been one of support building and institutional
design  within  the prevailing balance of  power  and  responsibility  among  levels  of
government.  It has been necessary.  But there will be pronounced shift  in philosophy
toward the "polluter pays" approach. Those who stand to gain from abatement, a broadly
defined community of water users, will be unwilling to leave all of the  land use discretion
in the  hands of those  who cause the problem.  Epp  and Shortle suggest that given this
inevitable policy shift,  the water quality agencies and  research community should focus
on design of effective and economically efficient mandatory  programs (Epp and Shortle,
1985, p.111). That is sound advice.

     Distribution of  burden  within  agriculture  deserves   policy attention  as well.
Farmers  on  erosive land will find it particularly painful to meet mandatory limits. They
suffer  the  initial disadvantage of  less  productive  land and then the  higher cost  of
compliance.    The  inevitable  result  will  be  further  concentration  of  agricultural
production  on the most responsive and easily protected lands  of the country.  Additional
people and areas will be displaced  from agriculture.  There must  be special attention to
these poeple to facilitate transition and limit the hardship involved.  But attempting to
mask the inherent disadvantage of  farming these lands  through special credit or disaster
payments would  be an expensive  and  apparently  inappropriate policy  response.   With
mandated erosion and  run-off standards,  the inherent advantage of  some lands would
likely be reflected in land price. Technology may reduce the differences between erosive
and non-erosive lands, however.

Directions for Policy Innovation

     The major needs in nonpoint involve, first, new organizational strategies for getting
those who cause the problem  together with those who gain from its  reduction.  As noted,
costs of  pollution and  costs  abatement are distributed differently. Only by creating a
political  environment conducive to negotiations between these categories of participants
can  lasting  policy  change be accomplished.  The  second  major need  is  for  financing
schemes  that  take advantage of  the economic facts of  pollution  and abatement.
Economists  have long been reviled for seeking social purpose through inherent selfishness
of the individual. That  is an unfair  assertion, of course.  The more positive perspective is

                                      K-X-9

-------
to facilitate expression of the true consequences of alternative actions so that private
choices will yield results that are socially responsible.

      1.  The  first suggestion is  not new at all but  entails giving  new  life and new-
missions to an old institution ~ the river basin commission. There is logic in hydrologic
units!  As water resource professionals have known for many years, the river basin unit
can  overcome  much of the transaction cost associated with bargaining between gainers
and  losers from a pattern of water use.  This capacity can be particularly important in
nonpoint  abatement where  the  source  of the problem is  diffuse,  benefits  widely
distributed in small increments. Some of these inter-state linkages have been established
for the Chesapeake.  Experience in the Deleware and Susquehanna River Basins has been
well  documented  (Libby,  1970).   Bay   area  regional  commissions  with multi-state
memberships could be established or focused more clearly on nonpoint problems. Purpose
of the  water  shed, or more generally  "problem-shed", is to  establish  the setting for
bargaining over the terms of nonpoint policy.  There should be no claims of designing the
optimal water management program.  Instead,  interested parties could iterate toward
acceptable levels  of  pollution and  abatement, and  a  distribution  of  burden that is
tolerable.  With focus  clearly on nonpoint,  voting and membership rules would have to
encompass the essential parties involved.  A different organization might be appropriate
for other resource problems in the basin.

      Hydrologic units are  awkward to  adminster.   They cross many  jurisdictional
boundaries  and  lack  political legitimacy.    The administrative  overhead can  be
substantial.   The  problems  are  not much  different  from  those  of multidisciplinary
research is an acedmic institution.  But  there can be no lasting reduction of nonpoint
pollution  without  some  mechanism  of   forcing confrontation among  the interested
parties.   Provision  must be made  for beneficiaries outside the basin and some basic
understanding  of problem source and benefit of improvement. Water quality standards
would help establish targets.  Even with mandatory controls  on erosion, the mechanism
for bargaining would be valuable.

      2.  Within a decision or bargaining  unit such  as a  river basin,  a  mechanism  for
redistributing  the  benefits  of abatement  could   improve  chances for  incremental
resolution  of the nonpoint problem.  Such an innovation has been analyzed by Park and
Shabman (1981) for the Occoquan River Basin in Virginia. Under  such a scheme, those
who gain from abatement could  spread  the benefits  sufficiently  to generate support
necessary  for  a compensation scheme  that will  bribe  the  polluter  not to pollute.
Alternatively,  this system might  produce the plurality to enact  mandatory controls.
Areas receiving major measurable  benefit from  abatement could compensate areas that
would benefit less to help finance installation of BMP's with the basin commission acting
as banker.  As long  as net benefits to all  parties benefits  are positive,  it is in their
interest to share to build support.  The basin commission's role is to reduce transaction
cost associated with bargaining both among beneficiaries and between those who  cause
and those hurt  by nonpoint pollution.

     The Chesapeake Bay situation would seem  to  lend itself to the  type  of innovation
suggested by Park and Shabman. Any such system would require improved information on
water quality benefits and impacts of alternative abatement techniques.

     3.  Another device for employing  an economic  incentive to encourage  private
actions  with a social purpose  is installation of a tax in pollution.   Economists like this
idea particularly well because it forces the producer  to internalize social costs  of  the
production process.  The  tax also  creates  the incentive to reduce pollution to avoid the
tax.  While there apparently is no real experience with  a tax on nonpoint pollution, Seitz

                                      K-X-10

-------
ejt al used a linear programming model of hypothetical farms representing western Illinois
agriculture to  conclude that a soil  loss tax would  be more  efficient than an erosion
restriction in reducing nonpoint (Seitz, et al, 1979, p. 375-376). Under a tax scheme, the
farmer  would  seek  the least  costly means  for  reducing run-off, unique  to  the land
conditions on that farm.  The  erosion standard or restriction, on the other  hand,  might
impose a more costly technique than is necessary.
     The obvious difficulty with  tax schemes is absense of specific  information  on
performance  of practices,  cost of  pollution  and other essential variables  for a  given
farm. It may be that trial and error would be a reasonable research design in this area of
building institutions. The theory of tax schemes is well developed, but implementation is
lacking.  Selected pilot studies could be a reasonable use of scarce research funds.

     4.  Marketable or tradeable rights to pollute constitute a third type of innovation
for marshalling the forces of greed in the interest of society.  The sediment and  waste
assimilative capacity of a stream or lake  might be allocated among farmers on the basis
of some readily determinable index such as frontage on water course or acreage adjusted
by an indicator of accessibility to water.  Then the farmer would have the right  to sell or
trade those rights when their value exceeds the cost of reducing run-off. Implementation
of such a scheme would  require monitering  to determine compliance for  a particular
farm.  There might  also be a  problem of localized  pollution, where farmers decide to
acquire  pollution  rights and let the pollution occur. Those farming areas that can reduce
run-off most readily will have much cleaner water. The average , or overall assimilative
capacity, could hide extreme  situations  that impose  local hardship.   There  would  be
overhead cost involved for the basin commission or other managing unit in defining  terms
of trade and upholding the result.

     After thorough review of  the economics and institutional  literature related  to
nonpoint pollution one could easily conclude that  there is really very little  that is new.
There are few  revolutionary new ideas. In fact,  there is  a fair amount of repetition at
the conceptual level.   What is lacking is experience,  empirical observation of various
means of coping with disparate patterns of distribution of the cost of pollution and cost
of abatement.  There are limits on acceptable means for redistributing those  burdens, but
the limits  are  less severe  than traditionally assumed.  Acceptability is a function  of
prevailing attitudes about who can or should shoulder more of the burden for meeting a
widely held social objective for clean water.
                                      K-X-11

-------
                                   References
 1.  Alt, K.  F.,  3.  A.  Miranowski and  E. O.  Heady.   1979.   "Social Costs  and
               Effecitveness of  Alternative  Nonpoint Pollution  Control  Practices",
               Best Management Practices for Agriculture and Silviculture, Ann Arbor,
               Michigan:  Ann Arbor Science Publishers, Chapter 23, p. 321-328.

 2.  Bailey,  G. W. and  T.  E. Waddell.   1979.   "Best Management Practices for
               Agriculture  and   Silviculture:    An  Intergrated  Overview",  Best
               Management  Practices  for  Agriculture and Silviculture, Ann Arbor,
               Michigan:  Ann Arbor Science Publishers, p. 33-56.

 3.  Baker, 3. and 3. Laflen.   1983.   "Water Quality Consequences of Conservation
               Tillage". 3ournal of Soil and Water Conservation. 28(3);  p. 186-193.

 ft.  Black, 3. R., H. Muhtar, 3.  Posselius and  3. Schwab.  1984.   "Results  of an
               Economic  Comparison  of  Conventional and  Conservation Tillage
               Systems in the Southeastern Saginaw Bay  Coastal Drainage Basin",
               Proceedings  for a  System  Approach  to Conservation Tillage, East
               Lansing, Michigan:  Michigan State University Cooperative Extension
               Service.

 5.  Bishop, R. C. and T. A.  Heberlein.  1979.  "Measuring the Values of Extra Market
               Goods:   Are  Indirect   Measures  Biased?"   American   3ournal  of
               Agricultural Economics,  61 (1979): 926-930.

 6.  Braden, 3. E. and D. L. Uchtmann.  1984.  "Soil Conservation Programs Adrnidst
               Faltering  Environmental  Commitments  and  the  New Federalism",
               Environmental Affairs Law Review, 10(3):  689-696.

 7.  Christensen,  L.  A.   1983.  "Water Quality:  A Multidisciplinary Perspective",
               Water Resources Research;   Problems and  Potentials  for Agriculture
               and Rural Communities, Ankeny, Iowa:  Soil Conservation Society of
               America,  p. 36-62.

 8.  Clark, E. H.  1985a.  "The Off-Site Set Costs of Soil Erosion", 3ournal of Soil and
               Water Conservation, 40:1, p. 19-22.

 9.  Clark, E. H.,  3. A. Haverkamp and W. Chapman.   1985b.  Eroding  Soils;  The Off-
               Farm Impacts, Washington, DC: The Conservation Foundation.

10.  Dinehart, S. and L. Libby.  1981.  "Cross-Compliance:  Will It Work?  Who Pays?"
               Economics, Ethics,  Ecology;   Roots  of  Productive Conservation,
               Ankeny, Iowa: Soil Conservation Society of America, p.  407-415.

11.  Epp, D. and 3. Shortle.   1985.  "Commentary:   Agricultural  Nonpoint  Pollution
               Control,  Voluntary or   Mandatory?"    3ournal  of Soil  and  Water
               Conservation, 40:1, p. 111-114.

12.  Groszyk,  W.  S.  1979.   "Nonpoint  Source Pollution Control  Strategy",  Best
               Management  Practices  for Agriculture and  Silviculture,  Ann Arbor,
               Michigan:  Ann Arbor Science Publishers, p. 3-10.
                                   K-X-12

-------
13.  Holmes, B. H.  1979. Institutional Basis for Control of Nonpoint Source Pollution,
              Environmental Protection Agency, Washington, DC, WH-554.

14.  Kasal, 3.  1976.  Trade-offs Between Farm  Income and Selected Environmental
              Indicators, A Case Study  of Soil Loss, Fertilizer, and  Land  Use
              Considerations,  Washington,  DC:    Economic  Research  Service,  US
              Department of Agriculture, TB #1550.

15.  Konrad,  3. G., G. S. Baumann and  S.  E. Bergquist.  1985.  "Nonpoint Pollution
              Contgrol:   The  Wisconsin  Experience",  3ournal  of Soil and  Water
              Conservation, 40(1) p. 55-61.

16.  Kugler,  D. E.  1984.  "Variable Cost  Sharing Level Program Implications  for
              Kentucky's 3ackson Purchase  Area:  An Economic and Policy Study of
              Cash Grain and  Production Considering Soil Depletion."  East Lansing,
              Michigan:  Department of Agricultural Economics, Ph.D. Dissertation.

17.  Libby, L.   1970.  "The Political Economy of Water  Management:  Conceptual
              Model and Decision Strategy for the Susquehanna River Basin", Ithaca,
              New York: Cornell University Department of Agricultural Economics,
              Ph.D. Dissertation.

18.  Lindblom,  C.  1979.  "Still Muddling,  Not  Yet Through", Public Administration
              Review, 39:  517-526.

19.  Magette,  W. L.,  R. A.  Weismiller and K. C.  Gugulis.   1985.   "Saving  the
              Chesapeake:  Maryland's Agricultural Education Program",  3ournal of
              Soil and Water Conservation, 40(1) p. 79-81.

20.  Park, W. and L. Shabman.  1981.  Securing Support for Nonpoint Pollution Control
              with  Local  Compensation,   Blacksburg, Virginia:    Virginia  Water
              Resources Center.

21.  Pope,  A.,  S. Bhide and E. Heady.  1983. "Economics of Conservation Tillage in
              Iowa", 3ournal of Soil and Water Conservation, 38(4):  370-373.

22.  Reneau, D. R. and C. R. Taylor.  1979. "An Economic Analysis of Erosion Control
              Options in Texas",  Best  Management  Practices for Agriculture and
              Silviculture, Ann Arbor, Michigan:  Ann Arbor  Science  Publishers, p.
              393-418.

23.  Ribaudo, M., C. E. Young and  D.  Epp.  1984.   Recreation  Benefits From  An
              Improvement in Water  Quality at St. Albans Bay, Vermont, Washington,
              DC: Economic Research Service, US Department of Agriculture.

24.  Seitz,  W. D., C. Osteen  and M. C. Nelson. 1979. "Economic Impacts of Policies to
              Control Erosion  and Sedimentation in  Illinois  and  Other  Corn Belt
              States", Best Management Practices for Agriculture and Silviculture,
              Ann Arbor, Michigan: Ann Arbor Science Publishers, p. 373-382.

25.  Unger, D.  G.  1979. "Improving Water Quality in Agriculture and Silviculture",
              Best Management Practices for Agriculture and Silviculture, Ann Arbor,
              Michigan:  Ann Arbor Science Publishers, p. 11-16.


                                    K-X-13

-------
26.  United States Department of Agriculture and Environmental Protection Agency.
              1975.   Control of Water Pollution from Cropland;   A Manual for
              Guideline  Development,  Washington,   DC:    Agricultural  Research
              Service, Report No. ARS-H-5-1.

27.  United States Department of Agriculture.  1980.  Obion-Forked Deer River Basin,
              Tennessee -- A Summary Report.  Nashville, Tennessee.

28.  United States Department  of  Agriculure.   1983.   Pickaway Soil  and Water
              Conservation District Resources Inventory.

29.  Wade, 3. C. and E. O. Heady.  1977.  "Controlling Nonpoint Sediment Sources with
              Cropland Management:  A National Economic Assessment", American
              Journal of Agriculural Economics, 59(1) p. 13-24.

30.  White, G. B. and  E.  J.  Partenheimer.  1979.   "The  Economic Implications of
              Erosion and Sedimentation  Control  Plans for Selected Pennsylvania
              Dairy  Farms",  Best  Management  Practices  for   Agriculture   and
              Silviculture, Ann  Arbor, Michigan:  Ann Arbor Science Publishers, p.
              3*1-358.

31.  Young, C. E. and F. A. Teti.  198*.  The Influence of Water Quality on the Value
              of Recreational  Properties  Adjacent  to St.  Albans Bay,  Vermont,
              Washington,  DC:     Economic   Research   Service,  United  States
              Department of Agriculture.
                                   K-X-1.4

-------
Panel Reports

-------
                               SUMMARY
                               Panel I
     Planning Processes for Nonpoint Pollution Abatement Programs
PANEL MEMBERS

Moderator:  John Konrad, Chief, Nonpoint Pollution Section, Wisconsin
Department of Natural Resources, Madison, WI

Panelists:

Richard Homer, Research Assistant Professor, University of Washington,
   Seattle, WA
Erhard Joeres, Professor and Chairman, Water Resources Management Program,
   University of Wisconsin, Madison, WI
Walter Knisel, Agricultural Research Service, U.S. Dept. of Agriculture,
   Tifton, GA
Gary Oberts, Environmental Planner, Metropolitan Council,
   St. Paul, MN
Michael Smollen, North Carolina State University at Raleigh,
   Raleigh, NC
Jy Wu, Associate Professor, University of North Carolina,
   Charlotte, NC

Recorder:  David Lee, Marquette University, Milwaukee, WI

INTRODUCTION

Panel discussion and audience participation for Panel  I involved plan-
ning processes for Nonpoint Pollution Abatement programs.  Although well
documented scientific data exist to support the fact that pollution other
than from "end-of-pipe" or point sources exists, defining what is con-
sidered nonpoint pollution and a plan to remedy its consequences is still
largely undefined.   Identifying problems and goals, preparing a technical
and economic attack, and drawing together conclusions  to determine the
effectiveness of the program, are essential to the overall success.  The
engineering, scientific and planning communities are dealing with such
immense pollution problems taking place in broad, complex ecosystems,
that no one person  or approach can completely solve them.  This panel
has quickly concluded that general planning steps (see attached flow
sheet) are necessary to implement a regional nonpoint  pollution abatement
program.   However,  one must be aware that program needs will vary greatly
from one  region to  the next and, therefore, a wide variation in applica-
tion and achievement in different programs can be expected.

FINDINGS

Many extremely important issues were raised concerning proper and effective

                                P-I-1

-------
 program planning and  implementation.  The following issues, aspects and
 conclusions were found to be the most significant:

 1.  Widely accepted means of defining nonpoint pollution problems are
    missing.  Water quality standards, per se, such as are employed in
    point source work, do not exist and probably will not exist for some
    time to come.  Criteria based on receiving water, beneficial use
    preservation, may represent the best strategy but more understand-
    ing of the effects of nonpoint discharge  is required to implement
    it.

 2.  Substantial nonpoint source research findings are already available
    to partially support the development of problem assessment methods,
    although little effort has been made to do so.  Further progress in
    this area will require much additional study.  Success in these forth-
    coming studies will require programs in which great attention has
    been given to definition of general  and specific objectives, and
    prioritization of program elements to achieve the maximum gain in
    knowledge for the resources expected.

 3.  In the nonpoint source water pollution field, unlike many other
    scientific and technical disciplines, there is little standardiza-
    tion of investigation methods.  Greater standardization should be
    promoted to allow more interchange among investigators and reduce
    at least some of the duplication of effort now occurring.

 4.  Dealing with superimposed point and nonpoint loadings to remedy the
    cumulative, concurrent pollution burden in receiving waters must
    be addressed.

 5.  Nonpoint pollution assessment via the use of mathematical  models
    is an indispensible tool, but is rendered useless without well or-
    ganized sampling and data collection efforts.  In addition to the
    site monitoring that takes place preceding model development and cali-
    bration, a continued long-term monitoring effort must be made to
    verify the model and evaluate any control techniques that are im-
    plemented.  The time lag between project implementation and expec-
    ted results is politically unattractive and can be a very sensitive
    issue.

6.  Effective, well-defined legislation  regulations are imperative to
    research project goals,  but the effort is merely rote if proper
    enforcement does not exist.

7.  The effects of nonpoint pollution loadings to groundwater must be
    assessed.   This investigation must be concerned with trade-offs
    between surface water improvement and increased groundwater con-
    tamination resulting from strategies promoting infiltration of non-
    point runoff.   More research is needed to determine the extent of
    chemical  and biological  pollutant reduction processes occurring with-
    in the  reduction processes occurring within the soil  layer.  If these
    mechanisms are overlooked,  erroneous conclusions may be reached
    regarding  hydraulic and pollutant loading rates that can be permit-
    ted on  land.


                                P-I-2

-------
 8.  The impact of nonpoint loadings from the atmosphere appears to be
     an area in which little research exists and which poses a great
     problem to nonpoint water pollution managers.  The political and
     institutional implications are many and could be extremely complex.

 9.  Our society works on an action/reaction basis.  The professional
     community must determine what action(s) must be undertaken in order
     to make the public aware of the extent of nonpoint water pollution
     and to stimulate the reaction that is needed,  i.e., support for
     study and abatement programs.

10.  It appears at this point that the engineering and technological
     hurdles are lower than those encountered by planners.   Identifying
     and researching nonpoint pollution problems in "critical areas" is
     essential, but projects based in these areas are doomed to fail if
     public participation and cooperation are not acquired.

CONCLUSIONS

It is the consensus of the panel that, although researchers  are well aware
of nonpoint pollution problems and many of their implications, very little
planning and program implementation has taken place.  Solving our point
source pollution problems has been difficult but the challenge, politi-
cally, technically, and economically, to improve the overall quality of
our environment has just begun.  Public awareness and cooperation is essen-
tial and the overall challenge will stem more from sociological than tech-
nical feasibility problems.  Essentially, the nature of the  problem stems
from the fact that water moves.  In its movement, it does not recognize
municipal,  state or any other political boundaries.  Truly,  what happens
in Minneapolis affects those living hundreds of miles away in New Orleans.
What programs will be implemented to halt or reduce these problems, who
will pay and how much and what level of political authority  will take
precedence in assuring continued environmental  regulation,  compliance
and enforcement,are all  issues that were examined in our panel discussion
and must continue to be addressed.
                               P-I-3

-------
      »!  Problem Identification
            Goal Definition
           Target Selection
           Plan  Preparation
Data Collection
\
f
         Problem Reevaluation
                                        Management Practice
                                             Evaluation
Institutional and Financial
      Formulation
           Goal Realization
Planning Process Steps For Abatement  Of Nonpoint Pollution,
         From "Fundamentals of Watershed Management",
         Presented by Gary Oberts  at  the International
         Symposium on Lake and Reservoir Management,
         North American Lake Management Society, Oct.16-19,
         1984, McAfee, N.J.
                         P-I-4

-------
              IMPROVEMENT OF ENVIRONMENTAL IMPACT ASSESSMENT
                  OF NONPOINT SOURCES OF WATER POLLUTION

                             Richard R.  Horner
       Research Assistant Professor* Department of C1v1l  Engineering
                         University of Washington
                        Seattle,  Washington,  U.S.A.
                                 ABSTRACT

Abatement of nonpolnt source water pollution  problems relies upon the
ability to assess the dimensions and effects of those problems.  This
ability 1s not highly developed, and advancing 1t 1s hindered by difficult
problems.  Nevertheless, progress can be made by systematically building on
a research foundation to develop Impact assessment  strategies providing the
necessary basis for decision-making.   This  paper discusses examples  of such
strategies from our research.   Included are techniques to  design cost-
effective monitoring programs to gather nonpolnt source data and to  carry
out stormwater runoff sampling.  Also discussed 1s  a stepwlse guide  to
assess quantitatively the aquatic Impacts of operating highways and  to
allocate  resources to their mitigation.  These approaches  offer means of
overcoming a number of the problems  Impeding nonpolnt source Impact
assessment and have potential  application 1n other settings.
                                  P-I-A-1

-------
 INTRODUCTION

 The passage of federal and state water pollution control legislation 1n the
 past 15 years has spurred widespread actions to Improve surface water
 quality  1n the  United  States.  For much of this period the majority of the
 attention has been directed toward control  of point sources  of  water  pollu-
 tion.  Continuing water pollution  problems  1n  cases where some success has
 been achieved 1n  reducing  point  source contributions has suggested, how-
 ever,  that more distributed  sources also are major factors 1n surface water
 quality.  These distributed contributors have been termed nonpolnt sources
 and are associated with stormwater runoff from land surfaces put to various
 uses.

 A prerequisite of acting effectively to solve nonpolnt source water pollu-
 tion problems 1s to be able to assess their dimensions and  their Influence
 on affected ecosystems.  This process  of environmental Impact assessment
 has a fifteen-year history stemming from the passage of the  National  Envi-
 ronmental Policy  Act  (NEPA).  Despite  this  rather  lengthy period,  little
 agreement has developed concerning how to conduct assessments 1n nonpolnt
 source water pollution cases.  The underlying legal and regulatory frame-
 work provides little  distinct guidance, and practitioners have not had
 discussions 1n the breadth and depth needed to adopt and disseminate  proven
 techniques and to promote standards of quality.  As a  result, nonpolnt
 source Impact assessments, much too frequently, cannot provide  meaningful,
 objective problem analyses that  can support effective decision-making..  In
 fairness, 1t should be noted that our field  of Interest 1n this symposium
 1s certainly not  alone 1n  this state.

 Much of the applied environmental  research  performed since the adoption of
 NEPA has had as  an Implicit objective the  enhancement  of abilities to
 conduct environmental  Impact assessments.   While substantial data bases and
 many useful methods have resulted from these efforts,  there  have been few
 attempts to apply the knowledge gained to  developing comprehensive assess-
 ment protocols.   Even more rare has been Implementation of these research
 results 1n the practices of organizations that must prepare  or  evaluate
 environmental Impact  statements.

 It 1s the central premise of this paper that environmental  Impact assess-
 ment of nonpolnt sources of water pollution  can be  Improved by  developing
 systematic methods firmly rooted 1n research results.   Further, the
 research community should take the lead 1n  formulating these methods  and
 should work closely with  practitioners and regulators to Implement them.
At the University of Washington  we have been addressing these Issues  1n our
 nonpolnt source work.   This paper will discuss the problems  we  have recog-
nized,  the steps we have taken  to solve them, and  where we  see  the need for
more widespread efforts.   We  believe our general approaches  have potential
application 1n other locations,  to different nonpolnt  problems,  and 1n  many
other  situations.

PROBLEMS IN NONPOINT SOURCE WATER POLLUTION IMPACT ASSESSMENT

Developing means of assessing environmental Impact necessitates some  degree
                                  P-I-A-2

-------
of generalization* whereby knowledge gained 1n one Instance can be applied
with acceptable reliability at another time and,  often, at a different
place.  The diffuse nature of the nonpolnt  source water pollution problem
creates numerous and  significant  difficulties  1n  meeting this necessity 1n
the presence of the limitations on financial resources that always exist.
These difficulties arise at every step  1n the process, from data collec-
tion, through data analysis and Interpretation, to  Implementation of re-
sults 1n usable  form.

Probably, some progress could be made 1n systematizing nonpolnt assessment
by applying methods already developed and reconsidering existing data with
this objective 1n mind.   Great progress, however, will  require substantial
new data collection.   Planning effective programs for this purpose  1s the
first obstacle.  In any sizable nonpolnt source water  pollution problem,
the potential subjects for monitoring are so extensive that full coverage
1s Impossible.  In the past, there has been  little consideration of how to
select scientifically among all the possible tasks and how to allocate the
available resources.  Further, there 1s little standardization  1n the
nonpolnt source water pollution field of specific sampling and data
analysis techniques.   Therefore,  results frequently are not transferable
from one time and place to another.

Natural conditions encountered 1n nonpolnt  source water pollution work
reinforce and add to  these problems.  The extreme variability documented
(Wan1el1sta, 1978; Homer and Mar, 1982) points out that nonpolnt water
pollutant concentrations and mass loadings  are very site- and time-speci-
fic.   Part of this reported disparity undoubtedly is due to method  Incon-
sistency, but variations 1n topography,  meteorology,  climatology, soils,
geology, specific features of land use, and other factors seem to make
great dispersion 1n the results of nonpolnt monitoring an inherent charac-
teristic. The result is that storm runoff  usually has a greater diversity
1n contaminants   than Industrial  and municipal sewage effluents, with less
predictability of their levels.  In addition to the measurement of physical
and chemical constituents of runoff  1s  the  Issue  of whether receiving water
biota should be observed la sJt.u or la vitro.  All of these considerations
expand the 11st of operative variables beyond  a measurable  number.

Intertwined  with all of the problems noted  above are the issues of long-
term versus short-term Impacts of nonpolnt  water pollution sources and the
cumulative  effects of a number of activities 1n a watershed.  As 1n other
endeavors in the environmental sciences, most  of  the existing understanding
is based on  short-term observations of the  acute  effects on aquatic biota
of high levels of single contaminants.  Low-level,  longer-duration expo-
sures, synergistlc effects of  multiple  pollutants, responses of communities
of species,  and how human  actions  Influence the accumulating burden on
receptors are all subjects that have received  little study.

Assuming that these problems of data collection and analysis can be over-
come, putting results to use in impact  assessment  is stll 1 a formidable
problem. Not all occurrences will have significant Impacts, and analysts
should have means of distinguishing  those that almost certainly will not
from those that may.  Most of the analytical effort then could be directed
toward the latter.  In this task,  analysts  need techniques that can produce
                                  P-I-A-3

-------
an objective and fully justifiable assessment 1n a convenient fashion.   The
limited  resources available for Impact mitigation should be allocated to
actual problems Identified through such  a  process.

Although the difficulties  with  nonpoint  assessment described are daunting*
some strategies are developing to make them manageable.  The next section
will  present our major contributions  to  this progress and the remaining
needs that we bel1eve exist.

APPROACHES TO IMPROVING NONPOINT SOURCE WATER POLLUTION IMPACT ASSESSMENT

Monitoring of Nonpoint Sources

We have  just begun the development of a  methodology to guide the design of
monitoring programs for assessing nonpoint  source water pollution  problems
throughout Washington State.  This effort 1s an  extension of work  we  have
completed for the electric power Industry  (Mar et al.» 1985).   We contem-
plate creating  a conceptual  framework to assist a user 1n Identifying the
potential causes of Impact and subjects  for monitoring and 1n  formulating
hypotheses to test 1n monitoring.  A central feature of the framework will
be a procedure  (based on  eigenvalue mathematics) to rank  monitoring sub-
jects and hypotheses with  respect to designated  criteria* so that the many
possible variables can be  reduced to those most  likely to yield the needed
Information.   Another feature will be a computational  algorithm to deter-
mine the optimum sampling  program design (number of stations*  frequency*
and replication) for a cost constraint or,  alternatively, a constraint, on
statistical power.   The product of this research 1s expected to focus
Washington State nonpoint  source monitoring programs on definite objectives
and to Improve their cost-effectiveness.  We believe similar approaches are
possible and advisable elsewhere.

In the absence of widely agreed  upon specific nonpoint sampling procedures*
we have given considerable attention to this subject.   Upgrading the
quality  of the  methods used and promotion of 1nterchangeab1l1ty of results
will not occur until a standard-setting body outlines  and promulgates
standards that all will  adopt.   Increasing the standardization of experi-
mental practices 1s a major prerequisite  to the  Improvement of nonpoint
source Impact assessment.

The most common means of sampling stormwater runoff has been to  collect a
number of samples at discrete points throughout  the runoff period  using
automatic equipment.   This  strategy produces a relatively complete record
but 1s costly 1n both sampling equipment  and laboratory analysis.   We
developed an Inexpensive technique to fractionate and composite runoff  from
an entire storm, thus reducing both costs greatly (Clark  et al.»  1981).
Our strategy 1s to perform enough discrete sampling through storms to
understand maximum  concentrations,  followed by storm composite sampling to
assemble a large whole-storm data  base.   The first data set  allows assess-
ment of Intense, acute effects,  which  we  also Investigate with  laboratory
bloassays (Portele,  1982).  The  large  latter set assists  1n  resolving the
problems of temporal  and  spatial variability discussed earlier and permits
both deterministic  modeling and statistical analysis, as  described below.
                                 P-I-A-4

-------
Protocols for Impact Assessment

We conducted a five-year study of storm runoff from operating highways  1n
Washington State that concluded with the development of  a  stepwlse guide
for assessment of aquatic Impacts (Mar et al.»  1982;  Horner  and Mar,  1982;
1983).  This guide has several  features that Illustrate  our  approach to
systematic nonpolnt Impact assessment based on research  results.  First,
Its organization 1s  hierarchical,  so that cases having differing potentials
for significantly Impacting aquatic ecosystems can be distinguished and
proportionately allocated resources for problem assessment and solution.
An Initial screening process, based on specific criteria,  removes those
cases highly likely to create an Insignificant Impact from further consi-
deration.   For those cases  analyzed further, the  basic premise of the
assessment 1s that the highway Impact  on the receiving water can be
assessed most realistically  1n  the context of the aggregate burden created
by all activities 1n the watershed.  This approach attempts  to deal with
the Issue of cumulative effects, although much more Insight 1s necessary
before this phenomenon can  be represented  adequately.

The large storm composite data base resulting from the research was
employed to develop a model of accumulated  pollutant loadings delivered by
highway runoff to the receiving water over an  extended period (Asplund  et
al.,  1982;  Chu1  et a!., 1982).   Variability  among  Individual  storm events
was too great for deterministic modeling,  but  the data were  applied suc-
cessfully to develop a probabilistic method for assessing  short-term
effects (Little et al.,  1983).  This approach exploits the log-normal
distribution of the  individual storm  data and  permits the  impact analyst to
estimate the frequency with which a given pollutant concentration, such as
a water quality criterion,  would be exceeded in a  receiving  water as a
result of highway runoff.   A similar technique was developed  by the U.S.
Environmental Protection Agency (1982) for  assessing the effects of urban
runoff.   It also was recommended by Loftis  et  al.  (1983) in  a  general
review of statistical models that might be  applied 1n water  quality regula-
tion.   Modeling of cumulative pollutant generation and probabilistic
evaluation of Individual events allows estimation of both  long- and short-
term  Impacts.

The guide offers an  opportunity to forecast potential  aquatic Impacts of
highway projects at  an early  stage of  development  and to allocate resources
for impact mitigation on the  basis of need.  This advance  Improves the
cost-effectiveness of stormwater runoff management.  We  have worked with
the Washington State Department of Transportation  to Implement the use  of
the Impact assessment guide in Its practices.

SUMMARY AND CONCLUSIONS

Assessment of the impacts  of  nonpolnt sources  of  water pollution  is rather
poorly developed, and formidable problems  impede its advancement.  We
believe these problems can be overcome by systematically applying research
results to formulate assessment strategies  that will  yield objective,
decision-making criteria.   Such strategies  are needed in the areas of both
data collection and its application to analyze potential impacts.  In the
former category, we  are developing an instrument  to  identify all of the
possible elements of Washington State nonpolnt monitoring  programs and to


                                  P-I-A-5

-------
 select the priority  ones.  In previous work we devised a strategy and tools
 to collect representative  stormwater runoff data more cost-effectively than
 with traditional  techniques.  Greater standardization of methods is a
 definite  need  in  the nonpoint source monitoring field.  With regard to
 application of research  data to formulate impact assessment protocols* we
 have developed a  guide allowing Washington State highway designers to
 identify  and scope actual environmental problem areas and to apply abate-
 ment resources accordingly.  This  guide reflects concern  for  cumulative and
 both short-  and long-term  impacts to the extent supported by current know-
 ledge.  We recommend implementation of similar strategies in  other loca-
 tions  and in other nonpoint source  problem areas.

 REFERENCES

 Asplund, R.L., Ferguson, J.F.,  and Mar* B.W. (1982).  Characterization of
 Highway Runoff in Washington  State.  J.  Environ. Engin. Div.  ASCE,  10J?,
 391-404.

 Chui,  T.-W., Mar, B.W., and Homer, R.R. (1982).  A Pollutant Loading Model
 for Highway  Runoff.  J. Environ.  Engin. Div.  ASCE,  108, 1193-1210.

 Clark, D.L., Asplund, R.L., Ferguson,  J.F., and Mar, B.W. (1981).
 Composite  Sampling  of Highway  Runoff.   J.  Environ. Engin. Div. ASCE,  IfiJ,
 1067-1081.

 Horner, R.R., and  Mar, B.W. (1982).  Guide for Water Quality  Impact
 Assessment of  Highway Operations  and Maintenance,  FHWA WA-RD-39.14.
 Department of Civil Engineering,  University of Washington,  Seattle.

 Horner, R.R., and  Mar, B.W.  (1983).  Guide for Assessing  Water  Duality
 Impacts of Highway Operations and  Maintenance.  Iran sportation..Research
 Record, 9J£, 31-40.

 Little,  L.M.,  Horner, R.R.,  and  Mar, B.W.  (1983).   Assessment of  Pollutant
 Loadings  and Concentrations in  Highway Stormwater Runoff, FHWA WA-RD-
 39.12.1.   Department of Civil Engineering,  University of Washington,
 Seattle.

 Loftis, J.C., Ward,  R.C.,  and Smillie,  G.M. (1983).   Statistical  Models for
 Water Quality Regulation.  J.  Water Pollut.  Control Fed., 55,  1098-1104.

 Mar, B.W., Horner, R.R.,  Ferguson, J.F.,  Spyridakis, D.E.,  and Welch,  E.B.
 (1982).   Summary - Washington State  Highway Runoff  Water  Quality Study,
 1977-1982, FHWA WA-RD-39.16.  Department of Civil Engineering,  University
 of Washington,  Seattle.

 Mar, B.W., Lettenmaier D.P., Horner,  R.R., Richey, J.S.,  Palmer, R.N.,
 Mil lard,  S.P.,  and Mackenzie,  M.C.  (1985).  Sampling  Design  for  Aquatic
 Ecological  Monitoring, Volume  1:  Summary Report.  Report to  Electric Power
Research Institute by Department of  Civil Engineering,  University of
Washington,  Seattle.

 Portele,  G.J.,  Mar, B.W., Horner,  R.R., and Welch, E.B. (1982).  Effects of
Seattle Area Highway Stormwater Runoff on Aquatic  Biota, FHWA WA-RD-39.11.


                                  P-I-A-6

-------
Department of C1v1l Engineering, University of Washington, Seattle.

U.S.  Environmental Protection Agency (1982).   Preliminary Results of the
Nationwide Urban  Runoff Program, Vol.  1.  USEPA, Water Planning Division,
Washington, D.C.

Wan1el1sta,  M.P.  (1978).   Stormwater.Management, Quantity and.Qua!1ty.  Ann
Arbor Science Publishers, Inc., Ann Arbor, Michigan.
                                  P-I-A-7

-------
                 QUANTIFICATION  OF  NONPOINT  SOURCE  POLLUTION
                                  Jy S. Wu
                             Associate  Professor
                       Department  of  Civil  Engineering
                  University  of  North Carolina  at  Charlotte
                       Charlotte,  North Carolina 28223
                                  ABSTRACT
      Storm  runoff pollution  can  be  characterized,  in  magnitude  and  in
concentration of  pollutans, as  intermittent and  impulsive-type  discharges
into receiving  waters,  causing shock  loading  problems to the  ecosystem of
these water bodies.  The classical  approach, using  critical  low flow as the
design  criterion  for water quality management schemes,  must include  the
effect of  storm runoff.   This  paper presents state-of-the-art methodologies
for quantifying storm runoff loads.  Four categories  of assessment  methods
are  presented:   these   include   zero-order,   rational,  statistical   and
simulation methods, for  achieving different levels  of prediction,  i.e.  (i)
average annual  storm  load,   (ii)  storm  load  per event and (iii)  storm  load
distribution within events.
                                   P-I-B-l

-------
                 QUANTIFICATION OF NONPOINT SOURCE POLLUTION

                            Jy S. Wu, Ph.D., P.E.
 INTRODUCTION
       In  recent  years,  pollution  from  nonpoint  sources   has   become  an
 increasingly   important  consideration   in   water   quality   planning  and
 management.    Recognizing  that  point  sources  are  obvious  in  nature  and
 relatively  easy  to identify and characterize,  the remaining  problem is the
 evaluation  of storm-related  or nonpoint  source  pollution.
 naturally   arises:  what   techniques   of   varying  cost,
  The question
accuracy,  and
sophistication are available  for achieving a successful assessment program ?

      Characteristics  of   storm  runoff  from  various  sources  have  been
reported  extensively  in  the  literature  (McElroy  and Bell,  1974;  Wu  and
Ahlert,  1976  ;   Novotny  and  Chester,  1981).    In  genreal,   storm  runoff
pollution  can  be characterized, both  in  magnitude  and  in  concentration of
pollutants, as intermittent and impulse-type discharges into receiving water
bodies.   The  classical approach, using  critical  low  flow  as  the criterion
for  water quality management,  must take  the effect  of storm  runoff  into
account.   This paper  presents  state-of-the-art  methodologies for defining
and quantifying storm  runoff  pollution.
LEVELS OF PREDICTION

      Various  levels  of refinement  can  be employed to  define  storm runoff
pollution.  They  range  from  a relatively  simple  yearly  loading  to  quite
detailed descriptions   that introduce temporal variation into the analysis.
Three levels of detail  can  be  defined (U.S.  EPA,  1976; Wu and Ahlert, 1978a
and 1978b).

Level 1: Average Annual Storm  Load

      This level  of detail  defines the average annual  storm-generated loads
as if they occur  continuously, during both wet and  dry periods.   It is used
to  assess  the  cumulative  long-term  effects,  e.g.  sediment depostion  in.
reservoirs,  and  may  be used  to  make  comparisons with continuous municipal
and  industrial  point-source loadings  and  describe  the  relative magnitudes
from each source type.

Level 2; Storm Load per Event

      This level  of detail  considers the  actual  temporal  distribution  of
storm event,  including variability  of  storm  loads from  one event  to  the
other.

Level 3: Storm Load Distribution Within Events

      This level of prediction describes  the  actual runoff loading rate,  as
a  function   of time,  within  each  event.     It  allows   the  evaluation  of
transient and shock-loading impacts on receiving streams.
                                    P-I-B-2

-------
      In general, level 1 is an integral form of level 2 and, in turn, level
2  is  an  integral  form of  level  3.   These classifications  provide useful
guidelines  for water  quality  planners  to  define  the  necessary  levels  of
stormwater impact anlaysis, and to select appropriate assessment techniques.
For example,  in  characterizing storm loads, an  average  annual  loading rate
is useful as a measure of the  relative  impact of storm  runoff.   In the case
of  sediment  deposition,   the   cumulative   effect   of all  storms   is  more
significant than the effect of a  single  storm or a specific period within a
storm.   In the case of transient  water  quality  impact analysis, e.g. oxygen
depletion of receiving waters during storm conditions, the long-term average
loading  becomes  much  less  important,  because   it  does  not  provide enough
information  to  determine  how  often  stream water  quality  standards  are
violated.
ASSESSMENT TECHNIQUES

      Numerous methods  have been used  to  estimate the magnitudes  of storm
runoff loads, for both urban and nonurban areas and for long-term as well as
short-term predictions.  These methods can be categorized as follows.

Zero-Order Method

      This is the most crude and least expensive method.  Storm runoff loads
from  an  area are  estimated from  data in  the  literature.   The method is
inaccurate in most cases and applicable only to long-term predictions. .This
is due to  the  questionable transferability of data from  areas  of different
climate, topography  and management  practices,  etc.    The  term  "zero-order"
indicates the relative flexibility of estimation from literature data.

Rational Method

      The Rational equation, Q=CIA, for predicting peak runoff rate is known
by every water resoucres engineer.   The  rational  type of equation takes  the
form  of  a  product of  several  independent  variables.   If  storm  runoff  and
pollutant concentrations  are  independent,  the mean runoff  loading  rate, w,
will  equal  to  the  product  of  a  mean  concentration,   c,  and  the  mean
discharge, q :

   w = q .  c                                             (1)

Assuming  "q"  can be   estimated  from  basic  hydrology,   "c"   can  then  be
calculated from:

      (a)   published data from a matched watershed;
      (b)   the   computed   flow-weighted   average   concentration,   or   a
            representative concentration, obtained over a period of grab
            sampling (Whipple et al., 1976);
      (c)   an equation which relates "c" to land use, population density,
            and street sweeping effectiveness (U.S. EPA, 1976).   Street
            sweeping has been a subject of debate for its effectiveness
            in reducing nonpoint pollution (Novotny, 1983).
                                    P-I-B-3

-------
      Typical examples of  the  rational  type method  include the  Univeral Soil
Loss  Equation  (Wischmeimer and  Smith,  1965}  and its modified  form such as
the  loading  function approach  (McElroy et al.,  1976).   The latter approach
can  be  used to  predict  sediment yield from  a  watershed,  while other water
quality parameters are related as  fractions of  the  sediment load.

Statistical Methods

      Statistical methods  employ techniques such as regression, correlation,
frequency analysis,  etc.,  to predict  storm runoff loads.

      A  simple  multi-regression  equation  was  employed  to  predict  total
pollutant  loads  for single events   (Wells et  al.,  1971).    The  variables
included  in  the  regression equaiton  are  rainfall  volume,  storm  duration,
time  between  storms, volume and  duration of  last rainfall.   Heaney et al.,
(1976)    developed  correaltion   equations  for predicting  annual  average
loading rates for  each pollutants, as a function of land use,  precipitation
and population density.  This  method  is proposed as a desk-top  procedure for
estimating the  quality  of  urban runoff  from  combined,  storm,  and  unsewered
portions of urban areas.

      In most cases,  it  is  important  to consider high-flow pollutant laoding
as a  phenomenon  separate from low-flow  pollutant loading.   Paterson  (1977)
developed  the   idea  of  a   "concentration matrix"  for  estimating  in-stream
storm runoff  loading.   The matrix is  a  4x4 square  matrix  having 16 blocks.
These blocks  contain a  specific pollutant  concentrations  representative, of
the  4 seasons of  the year and  ,  also, the  4 flow ranges of  low, medium,
medium  high  and high.   The representative concentration  in  each  block  is
obtained by averaging a  large  number  of water quality data (from STORET, for
example)  in  that particular flow  range  at  the corresponding  season  of the
year.   For each day, the   storm runoff loads are  determined  by multiplying
the daily flow data  and  the respective concentrations  in the  matrix.   Daily
loads can be summed  to derive  the  seasonal or annual loads.

      Wauchope and  Leonard (1980)  employed regression  analysis to estimate
the maximum edge-of-the-field  concentration of  pesticides from  an individual
event.     The   regression  equation   relates   pesticide  concentration  to
application  rate, time  of  runoff since  application  and  an  availability
factor  describing   the   pesticide-soil  interaction.     The   accuracy  of
prediction using  regression   equations  is  usually  limited by  the  range  of
numerical  values  of the variables  originally   used  for   deriving  the
equations.

Simulation Methods

      Simulation methods deal  with deterministic  rainfall-runoff mechanisms
that affect the  quality of  runoff  within a watershed.

      Sartor and Boyd (1972)   defined dust-and-dirt accumulation  on  street
surfaces as a major  source of  pollution in urban runoff.  Later developments
involve the use  of  their results  as  basic guidelines for  the  prediction  of
                                     P-I-B-4

-------
pollutant  accumulation  rates on  street  surfaces,  prior  to a  storm  event.
Accumulated dust-and-dirt  is subject to a  first-order  washoff  mechanism by
rainwater.   Other pollutant  concentrations  are predicted, as  fractions of
the total solid loads, as described in the original version of SWMM (Metcalf
& Eddy et.al., 1971).  Sutherland and McCuen (1974) related the accumulation
rate of  each  quality  constituent  to land use,  pavement  type  and condition,
traffic  volume and  length  of time  since  rainfall  or street  sweeping.   The
rate constant  of  the first-order accumulation  equation  was correlated  with
pavement condition, for single family and muti-family residential areas, and
with traffic volume, for industrial and commercial areas.

      The  use of  homogeneous  land  use  approach   in  conjunction with  the
concepts of pollutant accumulation  and  first-order washoff was demonstrated
by  Wu  and  Ahlert  (1978a).   The  first-order  washoff  model   is  calibrated
against  the  various types  of homogeneous  or  single land-use  subareas, to
obtain estimates  of the two parameters, i.e.  initial  loading  and  washoff
rate constant.  The calibrated model can then  be  applied  to the mixed land-
use areas.  In this case,  the initial  loading  and  the  washoff  constant are
evaluated   as   composite,    weighted   values   representing   the   different
percentages of land uses for the  whole  area.   Output  will  be  estimates of
storm  runoff   loadings   for   single  events  and  the time  distribution  of
pollutants within  events.    This  approach  minimizes the  drawbacks  of dust-
and-dirt concept  and  sediment-related  loads; model parameters are estimated
independently  for  each  quality  constituent.    However,  it  requires  a
relatively  intensive  sampling effort to obtain a detailed time  history of
runoff flows and quality, from a  number  of  single  land-use subareas and for
a number of storms.

      Donigian  and  Crawford   (1976)  modeled  sediment  yield  from  land
surfacess  including  the  processes of net daily accumulation  of sediment as
dust-and-dirt, detachment  of particles  by raindrop  impact  on  fine sediment
and  soil,  and  transport  of  sediment   by   overland  flow.   Other  quality
parameters  are related  as  fractions  of  the  sediment  loads.   Wu  (1980)
modified  this  approach  and  assumed  the  BOD  load  consists of  two  parts,  a
sediment-related part and  a  soluble part.   The soluble  BOD load  is subject
to the first-order washoff by surface runoff.  A stormwater assessment model
was developed  to interface  with the SCS TR-20 hydrology program.

      A  storm runoff model  can  also be  constructed to  perform continuous
simulation.   The  use  of  continuous simulation  is  to  supplement the limited
capability of field  sampling programs  in  assessing  the watershed responses
to  hydrologic  conditions and management practices,  over  an extended  period
of  time.   One of the difficulties  encountered in  continuous  simulation is
the lack  of adequate data-base for  calibration and verification.   In  many
cases,  the collected  data  contain missing  records and  thus  the  use of
statistical technique will   be needed to  generate  these  missing records   (Wu,
1983).   Nevertheness, the  use of  continuous simulation  will provide results
that can  be used  to develop probabilistic pollutant loading   to  assess the
cost  effectiveness  of  various  storm  water management  alternatives  at an
acceptable level  of storm runoff  load.
                                   P-I-B-5

-------
 CONCLUSION

       Various  levels of detail  and  analysis of storm runoff  pollution  have
 been  highlighted in this  paper.   However,  as  new problems will be  emerged
 the assessment  methodology will  have to be refined.   One  such  example is the
 uncontrolled  discharge of  landfill  leachate that  represents  a  new  horizon
 of  nonpoint  source  pollution  (e.g.   see  Wu  and  Hilger,  1984).    Future
 research  on  nonpoint  source  pollution  should  include  the  development  of
 assessment  programs for   those nonpoint  sources other than  the  conventional
 ones  such as combined  sewer overflow,  agricultural or urban  runoff  etc.


 REFERENCES

 Donigian, A.S.  and  N.H.  Crawford  (1976).  Modeling  Nonpoint  Pollution  From
       the  Land  Surface.    EPA-600/3-76-083, U.S.  Environmental  Protection
       Agency.
 Heaney, J.P.,  W.C.  Huber and  S.J.  Nix (1976).   Stormwater Management Model
       Level  1;    Preliminary Screening  Procedures.   EPA-600/2-76-275,  U.S.
       Environmental  Protection Agency.
 McElroy, F.T.R.  and J.M. Bell (1974).   Stormwater Runoff Quality  for Urban
       and Semi-Urban/Rural  Watersheds.   PB-231,  NTIS.
 MeElroy,  A.D.   et al.  (1976).   Loading  Functions for Assessment  of Water
       Pollution  From Nonpoint  Sources.   EPA-600/2-76-151,  U.S. Environmental
       protection  Agency.
 Metcalf & Eddy, Inc.,  University  of Florida and Water Resources Engineers,
       Inc.  (1971).   Storm  Water Management  Model, Volume 3.  User's  Manual.
       WPCRS 11024 DOC  09/71,  U.S.  Environmental  Protection Agency.
 Novotny, V. and  G.  Chesters (1981).   Handbook of Nonpoint  Pollution:  Sources
       and Management.  Van  Nostrand  Reinhold Co.
 Novotny, V.  (1983).   Effectiveness  of Low  Cost Practices  for Urban  Runoff
       Pollution  Control"!A  minicourseinstructionmaterialpresentedat
       International  Symposium on  Urban  Hydrology,  Hydraulics and Sediment
       Control,  University  of Kentucky,  Lexington.
 Paterson, R.B.  (1977)  A Preliminary Investigation of the  Phosphorus  Laoding
       Characteristics  of Lake  Carnegie,  Princeton, New Jersey.   B.S.  Thesis,
       Rutgers University,  New  Jersey.
 Sartor,  J.D.  and   G.B.  Boyd  (1972).    Water  Pollution  Aspects  of   Street
       Surface  Contaminants.   EPA-R2-72-081, U.S.  Environmental Protection
       Agency.
 Sutherland, R.  and  R.  McCuen  (1975).   A  Mathematical  Model  for Estimating
       Pollution  Loadings  in  Runoff from  Urban  Streets.    in   Proceedings
       International   Conference  on  Mathematical  Models   for  EnvTronmental
       Problems, edited by  C.A. Brebbia, John Wiley and Sons.
 U.S.  EPA  (1976).    Areawide Assessment Procedure Manual.   EPA-600/9-76-014,
       U.S. Environmental Protection  Agency.
Wauchope, R.D. and  R.A. Leonard  (1980).   Maximum Pesticide Concentrations  in
      Agricultural  Runoff:  a Semiempirical  Prediction Formula,  J. Environ.
      QuaJL, 9(4),  665-672.
Wells, D.M.,  T.A. Austin  and  C.C.  Cook  (1971).   Variation  of Urban   Runoff
      with Duration  and Intensity of Storms.  PB-204  235, NTIS.
                                   P-I-B-6

-------
Whipple, W. Jr., J.V. Hunter  and  S.L.  Yu (1976).   Characterization of Urban
      Runoff:  New  Jersey.    Water  Resources  Research  Institute,  Rutgers
      University, New Jersey.
Wischmeier, W.H. and  D.D.  Smith (1965).   Predicting Rainfall-Erosion Losses
      from Cropland East of the Rocky Mountains.Agricultural Handbook 282,
      U.S. Department of Agriculture.
Wu, J.S.  and  R.H.  Ahlert (1976).   State-of-the-Art  Review:  Nonpoint Source
      Pollution.  Technial  Report WRE-76-3, College of Engineering, Rutgers
      University, New Jersey.
Wu, J.S. and R.C. Ahlert (1987a).  Assessment of Methods  for Computing Storm
      Runoff Loads.  Water Resources Bulletin, 14(2), 429-439.
Wu, J.S.  and R.C.  Ahlert  (1978b).Prediction  and Analysis  of Stormwater
      Pollution.     in   International   Symposium   on  Urban   Storm  Water
      Management, University of Kentucky, Lexington, 183-188.
Wu, J.S.  (1980).   Development  and  Application  of a  Storm Water Assessment
      Model.  Doctoral dissertation, Rutgers University,  New Jersey.
Wu, J.S.  (1983).  Data  Management for  Continuous  Hydrologic  Simulation,   in
      Proceedings  of Stormwater  and Water  Quality  Model,  edited  by  T.O.
      Barnwell, EPA-600/9-83-015, U.S. Environmental Protection Agency, 161-
      176.
Wu. J.S.  and  H.  Hilger  (1984).    Evaluation of  EPA's  Hazard  Ranking System.
      J. of Environ. Eng., ASCE,  110(4), 797-807.
                                    P-I-B-7

-------
                                SUMMARY
                                Panel  II
                   Legislative Means and  Financing of
                         NPS Control  Programs
 PANEL MEMBERS

 Moderator:   James Arts,  Dept.  of Agriculture, Trade and Commerce Protec-
 tion, Madison,  WI

 Assistant Moderator:   Jim  Peterson, Attorney, Milwaukee Metropolitan
 Sewerage District

Panelists:

Kathleen Falk,  Public Intervenor,  Department of  Justice,
   Madison, WI
 Lawrence Libby,  Professor  of Agricultural  Economics, Michigan State
    University,  East  Lansing, MI
 Kathleen Segerson, Assistant Professor of  Agricultural Economics, Uni-
    versity  of Wisconsin, Madison, WI
 Ann Weinberg, U.S. EPA,  Nonpoint Pollution  Section, Washington, DC

 Recorder:   James D'Antuono, Southeastern Wisconsin Regional Planning
 Commission, Waukesha,  WI

 INTRODUCTION

 The charge  of this panel was to investigate the means of  legislative
 treatment for nonpoint source  pollution control,  to suggest how to treat
 this with legislation, and to  suggest how  to finance the  control of non-
 point source pollution.  The discussions were wide ranging, expressing
 a  broad spectrum of  views.  The spectrum was so broad, in  fact, that
 distillation of the  matters discussed into findings agreed to by all
 is impossible.   Accordingly, this report must necessarily  reflect only
 the ideas that  received  the most uniform approbation.

 FINDINGS

 General

 1.  The panel and the  attendees uniformly  reaffirmed that  legislation
     is  the  means to  achieving  nonpoint source pollution control.

 2.  There is a  not-entirely-understood interrelationship  between ground-
     water protection and nonpoint source pollution control programs.


                                P-II-1

-------
  3.  Legislation seems most appropriate on a national or federal basis
     to avoid regional disadvantages.  EPA, however, eschews any national
     legislative policy, in favor of state and local control of the problem.

  4.  Substantial social benefits from nonpoint source control measures
     were perceived.  Society would be positively benefited from the crea-
     ation of incentives to control nonpoint source pollution.

  5.  Financial incentives for nonpoint source pollution control in volun-
     tary programs are generally inadequate; some form of regulatory im-
     perative is needed.

  Legislative Means

  6.  Legislation requiring nonpoint source pollution control is the pre-
     ferred means to generate a "community will" to achieve the desired
     levels of nonpoint source pollution control.

  7.  Legislation should avoid water quality standards as the medium for
     achieving nonpoint source pollution control because the benefits of
     any control program are quite difficult to identify in strictly
     water quality terms,and enforcement of a violated water quality
     standard against a particular source of nonpoint pollution is prob-
     lematic and ineffective.

 8.  Legislation should either identify best management practices or
     authorize administrative determination of best management practices
     that are not related to performance standards.  The authorizing
     legislation should also contain provisions requiring compliance.

 9.  Efforts should be made to attach nonpoint source pollution control
     onto legislative efforts aimed at other environmental  problems.
     For example, research money could be garnered from toxic control
     legislation.

1.0.  Legislation should require affirmative efforts toward  nonpoint
     source pollution control  as a condition of other environmental
     grant programs.

Financing

11.  Funding of nonpoint source pollution control is logically part of
     the public support for a  clean, safe and healthy environment.   Accord-
     ingly, nonpoint source control  programs should be attached to other
     environmental  projects well  received by the public for funding.  For
     example,  erosion control  was  added to park development on a recent
     Missouri  Constitutional  Amendment.

12.  Use of sales tax money garnered from food purchases, at rates  as
     low as .1%, can  generate  needed research and implementation moneys.

13.  "No-cost"  control  programs can  be found by merely reviewing present
     governmental  practices.   For  example, in new residential  areas,
     abandonment of curb and  gutter  requirements can alleviate storm


                                P-I1-2

-------
     water run-off.

A Contrary View

14.  A thought-provoking paper was presented which utilized a perform-
     ance standard predicated upon achieving designated water quality
     standards, with a responsible group being required to pay taxes
     when the standards are not achieved and being provided bonuses or
     rewards when the standards are achieved.  Although a majority of the
     panel and attendees did not concur, the approach illustrates that
     control of nonpoint source pollution may ultimately be found in
     creative use of economic reward and penalty.

CONCLUSION

The panel concludes, federal regulation which establishes best manage-
ment practices is needed.  Funding must come by "piggy-backing" nonpoint
source pollution control onto other programs and having funds earmarked
for nonpoint source pollution.
                                 P-II-3

-------
                          LEGAL/REGULATORY APPROACHES
                        TO CONTROL OF NON-POINT SOURCES
                          OF GROUNDWATER CONTAMINATION

                               Robert F. Ehrhardt
                       Applied Management Sciences, Inc.
                            Silver Spring, Maryland

                                    ABSTRACT

The U.S. Environmental Protection Agency Issued Us Groundwater Protection
Strategy 1n August 1984.  The strategy outlined the agency's proposed approach
to addressing both point and non-point sources of groundwater contamination.
Non-point sources receiving attention In the Strategy Include leaking
petroleum and chemical storage tanks and chemical use, handling, and storage
(Including pesticides) over certain groundwater formations.  The strategy
envisions restrictions or bans on such activities or substances particularly
over groundwater that 1s "ecologically vital" or an "Irreplaceable source of
drinking water."

Because state groundwater quality programs will be required to reflect the
provisions of EPA's Strategy to some, as yet undetermined extent, a survey was
conducted of existing legal/regulatory approaches to groundwater quality
management at the state level of government Including programs to control
potential non-point sources of groundwater contamination, primarily pesticide
and other chemical handling and storage.  Few state programs were Identified
which regulate such sources to a degree commensurate with the EPA Strategy's
approach.  Significant changes In state legal/regulatory approaches may be
required for states to conform to this component of EPA's Strategy.
                                   P-II-A-l

-------
 GROUNDHATER  QUALITY  MANAGEMENT  CONTEXT

 Groundwater  quality  management  and  protection  has been described as one of the
 top  two  priorities on  the  environmental agenda for  the 1980's and possibly
 beyond,  the  other  top  priority  most  often mentioned being add precipitation
 control.   The  Importance of  the Issue 1s Justified  by a battery of now-
 famlHar  statistics:   over 50 percent of the U.S. population depends on
 groundwater  as a drinking water supply, over 80 percent In many rural areas of
 the  country.   Groundwater also  supplies at)out.40 percent of the nation's
 Irrigation requirements and  approximately 25 percent of the requirements for
 so-called  "self-supplied"  Industrial purposes.

 In addition  to these characteristics of groundwater use, characteristics of
 "typical"  groundwater  contamination  Instances  also  give evidence as to the
 Importance of  groundwater quality protection.   Groundwater contamination has
 often been discovered  near heavily  populated and not coincidentally
 Industrialized areas.  The cost of  remedial actions to remove, replace, or
 renovate contaminated  groundwater can be measured In the millions of dollars
 at a single  site alone.  The Congressional Office of Technology Assessment has
 sounded a  further ominous note  In Us recent study of groundwater quality
 management by  observing that what we know today about groundwater
 contamination  problems nationally 1s at least  as reflective of how we have
 discovered or  looked for problems and what we  have been looking for as 1t 1s
 representative of the  major outlines of actual  groundwater contamination
 problems across the  country.  Almost all Interested parties agree that there
 1s at present  Insufficient groundwater quality  monitoring 1n most locations to
 truly characterize the nature and extent of groundwater quality contamination
 problems accurately  or comprehensively.  As monitoring programs Increase and
 more data  becomes available  1t  1s more than likely that more contamination
 problems may be found  as well.
U.S. EPA'S GROUNDWATER PROTECTION STRATEGY

In this context of groundwater management needs and problems, the U.S. EPA
Issued Its Groundwater Protection Strategy 1n August 1984.  In It the Agency
outlined Initiatives 1n four areas designed to Improve groundwater quality
protection efforts:

    •    Strengthen state groundwater agencies and Institutions

    •    Strengthen EPA's Internal organization related to groundwater
         management

    •    Assess problems from unaddressed sources of contamination

         Issue guidelines for EPA decisions affecting groundwater protection
         and clean-up to promote consistency In decision making and regulatory
         requirements.

The subject of this conference, non-point source pollution, was addressed more
than once by EPA's Groundwater Strategy.  Under the category of unaddressed
sources of contamination, the Agency proposed to Increase efforts to protect


                                    P-II-A-2

-------
groundwater from pesticide and nitrate contamination.  EPA committed Itself to
evaluate more precisely the danger of groundwater contamination from pesticide
use, to use labelling restrictions to restrict pesticide use over certain
types of aquifers, to provide more health advisories and technical assistance
to states, and other types of studies and analyses to Increase understanding
of the problem and Us possible solutions.

The guidelines the Agency proposed to Increase the consistency of Us own
groundwater-related decisions are based on the concept of classifying
groundwater formations according to their present quality, current or best
use, and vulnerability to contamination.  Depending on Us classification (the
EPA system has three classes) groundwaters would be protected to a greater or
lesser degree, but all groundwater In the same class would theoretically, at
least, be protected to the same degree.  The strategy outlined, for example,
how hazardous waste facility siting restrictions would differ according to
what class of groundwater underlled a particular site.  The degree of clean-up
required under Superfund could also vary according to what class of
groundwater was Involved.

Non-point sources were also addressed In this groundwater classification
portion of the strategy.  The highest class of groundwater 1n EPA's
classification system, "Special Groundwaters", Include groundwaters that are
"Irreplaceable source(s) of drinking water" or "ecologically vital."  The
Agency proposes to look Into developing additional restrictions on the use,
disposal, or storage of potentially threatening chemicals over this class of
groundwater.  This could potentially Include pesticides, fertilizers, road
salts, storage tanks of various types, and other substances typically Included
on the roster of non-point sources of pollution.

While the groundwater classification guidelines were proposed to apply to
EPA's own decision-making they have Implications for states as well.  States
with EPA-delegated authorities 1n groundwater-related program areas (such as
RCRA, Superfund, Safe Drinking Water, etc.) will be required to show that
their programs are no less strlgent than Federal requirements 1n those areas
which will result from Implementation of the Agency's groundwater strategy.
EPA has said 1t will keep regulatory requirements for states general and
performance-oriented to the extent possible In these areas.

The strategy, particularly In Us groundwater classification components, has
generated great Interest, particularly among some segments of the regulated
community, as to how states regulate groundwater quality now and how they
might be required to change their approach to conform to EPA's Strategy.  The
General Accounting Office, the Office of Technology Assessment the
Environmental Law Institute, and the American Petroleum Institute have all
completed studies on state groundwater protection programs.  EPA's Office of
Groundwater has a similar study scheduled for completion In Spring 1985.
Various task forces and commissions have been formed to address how best to
protect groundwater quality, and national groundwater protection legislation
1s an almost sure topic for congressional debate 1n 1985.

A groundwater quality management study recently completed by the author
compared existing state groundwater programs with the provisions of EPA's
Groundwater Strategy.  The purpose of the study was to Identify the nature and

                                   P-II-A-3

-------
 extent  of  changes  that  might  be required  of  states  1n  order  to  bring  their
 programs  1n  line with  the Strategy.   One  of  the  topics  of  this  study  was  the
 non-point  source-related  provisions  of  the strategy mentioned earlier  1n  this
 paper.  Framed  as  a  research  question,  this  portion of  the study asked:   Do
 any  states presently regulate pesticide,  fertilizer, or  other chemical use,
 handling,  storage, or disposal  on  an aquifer specific  or groundwater-speclflc
 basis as proposed  In EPA's Strategy? Legislation,  regulations, policy
 statements and  related  program documents  from all fifty  states  were reviewed
 and  environmental  agencies In all  states  were contacted  for  answers to this
 question.  Some example approaches and  general themes  from the  state-by-state
 regional analyses  are presented In the  next  section.


 STATE LEGAL/REGULATORY  PROGRAMS

 With regard  to  non-point  source regulation,  most states  have not gone beyond
 the  requirement of best management practices (BMPs) as part  of  their  general
 water quality programs.   As has been pointed out by EPA  Hselt  recently,  BMPs
 designed primarily for  the protection of  surface water quality  may actually
 have resulted 1n Increased degradation  of groundwater quality,  e.g.,
 percolation  "catch basins" for  stormwater runoff which may result 1n
 concentrated pollutant  streams  reaching underlying  aquifers.

 Many, 1f not most  state agency  contacts consulted 1n the course of the study
 reported the perceived  political Impossibility of generating necessary levels
 of support 1n their  state legislatures  for any regulation of non-point sources
 of groundwater  contamination  beyond  BMPs.  Agency contacts offered this
 assessment while at  the same  time reporting  that their existing statutory
 authorities probably provided a sufficient basis for developing and
 Implementing such programs.

 Some states with otherwise comprehensive and  Inclusive programs to regulate
 point sources of potential  groundwater  contamination, explicitly exclude
 non-point  sources from  their  regulatory requirements.  Montana, as Just one
 example, Is a state which  employs a  classification  system similar to that
 proposed by EPA's Strategy and  requires permits for most point  source
 discharges to groundwater:  a combination of management approaches employed by
well less  than  half of  all  the  states.  The  state has a relatively strict
 non-degradation  policy  for groundwater  quality, yet expressly excludes from
 degradation restrictions  changes In  groundwater quality due  to non-point
 sources "where all reasonable land,  soil and water conservation practices have
 been applied."

This Is reflective of the  comments received  from most state agency contacts,
 that given the present  state of knowledge with regard to groundwater
contamination dynamics and the  expected high costs of thoroughly and
positively regulating diffuse non-point sources,  recommended management
practices and guidelines are the only feasible management option for non-point
sources capable of being  Implemented on a broad scale.   Some states  and
localities have begun to address at  least In part the non-point concerns of
chemical storage use, and  handling as outlined 1n the EPA's Groundwater
Strategy.   Approximately fifteen states have enacted specific regulatory
programs to control leaks  from above and below ground tanks for storing
petroleum products, chemical feedstocks, and other substances.   Over half of

                                   P-II-A-4

-------
the states with such programs are 1n the Northeast region of the country.
Host states have general authorities to respond to tank leaks and require
clean-up.  State agency contacts report that these authorities generally do
not provide a sufficient basis for an effective, preventive, regulatory
program.  Some existing tank leak prevention programs at the state and local
level consider the hydrogeologlcal formation over which a tank will be sited;
more often than not this 1s not an explicit consideration.  Many agency
contacts 1n states without tank regulation programs have reported that they
are waiting to see what requirements will be forthcoming under EPA's new tank
authorities from the 1984 RCRA amendments before proceeding further on program
development of the state level.

Apart from storage tank programs, few states have taken steps to restrict
chemical use, handling, or storage, over certain groundwater formations which
could be susceptible to contamination from such sources.  Some states' major
facility siting laws perhaps come the closest to embodying the types of
authorities and restrictions apparently envisioned for pesticide and other
chemical handling and storage (at least for Class I aquifers) by EPA's
Strategy.  Even states which currently classify aquifers or groundwater
Include restrictions only on certain point sources, such as hazardous waste
landfills, and even these are allowed to locate over aquifers 1f certain
design and operating conditions are met.  Seldom 1f ever 1n such state
programs are non-point sources singled out or Included for regulatory
attention.

New Jersey 1s an example of a state which has explldty addressed chemical
handling and storage as a possible source of groundwater contamination In Us
water quality regulations.  Major chemical storage facilities (greater than
400,000 gallons capacity) must provide for secondary containment structures
lined with Impermeable materials wherever hazardous substances are handled.
Leaks must be able to be detected by Inventory controls, automatic leak
detection devices or groundwater monitoring wells.  Regulations are applied to
above and below ground storage of petroleum and other hazardous substances,
drum storage areas, transmission pipelines and 1n-plant piping.

In Massachusetts, many local communities have enacted recharge area zoning
ordinances which regulate and restrict the handling, storage, use, etc. of
chemicals over particular aquifers and types of hydrogeologlc formations.  The
state also has an aquifer/land acquisition program through which localities
are assisted to purchase and thereby remove or prevent all threatening land
uses from property overlying their local groundwater supplies.

In Virginia, a county government has created a special Natural Resource
Conservation Overlay zone 1n the recharge area of one of Its Important
drinking water aquifers.  This special zoning legislation Is directly aimed at
restricting the use of agricultural chemicals 1n the recharge area to protect
groundwater quality
CONCLUSION

These are a few examples of Initiatives at the state and local level to
control one category of non-point sources with the potential to affect
groundwater quality:  chemical use, handling and storage.  These approaches

                                   P-II-A-5

-------
are neither  typical nor wide-spread among  the states.  The most prevalent
approach to  non-point source pollution of  groundwater continues to be
recommended  practices and guidelines.  Many states have requirements that
pesticides be registered for specific uses and that applicators be certified.
But few of these programs consider underlying groundwater explicitly as Is
proposed 1n  EPA's Groundwater Protection Strategy.

The EPA Strategy's proposed Initiatives relating to pesticide and other
chemical handling, use, and storage has the least number of analogies among
state programs of any of the Initiatives proposed 1n the Strategy.  The number
of local government programs of this type  Is not surprising, since what 1s
being proposed In the Strategy Is a variety of restrictive zoning authority
usually associated with traditional local  government powers.  It appears
likely that  any Initiation of even a limited regulatory program under this
component of EPA's Strategy will require significant changes 1n existing state
approaches to protection of groundwater quality from these and other non-point
sources of pollutants.
REFERENCES

Ground-Water Protection Strategy for the U.S. Environmental Protection Agency
    (U.S. EPA, August 1984).

Groundwater:  Strategies for State Action (Environmental Law Institute, 1984).

Protecting the Nation's Groundwater From Contamination (U.S. Congress, Office
    of Technology Assessment, October 1984).

Pye, V.I., Groundwater Contamination 1n the United States (University of
    Pennsylvania Press, 1983).

Trends In U.S. Groundwater Law, Policy, and Administration (Edison Electric
    Institute, January 1984).
                                   P-II-A-6

-------
                          POSITION PAPER

              URBAN WISCONSIN SHOULD JOIN THE FIGHT
                    AGAINST NONPOINT POLLUTION
             Kathleen M. Falk - Wisconsin Public Intervenor

     The  Wisconsin  Public  Intervenor,  an  Assistant  Attorney
General  in  the  Wisconsin  Department  of  Justice,  is  expressly
charged by  statute  with the duty  of  intervening where necessary
to protect natural resources in the State of Wisconsin.  The need
for protection of  these  natural resources from pollution through
nonpoint sources is  now  at a critical stage, yet there are still
no comprehensive programs  to either prevent or clean-up nonpoint
pollution.   Urban  Wisconsin must do  its part to ensure that the
future  will  be  different  from  the  past  by   regulating  urban
nonpoint pollution.    My  intention  is to  summarize some  of  my
observations as  to  why there are  still no comprehensive programs
to prevent or clean up nonpoint  pollution  and  why legislation
must be enacted  to  change  this.  Legislators must recognize that
one  of  the  most   promising  ways  to prevent   future  nonpoint
pollution is through the control of urban sprawl.

     In the  early  1970's,  a rash  of fundamental changes occurred
in federal  and  state environmental laws,  turning around the way
the country  was  polluting  its  resources.   The immediate focus of
the new laws was on  pollution  that one observed spewing out. of a
smokestack  into  the air,  or pouring  out  of an  industrial pipe
into a  river.   Very little  attention  and  effort was directed at
nonpoint pollution.

     We are at  the  point,  historically,  where  most  agree that
tremendous  advances  have  occurred  in  meeting many  of  the goals
and mandates  contained in the  1970's  laws  to clean  up our state
and nation  from point source pollution.  It  is imperative that we
now turn  our political  and  technical  skills  forward  to solving
the problem we  let  lie  underneath the rug  for  the  past decade.
Instead  of  cities   simply  pointing  their   fingers  at  pollution
caused  by   their rural neighbors,  and rural Wisconsin citizens
pointing   their  fingers  at   the  city   culprits  of  nonpoint
pollution,  and instead  of  debating who causes more pollution and
whose  pollution  is  the  worse,  it is  time  to  join  together  to
remedy  and  prevent  future nonpoint   pollution.    Of   the many
sources of nonpoint pollution  —  agriculture,  urbanization and
transportation,  to  name  a few  —  one  specific  cause of nonpoint
pollution gets too little  attention:  urban  sprawl.

     In a  broadly  historical sense, urban sprawl is a relatively
new  problem.    After  World War  II,  suburban  development  was
considered  a  goal,  almost an ethic;  paving  over our countryside
was equated with  "progress" and   "growth."   Numerous government
programs subsidized  suburban sprawl - many still do.

     Urban  sprawl   causes  enormous  environmental  and  economic
problems for Wisconsin.   One of the major environmental harms is

                             P-II-B-1

-------
urban  runoff,  one kind of nonpoint pollution.  A recent  study  of
the  Menominee  River Watershed  reveals  that approximately 75%  of
the  total  pollutant  loadings  in  the  river  comes  from urban
nonpoint  pollution.  Urban  runoff contains almost every  type  of
pollutant:   suspended  solids  and toxic substances, particularly
heavy  metals,  bacteria, nutrients,  asbestos,  oil and grease and
other  pollutants.  One study in Washington, B.C., found  that,  in
water  from  streets,  the concentration of suspended solids was 104
times  higher  than  effluent from  a secondary  sewage  treatment
plant,  and the   lead  concentration  was   1,015  times   higher.
(President's Council on Environmental Quality, 1980.)

     A  recent  study   by   Bauman,  Domanik  and  Konrad,  1980,
identifies  the two  major  types of nonpoint pollutants generated
by  urban land use:   1)   nutrients  and  sediments,  and  2) toxic
materials*

     Urban  sprawl is a major  nonpoint  source of these two kinds
of   pollutants  because  of  the  widespread  land  development,
increased   traffic  density  and  deficient  wastewater  treatment
facilities  that are  associated  with  it.

     The  magnitude  of  the  nonpoint  pollution generated  by  new
land development  is reflected  in  the  following  statistic on the
Menominee  River  Watershed,  taken  from  the Bauman,  Domanik  and
Konrad  study.   Developing  urban areas represented  only 2.6%  of
the  total  area  of the watershed,  but contributed  37%  and 48%,
respectively,  of  the   suspended  solids  and  total  phosphorous.
Subdivision construction  was  cited  as  a  major culprit  in  the
generation  of  nutrient and  sediment pollutants.

     To   further   underscore  the  excessive  pollutant   amounts
contributed by land under development,  it  is helpful to  compare
it with  industrial nonpoint pollution.   In 1978, industry put  an
estimated 5,100  kg/ha/yr  of suspended solids  into the Menominee
River Watershed.   Land under development was the source of 43,700
kg/ha/yr.   That means new  urban development,  such  as that which
makes  up   urban   sprawl,  generates  roughly  eight   times  more
suspended   solid  pollutants than   industry  (Bauman,  Domanik and
Konrad,  1980).    The  need  for  regulation  of  development   is
apparent.

     Another   by-product  of urban  sprawl  is  an  increase   in
nonpoint  pollution  generated   by  transportation.     Greater
commuting  distances  traveled   by  each  car  will influence  the
amount of pollutants emitted into  the air.

     Automobiles  are the  primary  source of  airborne lead in  the
environment (Carberry, 1980).   As a toxic, lead is one of the two
major  types  of  urban  nonpoint  source   pollution.    A recent
compilation of studies on the source of lead and its transport  in
urban  run-off  indicates that  emission of  lead  into  the air by
automobiles leads  to fall-out  of the lead  particule  in rainwater
and soil.   One study measured lead concentrations in rainwater  in

                             P-II-B-2

-------
Pittsburgh, Pennsylvania,  and  found average concentrations of 41
ppb  for  urban  samples.   This was compared  with  19 ppb for rural
samples.   The maximum  allowable lead  concentration  in drinking
water  at the  time  of  this study was  set by the  EPA at  50 ppb.
The  conclusion  drawn  by   J.B.  Carberry,  the  compiler   of  the
studies,  was   that   "rainwater  analysis    indicated   that  urban
regions  were  more  highly  contaminated  from  automotive exhausts
than rural  regions."  To  increase  the  emission  of airborne lead
as a by-product  of urban  sprawl  is not  only  senseless,  it is a
threat to  health.   In man, accumulation  of lead is faulted with
hemoglobin  interference  and kidney  function   impairment.    The
increase  in traffic  density  caused  by  commuters  from suburban
areas can only add to the nonpoint pollutants already threatening
the environment and human health.

     Along with the sprawl  of urban development  into  former rural
areas  comes  the  need for  wastewater treatment  facilities.   The
cost of  wastewater  treatment facilities  borne by each household
and  subsidized  through  grants  is   much  greater  for  these  new
facilities than for those  in established  urban areas.   Regardless
of  these prohibitive  costs,   facilities  must be built  so that
adequate  treatment  of  waste  can  be   achieved  before   it  is
discharged into the environment.

     An  example  of  the  cost discrepancy  between the  old  and the
new  facilities  is  the  following   statistic  from  a  study  of
wastewater  treatment  costs  done  by  James   A.  Hanlon   (1980).
Wastewater  treatment costs in  Greater Chicago,  Illinois,  were an
estimated $45.00 per household/year  in  1980.  The  treatment costs
for  the  Village of  Elwood, Illinois,  with a population  of 783
were an  estimated  $215  per household/year.   Urban Wisconsin can
no longer  afford to allow  new  sprawling  developments to consume
private  and public  monetary resources  in addition to compounding
the nonpoint pollution problem.

     The  state  can  prevent  multiple  ills by  preventing urban
sprawl.  Wisconsin should get to work by  beginning to get its own
house  in  order.    Each  day   state  agencies   make  many,  many
decisions  which  actually  cause,  not  prevent,   urban  sprawl  and
nonpoint pollution.  These  agencies should  redirect their efforts
and prevent suburban development and control this major source of
nonpoint pollution.   Urban areas should  start  reducing nonpoint
pollution  by  regulating  urban  sprawl.   Legislation is  long
overdue.
                             P-II-B-3

-------
References

Bauman, J., Domanik, A., and Konrad, J.  (1980).
     Nonpoint source pollution  In Urban Areas.  Seminar on
     Water Quality Management Trade-Offs.  U.S. EPA, 309-322.

Carberry, J.B.  (1980).  Water quality degradation due to nonpoint
     pollution  from urban sources.  OWRT-B-018-DEL/14-34-0001-
     8070, University of Delaware.  1-53.

Hanlon, James A. (1980.  Costs  for Wastewater Treatment,
     Seminar on Water Quality Management Trade-Offs,
     U.S. EPA,  81-86

llth Annual Report of the President's Council on Environmental
     Quality (1980).

Southeastern Wisconsin Regional Planning Commission  (1980).
     A nonpoint source water pollution control plan  for the Root
     River Watershed.  19-24.
                             p-H-B-4

-------
                   ECONOMIC INCENTIVES FOR CONTROL
                         OF  NONPOINT POLLUTION

                             Kathleen Segerson
            Asst.  Professor,  Department of Agricultural Economics
                           University of Wisconsin
                          Madison, Wisconsin,  USA
                                ABSTRACT

Attempts to control agricultural nonpoint pollution  (NPP) have focused on the
strong encouragement of the use of "best management practices  (BMPs)", with
cost sharing used as an inducement to participate.  However, few farmers have
chosen to implement approved BMPs, and it  appears that these voluntary
programs  are unlikely to result in levels of participation that are sufficient to
combat NPP adequately.  In addition,  imposing mandatory practices to reduce
NPP is unlikely to result in the use of least cost abatement techniques since
the effectiveness of alternative techniques will vary across farms.

An alternative to the  use of mandatory farming practices is the  use of economic
incentives to induce pollution  abatement. This paper suggests one possible
incentive mechanism that could be used when the suspected sources of the
pollutants can be identified (e.g.,  a group of farms bordering a waterway) but
the polluting activity  (e.g., pesticide application or runoff) cannot be directly
observed and/or  monitored.  The general mechanism  combines a system of
rewards for water quality above  a given standard  with a system of penalties
for sub-standard water.  Both the advantages and disadvantages of this
incentive scheme are discussed.

Although the discussion is in  the context of agricultural pollution, a  similar
incentive mechanism could also be applied in cases of non-agricultural NPP of
either surface or ground water.


Keywords;  Nonpoint  pollution, pollution abatement, agricultural runoff,
           economic incentives, pollution taxes.
                                   P-II-C-1

-------
INTRODUCTION

Twelve years ago the U.S.  Congress passed the Federal Water Pollution
Control Act Amendments (P.L. 92-500),  which were intended, among other
things, to address and control nonpoint water pollution. Yet, in its assessment
of the success of existing pollution control programs, the General Accounting
Office (1982)  identified NPP as one area in which little progress could be
claimed.  Since agricultural runoff is one of the main sources of NPP, the
problem is particularly acute in the largely agricultural North Central Region.

Attempts  to control agricultural NPP have  focused on the strong encouragement
of the use of "best management practices (BMPs),"  with federal or state
cost-sharing used as an inducement to participate.  However, few farmers have
chosen to implement approved BMPs.  For  example,  in Wisconsin, the
Department of Natural  Resources (DNR)  set a  goal of 70% participation after
two years by landowners in priority management areas, while actual
participation rates have ranged only between  11% and 38% (DNR,  1984).

There are at  least two possible reasons for low participation  rates.  The first
is that, even with cost - sharing, individual farmers  might expect that the costs
to them of implementing BMPs are not justified on the basis of their expected
returns.  For example, if the benefits of the  use of BMPs result primarily from
reduced downstream sedimentation and pollution rather than from increased
productivity or reduced input costs, these benefits  will not accrue directly  to
the farmer and it is unlikely that  farmers  will  be willing to bear the associated
costs voluntarily.

A second possible reason for low participation  is that farmers tend to be very
independent and object to anything they consider to be an infringement on
their freedom to operate their farms as  they choose.  Since participation
involves a commitment to certain operating practices, it limits a farmer's
freedom to operate independently.

Given these disincentives to participation,  it is not  surprising that the U.S.
Environmental Protection Agency (EPA)  should conclude that "In such cases,
sole  reliance on voluntary programs is not likely to  accomplish adequate
reductions in  pollutant loads and, as a result, other approaches may be needed
(e.g., economic incentives  or regulation)."  (EPA,  1984, p.  xv).   The most
likely  form of regulation to reduce NPP would be  some form of mandatory  use
of BMPs or other farming practices.   This approach has  two  drawbacks.
First, it does not in general provide sufficient flexibility to control NPP in  a
site and source-specific manner that guarantees the most efficient  use of
available  funds.  For practical reasons, direct regulation or control is
generally  applied across-the-board, without regard to differences between
polluters  in terms of either costs of abatement or damages imposed.  In the
case of NPP,  since the effectiveness of  alternative BMPs depends on conditions
that  vary across farms (i.e.,  any individual BMP is not necessarily  "best" for
all farms), mandating the use of these practices across-the-board is unlikely  to
result in  use  of least-cost abatement techniques.  In other words, resources
will be wasted unless site-specific controls can be imposed, since farmers
would be  forced to use practices that are inappropriate for their specific
conditions.

                                  P-II-C-2

-------
     Direct regulation has a second  potential drawback as well,  namely the
difficulty of monitoring to ensure compliance with the regulations.  Although
the use of some practices (such as terracing)  can be easily monitored, others
(such as  a  modification of fertilizer  or pesticide  application techniques to
reduce runoff) would be  difficult to monitor on individual farms.  Since
farmers would often have little incentive to comply with regulations, voluntary
compliance cannot be assured.  Thus, in the absence of substantial
enforcement efforts, the  effectiveness of direct regulation may be further
reduced.

As suggested by EPA,  an alternative to direct regulation is the use of
economic incentives to induce  participation.  This paper  describes one possible
incentive  mechanism that  could be used under certain conditions to reduce
NPP.   This mechanism combines a system of rewards for water quality above a
given standard with a system  of penalties  for sub-standard water.   The
possible use of this incentive  scheme and its advantages and disadvantages  are
the subject of this paper.


IMPORTANT CHARACTERISTICS  OF NPP

An important characteristic of NPP that  makes the standard solutions  that have
been successful in controlling  point  source problems unworkable  for NPP is
that,  although the likely  polluters can often be identified, it is generally  not
possible to  identify a one-to-one  relationship between the level of abatement or
discharge and the  damages from pollutants in the water system.  The reason
for this is twofold: (1)  given any level of abatement, the resulting water
quality effects are uncertain due to the contributing effects of stochastic
variables,  and (2) the level of abatement or discharge often cannot be directly
monitored by the overseeing authority without excessive costs.

More specifically,  the pollutant loadings that result from any given operating
practice depend on a number of climatic and topographic conditions in a
manner that cannot be predicted  with  certainty.   For example, as noted by  the
EPA,

     The  timing (intermittent discharge caused by rain or snow),
     concentration, and  dilution of the pollution  from nonpoint sources
     constitute only part of the picture  when one considers the nature of
     associated water quality impacts; the transportation and ultimate  fate
     of the pollutant constitute the other part.

     .  .  .The potential dilution of pollutants during high flow  must be
     considered against the velocity with which pollutants are dislodged
     and  transported.  Thus,  it is difficult to make  generalizations about
     the  concentration of loads from nonpoint sources.  (EPA,  1984,
     p. 1-8)

The result is that  associated  with any given abatement practice or  discharge
level at any given time is a range of possible loadings for each pollutant.
(More generally,  there is a range of possible damages in terms of the impacts

                                   P-II-C-3

-------
 on human health  and welfare that depend not only on loadings but also on
 factors such  as stream flow  and exposure risks.  Although conceptually the
 analysis could be applied to this broader range of impacts, for  simplicity  we
 focus  here only on the range of possible loadings.)  This  range can be
 represented by a probability density  function (p.d.f.) that is conditional on
 the abatement practice.   The p.d.f. gives the probability  that loadings of a
 given  magnitude will occur at the specified  time, where the probability depends
 on the abatement practices being used.

 The objective of policies to reduce NPP is then to  shift the distribution
 represented by the p.d.f. to the left, as illustrated in Figure 1, to increase
 the probability that  actual loadings will fall below some tolerance level.
 Probability
Tolerance Level
               distribution  with
                  abatement
                  distribution without
                       abatement
                            mean
                         loading with
                          abatement
             mean
         loading without
           abatement
pollutant
 loading
             Figure 1:   Distribution of Pollutant Loading With and
                       Without Abatement

If direct monitoring of all farm operations were economically feasible or
voluntary compliance with regulations were guaranteed,  then the distribution
could be shifted through mandatory  abatement practices.  However, as noted
above,  achieving this at least cost would in general require mandating
site-specific practices.   In addition,  monitoring certain farm activities such as
irrigation and pesticide or fertilizer  use on a  continual basis is  generally
impractical,  and voluntary compliance with  restrictions on use in the absence
of monitoring is unlikely.  Thus, the use of direct regulation to try to shift
the distribution seems  inappropriate;  the use  of a mechanism that provides an
incentive (either positive or  negative) for compliance is  likely to be more
effective.

In addition  to shifting the  p.d.f. of pollutant loadings,  incentives can also
result in risk-sharing.   The existence of a distribution  of possible loadings
implies  that  there  are certain risks associated with any given level of
abatement.   For example, if  the  actual level of abatement is set  so that
expected marginal costs  equal expected marginal benefits, there  is the risk
that mitigating  factors  will be favorable and therefore that this  level of
abatement will have been higher  than that necessary to  achieve  a target level
of environmental quality, i.e., actual loadings will be even less  than expected.
                                    P-II-C-4

-------
Conversely, under adverse conditions  this level of abatement could turn out to
be insufficient to meet the environmental quality goal.   To date,  U.S. pollution
control policy has in general been structured so that polluters bear none of
this  risk associated with  a given level of abatement and society bears the full
risk.  Since abatement requirements do not depend on  the realized
environmental conditions, i.e. the realized damages, the cost to the polluter is
the same regardless of those actual conditions.  His payoff from pollution
abatement is  nonstochastic.   However, society's payoff in  terms of reduced
damages is a random variable depending on the realized environmental
conditions.  If mitigating factors are favorable, society will "win"  the gamble
in the sense  that the  actual damages will be lower than expected.   Likewise
adverse factors imply  that society "loses" in that  it suffers damages that are
greater than expected.  If  society is  adverse to taking risks, it can increase
its expected well-being by shifting some of the risk to the polluters through
the use of policy instruments that result in risk-sharing.   This could benefit
polluters as well if sharing risks induced victims  to be  content with more
lenient  target levels.

The  use of incentive mechanisms and  risk  sharing to induce desired behavior
has been studied by many previous authors, e.g.  Stiglitz (1974),  Holmstrom
(1979), Shavell (1979), and  Mookherjer (1984), in the context of what has
come to be known in the  economics literature as the "principal-agent" problem.
A principal-agent problem exists when the welfare of one  party, called the
principal, depends directly on actions  taken by another party, called the
agent.  The principal's challenge is to devise a payment schedule  for the agent
to induce the  agent to take  those actions  that  best serve  the interests of the
principal.  However, the  principal is  unable to control the agent's actions
directly and cannot even  observe them.  He can only observe his  payoff that
results from the agent's actions, which depends not only on those  actions but
also  on stochastic conditions  or  events.  The following question is  then  asked:
what form of incentive mechanism should the principal use  to induce the agent
to take those  actions that best  serve  the interests of the  principal?  The
question can also be asked in the context  of a more general model  where there
are many agents instead of just one.

If we interpret  society as the principal and the farms generating NPP as the
agents, the principal-agent problem describes the  challenge faced  by society to
induce independent  farmers  to take steps that  will contribute to improved
water quality  and thereby serve  the interests of society.   Viewing NPP  in this
way  allows the insights that have been gained  for solving general
principal-agent  problems  to be  applied in  solving  problems of NPP.


AN INCENTIVE MECHANISM

Single Polluter Problem

In devising an appropriate incentive scheme, the  first  point to note is that
with regard to NPP we are interested in a reduction of loadings (or, more
generally,  the damages that result from them)  rather than a reduction in, for
example, soil  erosion per se.  This is  emphasized by the EPA (1984), which
notes that  lands with  high erosion rates are not necessarily those causing
significant  NPP problems, and vice versa.   Thus,  incentives designed solely to

                                  P-II-C-5

-------
 reduce soil loss  (such as soil loss taxes) are inappropriate for control of NPP.
 Instead, an incentive scheme  designed to shift the distribution of loadings is
 needed.

 Consider first the case of a single suspected polluter, e.g. a single farm
 whose land drains into a  nearby stream.  Let x  be the level of actual loadings
 of a given pollutant  in the stream, and let  x be a specified target or tolerance
 level of loadings, which is set by authorities and could, for example, be
 adjusted seasonally.   Actual loadings x will depend upon both the abatement
 actions taken by the polluter  (e.g., the use  of various BMPs)  and random
 variables reflecting unpredictable weather and stream conditions, as illustrated
 in Figure 1.

 A general incentive scheme designed to shift the distribution of actual loadings
 would take  the form  of automatic, required  payments  T(x)  that  depend upon
 the level of actual loadings as compared to  the target level x and are given by

                                       if x £ x
                                              __   I
                                       if x < x

 where t and k are constants  that can be set  by  the regulating authority to
 ensure that the payment scheme  provides the incentive necessary to induce the
 polluter to undertake the level of abatement that is deemed  socially desirable.
 The optimal values of t and k will depend upon  marginal abatement costs and
 benefits as  well as how increased abatement shifts the distribution of loadings.
 (See Appendix A for details.)

 The payment  scheme is composed of two parts.   The first, reflected by  t, is  a
 tax/subsidy payment that depends upon the extent to  which x  differs from  x.
 If actual loadings exceed  the target level,  the polluter pays a tax proportional
 to that excess, while actual loadings below  the target level result in a subsidy
 or credit to the  suspected polluter.  Note that actual loadings may differ from
 target levels because of either the abatement actions of the  polluter or the
 influence of the  random variables.  Thus,  the polluter may be liable for tax
 payments that result from influences outside his  control.  Likewise,  however,
 his liability may  be reduced (and he may even receive subsidies if x falls
 below  x) due to  favorable environmental conditions even if he has taken no
 action to control NPP.   Thus, in choosing  his level of abatement,  he gambles
 on what his actual tax liability will be and weighs the  additional cost of
 pollution abatement against the decrease in  expected payments that results
 from increased abatement.

 The same type of incentive is provided by the second  component of the
 payment scheme, reflected in k,  which is a fixed penalty imposed whenever
loadings exceed the target.   The amount of the penalty is independent of the
extent to which the target is exceeded.  Again,  the polluters can weigh the_
cost of abatement against  the  decrease in the probability that x will exceed x,
i.e.,  that he  will incur the penalty, that results  from increased abatement.
Note that the effect of this penalty scheme  is different  from that of penalties
                                   P-ll-C-6

-------
applied to actions (or inactions) that  are directly under the control of the
polluter (e.g., penalties for point emissions in violation of standards).   In the
stochastic case, additional abatement will in general always decrease  the
expected penalty by decreasing the probability that x will exceed x, whereas
under  penalties for emissions in excess of standards incentives exist to reduce
emissions to the standard level but not below.

Either  component of the incentive mechanism  can be used by itself to induce a
desired level of abatement,  or they can be used  in combination as  given  above.
(See Appendix A.)  Since a comparison of these alternatives is more relevant
when there are many polluters, they are discussed below in that context.

Multiple Polluters Problem

In many cases  of NPP, it is likely that several polluters, e.g., several farms,
will be suspected of contributing to the loadings of a  given waterway.  An
incentive scheme similar to  the one introduced above can still be used, if t  and
k are allowed to vary across polluters, i.e., if tax  payments of polluter  i are
given by

                  ft.(x-x) + k.        if x  £ x
          T.(x) = 1  l         l
                  I t.(x-x)             if x  < x

Again, t. and k.  can  be set to ensure  optimal levels of abatement by each
source.  (See Appendix A.)  Polluters with high marginal abatement costs or
whose  abatement  is likely to have a relatively small  effect  on  the distribution
of loadings will tend  to have lower optimal payments than other polluters.
Note, however, that each polluter's liability depends on loadings from the
whole group, not just his individual contribution, since at  any given time
individual contributions are not known or observable.

As in the case of a single polluter, possible  forms of the incentive scheme
include (1) a pure  tax/subsidy scheme where k.  = 0 and t. > 0 for all i, (2) a
pure penalty scheme  where t. = 0 and  k. > 0 for all i,  and1 (3) a combined
scheme where t. > 0 and k.  > 0 for all i.  Although the optimal abatement level
can be induced1 using  any \>ne of the three alternatives, their implications in
terms of total polluter/government payments are clearly different.  For
example, under the pure tax/subsidy scheme, combined subsidy payments to
all polluters when x < x could far exceed the benefits  of the reduced loadings,
since each polluter would in some sense get credit for  the entire reduction in
loadings.  In addition, it provides no  way to reward or compensate polluters
who  abate "more  than their share" and thereby  create  additional benefits for
all other polluters.  A pure penalty  scheme has the advantage of requiring  no
government outlays for low loadings, but  also suffers  from the inability to
compensate "good" polluters.  The combined  scheme avoids some of these
problems.  Under this scheme the t. values are  not constrained (see Appendix
A) and thus  can be chosen so that \he sum of subsidy payments when x < x
does not exceed the benefit of the reduced loadings.  Although this  still
requires government outlays when x < x,  those  outlays can theoretically be set
as low  as  desired (as long  as the k. are adjusted to maintain proper
incentives).  In addition to choosing the sum of  the t. to avoid excessive

                                  P-II-C-7

-------
 outlays, the regulatory authority  can also set the individual t.'s to reward
 "good" polluters.  Thus,  the combined scheme allows the distributional effects
 of the incentive mechanism to be adjusted to satisfy other,  non-efficiency
 objectives.


 ADVANTAGES AND DISADVANTAGES

 Regardless  of which form of the incentive mechanism is chosen, the use of
 such a mechanism has  several advantages.

 First, it involves a minimum amount of government interference in daily farm
 operations,  and farmers are  free to choose the pollution abatement techniques
 that are least cost for their  farms.   Since individual farmers are in a better
 position to  determine the  abatement practices that will be most effective for
 their land (and  will  have  an incentive to do so),  their freedom to choose the
 techniques  used provides the flexibility necessary to ensure that any given
 level of abatement is achieved at the lowest possible cost.

 Secondly, once in place, the incentive mechanism can be easily administered
 since it does  not require  continual monitoring of farm practices or metering  of
 runoff or soil loss.   It instead requires that the regulatory authority monitor
 pollutant loadings regularly and calculate the necessary tax or subsidy
 payment.  Accounts  can be cumulated over time with payments made
 periodically.   If, over the time period, tax liability exceeds subsidy  payments,
 then  no government  outlays would be necessary even under the pure tax/
 subsidy or  combined approaches.  The subsidies would simply  act as credits
 against tax liability.

 Thirdly, the  cost-sharing mechanisms of the  existing  NPP programs can be
 maintained to prevent placing excessive burdens on the agricultural sector,
 and  other considerations regarding an appropriate distribution of costs can be
 accommodated, as long as the parameters of the payment scheme are  adjusted
 accordingly to maintain proper incentives.  As mentioned above, under the
 combined approach the t.  values can  be chosen to reflect distributional
 considerations.  In  addition, federal  or state cost-sharing to cover a portion of
 investment or  operating expenses is consistent with use of the incentive
 scheme.  The  payment mechanism simply provides the incentive for
 participating in cost-sharing programs that seems to be missing under the
 current structure.

 A fourth advantage of the incentive scheme is that it focuses on water quality
 rather than erosion or runoff, which is more appropriate for controlling NPP.
 In addition, to the extent that some of the fluctuations in pollutant loadings
 can be  anticipated,  there  would be an incentive for farmers to try  to offset
 peaks by, for example, avoiding heavy pesticide or fertilizer applications prior
 to anticipated  rain or wind storms.

 Finally, these  incentives allow the risks associated with given abatement levels
 to be shared by victims and  polluters alike, which differs from the typical
 approach to environmental policy in the U.S. under which only victims are
 forced to gamble on  whether given abatement levels are sufficient.  This could
benefit  both polluters and victims if sharing risks induced victims to be
content with more lenient  target levels.
                                   P-II-C-8

-------
The  disadvantages of this incentive scheme include the information
requirements that are necessary  to set the levels of the t. and k. parameters
initially to provide the  correct incentive.  (In general, tliiis is a problem with
any  regulatory device  seeking to achieve socially optimal outcomes.)  The
necessary information includes abatement cost estimates, estimates of damages
from pollutant loadings, and estimates of how each  polluter's abatement affects
the distribution of those loadings,  which would require the use of individual
watershed models.

A  second  possible disadvantage of  the mechanism is its implications with regard
to distortionary taxation.  It  would have to be  structured so that allowing the
t.  and k. parameters to vary  across sources would  not be considered to be
d\stortionary taxation.


REFERENCES

Holmstrom, B.   (1979).  Moral Hazard and Observability.  Bell Journal of
     Economics, 10(1),  74-91.

Holmstrom, B.   (1982).  Moral Hazard in Teams.   Bell Journal of Economics,
     13(2).  324-40.

Mookherjee, D.   (1984).  Optimal Incentive Schemes with Many Agents.
     Review of Economic Studies, 51(3),  433-446.

Shavell, S.   (1979).  Risk Sharing and  Incentives  in the Principal and Agent
     Relationship.  Bell Journal  of Economics,  10(1),  55-73.

Stiglitz, J.   (1974).  Incentives  and Risk Sharing in Sharecropping.  Review
     of Economic Studies,  4,  219-255.

U.S. Environmental Protection Agency.   (1984).  Report to Congress: Nonpoint
     Source Pollution in the U.S.,  Office of Water  Program Operations, Water
     Planning Division, January.
U.S.  General Accounting Office.  (1982).   Cleaning Up the Environment;
     Progress Achieve!
     Washington, D.C.
Progress Achieved but Major Unresolved Issues Remain,  GAO/CED-82-72,
   ihir
Wisconsin Department of Natural Resources.  (1984).   Wisconsin Water Quality;
     Report to Congress, WDNR-SWQS-84-7, May.


ACKNOWLEDGEMENTS

The  author  would like to thank Michael  Carter, Jean-Paul Chavas,  and Daniel
Bromley  for comments on an earlier draft,  without implicating them for any
remaining errors.
                                   P-II-C-9

-------
                                APPENDIX A
This appendix describes  how the parameters of the incentive scheme  described
in the text could be set  to ensure proper abatement incentives.  The scheme is
similar to one suggested  by Holmstrom (1982) in the context of optimal
organizational structure.

Let the random variable x(a) be actual loadings of a given pollutant  which
depends on the level of abatement of a single polluter, denoted a.  Let C(a)
be the cost of abating  to level a, and let F(x,a) be the probability that x(a)
is less than a target level x,  given  a.  Finally, let B(x(o)-x(a)) be  the
benefit of increasing abatement from a base zero level to a.  Then  society
seeks the  level  of abatement that maximizes E[B(x(o)-x(a))]  -  C(a),  where  E is
the expectation operator  over  the random variable  x.  The optimal level of
abatement, denoted  a*, is implicitly  defined by  the first order  condition
E[B'-dx/da] + dC/da = 0.

Given  the  payment scheme T(x) in the text, the expected_value of those
payments, E[T(x)], is given by t-E[x(a)] - tx + k  (l-F(x.a)).  The polluter
thus chooses the  level  of abatement  that minimizes  E[T(x)] + C(a).   His
choice, a, is implicitly  defined by t-E(dx/da] - k(3F/3a) + dC/da = 0.  A
comparison of the optimality conditions for society  and  the  polluter  implies that
the polluter will bejjiduced to choose the level of  abatement that is optimal  for
society,  i.e., a = a , if:

     (1)  k = 0 and t = -(dC/da)/E[dx/da] evaluated at a*,
     (2)  t =  0  and k =  (dC/da)/(3F/3a)  evaluated at a*,  or
     (3)  t is arbitrary and k = (dC/da + tE[dx/da])/(3F/3a) evaluated at  a*.

Note that under the pure tax/ subsidy  case (case 1), the optimal tax  rate t is
equal to marginal benefits B' if B' is a constant.   Under a nonlinear  benefit
function, t ^ E(B').  However, E(B') may be a sufficient local  approximation to
the optimal t, or  serve as a guide in setting t.

If there  are multiple polluters, let a. be the abatement  level _of polluter i, let
C.(a.)  be i's abatement cost, and interpret a in x(a) and F(x,a)  as the vector
(a"..,1. ..,a  ) (where  n is the number of polluters).   Society then seeks the
vector a* = (a|,...,a*) that maximizes E[B(x(o) - x(a))]  - [C1(a1) + ...+
C  (a  )].  Individual polluters choose a.  to minimize E[T.(x)] + Cjta.) given a
set oi  expectations about the actions or other polluters.  An  analysis similar to
that for  the single polluter case indicates that,  in  equilibrium where  all
expectations are realized, the  possible optimal incentive schemes are given by:

     (1)  k. = 0 and t. = -(dC./da.)/E(3x/3a.] evaluated at a*,
     (2)  t. = 0 and kt = (dC./da. J/(3F/3a.) Waluated at a*,  or
     (3)  t* is arbitrary and k =l (dC./da.1 + tjEtSx/SajD/OF/Sap evaluated at
          a*.
                                   P-II-C-10

-------
           LEGISLATION,  NOT LITIGATION,  IS THE SOLUTION
                  TO NON POINT SOURCE POLLUTION
                                by

             James  H.  Petersen,  Senior  Staff Attorney
             Milwaukee Metropolitan Sewerage District
                             PREMISE

      Heightened recognition by society of problems arising from
non-point source pollution have raised an increasingly large
voice towards controlling such discharges.  Often times, however,
those adversely affected by non-point source pollution do not
have available the tools or knowledge of the problem sufficient
to obtain affirmative relief.  Indeed, it is the premise of this
paper that traditional legal remedies cannot adequately be relied
upon to remedy non-point source pollution.  Thus, it is important
that society's interests be protected by the legislature, the
source of our ordered form of government.

                           INTRODUCTION

      Recourse to legislation is particularly apropos in dealing
with non-point source pollution if for no other reason than non-
point source pollution does not respect traditional boundaries.
From any source of pollution, the effects pass on through the
watershed.  The resulting pollutant path, the pollution shed,
effects many jurisdictions, transcending municipal and county
lines and can, in the many cases, cross state lines providing a
federal cast to the problem.  Many of the traditional remedies
provided through the courts are frustrated by this concept of
pollution shed, and where not frustrated, the remedies are
rendered so expensive that only the most well-heeled of litigants
can realistically hope to finance a favorable result.

      A further reason for recourse to legislation is the
critical need to factor economic and societal values into the
calculus of environmental rights.  This analysis must of
necessity require society's judgments of the future and the cost
society is willing to bear today to ensure a sound environment in
the future.  As will be seen, this is particularly true when the
need for a solution of non-point source pollution points toward a
change in more or less traditonal uses of water resources to
remove and pass along non-point source pollutants.
                            P-II-D-l

-------
                                I.

                   TORT REMEDIES ARE INADEQUATE

      The  use  of  litigation as  a  vehicle to obtain relief  from
damaging non-point  source  pollution is  the  most  familiar vehicle
for people to  look  to  for  relief  and yet at the  same time  it is
probably the least  adequate way to obtain the  relief necessary.
This  inadequacy  is  a necessary  component of the  legal system from
which the  relief  must  be generated.  Relief would be based upon a
theory of  tort violation;  that  is, an  injury to  a protected
personal right caused  by another's acting in an  unreasonable
fashion.

      Two  aspects of traditional  tort  litigation prove to  be the
most  frustrating  in the pursuit of relief from non-point source
pollution:  the  need to establish a discharger's act was the
cause of damage  and that damage in fact occurred.  These
limitations in traditional tort litigation  create significant
problems with  the allocation of responsibility for it is the
plaintiffs burden in any litigation to  identify  a defendant whose
conduct was the  "but for"  or "substantial factor" cause of the
harm  done  to the  plaintiff person injured.

      With non-point source pollution,  identifying a "but  for"
cause of damage  is  not easy.  Perhaps a few statistics will
illustrate the point.  In  the United States there is enough
animal waste produced  annually  as would equal  the waste generated
by two billion people, nearly ten times the number of people
currently  living  in the united  States.   Each day, a head of
cattle produces  six times  the waste of  the  average human being.
Erosion control  in  and of  itself  is a major cause of non-point
source pollution  but the costs  of remedying it are astronomical
and not entirely  calculated.  The Environmental  Protection Agency
has made various  estimates in just three of those areas.

      EPA  has  estimated that highway erosion control could cost
anywhere from  a $130 million to $7 billion;  EPA  has estimated
that  urban construction erosion control would  cost anywhere from
$140  million per  year  to $1.4 billion per year.   EPA has
estimated  that sediment control in streams  would cost
approximately  $200  million per  year to  $3 billion per year.
These statistics  illustrate that  the problem of  non-point  source
pollution  is multi-faceted and  its effect on water is
pervasive.  Indeed, with such diverse sources  as animal waste,
erosion, and sedimentation causing non-point source pollution of
waters, it tends  to make the  allocation of  responsibility  in
traditional tort  litigation almost impossible.
                            P-II-D-2

-------
      A convenient starting point for review of the practical
effectiveness of the tort system to treat of non-point source
pollution is to review the source of the rights to be enforced.
The most obvious rights against non-point source pollution are
the rights of a riparian owner.  A riparian owner is the owner of
property bordering upon a water body.  The riparian owner has a
right to use of the water subject only to the use being
reasonable and to the return of the used water in as to close to
its original state as possible.  A riparian owner obtains rights
simply by virtue of the ownership of land bordering the water
body.  The owning of land bordering a water body brings with it a
common law right to use the water in common with all other
riparians.

      An action by a riparian to protect his rights is limited to
bringing an action against another riparian user.  If a riparian
owner seeks to obtain damages from a non-riparian, the riparian
is limited to suit alleging nuisance.

      In a suit to protect its riparian rights, the riparian need
only show that another riparian has interferred with the suing
riparian's use of the water.  It need not show negligence nor any
form of intentional invasion of a property interest.  While this
simplifies the burden of proof on the part of the riparian, it
does not entirely answer the non-point source pollution problem
since the polluter is subject to damages for its pollution
only.  Further, even a polluting riparian is allowed a reasonable
use of the water which may include some reasonable amount of
pollution caused by such use.  This may be compounded by a
downstream owner in effect granting an easement to pollute to an
upstream riparian by acquiescing in a course of conduct which is
known, and notorious.

      The other major source of rights is found in the law of
nuisance.  A nuisance is an unreasonable interference with a
person's enjoyment of land.  In the case of a riparian owner,
nuisance would be established if it could be shown that the
defendant did an act that polluted the water in a material
fashion and that the defendant's conduct was a proximate cause of
that unreasonable interference.

      It is the concept of proximate cause, however, which again
poses the most substantial impediment in an action for
nuisance.  Proximate cause is at best difficult to establish when
there are multiple discreet dischargers, making the tracing of
water quality at best unachievable.  It was the effective
inability of government to trace a water quality violation from
point source dischargers that led to the Federal Water Pollution
Control Act Amendments of 1972. Tracing non-point source
pollution dischargers presents an even tougher cause and effect
problem.

                            P-II-D-3

-------
      The point of the above discussion  is to show that
traditional tort remedies for relief from non-point source
pollution are simply inadequate to address the problem.  The
number of polluters are many, their pollution is diverse, and  the
cost of ascertaining proof to the requisite degree of certitude
is almost prohibitive.

      The law has attempted to find ways around these limitations
resulting from a "but for" causation standard of damage by a
defendant.  Most of these attempts have developed in the field of
products liability litigation.  This alone makes use of these
theories suspect when treating of non-point source pollution.
There are grave conceptual problems in superimposing these
doctrines upon non-point source pollution and it is accordingly
not a recommendation that this be done.  Nonetheless, it is
important that we recognize what those theories are so that by
understanding them we can appreciate why they should be rejected.

      The four prevalent theories are: 1) alternative liability,
2) market share liability, 3) enterprise liability, and 4)
concertive action liability.  In the State of Wisconsin, only  the
latter, concertive action liability, is a recognized theory for
allowing recovery in a products liability circumstance.  A brief
description of each of the four and its limitations:

     1.  Alternative Liability;  If more than one defendant could
         have caused the liability, the defendants may all be
         joined in the litigation and the defendants must prove
         among themselves who is not liable.  This theory of
         liability may not be used when a court does not have
         jurisdiction over all of the defendants.  This poses  a
         major problem with multiple jurisdiction pollution
         sheds.  It also causes problems with large, undefined
         groups of defendants as are likely in non-point source
         pollution.  Further, some of the defendants may be
         governmental bodies which are protected by dollar limits
         on their potential exposure.

     2.  Market Share Liability;  This is based upon the theory
         that when the entire market is divided among a fixed
         group of defendants, all participants receive benefit
         from the existence of the market and should share in
         liabilities arising from marketing.  All market
         participants are not needed in the litigation as a
         defendant, but any defendant in the litigation would  be
         responsible for only its pro rata share of the liability
         in proportion to its share of the market.  This is not
         readily available as an alternative for non-point source
         pollution because non-point source pollution is not a
         "market" oriented event similar to manufacture or
         distribution of a pharmaceutical product.

                            P-II-D-4

-------
     3.  Enterprise Liability;  This applies where there  is an
         industry wide practice which created the unreasonable
         risk of harm, which risk is not capable of attribution
         to a single defendant.  Like the alternative liability
         theory above, the defendants are required to prove among
         themselves who is not liable; unlike the alternative
         theory but like the market share theory the enterprise
         liability theory does not require that all defendants be
         in the suit.  What is required, however, is that there
         must be a well defined industry and it requires  an
         awareness of the risk and a joint liability to reduce
         the risk among all of the defendants.  These
         characteristics remove non-point source pollution from
         the enterprise liability theory of recovery since there
         is no well defined industry nor is there a joint
         liability to reduce risk;

     4.  Concertive Action Liability;  This theory of liability
         requires that there be a common design or agreement to
         act in a certain fashion.  Under this theory, all
         participants in the concerted action are responsible for
         any harm that is created.  The concerted action  theory
         does no require any written agreement, and can be
         established by a showing of conduct estabhlishing action
         in concert.  The fault with this theory in relationship
         to non-point source pollution is that the showing of
         concertive action across the broad cross sections of
         society that create non-point source pollution is
         virtually impossible.


      In a sense the harm caused by non-point source pollution is
a public harm, a public harm that cries for some form of
intervention by the government.  It is this no longer exclusively
private harm which should attract the attention of the respective
legislatures to conceive of means to address the need for a
remedy to non-point source pollution.

      The next problem, of course, is what form should that
legislation take.

             THE MANY FACES OF LEGISLATIVE SOLUTIONS
      Recourse to legislation as a solution to the problems of
non-point source pollution, while undoubtedly the preferrable
means, nevertheless creates the problem of what  form  the
legislation should take.  It must borne in mind  that  the
legislative forum has a number of limitations.   Legislation often
times requires political compromise and political compromise is


                            P-II-D-5

-------
not  susceptible  to  analytical review.   Further,  three  important
problems  arise with legislation.

      First,  it  must be  borne in  mind  that fixing  a problem is
always more expensive than creating  a  problem.   It is  human
nature that if something can be done cheaper  but create  a
problem,  it will have been done cheaper nevertheless.  Second,  it
is hard to assess environmental problems in terms  of unit
costs.  Thus, unlike fixing costs of pollution based upon
production, the  benefit  of fixing an environmental problem or the
harm caused by an environmental problem cannot be  reduced to
economically  meaningful  values that  fluctuate in a predictable
fashion.  Third, environmental problems themselves are subject  to
change by merely redefining what  the problem  is.  In the context
of non-point  source pollution, if stream erosion is eliminated
from the  definition non-point source pollution you have  created
an entirely different perspective on the problem.   By  like token,
if urban  non-point  source pollution  is eliminated  from the
legislation,  you have a  different problem than if  you  were
dealing with  all non-point source pollution.

      Bearing in mind these limitations on the legislative
process,  there are  basically four ways to achieve  a legislative
solution  to environmental problems.  These are:   (1) change the
legal status  of  the problem;  (2)  regulate the problem;  (3)  tax
the  problem;  and (4)  subsidize the solution.

      Legislation that changes the legal status  of an
environmental problem is easiest  to  proclaim, but  its
ramifications are often  the most  pervasive and hence disruptive
manner of dealing with the pollution problem.  The Federal Water
Pollution Control Act Amendments  of  1972 are  perhaps the best
example of legislation which changed the legal status  of a
pollution problem.   The  Federal Water  Pollution  Control  Act
Amendments were  predicated upon a change from water  quality
standards dictating what the effluent  limitation should  be to a
system whereby a legislatively determined effluent limitation
would be  achieved.   The  legislation  decreed a zero pollution
goal.  Notwithstanding the unreality of attaining  a  zero
pollution goal,  the idea behind it was essentially to  remove the
nation's  waterways  from  the order of public goods  free and
available to all.   In its stead,  Congress created  a  prohibition
against the use  of  those waterways as  discharge  points for
pollutants.

      The use of  legislative  change  in legal  status  to address
non-source point  pollution may not correct the problem,
however.  Politicians  are pragmatic  people in many ways.   It will
require more than just a definitional  change  in  legal  status to
work the  full change  in  legal status that is  necessary to control

                            P-II-D-6

-------
non-point source pollution.  The Federal Water Pollution Control
Act Amendments illustrate this point.  The Congress enacted a
complete package of changes, including a construction grant
program, a permit program, water quality standards and state
implementation plans program, and a creation of effluent
standards program.  A change in legal status for non-point source
pollution would require a similar, all pervasive response.  I
submit that such an all pervasive change in the status of non-
point source pollution, while possible, is not yet politically
expedient.

      A second legislative form to deal with non-point source
pollution would be to regulate the amount and means of non-point
source pollution discharges.  Regulation is favored in America
for a number of reasons.  Americans have a great love of quick
solutions and are willing to abdicate their ability to attain
greater societal change to a legal solution for a social problem
rather than develope other practical ways of dealing with that
problem.  Many social problem solvers are also lawyers, resulting
in a tendency on their part to create remedies with which they
are familiar.  Further, regulations alleviate the need for people
who are relatively unsophisticated in economic and social theory
to engage in intellectually painful analysis of how their acts
economically effect society as a whole.  Last, Americans favor
regulation of environmental problems because it permits them to
ease their consciences more easily.  This is important for many
because environmentalism has taken on virtually a religious
fervor.

      Regulation is, contemporaneous with its ease in being
understood, economically quite troublesome.  The relative ease in
understanding regulation of non-point source pollution comes from
the use in most regulatory shemes of the prescribed standard:
one standard for all like discharges.  It is this recourse to a
uniform standard that gives rise to the problems, however,
because this solution does not admit of any inefficiency in its
implementation and because the standards are more or less
rigid.  Coupled with this is recognition that there is never a
situation of governmental regulation without a cost of its own,
if in no other way than salaries and the like for administration
of the regulation, the creation of a bureaucracy.  Government
regulation creates distortions in economic efficiencies.  This
inefficieny is caused by the tendency to regulations to require
"equally proportionate reductions" of pollutants from each source
regardless of variance in control costs.*  Although it is easy to
     *Stewart and Trier, Environmental Law and Policy, 556
(Bobbs-Merrill, 1978)
                            P-II-D-7

-------
 prescribe a standard,  it is not so easy to justify the
 standard.  Thus,  determination of standards can be quite
 difficult.

       The third legislative avenue for  dealing with non-point
 source point pollution is to create a tax on discharges.
 Essentially, a tax would take the form of an emission fee per
 unit  of discharge.  While taxing a unit of discharge is a ready
 shibboleth  to the economic sloganeer it is not a concept that is
 free  of significant problems.

       Costs of abatement may vary per unit taxed, which is
 economically good. Taxes do provide an incentive for reduction
 and taxes are more adaptable and more easily administered.
 Nonetheless, the  standards to be achieved tend to be quite
 static.  Changing the  standard to reflect varying efficiencies is
 not easily  achieved.

       The tax concept  has many of the economic inefficiencies of
 the regulation except  the tax furnishes an economic incentive to
 abate the pollution.  Establishing a standard to be achieved is
 at best guesswork aided by political biases.  The unit of '
 detection,  the pricing mechanism, and monitoring are even more
 problematic.

       The last form of legislation that may be utilized to
 control non-point source pollution is the creation of a
 subsidy.    Subsidies take essentially two forms.  One is tax
 credits or  increased depreciation for capital expenditures
 related to  reduction in non-point source pollution discharges.
 Tax incentives and increased depreciation for capital
 expenditures create problems notwithstanding that the government
 in essence  underwrites the efforts being made to relieve non-
 point source pollution because there is a cost to the operation
 which had been making  the discharge. Depending upon the
 discharger,  that  cost  may be serious enough to turn what is
 otherwise a profitable function into a  nonprofitable function,
 primarily when these costs are not common across the competitive
 sphere.   For example,  where the government for any region less
 than  the  whole nation,  such as the State of Wisconsin,  requires
 its farmers to engage  in various non-point source pollution
 elimination activities,  including capital expenditures,  no matter
 what  tax  incentives are  given and no matter what increased
 depreciation is allowed,  there will nonetheless be a capital
 expense  to  the farm operation which farm operations in  other
 states  or regions  do not  have.   In a marketplace where  dairy
 products, grain products,  cattle products are essentially
 fungible, requiring but  one group of competitors in that market,
Wisconsin farmers,  to  take on the additional costs puts  an undue
 hardship  upon  them and puts them at a distinct market


                            P-II-D-8

-------
disadvantage.

      A second form of subsidy is the concept of buying up units
of non-point source pollution discharge.  The legislature would
establish an acceptable level of discharge that must be met and
create a fund with which to eliminate discharge to that level.
For every unit of historic discharge which is not discharged
following the enactment of the subsidy, the discharger would
receive a bounty from the government.  Those dischargers who have
relatively low costs in achieving discharge curtailment will then
rapidly move into that field and will curtail their discharge and
obtain the bounty money; those units with higher costs to curtail
discharging will not be able to partake of the bounty.
Nevertheless, this latter group would have the benefit of being
able to use more of the acceptable limit for their specific
discharges.

      Both subsidiary ideas, however, tend to break down in the
real world because the transference from theory to practice
contains a cost for government administration which cannot be
readily or accurately ascertained.  The cost for ascertaining the
historic level of non-point source pollution discharges is quite
illusive and not readily calculable to the requisite degree of
certitude.
      The four forms of legislative response to the problems of
non-point source pollution each have their own problems.  When
the public wants a solution to non-point source pollution, the
legislature must weigh which is more likely to meet the public's
demand.  The outside influence of the citizen demand for action
can force a choice.

                     III.  PROPOSED SOLUTION

      It seems that the most ready way to deal with non-point
source pollution is to regulate the discharge of pollutants into
the water.  Although it has severe disadvantages, the use of
regulation does have a simple, quick solution aspect to it which
tends to be favored in American society.  Regulation itself could
establish preferred pollution avoidance practices which could be
designated and enforced either by a state agency or as private
rights.

      Furthermore, once legislation takes up the subject matter
of non-source point pollution it may remove it from local
politics. Additionally, regulation need not address the whole
problem at one time.  Regulation would permit the legislature to
address a major segment at a time.  For example, it is not
inconceivable for the legislature to designate the Department of
Natural Resources or the Department Health, Social Services or
the Department of Industry, Labor and Human Relations as the

                            P-II-D-9

-------
authority to identify construction site pollution emission
discharge limitations.  The legislature could then delegate  to
the department the authority within another year to promulgate
regulations on sediment control and yet further regulations  on
highway rion point source pollution discharges.  In this fashion
the legislature would address the entire problem on a timetable.

      The State of Florida has began such efforts.  Its
Department of Environmental Regulation has enacted a number  of
non-point source pollution elimination steps concerning highway
erosion control which have subsequently lessened the discharge of
non-point source pollution to affected water bodies.  Such a role
could be achieved in Wisconsin and in other areas nationwide.

1791L
                           P-II-D-10

-------
                               SUMMARY
                              Panel  III
        Efficiency and Feasibility of Best Management Practices
                     For Nonpoint Source Control
PANEL MEMBERS

Moderator:  W. Trevor Dickinson, Professor of Watershed Engineering,
University of Guelph, Ontario,  Canada

Assistant Moderator:  Robert Biebel,  Chief Environmental  Planner,
Southeastern Wisconsin Regional Planning Commission,  Waukesha,  WI

Panelists:

Bradley Crowder, Institute for Research on Land and Water Resources,
   Pennsylvania State University,  University Park,  PA
John Harkin, Professor of Soil  Sciences and Water Resources,  University
   of Wisconsin-Madison
Jackie Robbins, Department Head and Professor of Agricultural  Engineering,
   Louisiana Technical University, Ruston, LA
Thomas Schueler, Department of Environmental Programs,  Metropolitan
   Washington Council of Governments, Washington, DC
Stuart Walesh, Associate, Donohue and Associates, Waukesha,  WI

Recorder:  David Kendziorski, Southeastern Wisconsin Regional  Planning
Commission, Waukesha, WI

INTRODUCTION

The panel discussed the effectiveness and feasibility of both urban and
rural BMPs, and addressed analytical  techniques for evaluating the prac-
tices.  The need to properly define nonpoint source pollution problems
and evaluation criteria was also noted.  It was acknowledged that  pro-
blems are multi-dimensional, and that implementation of remedies can be
greatly affected by social, cultural, religious, economic, and personal
preferences.

It was questioned whether too much emphasis in evaluating nonpoint source
control measures is placed on modeling, and whether more attention should
be given to monitoring the water quality effects of BMPs.  Panel members
felt that modeling was most valuable for guidance in selecting individual,
or an optimal mix of, BMPs. There should be good feedback between moni-
toring and modeling, with monitoring data being used to calibrate  and
verify models.  The most beneficial use of models is in the insight they
provide, rather than in the absolute quantification of variables.   Models


                            P-III-1

-------
also provide an economical means for assessing alternative conditions
which could not realistically be evaluated with monitoring data alone,
It was concluded that modeling and monitoring programs for nonpoint
source control are not substitutes; rather, they complement each other,

RURAL NONPOINT SOURCE CONTROL

Concerns

1.  Voluntary NFS control programs may not be effective and achieve
    successful implementation in areas with intensive agricultural
    production, particularly livestock raising.

2.  BMPs may be applied for water quality improvement purposes without
    regard    to cost-effectiveness and pollutant removal capability.

3.  BMPs which effectively reduce soil loss by reducing surface runoff
    may not sufficiently control nitrate leaching from excessive
    fertilizer use of manure applications.

4.  Manure storage facilities may simply encourage farmers to expand
    livestock operations.

5.  Cost-effective waste control measures for unconfined livestock
    operations still  need to be demonstrated.

Recommendations

1.  Cost-effective soil conservation measures include contouring, vege-
    tative cover, conservation tillage, and strip-cropping.   Terraces
    are effective, but not relative cost-effective.   Waterways and
    diversions cost less than terraces, but are also less effective.

2.  Nitrate leaching into the groundwater can be controlled by reducing
    heavy applications of manure and fertilizers, (especially in highly
    permeable soils), and by increased education of  farmers.

3.  The most efficient water quality improvement programs would utilize
    the most cost-effective BMPs in critical  NPS problem areas.

4.  Confined livestock waste discharges should be collected and treated
    as a point source.   Ultimate disposal  of the animal  waste should  be
    land application or use as a feed source.   Lagoon treatment and
    direct discharge to surface waters is unacceptable.

5.  Unconfined livestock waste discharges should be  treated  with practices
    consistent with optimal  forage production on a long-term basis.   Con-
    gregations of animals in critical  areas should be prevent.   Grazing
    areas should be maximized and rotated.
                             P-III-2

-------
URBAN NONPOINT POLLUTION CONTROL

Concerns

1.  Proper selection of urban BMPs is difficult because of the wide  range
    in the ability of the BMPs to remove various pollulants.

2.  There is a lack of available BMPs which can achieve a  reasonably
    high level of pollutant removal  in fully developed urban  areas.

3.  Stormwater storage in wet basins, one of the few highly effective
    urban BMPs, cannot be used in areas with high slopes,  low perme-
    ability, or limited space.

4.  Because of the limited knowledge available concerning  the design,
    operation, and effectiveness of urban BMPs, mandatory  specifications
    of BMPs would be premature and would result in "action",  but little
    progress.

Recommendations

1.  Effective urban BMPs include wet basins, porous pavement,  infiltra-
    tion systems, and artificial wetlands.   Dry basins, grassed swales,
    and street sweeping have not been found to be very effective.

2.  Among stormwater storage measures, dry basins cost least; extended
    dry basins cost about 10 more, wet basins 20 to 45 percent more.
    Infiltration systems are cost-effective only in smaller residential
    areas.

3.  Urban nonpoint source control programs are best implemented under
    the guidance of a comprehensive stormwater management  plan which
    addresses both water quantity and quality concerns.

4.  Screening procedures should continue to be developed to assist local
    developers in selecting those urban BMPs which are most appropriate
    within a given range of site characteristics.

5.  Stringent inspection and maintenance programs must be  developed  to
    help ensure the continued effectiveness of urban BMPs  especially
    storage facilities.

CONCLUSIONS

Following review of the urban and rural BMPs available, some  panel members
expressed concern about the inability to specify the relative cost-effec-
tiveness, and to quantify the pollutant removal effectivenss  of individual
BMPs.  The panel concluded, however, that many excellent studies and de-
monstration projects have been completed, and implementation  of both
urban and rural practices should proceed.  More will be learned as these
practices are implemented on a wide scale.   The advancing  state-of-the-
art have provided a strong direction toward which types of BMPs work best

                              P-III-3

-------
under certain conditions.  The panel expressed confidence that the techr
niques are available, and the knowledge exists, to sufficiently control
both urban and rural nonpoint source pollution in most areas.  New and
emerging technologies will likely "fine tune" the system, enhancing
water quality improvements achievable with the present range of BMPs.
Inability to guarantee that any BMP selected for implementation is the
wisest choice should not be used as an excuse for inaction.   Immediate
prudent use of existing patently cost-effective BMPs should be encour-
aged because they are all steps in the right direction.
                             P-III-4

-------
                EVALUATING BMPs IN PENNSYLVANIA'S CONESTOGA
                    HEADWATERS RURAL CLEAN WATER PROGRAM

                   Bradley M. Crowder and C. Edwin Young
          Agricultural Economists, USDA-Economic Research Service
                     The Pennsylvania State University
                     University Park, Pennsylvania, USA

                                  ABSTRACT

This paper discusses the relative cost effectiveness of best management
practices (BMPs) for controlling agricultural nonpoint pollution.  The
discussion herein applies directly to the control of soil and nutrient losses
resulting from crop production, and is derived from the economic evaluation
of the Conestoga Headwaters Rural Clean Water Program (RCWP) project in
southeastern Pennsylvania.  The purpose of the project is to provide
assistance for the installation and maintenance of BMPs to control nonpoint
pollution and improve water quality.

Due to the current status of BMP implementation in the project, monitoring
data are not available to evaluate the impacts of BMPs on water quality.
Physiochemical and economic modeling of BMP impacts was undertaken to
characterize the cost effectiveness and overall effectiveness of BMPs for
controlling agricultural nonpoint pollution.  CREAMS, a field scale model for
Chemicals, jUmoff, and Erosion from Agricultural Management ^Systems, is a
continuous computer simulation model that was used to estimate field-edge
losses of soil, surface runoff losses of N and P, and NO^ leached out of the
root zone.

Contouring and stripcropping (where the crop rotation makes it feasible) were
the most cost-effective practices for controlling erosion and surface runoff
losses of nutrients.  Conservation tillage practices were also cost-effective
alternatives.  Terrace systems were highly-effective runoff control measures
but were generally the least cost effective.  Other structural practices,
waterway systems and diversion systems, were found to be less effective than
terraces for soil loss control, but were more cost effective.

NO^ leaching, perhaps the project's most significant water quality problem,
may not be alleviated and may increase when soil conservation practices are
implemented.  Sound nutrient management practices are necessary to control
subsurface and surface deliveries of nutrients to water supplies.  Nutrient
inputs from animal wastes, commercial fertilizer, and crop residues must be
managed in a way that prevents excessive inputs and limits movement off the
field.  Manure storage and application techniques should be consistent with
maximizing the economic value of nutrients while minimizing field losses.


Keywords:  Agricultural nonpoint pollution, best management practices,
           economic modeling, cost effectiveness, water quality, CREAMS.
                                   P-III-A-1

-------
 INTRODUCTION;

 Increasingly  intensive  crop and animal production have characterized
 southeastern  Pennsylvania's Conestoga Headwaters Watershed and the remainder
 of  Lancaster  County.  This production intensity has been motivated by:
 (1) urban  pressures on  agricultural land, especially near the city of
 Lancaster;  (2)  Amish and Mennonite family farming traditions, which have
 resulted in the division of farmland into small parcels among children; and
 (3) proximity to large  East Coast markets.  The watershed is 1 of 5
 experimental  RCWP projects nationwide to be selected for comprehensive
 monitoring and  evaluation.  There are 44,516 hectares of land, of which 26,305
 is  in farms.

 The "average" farm in the watershed has 21 hectares of land, almost entirely
 used to produce feeds for animal operations that usually exceed 5 animal units
 (AU) per hectare of farmland  (USDA, 1982).  Over a 13-year period from 1970 to
 1983, the  number of milking cows increased by 67 percent, swine production
 increased  by  189 percent, and poultry production increased by 145 percent.
 During the same period, the percentage of total field and forage crops that
 was in corn grain and silage increased from 33 and 12 percent in 1970 to 36
 and 24 percent  in 1983, respectively (Pennsylvania Crop Reporting Service,
 1970, 1983).  Commercial N fertilizer purchases increased over the same period
 to  meet the perceived needs of the additional corn.

 Manure is  applied at an average annual rate of 90 MT per hectare in the
 watershed.  The average fertilization per hectare for corn, from animal manure
 and commercial  fertilizers, is:  482 kg N, 304 kg ?2Q5' and 3Ql* k8 K2°
 1983).  These high nutrient inputs and the intensity of agricultural
 production have resulted in local water quality degradation.
WATER QUALITY PROBLEMS

Agricultural nonpoint pollution in the Conestoga Headwaters Watershed affects
the water supplies of 175,000 people.  The water quality parameters of
greatest concern are:  (1) high NO^ concentrations in groundwater,  especially
in private wells; (2) fecal coliform (FC) bacteria in surface and groundwater
supplies; (3) triazine pesticides in both surface and ground waters; and (4)
sediment and turbidity in streams.

Most private wells exceed the Environmental Protection Agency (EPA) drinking
water standard of 10 ppm NO^-N, with many exceeding 30 ppm N03~N concentration
during the year.  Abnormally high levels of N, P, pesticides, and FC bacteria
also have been found in streams and wells (USDA, 1983).  Average annual soil
losses are estimated to be 19.7 MT per hectare in the watershed, with annual
losses from individual fields often exceeding 100 MT per hectare.  EPA has
identified southeastern Pennsylvania agriculture as a major contributor to the
Chesapeake Bay of N (especially N03 during spring runoff events), toxic
substances, sediment, and to a lesser extent P (EPA, 1983).

It is the goal of the RCWP project to substantially reduce the delivery of
agricultural pollutants to receiving waters by implementing conservation and
                                  P-III-A-2

-------
other BMPs on farms.  The purpose of this paper is to discuss the effective-
ness and cost effectiveness of selected management practices for controlling
field losses of soil and nutrients and ultimately for improving water quality.
This discussion is directed toward an area characterized by erosion problems
and high manure inputs.
BMP EFFECTIVENESS FOR CONTROLLING POLLUTION

Farmers in the RCWP project area currently are implementing contracted BMPs.
Water quality is being monitored and data are available only for current
management practices.  It could be years before newly installed RCWP practices
cause measurable changes in water quality.  As a result, computer modeling was
chosen to evaluate selected RCWP practices.

Modeling soil and nutrient losses

In order to isolate the effects of BMPs, both individually and in combination,
the CREAMS model was used to obtain estimates of field-scale soil and nutrient
losses.  CREAMS is a computer simulation model, constructed by USDA's
Agricultural Research Service to compare relative field losses of soil and
selected chemicals among different management practices (Knisel, 1980).
Computer modeling provides a method of estimating BMP effectiveness for
pollution control without incurring the financial and time costs of BMP
implementation and monitoring.

CREAMS estimates of field losses were used with economic optimization and
budgeting techniques to evaluate the cost effectiveness of:  (1) field BMPs;
(2) manure storage and handling systems (and their effects on nutrient
conservation/availability and field losses); and (3) manure transport (in
order to reduce high rates of manure application).  Surface runoff losses of
soil, N, and P, and subsurface losses of NOj leached below the root zone, were
estimated using CREAMS and published data (see USDA, 1984).  Pesticide losses
in surface runoff also can be modeled with CREAMS, but are not addressed in
this paper.  Deliveries of soil and nutrients to surface and ground waters can
be assumed to be some fraction of the CREAMS estimates of field losses.

Soil losses

Table 1 shows the field BMPs that were modeled, their approximate costs, and
the reductions in losses of soil and nutrients associated with the BMPs.  The
baseline soil loss was 24.3 MT per hectare per year.  This represented the
soil loss for conventional practices — tillage up and down the slope with a
moldboard plow followed by 2 passes with a harrow.  A distinction is made
between nonstructural management practices and structural practices.
Permanent vegetation, hay or pasture, is a highly-effective soil conservation
practice.  It also appears to be relatively cost effective, with cost
effectiveness defined here as the percent reduction in field loss per dollar
of expenditure.  Permanent vegetation is applied with RCWP cost sharing only
on severely erosive land that generally is adjacent to waterways.  The
opportunity cost of removing land from row crop production is prohibitive

                                  P-III-A-3

-------
without 50-percent government cost sharing, and therefore the practice has
limited applicability.

Contour stripcropping involves performing tillage operations along the
contours of the fields, and alternating row crops with close-seeded hay or
small grain crops in strips (modeled to be about 37 meters in width using
CREAMS).  The effectiveness of stripcropping is primarily due to the reduction
in average surface runoff caused by the strips of close-seeded crops, not
accounted for by modeling the strip of corn in Table 1.  Soil loss was reduced
by 39 percent at no cost, making contour stripcropping a highly cost-effective
practice for erosion control.

Conservation tillage practices — reduced tillage (chisel plowing followed by
harrowing once) and no-till — were found to be both effective and cost
effective for erosion control.  No RCWP cost sharing is provided to convert
tillage practices.  This is consistent with findings by Crowder, et al. (1984)
that conservation tillage practices reduce ownership and operating costs,
resulting in greater profitability.

Winter cover and residue management are important components of any
conservation, tillage system.  When combined with the zero-cost practices of
conservation tillage and contour stripcropping, the overall management system
provides significant erosion control for the cost of winter cover and residue
management, $0 to $50 per hectare annually (depending on the crop residue).
The more important structural practices being implemented with RCWP cost
sharing include:  (1) pipe-outlet terraces; (2) diversion systems;  and (3) sod
waterway systems.  The first column of Table 1  shows the approximate annual
costs per hectare served for structural practices, amortized at 12 percent
interest over 10 years.

Terraces are highly effective for erosion control, but at relatively high cost.
If a 10-year useful life is assumed, the annual cost per hectare is $163.70.
With 75-percent cost-sharing rates, however, terrace systems can be made
attractive to farmers.  Sod waterways and diversions may be more effective for
preventing soil loss to streams than for reducing erosion.  Given these
circumstances, they are probably more effective for protecting water quality
than soil productivity.  Because they are considerably less expensive than
terrace systems, sod waterways and diversions appear most cost effective than
terraces for protecting water quality.

The last entry in Table 1 illustrates the reductions in pollution resulting
from a typical combination of soil conservation practices.  An 89 percent
reduction in soil loss is possible at an approximate annual cost of $225 per
hectare.  The cheapest combination of management and structural practices can
be chosen that will meet targeted soil loss reductions on a given field.  In
this way, soil losses can be reduced at minimum cost to both farmers and
taxpayers .
                                  P-III-A-4

-------
     Table 1.   Annual  costs  and effectiveness of conservation practices for continuous  corn  silage,  daily spread
               45  metric  tons manure per hectare per year, 5 percent slope, Duffield silt loam.
 I
M
H
M

>

Ul

Conservation
practice
Conventional
practices
Permanent vegeta-
tive cover
Contour strip-
cropping
Winter cover and
residue manage-
ment
Chisel plowing/
reduced tillage
No-till
Terrace system
Diversion system
with 6-meter wide
sod filter strip
Sod-waterway system
Reduced tillage
TJ*Q 4~ 1 m at" a f\
CiSt/imaueu
annual cost
per hectare
N/C

$83*

N/C

$0-50


N/C**

N/C**
$163.70*
$30.09*


$20.35*
$225

Soil
__

95

39

14


44

68
72
43


64
89

— — — rGFCGnU PGQUCulOn
N in
surface runoff
	

87

30

11


34

50
56
42


49
78
1 n f* \ ^1 t\ 1 f\QQ&Q
in i iej.u losses
Total N
__

45

18

9


20

38
32
24


29
43

Total P
_»__

91

33

13


37

54
62
46


52
82
       with terraces,
       contouring,
       stripcropping,
       residue manage-
       ment ,  and sod
       waterway	
      ^Permanent vegetation was  amortized over  5  years  at  12 percent  interest,  and  is  cost  shared at  rates  up to
       50 percent.   Structural practices were amortized over 10  years at  12  percent interest.   Cost-sharing rates
       of 75 percent on structural  practices reduce  farmers' annual before-tax  costs to  25  percent of those shown.
     **Cost of winter cover and  residue management must be  included for continuous  corn.
       Source:  USDA (1984).

-------
 Nutrient  losses

 As  part of  the RCWP economic evaluation, *J approaches to farm nutrient
 management  and conservation were identified:

      1.   Use field conservation practices (BMPs) to control surface runoff
          losses of N and  P.
      2.   Reduce animal manure production and/or commercial fertilizer
          applications .
      3.   Improve the allocation and application of manure and nutrients
          throughout the farm.
      4.   Export manure from farms with excess nutrient problems.
Nutrient losses in runoff can be reduced by implementing conservation
practices which restrict surface runoff and soil loss.  The last 3 columns of
Table  1 show the reductions in nutrient losses for the BMPs discussed earlier.
Annual losses for conventional practices were 58 kg N per hectare in surface
runoff (of which almost 90 percent was sediment associated), 97 kg total N per
hectare (39 kg of NO^-N per hectare was leached through the root zone), and
28 kg total P per hectare in surface runoff.

The effectiveness and cost effectiveness of soil conservation practices for
controlling surface runoff losses of nutrients parallel those for controlling
soil loss.  This follows because most nutrient losses in surface runoff were
sediment-associated for the baseline conditions modeled.  However, soil
conservation practices have little effect on controlling NC>3 losses in
percolate.  In fact, such practices may actually increase the load of NO^ lost
by leaching because of the increased water percolation through the soil
profile, caused by reductions in surface runoff.  From the perspective of N
mass, the gains in N control are significant although not proportionate to the
reductions of soil and P losses.  However much, and perhaps most, of the
sediment-associated N will not become biologically active, so the efficiency
of erosion control methods for controlling biologically active N losses is
even less than projected in Table 1.

As for other water quality parameters discussed, permanent vegetative cover
and the combination of BMPs (last entry in Table 1) are the most effective for
controlling total N losses, followed by no-till, terraces, and sod waterways.
Diversions, reduced tillage, and contour stripcropping estimates indicate that
they can significantly reduce total N losses.  Among individual practices
(ignoring permanent vegetative cover due to its limited applicability and
effects on farm forage production), contour stripcropping, no-till, and then
reduced tillage are the most cost-effective nonstructural practices for
controlling total N losses.  In addition, nonstructural practices have the
advantage of keeping soil and nutrients in place on farm fields.  For
structural practices, total N loss reduction per dollar of cost is greatest
for sod waterways, then diversions, and last terraces — the same cost-
effectiveness ranking as for soil and P losses.

Decision makers should be aware that groundwater and stream pollution problems
are not likely to be alleviated by the implementation of soil conservation
practices alone if NOg is a critical water quality parameter.  Improvements in

                                  P-III-A-6

-------
surface water quality from surface runoff control may be offset by increased
NOg loads in groundwater and base flow.  When groundwater NO^ levels are a
concern, as they are in the RCWP project area, nutrient management practices
probably will be necessary to improve water quality  substantially.

jteducj^_nut£ient^ Applications

The second way to reduce farm nutrient losses is to match manure and
fertilizer nutrients more closely to crop needs.  Available nutrients in
manure and legumes should be credited toward the fertilizer needs established
by soil test recommendations.

Part of the over-application of N could be prevented if farmers had
inexpensive, accurate estimates of available soil N.  Unfortunately, such
estimates are not readily available, and the problem is further compounded
because the N that becomes available from current and previous manure appli-
cations is not easily predicted.  Due to these uncertainties, it makes
economic sense to apply excess N commercial fertilizer as insurance to
guarantee crop yields.  Over-application of nutrients would be expected to
decrease as the degree of uncertainty is reduced.

An example of the effects of over-application of nutrients, due to excess
manure (which is responsible for much of the nutrient problem in the project
area), is shown in Table 2.  A manure application of 45 MT per hectare results
in substantially less nutrient loss than the average watershed application of
90 MT per hectare.  Fertilizer and manure management plans are being imple-
mented through the RCWP program.  Better timing of manure applications to
enhance crop uptake and prevent field nutrient losses, and eliminating the
large commercial N fertilizer applications that exceed crop requirements, can
reduce the stream delivery of nutrients as well as other pollutants.  Reducing
fertilizer purchases also reduces farmers' costs.  Reducing the application of
manure nutrients is difficult when excess manure is produced on a farm.
Storage, handling, and application systems which encourage N volatilization
can be used, or the number of animal units on the farm can be reduced — not
an attractive option for farmers.

                            f animalmanur
Significant reductions in farm nutrient losses may be obtained by applying
manure evenly throughout farm fields, and by properly storing manure during
crop growth periods and the winter months.  For example, variation in soil NOj
levels within a single field varied from 3.3 to 230 kg of N03~N per hectare.
This was attributed primarily to "a relatively poor job of spreading nutrients
from manure."  It was predicted that most groundwater pollution results from
those sites on fields with extremely high soil NOg levels (USDA, 1984).  It is
believed that significant nutrient conservation and pollution reductions are
possible when manure nutrients are spread evenly on fields.  Such a claim,
however, is difficult to quantify.

Table 2 provides an illustration of the estimated nutrient losses for
different manure storage periods and application methods.  Losses shown are
field losses only, and do not account for barnyard runoff losses (which
usually are significantly greater on farms with unimproved barnyards, and
on farms where manure is stacked and spread periodically — a practice common


                                  P-III-A-7

-------
 Table 2.  Effects of storage period on nutrient losses (kg/ha)  for  corn silage
          following corn silage, 5 percent slope, Duffield silt loam,  45 and
          90 metric tons of manure annually per hectare.
Application
          Nitrogen Losses
Percolate     Surface        Total
                           Phosphorus  Losses
                                 Total
Daily spread
on surface
   41
13
                                            1,5 MT/ha
Six-months
storage with
plowdown

Six-months
storage with
injection

Twelve-months
storage with
plowdown

Twelve-months
storage with
injection
   45
   33
   20
12
                13
13
15
 57
               58
 35
Daily spread
on surface

Six-months
storage with
plowdown

Six-months
storage with
injection

Twelve-months
storage with
plowdown

Twelve-months
storage with
injection

66
82

25
21
yu ni/na - 	
91
103

9
8
   86
   62
   72
24
24
28
110
 86
100
 Practices include reduced tillage,  winter cover,  contouring,  stripcropping,
 with terraces and sod waterways installed.
 Source:  USDA (1984).
                                 P-III-A-8

-------
among farmers in the RCWP project area).  Manure storage systems conserve
nutrients, especially N, which is undesirable when plant-available N from
animal manures alone exceeds county crop nutrient requirements by 35 percent.
The P and K available from manures exceed crop nutrient requirements by  170
and 50 percent, respectively (USDA, 1983).

Field nutrient losses were similar for daily spreading and 6-month storage
systems at an annual loading rate of 45 MT per hectare of manure (Table  2).
Percolate NC^-N and total N losses were substantially less for 12-month
storage, an effective strategy for reducing nutrient loss.  Also, nutrient
conservation by manure storage can offset most, if not all, of the amortized
costs of a storage structure.  The economic modeling results indicate that,
for a small 45-cow dairy with 24 hectares of tillable land (resulting in
slightly less than 45 MT of manure per hectare applied to corn land), a
6-month earthen basin liquid storage system can increase annual net returns to
the farmer by $750 (with no cost sharing for the structure) compared to  a
daily spread system.  A 6-month solid storage system was found to decrease
annual net returns by $700 compared to daily spread.

If, however, a farmer in the RCWP project area stores manure and applies it at
the average rate of 90 MT per hectare, total N losses are about the same or
substantially greater than the daily spread system (Table 2).  If a farmer has
excess manure nutrients, a storage structure no longer pays for itself because
the marginal productivity of the conserved nutrients is no longer positive.
Some reduction in total nutrient losses is possible by storing manure, but at
substantially more cost than for farmers who do not have excess nutrient
problems.

A critical concern for farmers is the effect of on-farm nutrient-loss
abatement on net income.  For the representative small dairy farm mentioned
above, the results indicated that it was possible to reduce field losses of
nutrients by storing manure, by uniform spreading of manure, and by growing
less erosive crops or rotations on erosive fields.  Further low-cost
reductions in nutrient losses can be obtained by spreading manure evenly
within the field borders, by performing conservation tillage operations  along
field contours, and other conservation management practices discussed before.
These more cost-effective practices should be pursued to reduce field losses
before those which significantly reduce income and cost the government large
amounts of cost-sharing funds.  In addition, if manure storage is planned on
farms with high animal-to-land ratios, it should be designed not to conserve
but to release unwanted nutrients by chemically degrading them to non-noxious
or biologically unavailable compounds.

_0f£-f_ar_m_di_s£os_al_ of_manure_

The last approach to reducing farm nutrient losses involves transporting
excess manure to other locations, where its application will not be environ-
mentally harmful and where its nutrients can be utilized more fully.  With the
volume of excess manure that exists in Lancaster County and surrounding areas,
it is not anticipated that profitable manure marketing (for farmers with
excess manure) is possible on a wide scale.  The trend toward more intensive
dairy and animal production will exacerbate the problem, and probably will
require future manure hauling over longer distances than those required
currently for safe disposal.

                                  P-III-A-9

-------
Off-site manure disposal is costly for a number of reasons.  First,  manure is
a bulky material to handle and transport.  Second, ensuring safe application
to land, so as to minimize nutrient losses, limits the time period that manure
can be applied.  Manure applications must be compatible with crop growth
cycles.  Thus, manure must be stored for up to 10 months, to allow most of the
manure to be applied prior to spring planting of row crops.  Last, sufficient
hauling and application equipment should be available to apply manure during
this limited time period.  Such equipment will stand idle during the remainder
of the year.,  Therefore, manure hauling over long distances is not a low-cost
method for reducing nutrient losses.

An option for hauling manure up to 10 miles, from the representative small
dairy farm discussed earlier, was modeled using linear programming.   It was
assumed that farmers would pay for hauling and give the manure free to farmers
receiving it in order to reduce manure applications on the originating farm.
The receiving farmer would pay for field application.  Using the CREAMS
estimates of N loss, it was found that for a farmer using a daily spread
system, income was reduced by about 20 percent when manure was hauled
40 miles to reduce total N losses by 30 percent.  It was found under the
assumptions employed, that if a farmer stored manure for 6 months in an
uncovered solid storage structure, income was reduced by only about  8 percent
while achieving the same 30 percent reduction in total N losses.  Thus, manure
storage structures may be necessary to cost-effectively institute off-farm
manure disposal in the RCWP project area and elsewhere.  Cost-sharing funds
provided by RCWP and other programs for storage structures may make manure
export economically viable for farmers with manageable manure problems.  It is
expected that manure export will impose substantial income penalties on
farmers who have substantially more manure than can be used on their farms,
unless government subsidies are provided.

Uses of manure other than fertilizer, most notably as fuel for power
generation, usually require economies of scale at the farm level to  be
profitable.  Refeeding manure to livestock causes health problems when done at
significant levels.  Ongoing research on these and other alternatives may
provide hope for profitable off-farm manure utilization in the future.
CONCLUSIONS

It is apparent that while reductions in soil and P losses from farm fields  can
be achieved with soil conservation practices, the control of N losses  (as well
as FC bacteria) is more complicated.  Reducing surface runoff and its
constituents results in increased percolation of water through the soil.  This
potentially increases leaching of N03,  pesticides, FC bacteria,  and some small
amount of P to groundwater.  In short,  solving one problem likely aggravates
the other.

A comprehensive plan to promote sound management of manure,  fertilizer, and
pesticides is necessary to improve water quality in the RCWP project area.
The rigorous application of manure management and soil conservation practices
could impose severe financial hardship on farmers if government  assistance  for
such practices is not continued.  Even with the traditional  forms of
assistance provided through RCWP and other programs, farmer  participation in

                                 P-III-A-10

-------
conservation planning has been poor in the project area due to the religious
and cultural mores of the farm population.

A number of problems must be addressed to solve southeastern Pennsylvania's
agricultural pollution problems.  First, the level of pollution reductions
necessary to meet water quality goals must be identified for critical water
quality parameters.  Second, those watersheds which have pollution problems,
and the critical areas within the watersheds, should be identified.  The most
cost-efficient combination of practices should then be identified and selected
to meet the pollution reductions.  Third, implementation policies are required
that will achieve the necessary farmer participation to efficiently and
equitably obtain the desired pollution reductions.  Finally, the trend toward
increasing the intensity of land use should be constrained.  Future expansions
of animal operations should be conditional on the adoption of adequate
nutrient control technologies.  Gains made through better farm management
could be negated if cropping intensity and animal intensity continue to
increase without these controls.

REFERENCES

Crowder, B.M., Epp, D.J., Pionke, H.B., Young, C.E., Beierlein, J.G., and
     Partenheimer, E.J. (198M).  The Effects on Farm Income of Constraining
     Losses of Soil and Plant Nutrients:  An Application of the CREAMS
     Simulation Model.  Bulletin 850, Pennsylvania Agricultural Experiment
     Station, The Pennsylvania State University, University Park, Pa.

Knisel, W.G., Ed. (1980).  CREAMS:  A Field Scale Model for Chemicals, Runoff.
     and Erosion from Agricultural Management Systems.  Conservation Research
     Report No. 26, U.S. Department of Agriculture-Agricultural Research
     Service, Washington, D.C.

Pennsylvania Crop Reporting Service (1970, 1983).  Pennsylvania Crop and
     Livestock Annual Summary.  Harrisburg, Pa.

U.S. Department of Agriculture, Agricultural Stabilization and Conservation
     Service (1982).  Conestoga Headwaters Rural Clean Water Program - Plan of
     Work.  Harrisburg, Pa.

U.S. Department of Agriculture, Agricultural Stabilization and Conservation
     Service (1983, 1984).  Conestoga Headwaters Rural Clean Water Program
     Progress Report.  Harrisburg, Pa.

U.S. Environmental Protection Agency (1983).  Chesapeake Bay:  A Framework for
     Action, Appendices.  Philadelphia, Pa.
                                  P-III-A-11

-------
                          BEST MANAGEMENT PRACTICES
                            FOR ANIMAL PRODUCTION

                            Jackie W. D. Robbins
                             Professor and Head
                     Agricultural  Engineering Department
                          Louisiana Tech University
                           Ruston, Louisiana, USA
                                  ABSTRACT

The environmental  effects of grazing animals and how to control  these are
presented.  In general, pollutant yields from pasture and range!and operations
are very low.  They are not directly related to the number of animals or
amounts of wastes  involved.  Rather, they are intimately related to hydro-
geological and management factors and are best described as the results of
the erosion/sediment phenomenon.  Control is related to following practices
that are consistent with optimal forage production on a long-term, sustaining
bases and regulating animal movement/congregational  patterns to minimize
development of highly impacted areas.


Keywords:  Animal  wastes, Pasture, Water pollution, Agricultural  wastes,
           Agricultural land runoff, Nonpoint pollution, Rangeland,
           Livestock, Animal husbandry
                                   P-III-B-1

-------
 INTRODUCTION

 This  report summarizes state-of-the-art knowledge related to the yield and
 control of water  pollutants  from animal production systems.  Using findings
 from  recently  completed  research projects, this report updates earlier work
 by the author  that emphasized the character of the problem (Robbins, 1979).
 Herein, emphasis  is on control techniques or best management practices (BMP's)
 aimed at  keeping  wastes  from reaching surface waters.

 By way of orientation, the difference between confined and unconfined animal
 production is  that wastes generated in confined systems are subject to
 handling/diverting for conventional control or treatment, while those in
 unconfined systems cannot be handled and, thus, must be controlled through
 the management/operation scheme.  Correspondingly, the contrast between
 confined  and unconfined  animal production systems is very similar to the
 differentiation between  point and nonpoint sources of pollution, where a non-
 point source is one whose specific point of generation and exact point of entry
 into  the  environment cannot  be defined.  All grazing systems—where livestock
 have  free access  to pasture, range, woodland, or cropland and utilize the
 associated forage/residue as a major feed source—are unconfined systems.

 About 40% or 360  x 106 ha of the land area of the United States is used for
 grazing livestock.  Forages, which in the main are grazed, account for the
 production of more than  50% of the nation's milk, nearly 80% of the total feed
 for beef  cattle,  and 90% of  the total  feed for sheep.  Less than 5% of swine
 feed and  almost no poultry feed comes from grazing.

 While approximately 50% of all livestock wastes in the United States are
 estimated to be produced in confinement, about 75% of cattle, 85% of sheep,
 and 10% of hogs are in unconfined systems at any given time.   Table 1 lists
 the numbers of animal  production systems and animals maintained in unconfined
 production systems.  The vast numbers of individual  units involved in
 unconfined livestock activities are reflected by the values given for numbers
 of production units (about one million).  Although specialization, labor
 reduction, and improved efficiency have caused a trend toward confined
 production systems, unconfined production is expected to continue to
 predominate the beef and sheep industries.

 IMPACT ON PASTURE AND RANGELAND

 As grazing animals traverse pasture and rangeland, the stresses applied to
 the soil  beneath  their hooves often exceed the strength of the soil.
 Trampling may cause poorer water infiltration.   Reductions in infiltration
 rates due to grazing correlate with increases in runoff as indicated by
 numerous studies.   For example, White et al. (1982)  found that where winter
 feeding of cattle occurred, the runoff volume was three times that of normal.
Winter rotational  grazing increased runoff by 75% and summer  rotational
 grazing only,  increased runoff by 17%.   The increased runoff  volumes  were
 primarily associated with the dormat season, winter  feeding and pasturing.
Trimble et al.  (1951)  measured average infiltration  rates of  52 and 11  cm/h
on ungrazed and grazed pastures, respectively.   In a New Mexico study of
grazing intensity-infiltration relationships, infiltration rates were
determined to  be  10.5, 5.5, and 2.1 cm/h on undergrazed, overgrazed,  and
depleted  ranges,  respectively (Flory 1936).

                                   P-III-B-2

-------
           TABLE 1.   DISTRIBUTION OF LIVESTOCK (After Robbins, 1979)
   Type Animal
  No.  of
Production
  Units
 No. of
Animals
  No.
Confined
No. Un-
Confined
  Percent
Unconfined
Cattle and Calves
   Beef Cattle
   Dairy Cattle

Sheep and Lambs

Goats and Kids

Hogs and Pigs

Horses and Ponies
   490
   220

    50

    10

   260
                                     Thousands
123 000
101 500
 21 000

 12 700

  1 300

 55 000

  8 000
 30 000
 20 000
 10 000

  2 000

    200

 50 000
 93 000
 81 500
 11 000

 10 700

  1 100

  5 000
     76
     80
     52

     84

     85

      9
                                   P-III-B-3

-------
 After  a  comprehensive  review of over a dozen studies, Moore et al.  (1979) were
 able to  summarize  that light and moderate intensities of grazing during
 unsaturated  soil conditions will not decrease infiltration rates.  However, a
 concentration of livestock is likely to produce significant decreases in
 infiltration.  Adverse impacts on sheet and rill erosion and water quality
 constituents are an outcome of these changes.  The phenomena of heavy
 trampling damage to infiltration rates may be an insignificant problem on fine-
 textured high clay content soils where undisturbed rates are less than 25 cm
 per hour.  It is well  documented that periods of non-grazing, however built
 into the grazing system, allow certain natural processes to progressively
 improve  soil water infiltration rates to predisturbance levels.

 A decrease in infiltration and increase in runoff on grazed lands may lead to
 more arid conditions than normal.  For example, Hanson et al. (1970) found
 that the reduction in  available moisture on heavily grazed South Dakota
 ranges averaged 2  cm a year, or about 8% of the annual precipitation available
 for plant production.  A decrease in moisture available for forage  production
 is usually accompanied by an increase in soil erosion.

 Selective defoliation  by grazing animals decreases the preferred forage species
 and allows the proliferation of less desirable species.  Close grazing of
 desired  plant species  during critical stages of growth, such as before seed
 maturation, hinders their competitive capacity.  Reduction of the leaf area of
 forages  by close grazing reduces their photosynthetic abilities and may affect
 plant succession (Stoddart et al., 1975).  Heady (1975) attributes  the most
 prevalent cause of range!and retrogression to overgrazing and other faulty
 management of livestock operations.

 The general congregational and waste elimination behavior patterns  of live-
 stock follow ancestral instincts, but may be influenced by management
 practices.  Behavior patterns and preferences of livestock species  tend to
 result in certain  areas showing intense environmental  stresses.  These
 "problem areas" are usually only a small fraction of the total  unconfined
 production site and are caused by unregulated animal  behavior that  leads to
 selective—and over—grazing (of sites as well as plant species) and to animal
 congregation and waste elimination in or near water sources or other critical
 areas.   While only a small portion of the total  production area, these problem
 areas often contribute the major portion of pollutants and constitute the only
 sources that when  controlled results in a discernable improvement in runoff
 quality.

 BACKGROUND LEVELS OF POLLUTANTS

 In assessing the environmental  impact of unconfined animal  production, the
 concept of "background (natural) levels" of pollutants is important.   From
 a study of 12 agricultural watersheds, Robbins et al. (1971) found that dis-
 tinguishing between pollutants  from farm animal  production units and natural
 pollutants in receiving streams is difficult or impossible.  More recent
 studies designed to establish the contributions of pollutants from  unconfined
 animal  production to surface waters by Moore et al.  (1979), Doran et al.
 (1982), White et al.  (1982),  Powell  et al.  (1982),  and Saxton et al.  (1982)
all  reemphasize that control  of pollutants  from unconfined animal production
units may be to no avail  unless other pollutant sources that naturally occur
 in the  watershed are controlled as well.   When management is directed to

                                   P-in-B-4

-------
optimizing forage production, pollutant yields from unconfined animal
production systems are not more than would occur under native conditions.
Only problem areas as noted earlier and areas used more to confine animals
rather than to graze them would seem to be subject to corrective actions.

EROSION/SEDIMENT

Sediment is both a pollutant and a carrier of pollutants.  As a first
approximation, factors that govern erosion and sediment yields are the same
factors that control pollutant yields from unconfined animal  production
systems.  Numerous studies including Robbins et al. (1978), Smeins (1976),
Moore et al. (1979), Saxon et al.  (1982), White el al. (1982), and Powell
et al. (1982) have pointed to the  need for good soil  and water conservation
practices to minimize the movement of pollutants from animal  production units
into streams.  Excellent, detailed and thorough reviews and discussions of
erosion, erosion models, and sediment yields as related to agricultural non-
point sources have been prepared by USEPA (1973), Stewart et al. (1975),
and Sweeten and Reddell (1978).  Sediment yield to streams and lakes exceeds
2 t ha'lyr"! on the average.  On-site erosion is estimated to be twice this
value for an average of more than  4 t ha~lyr~l or a total of 3.6 x 10^ t/yr
for the total U.S. land area.  According to Froehlich (1976), the smallest
levels of sediment yield are from  certain undisturbed forestlands in the
Rocky Mountains where sediment reaching high elevation streams may range up
to 0.36 t ha-lyr'l.  Other examples of background sediment yields are  0.59,
0.90, and 1.10 t ha-1yr~^ from three neighboring watersheds in western Oregon.
Well stocked southern pine forests my yield 0.7-1.1 t ha"lyr~l.  In south-
western forests with low precipitation, yet high intensity storms, sediment
yields average 1.1-1.4 t ha~lyr~l.

Background sediment yields may be  compared with yields from cropland.   Crop-
land has been credited with responsibility for 50% or 0.9 x 10^ t/yr of the
1.8 x 10^ t/yr total sediment delivered to U.S. streams and lakes. As noted
by Sweeten and Reddell (1978), 70% of the nation's cropland yields more than
6.7 t ha'iyr'l.  Representative values of on-site erosion from several sources
as reported by USEPA (1973) are given in Table 2.  Here, grassland includes
pasture and rangeland.  While the  erosion rate from grassland is 10 times
that from forestland, it is considerably less than the average rate of
4 t ha~lyr~l for all land; and it  represents the average background or
natural rate for grassland ecosystems.

Sediment yields are very erratic.   The largest annual sediment loads for a
given stream are often 20 times greater than the smallest sediment load and
generally can be correlated with those years of greatest runoff.  Large
differences in sediment yield can  exist on adjacent streams discharging at
the same rate.  And, even in the same stream, suspended sediment concentration
can vary tenfold at a given discharge rate, depending on many factors
(Froehlich, 1976).

Most sediment from a watershed may come from a relatively few small areas
needing corrective attention.  Stewart et al. (1975)  listed conditions
indicative of high sediment yield  potential that usually can be identified
by observation.  Overgrazing and resultant loss of groundcover can increase
greatly the credibility of pasture and rangeland.


                                   P-III-B-5

-------
   TABLE 2.   REPRESENTATIVE RATES OF ON-SITE EROSION FROM VARIOUS LAND USES
                                       (USEPA 1973)
                        Rate,      Relative Rate,     Total,     Relative Total
    Land Use	t/ha yr	Forest = 1	t/yr	Forest = 1

Forest                   0.085            1             16.8            1

Grassland                0.85            10            185             11

Cropland                17              200           2840            168

Harvested Forest        42              500            187             11

Construction           170             2000            100              6
                                   p-iil-B-6

-------
Results from four test sites that exemplify water quality effects resulting
from suspected problem sites and unimpacted sites are given in Table 3.   These
results vividly emphasize the need to consider problem areas separately  from
the remainder of the unconfined production system and to design corrective
actions to effectively control  the impact of these.  Yields of pollutants from
heavily stocked pastures are exemplified in Table 4, along with that from
other land uses.  As indicated, heavily stocked pastureland contributes  less
than cultivated fields but more than permanent grasslands.  Thus, improved
pasture management may be called for to control the pollutants from
such heavily used areas.  The pollutant levels from other than problem areas
and heavily stocked pastures are seldom discernible from background levels and,
on an area! basis, are of the same magnitude as yields to land from rainwater
and yields to surface waters from undisturbed lands.

Prediction of increases in pollutant yields to receiving waters due to
unconfined animal production is not possible with existing technology.  As
measured in receiving streams,  these increases usually are not only small in
value but extremely erratic as  well.  As a first approximation, models for
erosion/sediment yields may be  adopted/used to predict pollutant yields  from
unconfined livestock systems.

BACTERIAL COUNTS

Bacterial  indicators are reputed to be the most sensitive index of water
quality impacts resulting from  unconfined livestock.  But the actual signi-
ficance or meaning of elevated  bacterial counts in this case is not clear.
While the yield from pastured areas may be sometimes several fold that from
control areas (White et al., 1982), the opposite is also not uncommon (Saxton
et al., 1982).  Their use in evaluating the effect of unconfined livestock
production on runoff is certainly difficult, if not impossible.  A growing
consensus among researchers is  that their use in evaluating nonpoint pollution
is inappropriate.  These organisms are properly used to designate the
bacteriological safety of potable water.  Their presence in water is accepted
by public health authorities as an indication of fecal pollution.  Whenever
water is polluted with fecal material, it poses a potential health hazard if
used untreated for drinking or  body contact recreational purposes.  There is
a dearth of documented cases of health problems correlated with bacterial
indicator counts in water related to unconfined animal production systems,
but this may be due solely to the lack of intensive and systematic research in
this area.

CONTROLS

While certain specific questions such as desired level of control and best
methods to achieve water quality goals remain unresolved, the basic concepts,
methods, processes, and procedures needed to reduce pollutant yields from
unconfined animal production systems are available in principle.  Since  the
pollutants are nonpoint in origin, control methodology is effected through
changes in the production management system; i.e., by incorporating practices
involved in controlled grazing, regulated animal congregation patterns,  sus-
tained forage production (establishment and protection of desired species),
intensive erosion control, and  proper land use.  This may, under extremely
adverse site conditions, call for complete exclusion of livestock.


                                   P-III-B-7

-------
      TABLE 3.   WATER QUALITY BY LAND USE (SEWELL AND ALPHIN 1972)
Land Use
                   BOD5
DO
NOa-N ,     POa-P
     ppm
          Total  Coliform,
             cts/100 ml
Ungrazed
Woodland

Heavily-Grazed
Pasture

Farm Pond in
                    2.5
                   13.8
8.6
6.1
 0.05
 4.5
0.05
7.1
  1500
330000
Pasture
Farm Pond in
Woodland
10.0

3.1
7.9

7.6
0.20
i

0.12 '
0.05

0.02
600

1500
TABLE 4. POLLUTANT
Pollutant
Total Residue
Suspended Solids
Total Phosphorus
Nitrate-Nitrogen
Total Kjeldahl Nitrogen
Chemical Oxygen Demand
YIELD BY LAND USE

Cultivated
334
286
0.30
0.37
0.91
48
(DORNBUSH ET AL.
Yield, kg ha" yr
Pasture
58.2
11.8
0.25
0.40
1.12
28
1974)
-1
Grassland
32.4
4
0.1
0.24
0.73
13
                                P-III-B-8

-------
Pasture and rangeland management practices leading to optimal  forage pro-
duction on a long-term sustaining basis, such as that resulting from maintaining
proper stocking rates and forage production practices with erosion control,  are
also those practices which minimize environmental  impacts.  On the other hand,
when exploitive (poorly managed) production is followed, unnecessary detri-
mental  effects on surface water quality are likely to occur.  These systems
may require major modification/adaptation to meet environmental  goals as well
as to achieve other objectives which govern pasture and rangeland use.

As detailed by Robbins (1978), when developing or modifying management practices
for unconfined animal production systems, consideration should be given to the
following concepts:

1.  Multidisciplinary and multiagency review teams should be used to evaluate
    specific environmental problems resulting from unconfined animal production
    and to recommend changes in management practices to correct the problems.
    Their recommendations should be applied in concert with established
    regulatory procedures.

2.  Regulatory programs should account for local (site specific) conditions.
    Regulated changes in management practices should be restricted to those
    documented to have measurable water quality benefits for receiving waters.
    Due to the low level  of pollution associated with unconfined animal
    production, regulatory programs that would in general discourage or
    restrict livestock production on pasture and rangeland should be avoided.

3.  Install and maintain an effective and complete program of soil erosion
    control.

4.  Follow stocking rates and controlled utilization of forages (e.g.,
    rotation, deferred, and seasonal grazing) that reduce erosion and minimize
    waste accumulation.  Reduce stocking rates in problem areas and at
    critical times or seasons.  Stocking rates and grazing programs should
    be tailored to the soil, vegetation, topography, hydrogeology, and micro-
    climate of the particular site.

5.  Avoid animal stocking rates and other practices that create holding areas
    rather than grazing areas.  Promote necessary animal congregation in
    areas that are hydrologically remote from streams and other major drainage
    channels.  Periodically move bedground, shelter, salting, feeding, and/or
    watering areas to control waste accumulation,  soil compaction, and
    erodible paths and areas.

6.  Maintain to the extent ecologically feasible,  highly productive forage
    and dense ground cover on the land to decrease volume and rate of runoff,
    to entrap and hold animal wastes, to utilize fertilizer nutrients, and to
    prevent erosion.  Increase herbaceous cover in proximity of stream banks,
    downslope from animal congregational areas, and on other critical areas.

7.  Where the number of animals per unit area or the characteristics of the
    site present pollution problems, appropriate drastic management alternative/
    practices may include:
                                   p-lii-B-9

-------
        (a)  Restrict animal access to critical areas.  Use fencing to prevent
            livestock from entering highly credible areas and critical  stream
            or pond reaches and to prevent animals from wading in water.
            Provide summer shade (trees or artificial shelters) and insecticides
            to lessen the need for animals to enter water for relief from heat
            and insects.  Block erodible paths with physical  barriers and
            revegetate eroded paths.  Move drinking facilities outside
            critical areas.

        (b)  Increase rate of fecal  degradation and incorporation.  Modify feed
            formulation and/or texture.  Use tillage to break up, manipulate,
            and incorporate wastes in particular problem areas.

        (c)  Use land forming and diversions to modify drainage patterns.

Total stream fencing and other drastic controls will usually be both unnecessary
and impractical except for a few problem areas within an operation.

CLOSING REMARKS

One major challenge still remaining is to demonstrate cost-effective routes
toward achievement of various levels of pollution control for unconfined animal
production systems.  Whether resources should be used to reduce pollutant yields
from a given production system really requires an on-site evaluation of whether
the pollutants have a real adverse effect on the quality of the receiving water
and whether the yields can be significantly reduced.  In view of the many
desirable characteristics of unconfined animal production compared with the
problems associated with alternate livestock production techniques, adequate
research, management, and educational  resources should be allocated to assessing
the importance and significance of increased levels of pollutants (indicator
bacteria, plant nutrients, and oxygen demands) in surface waters (streams and
impoundments), with consideration given to fate of the pollutants and uses of
the waters.  In cases where pollutants from livestock activity are not dis-
cernible from background levels, little or no change may be noted in stream
quality even when significant efforts and resources are expended on control.
When grazing is approached using well-planned management practices that include
pollution control  as an integral part, rather than a laissez-faire approach,
the impact of unconfined animals will  be well within acceptable limits  and in
keeping with water quality goals.

ACKNOWLEDGEMENTS

This paper is based on work supported in part by funds from the U.S.
Environmental  Protection Agency.  The grant, No.  R805559, was administered
through the Robert S. Kerr Environmental Research Laboratory, Office of
Research and Development, Ada, Oklahoma 74820.
                                   P-III-B-10

-------
REFERENCES

Doran, J.W., Schepers, J.S., Swanson, N.P., and McCalla,  T.M.   (1982).
Chemical and bacteriological quality of runoff from a  central  plains  pasture-
land.  Unpublished report for IAG 78-D-X0088, U.S.EPA, Ada,  OK.
Dornbush, J.N., Andersen, J.R., and Harms, L.L.  (1974).   Quantification of
pollutants in agricultural  runoff.  EPA-660/2-74-005.   Superintendent of
Documents, U.S. Government Printing Office.
Flory, E.L.  (1936).  Comparison of the environment and some physiological
responses of prairie vegetation and cultivated range.   Ecology 17:67-103.
Froehlich, H.A.  (1976).  Inorganic pollution from  forests and rangelands.
Publication No. SEMIN-WR-021-76.  Water Resources Research  Institute, Oregon
State University.
Hanson, C.L., Heinemann, H.G., Kuhlman, A.R., and Neuberger, J.W.   (1970).
Grazing effects on runoff and vegetation on western South Dakota  rangeland.
J.Range Management 2^:418-420.
Heady, H.F.  (1975).  Rangeland Mangement.  McGraw-Hill  Book Company, New  York.
Moore, E. (ed.) et al.  (1979).  Livestock grazing  management and  water quality
protection.  EPA-910/9-79-67.  National Technical Information Service,
Springfield, VA.
Powell, J., Crow, F.R., and Wagner, D.G.  (1982).  Rangeland watershed water
budget and grazing cattle waste nutrient cycling.  Unpublished report for  Grant
No. R-803735.  U.S.EPA, Ada, OK.
Robbins, J.W.D., Howells, D.H., and Kriz, G.J.  (1971).   Role of  animal wastes
in agricultural land runoff.  EPA-13020DGX.  Superintendent  of Documents,  U.S.
Government Printing Office.
Robbins, J.W.D.  (1978).  Environmental impact resulting  from unconfined
animal production.  EPA-600/2-78-046.  National Technical  Information Service,
Springfield, VA.
Robbins, J.W.D.  (1979).  Impact of unconfined livestock  activities on water
quality.  Transactions of the ASAE (22|6:1317-1323.
Saxton, K.E., Elliott, L.F., Papendick, R.I., and Jawson,  M.D.  (1982).  Effect
of animal grazing on water quality of nonpoint runoff  in  the pacific  northwest.
Unpublished report for IAG D6-0030 and 78-D-X0249.   U.S.EPA, Ada,  OK.
Sewell, J.I. and Alphin, J.M.  (1972).  Effect of agricultural land use on the
quality of surface runoff.   Progress Report 82.  Tennessee  Farm and Home
Science, University of Tennessee.
Smeins, F.E.  (1976).   Influence of vegetation management on yield and quality
of surface runoff.  Annual  Report No. C-6310.  Texas Water  Resources  Institute,
Texas A&M University.
Stewart, B.A. (ed.) et al.   (1975).  Control  of water  pollution from  cropland:
volume I.  EPA-600/2-75-026a or ARS-H-5-1.  Superintendent of Documents,   U.S.
Government Printing Office.
Stoddart, L.A., Smith, A.D., and Box, T.W.  (1975). Range  Management, 3rd
edition.  McGraw-Hill  Book Company, New York.
Sweeten, J.M. and Reddell,  D.L.  (1978).  Nonpoint  sources:   state-of-the-art
overview.  Transactions of the ASAE (21)3^:474-483.
Trimble, G.R., Jr. and Weitzman, S.  TT951).   Effect of soil and  cover conditions
on soil-water relationships.  Station Paper No. 39, U.S.  Forest Service, North-
eastern Forest Experiment Station.
USEPA.  (1973).  Methods for identifying and evaluating the  nature and extent of
nonpoint sources of pollutants.  EPA-430/9-73-014.   Superintendent of Documents,
U.S. Government Printing Office.


                                   P-III-B-11

-------
White, R.K.,, VanKeuren, R.W., Owens, L.B.,  Edwards,  W.M.,  and Miller, R.H.
(1982).  Effects of livestock pasturing on  nonpoint  surface  runoff.  Unpublished
report for Grant No. R-804631.  U.S.EPA, Ada,  OK.
                                   P-III-B-12

-------
            COMPARATIVE POLLUTANT REMOVAL CAPABILITY. ECONOMICS
        AND PHYSICAL SUITABILITY OF URBAN BEST MANAGEMENT PRACTICES
                 IN THE WASHINGTON B.C. METROPOLITAN AREA
  Thomas Schueler, Robert Magill, Michael P. Sullivan and Cameron Wiegand

                   Department of Environmental Programs
              Metropolitan Washington Council of Governments
                     1875 Eye St. Washington DC 20006
                                 ABSTRACT

Recent  field  studies  have  assessed  the  comparative  pollutant removal
capability  of  six  kinds  of  urban  stormwater best management practices
(BMPs)  in  the  Washington  metropolitan area.  While some urban BMPs were
found to be very effective in removing pollutants entrained in urban runoff
(wet  ponds, extended detention, porous pavement), the most frequently used
stormwater  BMPs,  dry  ponds  and  grassed  swales, exhibited little or no
pollutant  removal  capability.  Local  governments have recently sought to
incorporate  water  quality  objectives into existing stormwater management
programs  by encouraging BMPs with high pollutant removal capability.  This
strategy  has not led to widespread implementation because of uncertainties
about  the  physical, economic, and environmental constraints of these BMPs
at the site level.

A  systematic  approach for screening urban BMP options is presented, based
on  a  comparative  review  of  the  suitability  of  BMPs  with respect to
pollutant  removal,  cost,  and  site/environmental  suitability  for  nine
hypothetical  development  scenarios.  A  three  level  screening matrix is
developed which provides rapid guidance to site planners on the suitability
of  seven urban BMP options. Major elements evaulated in the matrix include
site  restrictions  such as soils, slope and site area, space requirements,
pollutant  removal,  local and downstream habitat modification, landscaping
and   aesthetic   value,   total   cost,   and  operation  and  maintenance
requirements.   The  use  of the screening matrix for developing effective,
low cost urban nonpoint source control strategies is also discussed.
INTRODUCTION

Stormwater  management programs have been in effect in the Washington, D.C.
area  since  the early 1970s with the objective of controlling increases in
post-development  peak runoff rates.  To date, nearly 3,500 structures have
been  constructed.   At  least  75%  of all new regional development is now
serviced  by  some kind of stormwater control, with dry ponds being applied
in about two-thirds of all development situations (MWCOG,1984a).
                                 P-III-C-1

-------
As  more  information  has  become  available on the impact of urban runoff
loadings   on   important  area  receiving  waters,  Washington  area  local
governments  have actively sought to incorporate nonpoint pollutant removal
into their existing stormwater management programs, chiefly through the use
of   alternative   urban   "Best   Management   Practices"  (BMPs).   These
alternatives  rely  on  either  greater retention of stormwater (wet ponds,
extended  detention  ponds),  or  enhanced  infiltration  (porous pavement,
grassed swales, infiltration basins and trenches).

Despite  strong  local institutional and legislative commitments to improve
urban  stormwater  quality,  less than 20 percent of all stormwater control
structures  installed  or  designed  during  1982  and 1983 had significant
pollutant removal capability. (MWCOG,1984a) At the site planning level, the
most  frequently cited reason for the lack of widespread implementation has
been  uncertainty about the cost, efficiency and suitability of alternative
BMP options.

In  1983,  the  Metropolitan  Washington Council of Governments completed a
four  year  BMP  monitoring  and cost survey under the EPA Nationwide Urban
Runoff  Program  (NURP).  The  objective  of  the research was to provide a
comparative  assessment  of the relative suitability of selected urban BMPs
for  nonpoint  source control.  Subsequent research has examined the degree
to  which  urban  BMPs  can  be  applied  over the range of site conditions
encountered  in  the  region.  This  paper summarizes the major findings of
these efforts, and presents a systematic screening procedure for evaluating
alternative BMP options at the development site level.

URBAN BMP ASSESSMENT RESEARCH

BMP Pollutant Removal Field Studies.

The  pollutant removal capability of six different urban BMPs was monitored
over  a  12  to  18  month  period  in  suburban Maryland and Virginia. BMP
monitoring  sites included dry ponds (1), extended detention dry ponds (1),
wet  ponds  (3), grassed swales (3), infiltration pits and trenches (2) and
porous  pavement  (1). Flow-activated automated water quality samplers were
installed at the inlets and outlets of 11 BMP sites, which drained small (3
to 90 acre), single land use watersheds. Flow weighted composite samples of
nutrients,  sediments,  COD  and  trace  metals were collected for 20 to 50
storm  events at each site.  Pollutant removal efficiency was calculated in
terms  of  total  mass outflow load/total mass inflow load for volume BMPs,
and outflow median loading rate/inflow median loading rate for infiltration
BMPs.  More  detailed  discussion  of  monitoring and data analysis methods
employed   in   the  study  can  be  found  in  the  final  project  report
(MWCOG,1983a)

BMP Cost Surveys

Construction  costs for pond BMPs were evaluated in an extensive local cost
survey   (MWCOG,1983b).   Construction  components  and  requirements  were
identified  for  31 representative SWM ponds in the Washington metropolitan
area  (e.g. cubic yards of excavation, linear feet of pipe etc.).  In-place
unit  cost  estimates  for  each pond component or construction requirement
were  then  independently solicited from 12 public and private construction


                                 P-III-C-2

-------
and  engineering  firms.  By applying  the unit cost  estimates to  each of  the
31   ponds,   372   separate  estimates   of  construction cost were generated.
Subsequent  statistical  analysis of the dataset yielded a  strong  (^=0.84),
curvilinear  relationship between pond storage volume and construction cost
(Table 1).

No   systematic   cost  surveys  of infiltration BMPs  could  be developed since
very  few   have   been installed in the region.  As an indirect approach,  the
basic  geometry  of infiltration BMPs  was inferred from design standards  for
each  site  (MD DNR.1984).  Construction requirements  for each site were then
estimated   based on site  geometry, and construction  costs  were estimated by
multiplying  construction  requirements  by   unit cost data obtained in  the
pond cost study  (Table  1).

Estimates   of  average  O&M costs for  pond BMPs  were  obtained in  a survey of
O&M  budgets  and  component   cost  records   of  six  public works agencies
(MWCOG,1983b).     Typical  O&M activities reported included  inspection, grass
mowing,  debris   removal,   bank  stabilization,  structural  repair,  nuisance
control  and  sediment  removal.   Analysis  of   O&M   budgets suggested that
regular  performance  of  these  tasks  would entail an annual cost  equivalent
to    3  to   5%   of  pond   BMP   construction   cost.    O&M   requirements   for
infiltration  BMPs are  generally different in scope  and frequency;  the best
available methods  for evaluating these costs  are shown below.
  Table 1. METHODS USED TO CALCULATE COSTS FOR SELECTED URBAN BMPI IN THE WASHINGTON DC METROPOLITAN AREA
  BMP TYPE
                 CONSTRUCTION   CONTINOfNCY    0 fc H COST
                 COST    (C)   COST  (CC)    (annual|
                                                         COMMENTS/NOTES
  DRY POND
  EXTENDED
  DETENTION
  DRY POND
  GRASSED ROADSIDE
  SWALES
                 82.51Vs*«0.51  C * 0.25
                                      C • 0,05
 82.5Ws"»0.51  C • 0.25
                      C • 0.05
                 82.51Vs«0.5l  C « 0.25
                                      C • 0.05
Equations derived in Washington DC Stonawater
Pond Cost Study (MWCOG,1983).

O&M Costs were computed assuming a conservative
value of 5%
Contingency costs are defined as costs Incurred In
planning, designing and overseeing construction of
BMP
                                                 Vs
                                                     volume of pond up to the crest of the
                                                     emergency spillway (in cubic feet)
INFILTRATION
BASIN
INFILTRATION
TRENCH
POROUS PAVEMENT
(PPfl)
Not Space
Limited
POROUS PAVEMENT
(PPI2)
Space
Limited
66.23Vs*"0.51
Construction
Requi rements
Determined by
Design Geometry
Increments I
Cost over
Conventional
Pavement
As above, +
cost of extra
reserve! r base
under asphalt
for runoff
storage
C • 0.25
C • 0.25
C * 0.35
Includes
extra cost
for tddl.
site surveys
soil testing
inspection
C • 0.35
C • 0.05
$286 • SA
of trench
* Buffer
SA in acres
SO. 003 «
SA of PPf 1
(commerc la 1
rate for
vacuum street
sweeping
it times/yr
SO. 003 • SA
of PPf 2 site
(SA In sq.ft)
Equation derived by factoring out the costs of
outlet works and structures from the pond cost
equation.
OlcM requirements assumed to be similar to ponds.
Construction requirements Include: excavation,
fill of coarse stone or gravel, filter cloth,
test wells, and seeding/mulching for buffer strip
OteM based on cost/maintained acre In (MWCOC,1983)
Incremental costs include: higher unit costs of
porous asphalt relative to conventional asphalt,
extra costs for filter cloth, and test wells.
SA for PP|1 * 60 to 70% of total site area
Depth of extra base course computed from State of
Maryland Design Specifications. Associated extra
cost for excavation and coarse stone fill derived
from unit cost data contained in pond cost study.
SA for PP4>2 • impervious site area/2
(13 to M3X of total site area)
Not         Not          Not          Grassed roadside swales generally assumed to be
Calculated     Calculated     Calculated      loss expensive than the curb and gutter convey-
                                  ance systems they replace.
  TOTAL BMP COST It defined (C + CC) + OHpv; where OMpv Is the present value of all future O&M payments, assuming a twenty
            year project life and an 8{ discount rate.  Land costs were assumed to be zero for all BMP*.

  ANNUAL BMP PAYMENT was computed assuming a twenty year loan at 6% interest rate
                                     P-HI-C-3

-------
Physical and Environmental Suitability Surveys

Physical  or  environmental conditions which might restrict the application
of  a  particular  BMP  were  evaluated  in a review of local and state BMP
design  manuals  and  other relevant literature.  Approximately thirty site
planners, engineers and public works officials involved in BMP construction
were also interviewed to obtain their perceptions and experiences about the
limitations and attributes associated with each BMP option.

Development of BMP Screening Model

To  arrive  at  a  comparative assessment of BMP options at the site level,
nine  hypothetical  development  site  scenarios  were  examined in detail.
These   scenarios   included   single   family  residential  (2  DU/Ac--25%
impervious),   townhouse   (8   DU/Ac--65%   impervious),   and  commercial
development  (85%  impervious)  situated  on  1, 10, and 25 acre sites. The
TR-55  tabular  method  (SCS,1975)  was  used  to determine the increase in
post-development  peak  discharge and runoff volume associated with the two
year  design  storm at each site.  Curve Numbers (CN) of 70, 85 and 92 were
assumed  for  the  single  family  residential,  townhouse,  and commercial
developments,  respectively.  All  sites  were  assumed  to  have  soils in
hydrological  group  "B",  and  slopes  of  3-5%. The stormwater management
criteria  for  all  sites  was to maintain post-development peak discharges
(pond  BMPs)  or  increases in upland runoff volumes (infiltration BMPs) to
pre-development  levels  for  the two-year storm (defined as meadow,CN 58).
Required  runoff  storage  volume was determined for each BMP by either, the
short-cut  routing method (pond BMPs) or State of Maryland infiltration BMP
design standards (MD DNR, 1984).

The  total  cost  of  installing  and  maintaining  each  BMP  on  the nine
development  sites  was  computed  using  the  methods outlined in Table 1.
Physical constraints, such as space requirements, were evaluated assuming a
standard  geometrical  design  for  each  site.  Environmental factors were
assigned  to  each  BMP site based on the consensus among site planners and
engineers surveyed.

RESULTS OF THE COMPARATIVE URBAN BMP ASSESSMENT

Comparative Pollutant Removal

Long  term  pollutant  removal  observed  at  the  BMP  sites  monitored is
summarized  in  Table  2.  As can be seen, the degree and kind of pollutant
removal  achieved  varied  widely among the BMPs studied.  For example, the
dry  pond  investigated was not effective in removing sediments, nutrients,
trace  metals  or  oxygen-demanding  materials.  Urban  stormwaters  passed
through  the  dry  pond  too  rapidly  to  achieve  significant removal via
settling.  During some storms, the dry pond actually became a net source of
pollutants,  as pollutants were either scoured or resuspended from the pond
bed.

Extending  detention  time  in  dry  ponds  by  6  to 12 hours proved to be
effective in removing sediment, organic nutrients and trace metals. Removal
of  biologically available nutrients and other soluble pollutants, however,
was  negligible.   Wet  ponds  were found to be the most effective pond BMP


                                 P-III-C-4

-------
studied.   Removal  of   sediment,  trace  metals   and  organic  nutrients was
comparable  to  rates observed at the extended detention pond.  In addition,
biological  processes   occurring  within  the pond effectively removed the
soluble nutrients most  readily utilized in  algal growth.

No  statistically  significant  improvement  in   runoff  quality  could  be
attributed  to  grassed swales when compared to  traditional  curb and gutter
conveyance  systems.  The   poor performance of swales appeared to be due to
the  characteristics  of  the  sites  monitored--moderate slopes (4 to 6%),
relatively  permeable   soils,  and short grass.   It is possible that better
removal   might  have  been  achieved  by  swales  with check dams on sandy,
relatively  flat  sites.   However,  these   conditions are not  common in the
Washington metropolitan area.

Monitoring  difficulties and site limitations prevented a valid statistical
assessment  of pollutant removal by infiltration trenches.   Trench BMPs did
appear  to  be  capable of removing sediments and trace metals, but did not
show  any  capability   for  removing nutrients.  It should be noted that the
trenches   monitored  had  outlet  structures, and thus did  not function as
"true"  infiltration BMPs  (which infiltrate incoming runoff  completely into
the  soil  profile).  If  complete  infiltration  is  accomplished,  as  is
currently  recommended,  it  is likely that infiltration BMPs  would exhibit
much  higher  removal   efficiencies.   This,  in  fact, was  the case at the
porous  pavement  site   monitored.  Although  the  site did  have perforated
underdrains (and therefore could be monitored),  significant  infiltration of
runoff  into the underlying soil was achieved. As a result,  removal of-most
pollutants at the porous pavement site approached 90%.
             Table 2.   COMPARATIVE POLLUTANT REMOVAL OF URBAN BEST MANAGEMENT
                      PRACTICES IN THE WASHINGTON D.C. METROPOLITAN AREA NURP
                      FIELD MONITORING STUDY
                      DRY   EXTENDED   WET     INFILTRATION    POROUS   GRASSED
                      POND  DETENTION  POND   BASIN   TRENCH   PAVEMENT SWALES
                           DRY POND
             PARAMETER
                             PERCENT REMOVAL OF POLLUTANTS
              SEDIMENT
                       14
                                   55
                                                  50
                                                         >95
                                                                NSD
             TOTAL
             PHOSPHORUS

             ORTHO-
             PHOSPHORUS

             TOTAL
             NITROGEN

             NITRATE &
             NITRITE

             TKN

             EXTRACTABLE
             LEAD (Pb)
20


 0


10


 9

11


-5
             EXTRACTABLE
             ZINC (Zn)   -10

             COD         0
      10
24


10

15


84


57

40
66


84


28


60

15


65


51

28
                           60
                      -8
                           48

                           37
 60


>50


 88


>70

>90


>95


>99

 83
NSD


< 0


NSD


NSD

NSD


NSD


< 0

< 0
                                  P-III-C-5

-------
Relative Cost of Urban  BMPs

Figure  1 displays the  computed total  costs  associated with seven urban BMP
options  for  several   land   use  and   site   area combinations.  In general,
conventional  dry  ponds   represent  the least cost BMP investment.  Extended
detention ponds involved  an  additional cost  of 5-10% at most sites,  whereas
wet  ponds and infiltration  basins were 25 to 45% more expensive, depending
on land use and area serviced.  Infiltration  trenches cost  approximately 50%
more   than  dry  ponds   on   small  residential  sites, but  were  clearly
uneconomical on larger  or more  intensively developed sites.  Porous pavement
was  generally  competitive  with pond  BMPs when space requirements were not
limited.  When  available space  was  limited,  however, porous pavement was
only economical for small developments sites.
*» 2.5 .
TOTAL BMP COST , ANNUAL PAYMENT (in thousands of dollar
-» M CO ^. Ol O> —
oooooo cn_«io
	 1 	 L _L.. 1 i ,
HI! DRV POND [ • \ IN BASIN I • • 1 POROU842
K'i'»/l EX.DET P E=i (-TRENCH
K\\| WET POND | | POROUS* 1

ONE ACRE SITES

cr
	 rr ' ::.v • r''
5 n •'•' ^
1 n lltV::
~- 11 ^ ''''L. >^

3
3
|
i
In







25 ACRE SITES


P^FIn

1 n
In 1 Ifl^FI




!ja
i
1 ;
' *1




1

\









_^J
1
| 	
in
—



i


Hi

i

1

~

=5
3
3 	
a
i
In
^y
13
3
3
3
=3
^

a
—4 	
3 :
IT
3

-j


'
                    Single Family
                     Residential
                     (2 DU/AC)
 Townhouse
Development
 (8 DU/AC)
  Commercial/
 Light Industrial
(85% Impervious)
 Figure 1.  COMPARATIVE TOTAL COST OF SEVEN DIFFERENT URBAN BMP OPTIONS IN THE
           WASHINGTON METROPOLITAN AREA, 1984 DOLLARS.  Top Panel: Cost Estim-
           ates for One Acre Sites. Bottom Panel: Cost Estimates for 25 Acre
           Sites.   Total Cost Defined as Construction Cost + Contingenency Cost
           + Annual O&M Amortized at 8% for 20 years + Financing Cost.
                                 P-III-C-6

-------
The  higher  cost  of  infiltration  BMPs can be attributed to two factors.
First,  unlike pond BMPs which only control the increase in peak discharge,
infiltration  BMPs  must  have  sufficient volume to accommodate the entire
increase in upland runoff volume.  As a result, infiltration BMPs generally
have  greater  excavation requirements because of the larger storage volume
needed.  In  the  case  of  infiltration trenches and porous pavement,  even
greater  storage  volumes  are  required to allow for non-void space in the
gravel  or coarse stone filter.  The cost of filter materials can also be a
major  factor  which  drives  up  infiltration  trench  and porous pavement
construction costs.

One  advantage  of  infiltration  BMPs  is  their  lower  cost  for  annual
operations  and  maintenance,  relative  to  pond  BMPs.  Because  of their
inherent design, many common pond O&M tasks, such as structural repair, and
sediment  and  debris  removal,  are  either  unnecessary  or impossible to
perform  on  infiltration BMPs.  Porous pavement, for example, was found to
be  cheaper  to maintain than any other BMP on commercial sites. Similarly,
infiltration  trench  maintenance  typically  was  less  than  one-third as
expensive  as  annual  O&M  for dry ponds on residential sites. As might be
expected,  dry  ponds typically involved the lowest O&M cost for pond BMPs,
ranging  from $6 to $60/DU/yr on residential sites, and $62 to $232/Acre/yr
on  commercial  sites.  Extended  detention  pond  O&M  involved  a  modest
additional  O&M  cost of 5 to 7%, whereas wet ponds and infiltration basins
involved 16 to 70% more O&M cost when compared to dry ponds.

Physical Suitability of BMPs

The  most  frequently  cited  physical constraints to urban BMP application
which   emerged  in  the  suitability  surveys  included  topography,  soil
permeability,  space  requirements and site size.  In general, infiltration
BMPs  were found to be subject to the greatest number of site restrictions.
Current  State  of  Maryland  Infiltration  Standards prohibit their use on
sites  with  soil  profiles  of  low  to moderate permeability ("C" and "D"
soils),  slopes  greater  than 5%, and a depth to bedrock or groundwater of
less  than  2  feet. Also, infiltration trench and porous pavement BMPs may
not  be  feasible  on  larger, more intensively developed sites, because of
their  significant  space  requirements  (which may often exceed 25% of the
total  site  area).  However,  the  greater  space  requirements  of porous
pavement  may not be restrictive on some sites, particularly if significant
portions of the site are devoted to parking.

Application  of  extended detention ponds in the region is normally limited
to  sites with a minimum size of about 20 acres. In smaller sites, required
orifice  size in extended detention ponds becomes so small (<2 inches) that
the  outlet  is susceptible to chronic clogging. Wet ponds can usually only
be  applied  to  development  sites  in  excess  of  10 acres, as a minimum
watershed  size  is  needed  to  maintain  a permanent pool of water during
extended  periods of dry weather. Although pond drawdown does not adversely
affect  a wet pond's runoff control or pollutant removal properties, it may
reduce or eliminate desired habitat, aesthetic or landscape benefits.
                                 P-III-C-7

-------
Environmental Considerations in BMP Selection

Natural Environment

As  urban  BMPs  nearly  always  represent  a  modification  to  the  urban
landscape, it is often necessary to select or design BMPs to minimize their
impacts  on  the  environment. Recent research has shown, for example, that
BMPs  can  have a substantial influence on the quality of wildlife habitat.
A  field  survey  conducted  in Columbia, Md. (Adams et al, 1983) suggested
that dry ponds had little habitat value. The same study also concluded that
wet  ponds  enhanced both aquatic and terrestrial habitat value.  A variety
of  waterfowl, marsh and shore birds, and terrestrial wildlife utilized wet
ponds,  presumably  attracted  by  the  food  and cover supplied by aquatic
vegetation.  With proper design, it is likely that extended detention ponds
can  also function to create valuable wetland habitat.  Infiltration trench
and  porous  pavement  probably have a negative influence on local habitat,
since they consume more area than pond BMPs, and are normally unvegetated.

Downstream   habitat  is  also  influenced  by  urban  BMPs.  For  example,
infiltration  BMPs  often can sustain downstream flows during extremely dry
periods  because  of  their  superior  groundwater  recharge, thus reducing
seasonal stress on downstream aquatic life.  Infiltration BMPs and extended
detention ponds also function to reduce the loss of riparian habitat caused
by downstream bank erosion (since these BMPs attenuate peak flows to a much
greater degree than either wet ponds or dry ponds). Some concerns have been
expressed  that  under  some  conditions  wet  ponds  may  produce  thermal
discharge which could impact trout fisheries in cold-water streams.

Human Environment

An  often  overlooked  aspect  of  urban  BMPs  is  their role in the human
environment.  Urban  BMPs  can  function  as  a  landscape amenity or as an
unsightly  and  discordant  community  feature.  Similarly,  urban BMPs can
support  limited recreation use or be under-utilized and possibly dangerous
empty space, depending on the type of BMP employed, and its subsequent O&M.
As  resident's  perception  of aesthetic value prominently figures in their
acceptance  of and support for BMPs, these considerations can be important,
particularly   if  the  same  residents  will  be  responsible  for  future
maintenance.

   In  this  respect,  wet  ponds  appear  to be superior to other BMPs, if
properly designed and maintained.  In a recent survey of resident attitudes
towards  stormwater  ponds  in  Columbia,  Md.  (Adams  et al,1983), 75% of
residents  surveyed  preferred  wet  ponds over dry ponds. Most respondents
felt  that  wet  ponds provided a more pleasant and natural environment, as
well more diverse recreational opportunities.

CONCLUSIONS

Comparative   data  on  the  cost,  efficiency  and  physical/environmental
feasibility  of six urban BMPs has been integrated into a summary screening
matrix  shown  in  Table 3. The matrix is intended for use as a preliminary
screening  procedure for rapid evaluation of available urban BMP options at
the site planning level.


                                  P-III-C-8

-------
Table 3. Generalized Screening Criteria Model For
        Selecting Urban Best Management Practices
SCREENING
CRITERIA
SITE SIZE
. 1-5 ACRES
£ 10 ACRES
» >25 ACRES
O SOILS
* "A" SOILS
C "B" SOILS
a "C" or "D" SOILS
g SLOPE
E 0-5%
w 6-20%
< >20%
U SPACE REQUIREMENTS
U3 < ^ C AT
Jj U < 5 %
g #< 6-15%
•?£: 16-25%
l£ 26%-75%
2 POLLUTANT REMOVAL
_ NUTRIENTS
f- ALGAL LIMITATION
«? SEDIMENT
j TRACE METALS
5 OXYGEN DEMAND
Z OTHER FACTORS
£ GROUNDWATER RECHARGE
Z WILDLIFE HABITAT
2 RECREATION
g AESTHETICS/LANDSCAPE
Z THERMAL DISCHARGE
LOW FLOW MAINTENANCE
a TOTAL COST OF BMP*
O within 10% of DP
C llto50%>DP
5 51 to 100% > DP
g over 100% > DP
53 O4M REQUIREMENTS*
Z less than DP
8 0 to 10% > DP
W llto50%>DP
5 ^lo/n^np
A ji tv^jjr
9 OTHER FACTORS
§ ECONOMIES OF SCALE
LIABILITY/SAFETY
RESIDENTIAL BMPs
(SINGLE FAMILY RESIDENTIAL -•• TOWNHOUSE)
DP XD WP IB IT
0 • • O O
o • o o o
0 O 0 O ®

0 0 0 O O
O O O 0 0
o o o • •

§ § 8 8 •
Or~\ n\ ^ ^
(J (£> w w
- - o o o

— — — — —

9 0 O O O
® o 8 8 8
• o o o o
• 0 0 O O

• ® 0 O O
* 0 O ® •
0 • O 0 •
® 0 O 0 0
o © • o o
• ® o o o

o • • •
-00*
A
~~~~ ~~ ~~~ ^0


• • • ©
o • • o
- 0 0 -


o o o © •
© © © 0 O
iirffNn n APPpriPBiATP f?\ OFTEN APPROPRIATE
LEGEND U APPROPRIATE (/_) W/DESION CHANGES
•NOTE: COMPARATIVE DP =• DRY POND (DETENTION BASIN)
BMP COST DATA HAS XD = EXTENDED DETENTION POND
BEEN COMPUTED WP WET POND (RETENTION BASIN)
USING DRY POND (DP) 1 B INFILTRATION BASIN
COSTS AS AN INDEX 1 T INFILTRATION TRENCH
BASE PP POROUS PAVEMENT
GS GRASSED ROADSIDE SWALES
PP GS
© 8
O 0

o o
0 0
0 9

: i

• 0
0 0
o -

o •
o •
o ®
o ®
o •

O 0
• 0
• ®
® 0
o o
o ®

• o
0 —
0 -

© o
o -


8 8
COMMERCIAL BMPs
DP XD WP IB IT PP
O • •000
O • O O ® 0
O O O 0 • 0

0 © © o o o
O O 0000
O O O 9 9 9

O O O O O O
O O O Q • •
O O O • • •

O O O 0 • •
_ _ — © © •
- - - O O O

• • 0000
• • O 0 0 0
• O O 0 O O
• O O 0 0 0
• © © 0 O O

• ® 0 O O O
0 • O 0 • •
® 0 00®®
O 0 9 0 U 0
• ® O 0 O O

- o • • • ®
- O O O • 0
— — 	 	 00
— — — — O O

- • • • o ®
- O • • - 0
- - O O - 0


§O 0 0 • •
© 0 0 O O
® ™PR.ATE • NOTAPPROPR.ATE
"A" SOILS = SANDS, LOAMY SANDS
"B" SOILS - SANDY LOAM, LOAMS
"C" SOILS - SILT LOAM, SANDY CLAY LOAMS
'.'D" SOILS - CLAY LOAMS TO CLAY



               P-III-C-9

-------
Use of the Screening Matrix

As  an  example,  suppose  a site planner is reviewing a proposal for a ten
acre  shopping mall development situated on a site of moderate (>5%) slopes
and  relatively  permeable  "B"  soils.  Apart  from  controlling  the peak
discharge  from the two year design storm, the planner is required by local
regulations  to  install a BMP capable of effectively removing sediment and
nutrients,  and  having,  at  a  minimum,  no  adverse  impact on the local
environment.  At  the  same  time,  the site planner wishes to minimize the
costs and space requirements facing the builder.

Consulting  the  commercial development column in the screening matrix, the
site  planner  can  quickly  eliminate the use of extended detention due to
site  constraints.  Likewise,  the dry pond and infiltration trench options
can  be  rejected  because of poor pollutant removal and prohibitive costs,
respectively.  If space is limited at the site, porous pavement can also be
ruled  out  on the basis of large costs and space requirements. However, if
the  site will have a large area devoted to parking, porous pavement may be
considered.   Wet  pond  and  infiltration  basins are also indicated to be
feasible  BMP  options.   Both  have  similar  cost  and  O&M requirements.
However,  the  matrix  indicates  that wet ponds would have a smaller space
requirement  and greater overall habitat and amenity value. As this example
illustrates, the screening matrix enables a site planner to select the best
overall BMP option given diverse and often competing objectives.

Implications for Developing Effective Urban BMP Strategies

The  matrix  also  offers  several useful insights for developing low-cost,
effective  urban  nonpoint  source  control  strategies. For example, it is
clearly  evident  that  there are no single, "all-purpose" BMPs that can be
applied over the entire range of site conditions. Rather, each urban BMP is
only  feasible  and  cost-effective in a rather narrow range of development
situations.  As  a  consequence, it would appear that stormwater management
programs  should  be both flexible and site-specific, as broad prescriptive
BMP  policies  will  rarely  be  sufficient to cover the entire spectrum of
development activities encountered in urbanizing areas.

For  example, if high overall pollutant removal is a desired objective in a
stormwater  management  program,  wet  ponds  would appear to be the BMP of
choice on most development sites greater than 10 acres in size. If nutrient
control  is  not a major concern, then extended detention ponds may also be
attractive  on larger (20+ acre) sites. Infiltration BMPs generally provide
the  greatest  overall  mix  of  pollutant removal, cost and feasibility on
smaller  sites  (1-10  acres).  In particular, porous pavement appears most
appropriate  in  small  commercial  development sites, whereas infiltration
basins  or  trenches may be more suitable on small residential sites.  When
infiltration  BMPs are not possible because of site constraints (a frequent
occurrence  in  the  Washington  metropolitan  area), there appear to be no
other  BMP  options  available  which possess a significant nonpoint source
control capability.

The  screening  matrix reveals that there are several other combinations of
site  conditions, environmental or cost restrictions for which no effective
BMP  can  be  applied for nonpoint source control.  One implication is that

                                 P-III-C-10

-------
 the   existence  of  these situations could substantially curtail success in
 achieving  basin-wide urban nonpoint source loading reductions. However, it
 is  possible  to  achieve  some  measure of control at these sites by other
 methods.  For  example,  some  local  jurisdictions  have  adopted off-site
 regional  stormwater  management  programs  which  make it possible to pool
 urban runoff control needs from several small sites in a single, larger wet
 pond  or  extended detent-ion pond further downstream. Developers who cannot
 build an effective BMP on-site are required contribute to the financing of
 the off-site facility.

 A second approach might be to separate stormwater runoff control from water
 quality  protection  needs.  Thus, low-cost detention BMPs could be used to
 control  runoff  from  the  two year design storm, while infiltration BMPs,
 designed  to  accommodate  much smaller runoff volumes, could be applied to
 achieve  pollutant  removal  for  smaller  and more frequent runoff events.
 Because  of their reduced size, these infiltration BMPs would presumably be
 subject to fewer cost and site restrictions.

 Future Directions

 While   the  matrix  can  only  evaluate  a  combination  of  nine  general
 development  scenarios  at  present,  future  research is being directed to
 develop an interactive software package capable of analyzing BMP options of
 any   individual  development  site,  given  simple input parameters readily
 available  from  site  plans.  Concurrently, research is being conducted to
 better  define  the  more uncertain elements of the matrix, particularly in
 regard  to  infiltration  costs.  By  making  better  BMP  screening  tools
 available  to  site  planners  and developers, it is anticipated that wider
 implementation of effective urban BMPs can be achieved, and consequently, a
 greater degree of urban nonpoint source control can be attained.

 REFERENCES

 Adams, L.W., L.E. Dove, D.L. Leedy and T.M. Franklin. (1983) Urban Wetlands
      for Stormwater Control and Wildlife Enhancement. Urban Wildlife
     Research Institute. 64 pp.

 Maryland Department of Natural Resources.(1984) Standards and Specifications
      for Infiltration Practices. Stormwater Management Division, Maryland
     Water Resources Administration. 180 pp.

 Metropolitan Washington Council of Governments. (1983a) An Evaluation of the
     Costs of Stormwater Pond Construction and Maintenance. Contract Report
      for the U.S. EPA Nationwide Urban Runoff Program. 98 pp.

	___. (1983b) Urban Runoff in the
     Washington Metropolitan Area: Final Report Washington DC Area Urban
     Runoff Project. U.S. EPA Nationwide Urban Runoff Program

	.  (1984) Potomac River Water
     Quality 1983: Conditions and Trends in Metropolitan Washington. Water
     Resources Planning Board. 94 pp.

U.S. Soil Conservation Service.  (1975) Urban Hydrology for Small Watersheds,
     Technical Release No. 55. U.S. Dept. of Agriculture. 75 pp.

                                  P-III-C-11

-------
                    IMPLEMENTATION OF URBAN NONPOINT SOURCE
                      WATER POLLUTION ABATEMENT PROGRAMS
                               Robert P. Biebel
                         Chief Environmental Engineer
              Southeastern Wisconsin Regional Planning Commission
                              Waukesha, Wisconsin

                                      and

                             David B. Kendziorski
                               Principal Planner
              Southeastern Wisconsin Regional Planning Commission
                              Wa u ke s ha, Wi sc ons in
                                   Abstract
Urban  runoff  management strategies  have shifted  from  exclusive flooding and
drainage concerns  to  an increased emphasis on water quality and comprehensive
stormwater management.   Recent studies, especially  those  conducted under the
Nationwide Urban Runoff Program (NURP), have improved the technical basis for
implementing  urban nonpoint  source  pollution  abatement  programs.  However,
there  are  still  questions regarding  the  practicality of  providing adequate
urban  runoff  pollutant  control in existing urbanized areas. Implementation of
an urban nonpoint  source control program should  ideally  include the prepara-
tion of a stormwater management plan.
                                 P-III-D-1

-------
 INTRODUCTION

 Historically, most urban runoff concerns were related to flooding and drainage.
 The concept  of the  full-flowing  storm drain designed by  the  Rational Method
 has been  the mainstay  of  urban storawater drainage practice  in this country
 since the late  1800s.  Its  objective of the rapid removal of surface water did
 not fully consider water quality effects, as well as downstream water quantity
 effects,  through the construction of culverts,  storm sewers, and open channels
 designed  to accommodate the increased flow rates from urban areas.  The effect
 of this  approach  has  been  to increase the velocity  of  runoff,  reduce  the
 natural storage of  water  in  the watershed, and  transmit  the Increased runoff
 more  quickly downstream.   It  is well known that the impervious surfaces of an
 urban area contribute  to  these increased flow rates:  within  a typical resi-
 dential area, streets  and  parking areas may contribute from 45 to 85 percent
 of the runoff volume, with roof surfaces contributing an additional 10 percent.

 Over  the past 20  years,  however,  water  quality  has  become an  increasing con-
 cern.   In the 1960s, studies  by the U.S. Public Health Service  and the Federal
 Water  Quality Administration  Indicated  that  urban runoff  caused water quality
 problems  in receiving  surface waters and posed a human health  hazard.  In the
 1970s,  water  quality management,   or 208,  plans were prepared  by regional or
 county planning agencies in most major urban centers.  Many of  the 208 studies
 concluded that  urban stormwater  runoff  was  a  water quality problem,  but  the
 208 studies lacked sufficient data on:

     o    Sources of pollutants

     o    Washoff and transport mechanisms

     o    Receiving water Impacts

     o    Effectiveness  and costs   of urban nonpoint source control  measures

 In the  1980s,  the  EPA  funded the  Nationwide  Urban Runoff  Program  (NURP)
 studies,  and  state  and  local  nonpoint source  planning  efforts—such  as  a
 Milwaukee  Harbor  Estuary Study and  the Milwaukee  River  Priority  Watersheds
 Planning  Program  in  Southeastern  Wisconsin—also began to address  control of
 urban nonpoint  source pollution.

 In order  to understand  how  these recent  studies have  helped in  the implementa-
 tion  of urban  nonpoint source  control  measures, it  is necessary  to  briefly
 review  the  technical findings  of these studies.

 CHARACTERIZATION OF URBAN STORMWATER RUNOFF

 Compared to rural  streams,  streams draining urban areas  frequently have similar
 concentrations  of  solids and  nutrients, and  higher concentrations of metals
 and bacteria.  No  only, as  stated  above,  are runoff volumes higher from highly
 impervious land  uses, such as  commercial  and industrial land  areas,  but pollu-
 tant concentrations are also  higher  from  these  land  uses.   Metals  in  urban
 runoff which often violate acute (short-term) and chronic (long-term)  criteria
 include copper,  zinc, lead, chromium,  and cadmium.  Within  the  Menomonee  River
 in southeastern Wisconsin, studies have  indicated that metal  concentrations in
 the water  column,  as well as  in the bottom sediments, Increase substantially
as the River flows  into the Milwaukee metropolitan area.


                                 P-III-D-2

-------
URBAN NONPOINT SOURCES

Street  surfaces  are the predominant source of pollutants In urban runoff.  In
typical urban  areas,  street surfaces may account  for 70 to 80 percent of the
total solids  load during average storm events, While pervious surfaces gener-
ally account  for less than 20 percent.  Pervious  surface loadings become more
significant  during larger  storm  events.   Traffic  emissions  and atmospheric
loadings may each account for up to 10 percent of  the total loading.

The  NURP  studies  indicated   that  pollutant  loadings  on  streets  generally
increase  as  the  tmperviousness  of the land area  Increases.   With respect to
pollutant  loading,  the  impervious surfaces which  are  directly connected to a
storm sewer  system are  considered to be most  important.   The composition of
the street loadings also varies for different  land uses:   transportation and
industrial  land  uses  generate higher concentrations of metals,  whereas resi-
dential areas have  the highest concentrations of fecal  coliform.

URBAN NONPOINT SOURCE CONTROL MEASURES

Street Sweeping
In general, the NURP studies indicated that street  sweeping is not an effective
measure for achieving high levels of nonpoint source control.  Street sweeping
was not very effective because:

     o   Pollutant  loading removals  from the  street  surface by  the sweeper
         were often less than 50 percent.

     o   Sweepers effectively remove the larger particles but do not generally
         remove  the smaller  particles—which contain higher pollutant concen-
         trations.

     o   Street  loads accumulate rapidly following  sweeping or a storm event.

     o   There is  too much variability related to wind,  traffic,  street, and
         rainfall conditions.

Nevertheless,  street  sweeping  programs  can be optimized, providing a modest—
up to 20 percent—level of nonpoint source control, particularly in commercial
and high density residential areas, by:

     o   Increased sweeping in spring and fall

     o   Increased sweeping of highly impervious areas

     o   Improved parking regulations

Stormwater Storage
The NURP studies Indicated  that atormwater storage could  be an effective way
to achieve relatively high pollutant  removals.   The  studies indicated that:

     o   Properly designed wet  basins  can remove  more  than 50 percent of the
         incoming particulate  loadings,  whereas,  dry  basins  generally do not
         remove significant amounts of pollutants.


                                 P-III-D-3

-------
     o   Some  dry  basins may be modified to  increase their detention time and
         pollutant removal effectiveness.

     o   Off-site  basins,  draining large drainage areas of perhaps  100 to 700
         acres  (0.4  to  2.8  km ),  are more  cost-effective than onaite basins
         draining 5 to 20 acres  (0.02 to 0.08 km ).

IMPLEMENTATION

The NURP and  other related studies have provided the information on the tech-
nical basis for  the  development of implementation programs for urban nonpoint
source  control.   It  may be  concluded  from these studies  that  there is not a
wide choice of urban control measures that can achieve relatively high levels
of reduction in pollutant loadings.  Generally, storrawater storage in properly
designed wet  basins  appears  to be the only  practical  method available which
can achieve removal  levels exceeding 30 percent for several pollutants.  How-
ever, there are  potential  limitations and problems associated with the use of
stormwater storage in many urban areas.  These limitations include:

     1.  Maintenance problems related to:

         a.  Lack of adequate access

         b.  Clogging of outlets

         c.  Litter accumulation

         d.  Weed control

     2.  Safety problems related to:

         a.  Stagnant pools  in  urban  areas which attract mosquitoes, rodents,
             and other pests;

         b.  High water velocities during storm events;

         c.  Attraction to children.

     3.  Inadequate  flood  control,  if  hydraulic design  and analysis  is  not
         considered.    There   frequently  is  a  large difference between  the
         volume  of  storage  and outlet  capacity needed   to  provide adequate
         flood control and that  needed  to  provide optimal water  quality pro-
         tection.

     4.   Aesthetic  problems,  since these  basins  accumulate  large  amounts of
         pollutants,  are  frequently sediment-laden, and have high  concentra-
         tions of nutrients which can support weed and algae growths.

     5.   Lack  of suitable  available open  land sites.  This  is  probably  the
         most  significant limitation for the use of  stormwater storage basins,
         particularly in developed urban areas.  Alternative storage  facili-
         ties,  such  as  subsurface  storage  and stormwater  treatment systems,
         which may be  more suitable  in  developed urban areas, are  also  very
         expensive.


                                 P-III-D-4

-------
While many  of  these  limitations to the use  of  stormwater storage facilities
can  be  mitigated  by  proper planning  and  design,  and  by  the  development of
effective management and maintenance programs, these factors, nevertheless, do
limit the wide application of stormwater storage as a water quality management
measure.

Other urban  nonpoint  source control measure  which  can  achieve limited effec-
tiveness include:

     1.   Street sweeping, which as noted above can often be designed to remove
         an  additional  10  to 20 percent of the loadings of various pollutants
         in some areas.

     2.   Construction site erosion control, which can effectively prevent high
         loadings of sediment.

     3.   Public  education   programs,  which could  particularly be  helpful in
         voluntary nonpoint source control programs.

     4.   Pet  waste control ordinances,  which can  help reduce high bacterial
         loadings especially in residential areas.

Finally, urban nonpoint source control programs, especially those that require
extensive design and  investment,  are best implemented under the guidance of a
comprehensive stormwater management plan.  A  stormwater management plan should
address  both water quantity  and quality  concerns  and contain  the following
elements:

     1.   Hydrologic/hydraulic  analysis of urban  runoff quantity  and quality
         under existing and future conditions.

     2.   Water quantity and quality objectives and design criteria.

     3.   Development and evaluation of the costs, hydraulic effects, and water
         quality  impacts  of  alternative  stormwater  conveyance  and  storage
         measures.

     4.   Selection of a recommended plan which is economically feasible, which
         provides  for  the  wise use of water  resources, and which meets, where
         practical, the objectives of the plan.
                                 P-III-D-5

-------
                             NFS POLLUTION CONTROL:
                           A CONSULTANT'S PERSPECTIVE3
                                       by

                                Stuart G. Walesh.1'
INTRODUCTION

Since the  early  1970's,  much of the non-point source (NPS) pollution manage-
ment activity in the U. S. has been driven primarily by government regulatory
and  funding  programs.   Examples of these programs,  which have  been  at  all
levels of  government  from federal  to local, are  the  federal  208 Program  and
local (county, city, village, and special, district) erosion and sedimentation
control ordinances.   Government programs  have  apparently been  based  on  the
conviction that NPS problems exist  or will develop.

In  other  words,  NPS management  activities to dale have  been  primarily  in a
"top-down" mode,  that is,  government  directing  individuals  and entities to
mitigate or prevent NPS pollution problems.  The historic "top-down" approach
to NPR management  is  characterized by emphasis on  research  and development,
large-scale  studies,   broad  recommendations,  and,  in  general, absence  of
specificity and little implementation.

The  historic  "top-down"  approach   contrasts  with a  "grass-roots"  approach.
With the  latter,  individuals and entities undertake NPS pollution management
because they  have  NPS  or NPS-related problems that must be solved now.  Some
consulting engineers have had opportunities to apply NPS pollution management
measures on a problem  resolution basis completely outside of, but consistent
with the intent of, government programs.

PURPOSE

Using a case  study approach, this  paper briefly describes several NPS pollu-
tion  control  projects  motivated  largely  by  local  need, not  by government
regulatory and funding programs.   The  case studies are intended to stimulate
thinking concerning the  value of encouraging and  supporting  a "grass roots"
approach to NPS  pollution management.   In the absence of, and perhaps unfet-
tered  by,  government  controls,  innovative,   cost-effective  NPS  pollution
control measures  have been  designed and  are  being implemented.  Experience
gained on  these  and similar "grass roots" projects  is  used  to draw conclu-
sions  and make  certain  recommendations  concerning  NPS  pollution  control.
a
 Presented  at  the Non-Point  Pollution Symposium,  Session  on Efficiency and
 Feasibility of  Best Management  Practices for  NPS Control,  Milwaukee, Wis-
 consin, April 23-25, 1985.

 Head-Water  Resources  Engineering,   Donohue  &  Associates,  Inc.,  Milwaukee
 Division, 600 Larry Court, Waukesha, WI  53186  (414/784-9200).
                               P-III-E-1

-------
CASE STUDY  1:  SEDIMENTATION BASINS IN SERIES WITH A DETENTION FACILITY

Physical Setting

The  setting for this  project,  as  shown  in Figure  1,  is a  640 acre almost
completely urbanized watershed in the Madison, Wisconsin area.  The watershed
contains parts  of three  government entities—the City of Madison,  the Town
of  Madison, and  the  Town of  Fitchburg—plus  the  University  of  Wisconsin
arboretum.  The arboretum is  a sanctuary  for native  plants  and wildlife and
is used for research and  teaching.

Problems

Two  growing problems  were of  concern to  area  residents several years  ago
at the beginning  of  the  engineering phase  of  this project.   First,  flooding
of increased frequency and severity was occurring in the peninsula of single-
family residential  development which  protrudes  into the  arboretum  from the
east,  as  shown  in  Figure 1,   Watershed  runoff  from the  entire  watershed
passing  through and  over  this  area  caused  surface and  basement  flooding.
Second, debris  and sediment carried from the urbanizing watershed were being
deposited  in and  causing damage  to  meadows, wetlands,  and lagoons  in the
arboretum.

Solution

Working together,  the three government entities and  the  Arboretum Committee
agreed to a  surface  water management  system  consisting of two sedimentation
basins and  a  sedimentation facility as shown in Figure 2.  The sedimentation
basins are  located on the upstream side of the arboretum and trap objection-
able sediment  and attached pollutants,  floatables,  and  other debris before
it enters  the arboretum.   The  detention  facility, which  is normally empty,
temporarily  impounds  stormwater  in the  arboretum  and  thus  provides flood
protection to the downstream residential area.

The  2,200  foot long  berm required for  the detention  facility was  designed
with a  curvilinear  alignment  and  irregular  cross-section   to  minimize its
visual impact  on the  aesthetically sensitive arboretum.   The sedimentation
basin-detention  system  was  constructed   in  1982  at  a   total  of  cost  of
$103,000 and is in operation (Donohue, 1981; Raasch, 1982).

CASE STUDY 2:  SEDIMENTATION BASIN IN SERIES WITH RESTORED WETLAND

Physical Setting

McCarron Lake  in  Roseville,  Minnesota receives runoff from a fully urbanized
1.6  square  mile watershed.   The 72 acre  lake is  heavily used for  fishing,
swimming,  and boating and provides a setting for a county park.

Problems

A  one-year  diagnostic  study  confirmed  that  McCarron  Lake  is  eutrophic.
Hypolimnetic oxygen  depletion releases nutrients  into the water column when
the lake is  stratified and there are  abundant  nutrients  in the water column
throughout  the  year.  Surface  water  runoff  into  the lake  is the principal
contributor of the troublesome nutrients and of sediment.

                               P-III-E-2

-------
                                   FinURK 1
                                 WATKKSHKD MAP
                                                        'I »>^Av """» •"""«•
                                                       thid I I V7/'\ ^S< \ Ml Milt MltAl
                                                       ^N^^^v'' *H"
Source:   Raasch, 1982
                              P-III-E-3

-------
                                       FIGURE 2
                   SEDIMENTATION BASINS-INTENTION FACILITY SYSTEM
.... J I ._.. J L_J t ..... -. J L ________
              J L ...... ____ I UH, ,H J I _ ...... _
   Wtbl
   SfcOIMtNTAIIOH  - -
   PONO
                       N
Source:   Raasch, 1982
                                  P-III-E-4

-------
The  solution  to the  nutrient  and sediment problem has  two  components.   The
first,  as shown  in  Figure  3,  is  a  sedimentation basin  in series with  a
restored  wetland  on  a drainageway  that  controls  80  percent of  the  runoff
into  the  lake.   The  1.5  acre  sedimentation basin bounded on  three  sides by
8.5  acres  of  available,  but  normally dry,  detention storage  area  is located
at  the upstream  end  of  the  system.   The  sedimentation basin and related
detention  area  will  trap  sediment  and control  flow  into  the restored  wet-
land.  The wetland would be restored to a size of about 8.0 acres by building
a  series  of  low berms across  the  existing channel,   as shown  in  Figures  3
and  4.   Vegetation such  as  cattails, sedges,  and willows will grow  in the
restored  wetland.   The sedimentation  basin-restored  wetland  is  expected to
remove  about  75 percent  of  incoming  total  phosphorus  and  suspended  solids
and have a total cost, including engineering and administration, of $250,000.

The  second component  of  the  NFS pollution  control  system is two large stomps
to  control  large  street  runoff from the  remaining 20 percent of  the  water-
shed.   The circular  sumps,  one  of  which is  shown   in Figure 5,  would be
large--both are  10 feet  deep  and one is  18 feet  in  diameter and the other
24 feet in diameter.   As  suggested  by Figure 6,  the  sumps are simple struc-
tures  designed  to   remove  the  suspended  solids  and buoyant  materials.   The
sumps, which are expected to achieve 65 percent reduction in suspended solids
and  a  significant  reduction  in adsorbed or absorbed phosphorus, have a total
cost of $75,000.

The  sedimentation  basin-restored wetland-sump  system is  under construction
and will be completed by late summer 1985  (Donohue, 1983).

CASE STUDY 3:  JiUTRFACK AND SIJBSIIRFACE DETENTION IN AN  URBAN AREA

Physical Sett  i
The 1,200 acre Skokie, Illinois Howard Street Sewer District (HSSD) , as shown
in Figure  7, is a  combined  sewer service area.  The  HSSD  is  completely and
densely developed and  over  80 percent of the district is single-family resi-
dential  with the  rest  being commercial and  industrial.  The  long,  narrow
HSSD is very flat having an overall slope of only 0.2 percent.

The combined sewer  system serving the HSSD discharges at the east end of the
district to an interceptor paralleling the North Shore Channel.  The start-up
of the  Chicago area Tunnel  And  Reservoir Plan  (TARP)  in 1985 will mitigate
wet weather overflow to the channel.

Problem

Athough solution of the  combined sewer overflow problem  is imminent, serious
basement  flooding  will  continue  to  occur in many of  the over 5,000 single-
family  residences  and other  buildingy ia  the  HSSD.  TARP  will improve the
outlet  conditions   for  runoff  events  from the  district,  but  it  will  not
resolve  deficient   sewer  capacities  throughout most  of the  HSSD.  Detiign,
construction, arid operation  of  a conveyance or detention system to  take full
advantage  of TARP  is ueceusary  to  solve  the  flooding problem and  is the
complete  responsibility  of   the:  Village of  Skokie.   Although  a relief sewer
system, such as  new separate storm sewers,  is  technically feasible, the cost
is prohibitively high.

                              P-III-E-5

-------
                                      FIGURE 3
                   SEDIMENTATION BASIN  IN SERIES WITH RESTORED WKTl.AWu
                                       10 ACRE SEDIMENTATION BASIN
                                       INCLUDING I5ACRE PERMANENT POND
       '    / x
   y    f, \
•»—-          /  i
   NORMALLY    'V
   DRY AREA  / N
   PERMANENT
   POND
                                                   SEDGE,WILLOWS ft CATTAILS
     LOW LEVEL BERMS
                    SEDGE a WILLOWS
                                                              OUTLET CONTROL
                   MIXED VEGETATION
                     SEDGE a WILLOWS
                        STREET  DITCH
                                                             MC  CARRON   LAKE
       LEGEND
       LOW LEVEL BERM
 7     SURFACE STREAM MONITORED (ID NUMBER)
 . - •*   DIRECTION OF WATER FLOW THROUGH  WETLAND
Source:  Donohue,  1983
                                   P-III-E-6

-------
                              FIGURE A

   LONGITUDINAL SECTION THROUGH  SEDIMENTATION BASIN AND RESTORED WETLAND
                   SEDIMENTATION
                   BASIN  BERM
 SEDIMENTATION
 BASIN   ^---''
                                                ;,
FILL
                                                 TYPICAL  SECTION OF  LOW  BERM
  NOTE  LOW BERMS TO BE CONSTRUCTED AT ELEVATION
        AND SPACING TO PROVIDE A WETLAND SLOPE
        OF LESS  THAN 0.002 FT/FT
        LOW BERM OVERFLOW  SECTIONS PROTECTED
        FROM EROSION BY PERMANENT PLASTIC
        REINFORCING MESH  IN  TOPSOIL.
                                                                                 NO SCALE
                                                                  WETLAND
                                                                  EXIT BERM
Source:  Donohue,  1983

-------
                                      FIOURE 5



                                   SUMP LOCATION
                                              MC CARRON   LAKE
     EXISTING TEMPORARY CURBING
     PROPOSED PERMANENT
                                                       EXISTING ASPHALT  SPILLWAY
                                                      PROPOSED STORMWATER SUMP
             FLOW  PATTERN OF
             STREET  RUNOFF
Source:  Donohue,  1983
                                  P-III-E-8

-------
                                   FIOURF. 6
                              SECTION THROUGH SUMP
                            FLOW OF STORMWATER RUNOFF
                                                PROPOSED  STORMWATER SUMP
   MC CARRON BLVD. S.
MC CARRON
  LAKE
Source:  Donohue,  1983
                              P-III-E-9

-------
Solution

After  extensive  field  and  office  investigations  spanning  several  years,
the  Village of Skokie  has begun  implementation  of a  runoff control  system
(RCS)  in  which stormwater  detention will  be  retrofitted into  the  existing
stormwater  drainage  and combined sewer system  (Donohue,  1982, 1984a,  1984b,
1984c; Walesh and Schoeffmann, 1984).  The concept is shown in schematic form
in Figure 8 for existing and proposed conditions.

Stormwater  will  be  intentionally  ponded  in streets in  a controlled fashion
and  then  gradually  released  to the  combined sewer  system at a rate that can
be  conveyed by  the  combined  sewer  system without  surcharging.    On  those
streets where insufficient surface storage is available, supplemental subsur-
face  storage  tanks  will  be  used.  More  specifically,   the  recommended RCS
consists of:

     0    431  flow  regulators  functioning  in  conjunction  with 210  street
          berms.

     0    Eight subsurface storage tanks.

     0    9,500  feet of  30-irich  to 72-inch separate  relief sewer for the
          commercial, downstream end of the HSSD.

The estimated 1985 cost for the recommended RCS, including engineering, legal,
and administration fees, is about $8.5 million or less than 20 percent of the
cost of conventional relief sewer system.

Most of the temporary street  detention will be accomplished by installation
of  simple  flow regulating  devices in  existing  catch   basins  as   shown  in
Figure 9.   Each  catch basin contains a sump which  traps both settleable and
buoyant material.    In  addition,  each subsurface  tank will,  as  illustrated
in  Figure  10,  contain  a  large  sump to trap  soluble  and buoyant  material.

Therefore,  essentially  all stormwater will,  while in route  to the combined
sewer, pass through and be detained in  a  sump where  settleable  and buoyant
material,   along with  adsorbed and  absorbed potential   pollutants, will  be
removed.   These  sumps will be  regularly  cleaned  and  maintained by Village
personnel.

Design of flow  regulators,  street berms,  and tanks  is  almost complete. Flow
regulators  installation  is underway  and  will be completed  this  year.   Tank
and berm construction will begin soon and be completed in 1985. Relief sewers
will be designed in  1985 arid are scheduled for construction in 1986.

Although  control  of suspended  solids  and  other  potential NFS  pollutant  is
not  a  primary  objective of the Skokie RCS,  this  system or a variation on it
has  the potential  to control  NFS pollutants near their source because of the
widespread  detention occurring  and the presence of numerous sumps and traps.
An engineered RCS  built into  an new separately sewered development or retro-
fitted into an existing  separately sewered or  combined sewered area  could
have the following water quality benefits:
                                P-III-E-10

-------
                                        FIOURE 7
               STUDY  ARKA:  HOWARD STREET SEWER DISTRICT IN SKOKIK, ILLINOIS
                                               STUDY
                                                HOWARD ST
                                                                            r> Mi
Source:   Waleeh  and Schoeffmann 198A
                                    P-III-E-11

-------
                                                        FIGURE 8
                             STREET PONDING,     REGULATED CATCH BASIN AND UNDERGROUND TANK
i
-TI
                         DOWNSPOUT
                         CONNECTED
                         TO SEWER
            DISCONNECTED
            DOWNSPOUT
                                                              TEMPORARY K>NDING
                                    TANK OPTION
                         HOUSE
                                                                   VtLLACE COMBINED  SEWER
                                                                                                      USED WHERE  STREET
                                                                                                      PONDING CAPACITY  IS
                                                                                                      INADEQUATE
                   EXISTING  CONDITIONS
RUNOFF   CONTROL  SYSTEM
    Source:  Walesh and  Schoeffmann,  198-

-------
                                FIGURE 9
                FLOW REGULATOR AND SUMP IN CATCH BASIN
                                         - CATCH  BASIN
        OUTLtT PIPE
                             -ORIFICI RESTRICTOR
Source:  Wnleah  and Schoeffraann, 1984
                              P-III-E-13

-------
                                    h:  10


                           UNDERGROUND TANK
V /
-- f
A
/
^
tXOAVAMON
1. 1 Ml IS
               COMhlM.lj---
               SE Wfft
K t UW
HEGULATOR
Ht I NTOHCI-C>
CONCHL 1C  tlOX
Source:  Walesh  and  Schoeffmann, 1984
                               P-III-E-14

-------
     0    Reduced  frequency  and  volume  of combined  sewer  overflows to  the
          surface water system.

     0    Control  of  NFS  pollution  near its  source  for subsequent  removal
          as a part of sump cleaning and maintenance.

     0    Reduced peak flows  at a wastewater treatment plant.

OBSERVATIONS

The following observations are made based on experience,  such  as the examples
presented  in this  paper,  with  engineering of  NFS  controls  at. the  "grass
roots" level in response to particular environmental problems:

     1.   Suspended Sediment  is Primary Target:   In  addition   to   being  a
          potential  pollutant,  suspended  sediment can  be a   tranporter  of
          potential pollutants such  as  phosphorus,  pesticides,  heavy metals,
          bacteria, and oxygen-demanding materials.   Therefore, the success-
          ful control  of  suspended  solids  should  indirectly result  in sub-
          stantial control of other non-point source pollutants.

     2.   Small Incremental Expense Concept:   In  terms  of  land  acquisition
          and construction  costs, relatively  little  additional expenditures
          are likely  to be  required to add effective water  quality control
          components  to planned  stormwater detention/retention  facilities.
          That is, control of the quality of stormwater can often be achieved
          for a small incremental cost above that needed  to control the quan-
          tity of stormwater.

     3.   Need for Design-Oriented Manual:    Research   results,  performance
          data,  and design  tools and  techniques  should be  integrated into
          a  design-oriented  manual.   Such  a  manual could  present  tools and
          techniques  to be used  by  engineers in  the planning,  design,  and
          operation of  temporary and permanent NFS pollution control facili-
          ties such as  sedimentation basins,  detention/retention facilities,
          enhanced wetlands and sumps.

     ^-   Stringent Inspection and Maintenance Requirement^:  The  principal
          purpose  of  NFS  control measures is removal and concentration of
          potential  pollutants  from stormwater   runoff.   At minimum,  sedi-
          ment and  other  debris  will be collected  in NFS control facilities.
          These facilities must be regularly cleaned and otherwise maintained
          if  they  are  to function  as  intended.   Designers should clearly
          indicate  the  type  and  frequency  of  inspection  and  maintenance and
          provide cost  estimates.

     5.   Resist Mandating Specific  NFS Measures:   Given  the  limited,  but
          growing,  knowledge  concerning  the  design  and  operation  of NFS
          pollution control  measures,  local,  state, and federal governmental
          units  should  "go slow" in mandating specific  means  of controlling
          NFS pollutants.   Such  agencies  should  focus  instead  on statements
          of intent and performance.  Premature rule-making and regulation is
                               P-III-E-15

-------
          likely to  result  in "action" but little progress.   More  is likely
          to  be  accomplished in advancing  the  state-of-the-art of achieving
          significant  control of  NFS  pollutants  by (a)  funding additional
          research  and development  projects and  pilot  studies and  (b)  by
          encouraging  the control  of NFS  pollutants  but not  dictating the
          means.

Acknowledgements
                                                                   ,'i.
The  author gratefully  acknowledges  the  ideas  and  information  provided  by
colleagues  and  drawn from  project experience at Donohue  & Associates, Inc.
However, the author is solely responsible for the paper's content.
                               P-III-E-16

-------
                                CITED REFERENCES

Donohue  &   Associates,   Inc.,   Lake Forest Storm Water Detention Facility -
Preliminary Engineering. 1981.

Donohue  &  Associates,  Inc.,  Me Carron Lake Diagnostic - Feasibility Study -
Roseville,  Minnesota. November,  1983.

Donohue & Associates, Inc., Preliminary Engineering - Runoff Control Program -
Howard Street Sewer District -  Skokie, Illinois, July, 1982.

Donobue  &   Associates,   Inc.,   Monitoring - 1983 - Runoff Control Program -
Howard Street Sewer District -  Skokie, Illinois, January, 1984a,

Donohue  & Associates,  Inc.,  Flow Regulator Pilot Study - Runoff  Control Pro-
gram - Howard Street Sewer District - Skokie, Illinois, February, 1984b.

Donohue  & Associates,  Inc.,  Preliminary Engineering - Addendum - Runoff Con-
trol Program - Howard Street Sewer District - Skokie, Illinois, March, 1984c.

Raasch,  G.E.,  "Urban Stormwater Control Project in an Ecologically Sensitive
Area,"  Proceedings - 1982 International Symposium On Urban Hydrology, Hydrau-
lics and Sediment Control,  University  of   Kentucky,   Lexington,  Kentucky,
July 27-29,  1982, pp. 187-192.

Walesh,  S.G.   and   M.L.   Schoeffmann,   Surface and Sub-Surface Detention in
Developed Urban Areas: A Case Study,  presented  at the  American  Society  of
Civil  Engineers  Conference "Urban  Water  -  84,"  Baltimore,  MD.,  May 28-31,
1984.
                               P-III-E-17

-------
                               SUMMARY
                              Panel TV
             Organization and Institutional Arrangement of
                   Nonpoint Source Control Programs
PANEL MEMBERS

Moderator:  Jim Bauman, Nonpoint Pollution Section, Wisconsin Department
of Natural Resources, Madison, WI.

Assistant Moderator:  Dean Massey, Economic Research Service, U.S.  Depart-
ment of Agriculture. Madison, WI.

Panelists:

Floyd Heft, Soil Conservation Society of America,  Columbus, OH.
William Horwath, National Association of Conservation Districts, Stevens
  Point, WI
William Katz, Professor, Marquette University, Milwaukee, WI.
Fred Madison, Associate Professor, University of Wisconsin, Madison, WI.
Glen Stout, Professor, University of Illinois, Urbana, IL.

Recorder:  Gary Korb,  Southeastern Wisconsin Regional Planning Commission,
  Waukesha, WI.

INTRODUCTION

Sources of nonpoint pollution are numerous and diversified and their control
would presently be the responsibility of a number of federal, state and
local agencies and private landowners and developers.   The present approach
to nonpoint pollution control relies on voluntary compliance of polluters
with little financial assistance available from federal and state sources.
The oversight of the nonpoint pollution control programs are carried out
by U.S. Department of Agriculture through its soil and water conservation
programs; by the U.S. Environmental Protection Agency which has very limited
resources to carry out any meaningful nonpoint pollution abatement programs,
and by states that have enacted nonpoint pollution control and abatement
laws.  The questions this panel was faced with were as follows:

1.  Is there an adequate organizational structure in place which would
    enable to carry out nonpoint pollution abatement programs?

2.  If the present organizational structure could be used how could it
    be improved?

3.  If the present organizational structure is not adequate what new
    organizations and institutions are needed to carry out the programs?

DISCUSSION

Panelists agreed that the present organizational structure and institutions
are not adequate however the panel did not   reach   a conclusion whether a
new organizational structure is needed or whether present organizations and
institutions should be improved.

                               P-IV-1

-------
 There    are a number of drawbacks  and  gaps  in  the present organizational
 structure which were identified  by  one  of  the panelists as follows:

 1.   Inadequate funding.

 2.   Lack  of joint  planning  for soil  erosion  and water quality control among
     the agencies.

 3.   Lack  of public accountability due to large  number of agencies involved.

 4.   Institutional  goals  are not  clear.

 5.   Dilution and shallowness  of  educational  programs due to voluntary
     approaches.

 6.   Uneven treatment and understanding  of  the nonpoint pollution problem
     across the state lines.

 7.   Targets  and  goals of the  programs have not been clearly specified.

 8.   U.S.D.A.  agency roles and roles  of Soil  and Water Conservation Districts
     have  not been  clearly defined.

 Another    member of the  panel pointed out  the legislative limits that have
 been imposed on  present  Soil and Water  Conservation Districts.  In 13 fiorth
 Central states  there are great differences in the regulatory powers of the
 SWCD's, and  subsequently, there are  substantial differences among the states
 as to:

 1.   Ease  of adoption of  ordinance.

 2.   Recommendations  for  soil conservation  practices.

 3.   Administrative  procedures.

 4.   Enforcement  provisions.

 Some  stateshave so  called  permissive regulatory powers, otherstateshave
mandatory  regulatory powers.  States with  permissive regulatory policies
may  have difficulties  in getting necessary regulations adopted.

 The availibiltyof fundsand  funding for NFS control programs was also ex-
 tensively  discussed.   Some  panelists and discussers had reservations against
 subsidies  since  they  could make the  polluters wait for the subsidy and do
essentially  nothing.   In general, financial aid (cost sharing) and subsidy
 should be  given  primarily to existing farms that have initiated a nonpoint
pollution  problem.    New  farms and fanning entreprenuers  should consider
the NFS control as a  part of their cost.

An innovative financial  arrangement was suggested by one panelist for lake
management districts whereby lake users (water quality beneficiaries) pro-
vide  subsidies to upstream  farmers to implement nonpoint pollution control
programs.

                                P-IV-2

-------
CONCLUSIONS AND RECOMMENDATIONS

1.  The panel concluded that the present lack of funding and somewhat
    inadequate organizational structure should not preclude the initia-
    tives to limit nonpoint pollution inputs.  The state-of-the-art of
    the technology and existing organizational structure are satisfactory
    for an initial effort to curb the nonpoint pollution in many watersheds.

2.  Lack of state and, primarily, federal  funding has been cited as the
    major obstacle for successful nonpoint pollution abatement programs.

3.  The role of existing agencies should be clearly identified to avoid
    overlaping and unnecessary competition.

4.  Cost sharing and subsidies should be used with caution in order for
    them not to become an obstacle and a factor causing slowdown of the
    efforts.  Cost sharing and other subsidies should be limited to
    existing farms and based on their potential to degrade water quality.
    New farming operators and construction developers should incorporate
    the nonpoint pollution control in their cost.

5.  New organizational structures may be needed if the present existing
    organizations  and institutions are unable to accomplish the goals.
    The future most optimal organizational structure is not known.

6.  The organizations and institutions should have some regulatory back-up
    if voluntary approaches fail.
                                P-IV-3

-------
              ORGANIZATION AND INSTITUTIONAL ARRANGEMENTS
                         OF NFS CONTROL PROGRAMS

                           William J. Horvath
                 North Central Regional Representative
             National Association of Conservation Districts
                        Stevens Point, Wisconsin
                                ABSTRACT

Under S304(4)(1)(A,B,C) of Public Law 92-500, the Federal Water Pollution
Control Act Amendments of 1972, EPA was directed, after consulting with  .
appropriate state and Federal agencies and other interested persons to:
     "issue to appropriate Federal agencies, the States, water
     pollution control agencies, and agencies designated under
     S208	information including guide lines for identify-
     ing and evaluating the nature and extent of nonpoint sources
     of pollutants resulting from
          (A) agricultural and silvicultural activities,
          including runoff from fields and crop and forest
          land;
          (B) mining activities, including runoff and siltation
          from new, currently operated and abandoned surface
          and underground mines;
          (C) all construction activity, including runoff from
          the facilities resulting from such construction. . ."

Both the Federal Government and state governments have struggled with the
complexity of nonpoint source pollution control.  Control raises complex
legal problems by the very definition of it.  Standards must be set and
identification of source becomes a burden to the legal framework of
enforcement.

More importantly—the number of actors necessary to control nonpoint
pollution becomes difuse and often confusing.  Institutional and legal
remedies can not be wrapped up in neat packages.  There are shortcomings
that can only be addressed below the Federal level.
                               P-IV-A-1

-------
                ORGANIZATION AND INSTITUTIONAL ARRANGEMENTS
                          OF NFS CONTROL PROGRAMS

BACKGROUND

     It is well established that until the Federal Water Pollution Control
Act  (FWPCA) was revised by PL 92-500 in 1972, water quality efforts dealing
with nonpoint pollution were fairly limited to research.  About the only
programs that dealt with nonpoint sources did so on an indirect basis.
Regular soil conservation programs and watershed protection activities of
SCS, cost-sharing programs such as AGP administered by ASCS, and pesticide
registration and labeling programs of ARS were the bulk of activities.
State efforts were generally conducted as an adjunct to these efforts.
Soil and water conservation districts and their associated agencies only
indirectly dealt with the problem in what was thought of as best a bootleg
operation.

     PL 92-500 redirected the nation's efforts in water pollution abatement
efforts.  These new efforts focused attention on controlling both point and
nonpoint pollutants at their source.  Obviously, nonpoint took a secondary
seat to the massive efforts undertaken to control point sources.  The
problem of point was more definable, the technology to a degree in place,
and the institutional arrangements more clearly defined.

     It was difficult to define nonpoint sources as well from the legal
standpoint.  The first consideration is the definition of pollution itself.
Legislation is designed to maintain water quality by controlling pollution
and that depends upon limiting contamination below some established
standard.

     Then what about identifying the pollutant itself?  If you identify it
does it become a source point and if you can't how can you legally impose
sanctions on the polluter?

     Such was the struggle in simplistic terms,  but as the entire water
quality program gained experience,  it became clear that pollution
contributed from nonpoint sources could easily equal or exceed efforts in
point source control.

     Nonpoint was never defined by statute but defined more by exclusion in
generally three categories.

     (1)  It was not susceptible to treatment at the "end of a
     pipe" however controllable by other changes generally in land
     management;

     (2)  It was convened by a natural process such as a storm or
     seepage rather than through a deliberate controlled effort; and

     (3)  It was generally generated by a diffused set of man's
     activities and not identified with any particular one.
                              P-IV-A-2

-------
     Refining was done to this identification for a number of reasons
including the well known Natural Resources Defense Council v. Train which
narrowed the definition of nonpoint sources.  That called for regulation
including the permit system to include smaller feedlots, irrigation return
flows and storm sewers in urban areas and certain activities of
silviculture such as rock crushing, gravel washing, log sorting and
storage.  A howl of protest and a flurry of activity arose over this
ruling.  Subsequent rules by EPA in 1976 softened some of the requirements
for irrigators through the issuance of general permits if in conformance
with 208 plans.

     Farm and some closely aligned organizations like NACD spent a
considerable amount of time trying to educate the actors to be involved.
First to educate EPA and others aligned with them that nonpoint source had
to be handled differently than point source.  A permit and enforcement
system just wasn't  feasible for the bulk of nonpoint problems.  It would
not work to place a policeman on every farm in America.

     Secondly, an educational process was needed in EPA as well as state
water quality agencies that a mechanism was in place to handle a good
portion of the probem since it became clear that soil movement was a major
culprit in nonpoint sources.  Nor was it as easy as establishing a single
set of laws or ordinances to handle problems associated with agriculture,
urban storm water, construction site run-off, mining run-off and
silviculture.

     NACD through a series of grants from EPA, conducted a number of
nonpoint source pollution projects including a series of 45 state sediment
control institutes, a series of nonpoint pollution control workshops and a
number of publications including a series of 48 Nonpoint Notes published
over a period of years designed to inform all who were involved in the
nonpoint source areas.  The Soil Conservation Service placed over 30 staff
people on IPA's in Washington and EPA regional offices to assist them in
establishing a nonpoint program.  Cooperative Extension placed people on
IPA in regional offices or contracted for conducting educational
activities.

     The Chicago office of EPA, because of the implications of nonpoint in
the Great Lakes Basin, effectively utilized Section 108 of the Water
Quality Act to fund several water basin projects to test certain
approaches.  The Red Clay Project along Lake Superior tested new
methodology to combat sediment in red clay soils, the Washington County,
Wisconsin project was designed to take a look at institutional
arrangements, and the Black Creek, Indiana project was designed to look at
a variety of best management practices (BMPs) and determine methods for
measuring the results of BMPs against an improved water quality.

     The appropriate roles of state and other management agencies received
their most significant direction by the issuance of SAM-31, a program
guidance memorandum establishing implementation criteria for EPA approval
of NPS elements of state water quality managements.  This guidance document
                              P-IV-A-3

-------
stated that  the appropriate  level of government and the form of a
regulatory program depended  on existing  laws, programs, and governmental
working relationships.  Conditional approval could be given by EPA if the
state could  develop changes  in legislative authorites or administrative
capabilities.

     Fully approved NFS regulatory programs had to include the following
elements:
     -enforcement authority
     -a designated management agency to  implement the program
     -authority to use the control tools necessary to implement the program
     -monitoring authority
     -authority to require application of BMPs
     -authority to control the activity> pollutant or geographical area
      to be  assigned

     Sam-31  also provided for approval of non-regulatory NFS control
programs but such programs had to include
     -a schedule for the program including implementation,
      monitoring, and progress evaluations
     -annual progress reporting
     -adequate technical and financial assistance
     -effective educational programs
     -identification of BMPs

REMAINING INSTITUTIONAL AND ORGANIZATIONAL GAPS

     While a decade has passed since the public attention was first turned
to nonpoint pollution, there remains a number of organizational and
institutional impediments to completing the task.

Lack of. reliable funding sources both at the state and Federal level,

     Even though it is recognized that nonpoint source is the remaining
significant problem in water quality, Congress has been unable to
appropriate the funds necessary to attack the problems.   At the state level
it is a mixed bag primarily because major emphasis in the last
several years has been on soil erosion control.  Water quality and offSite
benefits are secondary in concern.

Lack of combined planning and implementation for water quality and soil
erosion.

     Planning for soil erosion and water quality at the  state level as well
as the Federal level is done in different agencies which have different
objectives.  They are not necessarily coordinated at the state level  and
use a separate set of disciplines.
                               P-IV-A-4

-------
Lack £f enforcement and standards leads to the lack of accountability.

     With diffusion of best management agencies in the process including
Federal agencies, state agencies and local agencies, the public has a
difficult time of pin-pointing anyone accountable for program results or
expenditure of funds.

The institutional role for water quality control agencies in nonpoint is
not clear.

     They must rely heavily on other conservation agencies that may be
reluctant to play supporting roles in the overall effort.

A diffusion of. educational effort combined with voluntary application is; _a_
problem.

     Education of those necessarily involved in the process including those
landowners who apply them are diverse and numerous.  On a national scale
this involves literally millions of people.

Unequal ability between states t_o adequately develop a_ state mechanism is
still a_ problem.

     A river basin such as the Great Lakes or Susquehanna River has a bevy
of state capabilities which often do not mesh with the severity of the
problems.

An over dependency on data and data that may be flawed still exists.

     This unecessarily complicates implementation of water quality
projects.

The external influcence of factors in controlling nonpoint are ever
present.

     Weather, changes in technology in farming, state budget problems, all
are major factors in planning.

Conflict between old concepts £f treating every problem equally with some
effort rather than targeting the project i^s still a political problem where
responsibility is diffuse.

The relationship between USDA agencies and conservation districts which
most often provide technical assistance  and coordination and state water
quality agencies i^s not well defined.

     The providers and implementers have different expectations.
                               P-IV-A-5

-------
     Overall, the institutional arrangements are not as clear and concise
as we might like because of the nature of the problem which is diffuse and
requires a wide variety of approaches.  However, the funding for water
quality programs dealing with nonpoint supercede all problems.  When
society funds these programs most of the implementing problems will
disappear.
                              P-IV-A-6

-------
                NORTH CENTRAL STATES CONSERVATION DISTRICTS:
                     EFFECTIVENESS OF REGULATORY POWERS

                               Dean T. Massey
              Agricultural Economist, Economic Research Service
                       U.S. Department of Agriculture
                     Law School, University of Wisconsin
                           Madison, Wisconsin, USA
                                  ABSTRACT

Soil and water conservation districts are the only local  agencies primarily
responsible for controlling soil erosion and reducing sediment.   The 13 North
Central states have such districts or an equivalent.   Districts  in 9 of these
states are authorized to regulate land use.  Kentucky, Nebraska, North Dakota
and Wisconsin provide for permissive regulatory powers; other states require
more mandatory controls.  Iowa and Ohio regulations are based on soil loss
tolerances; Michigan's regulations are based on land-disturbing  activities,
and South Dakota's, on conservation standards.   Illinois has both permissive
regulations and mandatory standards for land-disturbing activities.   A
district's effective abatement of nonpoint pollutants depends on the ease of
adopting regulations, conservation practices to be included, procedures for
administering regulations, and enforcement methods.

Adopting land use regulations in states with permissive regulatory powers can
be difficult, since proposed regulations must often be approved  by a high
percentage of voters.  These states permit districts  to include  adequate con-
servation practices in erosion control regulations for agricultural  land.
None of these states specify how the regulations are  to be administered or by
whom, although most provide for enforcement power.

In states with more mandatory controls, state agencies have more control over
nonpoint pollution.  Regulations are handled in three ways:  (1) districts
are required to adopt regulations and have them approved by a state agency;
(2) districts are required to adopt conservation standards and regulations
based on state agencies' programs and guidelines;  or  (3)  state agencies are
required to adopt statewide control programs and regulations that are admini-
stered and enforced by state and local agencies.

Only Illinois, Iowa and South Dakota conservation districts administer land
use regulations.   Permits are required for land-disturbing activities in
Michigan and South Dakota.  Agricultural land in Illinois, Iowa  and South
Dakota is deemed to comply with regulations until  a complaint is made; the
districts then provide for enforcement.  No enforcement provisions exist in
Ohio under either agricultural or urban pollution control  rules.

Keywords:  Agricultural runoff pollution, conservation practices, land use
           regulations, nonpoint pollution, nonpoint  pollution abatement,
           regulatory powers, soil erosion, soil conservation districts
                                  P-IV-B-l

-------
 INTRODUCTION

      Nonpoint source water  pollution  is  a major environmental problem in much
 of  the  United States.   Sediment  from  erosion  is the major nonpoint pollutant,
 and agriculture contributes more than half of this, both by mass and volume.
 Almost  half  of the agricultural  sediment comes from cropland, and nearly half
 of  all  cultivated  cropland  consists of soils  with a high potential for ero-
 sion.

      Controlling sources  of nonpoint  pollutants is considered the best method
 of  preventing sediment  and  plant nutrients from reaching water resources.
 Possible  institutional  tools for controlling  sources of nonpoint pollutants
 include prohibiting  certain land uses at particular locations and regulating
 uses on land at the  source.  Most of  the responsibility for controlling non-
 point pollutants rests  at the  local level, with federal and state governments
 establishing water quality  standards,  guidelines, and soil loss limits, and
 providing  technical  and financial assistance.  Soil and water conservation
 districts  are the  only  local agencies whose primary responsibility is con-
 trolling  soil  erosion and reducing sediment and other resultant nonpoint
 source  pollutants.

      Section 208 of  the Clean  Water Act  provides for the identification of
 agricultural,  urban  and construction-related  nonpoint sources of pollution,
 and for the  development of  areawide water quality management plans containing
 regulatory procedures to  control  pollutants from these sources (Iowa L.R..
 1977).  (Citations to applicable federal and  state statutes and state agency
 administrative regulations  are provided  in the references.)  One or more
 state,  regional  or local  agency  must  be  designated to implement the areawide
 plans,  and a  possibility  may exist for designating soil and water
 conservation  districts  as management  agencies  (Davey).

      This  paper  deals solely with the  13 North Central states:  Illinois,
 Iowa, Indiana,  Kansas,  Kentucky, Michigan, Minnesota,  Missouri,  Nebraska,
 North Dakota,  Ohio,  South Dakota and  Wisconsin.  It describes the regulatory
 powers  of  soil  and water  conservation districts in these states,  and evalu-
 ates  the effectiveness  of these  regulatory powers in abating nonpoint sources
 of  pollution  in  both rural  and urban  areas; it finds that the effectiveness
 of  regulations  depends  on ease of adoption, conservation practices that may
 be  included,  administrative procedures, and enforcement ability.   District
 regulatory powers  are also  evaluated  for their effectiveness in  providing the
 local regulatory programs required to  implement Section 208 areawide water
 quality management plans.

 DISTRICT ORGANIZATION AND REGULATORY  POWERS

     All 13 of  the North  Central states  have soil  and  water conservation dis-
 tricts  or an  equivalent.  Nebraska eliminated conservation districts and
created 24 natural  resources districts that assumed their functions.
Wisconsin abolished  soil  and water conservation districts and transferred
 their functions  to county boards, requiring them to create land conservation
committees to  implement soil conservation programs.

     Soil  and water conservation districts and their equivalents,  except in


                                  P-IV-B-2

-------
Indiana, Kansas, Minnesota and Missouri, have some type of land  use regula-
tory powers to control soil erosion and sediment.   Illinois,  Kentucky,
Nebraska, North Dakota and Wisconsin give districts or counties  permissive
power to adopt land use regulations.  More mandatory controls are established
in Iowa, Michigan, Ohio and South Dakota, and in Illinois (under a second
statute).

OPERATION OF LAND USE REGULATIONS

Adoption Procedures

     District supervisors in four of the states with permissive  regulatory
powers—Illinois, Kentucky, Nebraska and North Dakota—may develop and  adopt
land use regulations after a public hearing and after the proposed regula-
tions have been approved by the required number of voters on  a referendum.
Approval in the four states requires a favorable vote of between 75 and 90
percent of the qualified voters voting in the referendum.  In Wisconsin,  the
county land use conservation committee develops the land use  ordinance  and
submits it to the county board, who may adopt it after a public  hearing.
However, the ordinance adopted by the board does not become effective until
approved by a majority of the voters on a referendum in the affected area.

     Even though district supervisors or county boards in these  five states
are authorized to adopt land use regulations, the requirements for approval
on a referendum as a prerequisite to adoption make adoption almost impossible
in these states, except Wisconsin.   Requiring a referendum and a favorable
vote of more than 51 percent can block the enactment of regulations, espe-
cially if voting is limited to landowners within the district (Holmes).
Because of such adoption procedures, these states cannot effectively provide
the regulatory programs necessary for nonpoint pollution control  under
Section 208 of the Clean Water Act.

     Adoption of the more mandatory regulations in Iowa, Michigan, Ohio and
South Dakota, and under Illinois' second statute,  involves a  joint effort  by
state agencies and local governments and soil and water conservation dis-
tricts (Davey; EPA, 1978).  Illinois amended its original statute requiring
the State Department of Agriculture to adopt guidelines, and  requiring  dis-
tricts to adopt erosion control programs in accordance with state guidelines.
Iowa district commissioners have classified land in the state's  100 districts
and established and implemented soil loss limits for each classification.

     Illinois and South Dakota require state agencies to prepare guidelines
for erosion and sediment control, and require conservation districts to adopt
regulations or standards based on the state guidelines (Holmes).   The
Illinois Department of Agriculture and South Dakota Conservation Commission
have developed comprehensive statewide erosion and sediment control  programs
and guidelines to be implemented and administered by the conservation dis-
tricts.  Each district in Illinois had until April 1982 to develop and  adopt
soil erosion and sediment control programs and standards consistent with the
state program and standards; the department reviewed and approved district
programs and standards, and also developed appropriate programs  and standards
for districts that failed to develop their own.  All except two  conservation
districts in South Dakota have developed and adopted conservation standards


                                   P-IV-B-3

-------
 in cooperation and  consultation  with  counties and municipalities; the commis-
 sion  reviewed  these standards  prior to adoption.

      State  agencies adopt  land use regulations that are enforced by state
 agencies  and local  governments in both Michigan and Ohio.  The Michigan
 Department  of  Agriculture,  with  the assistance of the conservation districts,
 prepared  and submitted  a unified statewide  soil erosion and sedimentation
 control program for approval by  the Water Resources Commission.  The commis-
 sion,  with  the assistance  of the Department of Agriculture, prepared and
 adopted rules  for the unified  program that  became effective on January 1,
 1975.  The  Ohio Division of Soil and  Water  Districts, with the approval of
 the Soil  and Water  Conservation  Commission, adopted standards to abate soil
 erosion and degradation of state waters from agricultural, urban and live-
 stock  feeding  activities (Holmes).

     Conservation districts in the five states with more mandatory controls
 come closer, as adoption procedures are concerned, to meeting the program re-
 quirements  of  Section 208  water  quality management plans.  Illinois, Iowa and
 South  Dakota give districts the  most  authority in land use regulation:  dis-
 tricts in these states  adopt their own soil erosion and sediment control pro-
 grams, regulations  or standards  and have them approved by a state agency to
 assure compliance with  statewide programs.  Problems would exist, however,
 for districts  in Michigan  and  Ohio, because rules and standards are adopted
 at the state level,  with districts playing  only an advisory role.

 Conservation Practices  to  be Included inRegulations

     Legislation should be  sufficiently broad to enable state and district
 regulations to  include  a variety of soil management practices from which can
 be selected the most appropriate combination for each situation or condition
 (EPA,  1973).   Criteria  for  selecting management practices include their abil-
 ity to manage  pollutants generated from nonpoint sources, compatability with
 water  quality goals, effectiveness in preventing or reducing pollutants,
 practicability,  and  technical  capability of preventing or reducing runoff.

     Illinois,  Kentucky, Nebraska, North Dakota and Wisconsin, with their
 permissive  powers, do not require the inclusion of specific provisions in
 their  land  use  regulations.  Enabling legislation in these states,  except in
 Wisconsin,  generally provides  that district regulations may contain any one
 or  a combination of  the following:  (a) requirements for performing necessary
 engineering operations; (b) requirements for observing particular methods of
 cultivation; (c) specifications  for cropping programs and tillage practices
 and (d) provisions for  removing  highly erodable areas from cultivation
 (Holmes).

     Enabling legislation  in these states is agriculturally oriented and
 applies to  soil  erosion control  (Davey; EPA, 1973).   Wisconsin's statute has
 the strongest provision prohibiting land uses and management practices that
 cause excessive  erosion, sedimentation and nonpoint source pollutants.   Land
 use regulations  adopted by conservation districts in the other four states
 would probably  be too narrow to  fulfill  the requirements of the Section 208
 areawide management  plans.   Such regulations, if adopted,  would adequately
manage only nonpoint pollutants derived from agriculture.

                                  P-IV-B-4

-------
     Conservation practices that may be included in regulations of the five
states with more mandatory programs vary.   Iowa's program is based on soil
loss limits, and Michigan's is based on land-disturbing activities.   A combi-
nation of soil loss tolerances and standards for land disturbances forms  the
basis of the Illinois, Ohio and South Dakota programs.

     Landowners in Iowa are required to employ soil conservation and erosion
control practices on their land, but district commissioners may not specify
particular practices to be employed as long as the owner voluntarily complies
with established soil loss limits.  Commissioners also may not require the
employment of erosion control  practices on land used in good faith for only
agricultural purposes and may not prohibit operators from fall  plowing.

     The Michigan Water Resources Commission adopted rules based on land-
disturbing activities, implementing the Department of Agriculture's recom-
mended unified statewide program that included provisions for land use plans
and permits.  Under the rules, all persons engaged in earth changes that  re-
quire a permit must design, implement and  maintain acceptable erosion and
sedimentation control plans that effectively reduce erosion.  Permits are re-
quired when disturbing one or more acre of land, or if the disturbance is
within 500 feet of a lake or stream.  In addition, approved soil  and water
conservation district erosion and sedimentation control standards and speci-
fications must be followed by those requiring a permit.  Agricultural  prac-
tices were included in the rules in 1979,  and now agricultural  earth distur-
bances, except for plowing and tilling, or unless the farmer has an agreement
with the district, are subject to permit and plan requirements.

     Conservation practices incorporated in land use regulations in Illinois,
Ohio and South Dakota are based on a combination of soil  loss tolerances  and
conservation standards for land disturbances.  Under Illinois Department  of
Agriculture guidelines, all conservation systems and practices applied to
agricultural land must seek to reduce soil losses over a period of years  to
levels at or below the soil loss tolerance ("T" value) established by the
Soil Conservation Service.  A "T" value is the average annual  soil  loss,  in
tons per acre, that a given soil may experience and still  maintain its pro-
ductivity over an extended period of time.  Policies and specifications for
various erosion and sediment control devices, structures and practices ap-
plied to agricultural land were also given in the guidelines.   Programs and
standards for various soils and land uses  adopted by the districts contain
criteria, guidelines, techniques and methods for controlling erosion and
sediment from land-disturbing activities (Holmes).

     District conservation standards or soil  loss tolerance limits in South
Dakota are to be consistent with the State Conservation Commission's guide-
lines.  The guidelines contain recommended soil loss limits and suggested
conservation practices and methods for controlling erosion and sediment
(Holmes; EPA, 1978).  Counties adopt ordinances incorporating district stan-
dards relating to agricultural and non-agricultural land-disturbing activi-
ties; non-agricultural activities subject  to permit requirements must adhere
to the district's soil loss standard and to a soil erosion and sediment con-
trol plan (Holmes).

     The rules adopted by the Ohio Division of Soil and Water Districts to

                                 P-IV-B-5

-------
 abate agricultural  nonpoint pollutants  provide for  the achievement over a
 period of years  of  the  applicable  soil  loss  tolerance or permitted soil loss
 values.   Farmers are  required  to practice  conservation and to follow a man-
 agement system,  so  that under  given  cropping and management conditions, the
 predicted soil loss (using  the Universal Soil Loss  Equation) from erosion is
 equal  to or  less than the  "T"  value  for that specific soil series, as speci-
 fied  in the  Ohio Erosion Control Guide  or  the SCS Technical Guide.  Farmers
 may not use  earth-disturbing practices, including tillage, on land immediate-
 ly  adjacent  to water, unless practices  are constructed or implemented in ac-
 cordance with proper  engineering designs.

      Conservation practices that may be included in the land use regulations
 of  the five  states  with a more mandatory approach are generally broader than
 those  of the states with permissive  regulations.  All are based on land-
 disturbing activities for  both agricultural and non-agricultural activities,
 and all  include  sediment control in  addition to erosion control.  One draw-
 back  to  the  Iowa and  South  Dakota  regulations is that they do not apply to
 agricultural land unless the erosion exceeds specified soil loss limits and a
 complaint is filed.   However,  considering  the management practices and land-
 disturbing activities that  may be  included in the content of land use regu-
 lations,  the statutes and regulations of these five states are probably broad
 enough to fulfill the regulatory requirements of Section 208 areawide water
 quality  management  plans.

 Administrative Procedures

     Nothing in  the Illinois,  Kentucky and North Dakota statutes states how
 land use  regulations are to be administered or who administers them.
 Nebraska  statutes provide only that  districts have power to administer the
 rules  and regulations;  Wisconsin provides  that the county board prescribe
 administrative procedures in the land use  ordinance.  In these five states,
 enabling  legislation  concerned  with  administrative procedures is probably in-
 adequate  to  effectively provide the  necessary regulatory programs for non-
 point  pollution  in  the  Section  208 areawide plans.   Such legislation should
 provide who  is to administer regulations,  the method for administering,
 whether  permits  are to  be issued, and,  if  so, who issues them, and if plans
 and specifications  are  to accompany  permit applications.   The powers and du-
 ties of  the administrator should be  specified along with the permit,  plan,
 and specification requirements.

     Of  the five  states  with more mandatory regulatory powers, only Michigan
 and South Dakota  provide much detail  on administrative procedures.   Regula-
 tions  are administered  by districts  in  Illinois and Iowa,  by counties in
 Michigan, by local  permit-issuing authorities in South Dakota, and  by the
 State  Division of Soil and  Water Districts in Ohio.

     Illinois does not  require  persons engaged in land-disturbing activities
 to apply  to soil   and water  conservation districts for permits or to submit
 erosion and sediment control plans.  Districts,  upon request,  make adequate
 information and technical assistance available to persons  engaged in  land-
disturbing activities.   Neither are permits or plans required  for land-
disturbing activities in Iowa.
                                   P-IV-B-6

-------
     The Ohio Division of Soil and Water Districts established procedures for
the districts to administer agricultural and urban pollution abatement rules
and to enforce animal waste management rules.  The division also recommends
criteria and procedures for approving urban sediment pollution abatement
plans and issuing permits.  Under the agricultural pollution rules, the divi-
sion will inform the districts of state standards, criteria and procedures
for pollution abatement; enter into agreements with districts for implemen-
ting agricultural pollution abatement programs; provide administrative guid-
ance for implementing and administering such programs; and implement agricul-
tural pollution abatement programs in districts failing to negotiate an
agreement with the division.  Permits, plans or specifications are required
for land-disturbing activities in Ohio.

     Counties in Michigan administer the state-adopted rules; county boards
designate a particular agency responsible for enforcing the rules.  Each
county enforcing agency adopts the district's standards and specifications.
Persons engaging in earth changes must obtain permits from appropriate county
enforcing agencies, and must submit soil erosion and sedimentation control
plans with permit applications.  District standards and specifications that
have been approved by county enforcing agencies must be followed and utilized
as they apply to specified earth changes.

     Conservation standards adopted by districts in South Dakota are adminis-
tered by local permit-issuing authorities; neither the State Conservation
Commission nor conservation districts may issue permits.   Each permit-issuing
authority within the district must include provisions in  its permit procedure
to ensure that any proposed action is in compliance with  district conserva-
tion standards.   Persons engaged in agricultural or minor land-disturbing
activities need not prepare a conservation plan, secure a permit, or report
their activities to the conservation district unless they violate the dis-
trict standards.

     Not only are Michigan and South Dakota alone in providing much detail on
administrative procedures, they are also the only states  requiring the issu-
ance of permits for land-disturbing activities.  Basically, only Michigan
statutes provide sufficient detail on who administers the conservation pro-
gram and how it is administered to be adequate for Section 208 water quality
management plans.  Administration in Michigan, however, is not by the soil
conservation districts, but rather by counties.

Enforcement Procedures

     A variety of methods are available in the states with permissive powers
to enforce the land use regulations.   Of these five states, only Nebraska
does not provide for enforcement powers.  Wisconsin goes  further by providing
that county ordinances should prescribe administrative and enforcement proce-
dures.  District supervisors have the authority in Illinois, Kentucky and
North Dakota to enter upon any lands affected by the regulations to determine
compliance.

     Illinois and Kentucky districts may provide in their regulations that
persons damaged by someone violating regulations may recover from the viola-
tor.  Wisconsin ordinances may be enforced through civil  forfeitures or by

                                  P-IV-B-7

-------
 injunctions  in  court actions  initiated by the county board or the land con-
 servation committee.  There are  no criminal penalties or forfeitures in
 Illinois, Kentucky or North Dakota.  District supervisors in Illinois,
 Kentucky and  North Dakota may initiate a court action compelling violators
 to  perform the  work, or  permitting supervisors to perform the work themselves
 if  the violators failed  to perform in a specified time.  Conservation dis-
 tricts or counties in Illinois,  Kentucky, North Dakota and Wisconsin have
 adequate available enforcement powers over their land use regulations to per-
 mit regulation  by such agencies  in a Section 208 areawide management plan.

     Enforcement authority also varies in the five states with a more manda-
 tory approach.  Agricultural  land in Illinois, Iowa and South Dakota is
 deemed to comply with regulations unless a complaint is made; the conserva-
 tion districts  are then  responsible for providing at least the initial phases
 of the enforcement process.   Counties and a state agency provide the enforce-
 ment in Michigan, while  there are no enforcement provisions for the Ohio
 agricultural or urban pollution rules.

     Any person in Illinois engaging in land-disturbing activities is encour-
 aged to comply with district  standards for erosion and sediment control.
 Complaints may  be filed  with  the conservation district by any person, the
 district, or the Department of Agriculture if it is believed that a serious
 problem exists.  If the  district finds that standards have been violated, it
 notifies the violator and seeks voluntary compliance, gives a deadline for
 compliance, and suggests practices enabling the person to comply with the
 standards.

     Landowners in Iowa  are required to employ either conservation or erosion
 control practices.  An owner or occupier of land being damaged from sediment
 files a written complaint with district commissioners, charging that soil
 erosion is occurring in  excess of the district limits.  If the commissioners
 find sediment damages and excessive soil  erosion, they are required to issue
 to the violator an administrative order stating the extent to which soil ero-
 sion exceeds the district's soil  loss regulations; if the land is agricultur-
 al, the order states when the needed work is to be commenced and satisfactor-
 ily completed.  Persons  failing to comply with the order may be held in con-
 tempt and fined or imprisoned.

     South Dakota does not require persons engaged in agricultural  or minor
 land-disturbing activities to prepare a plan or obtain a permit unless the
district determines that a particular land disturber is violating district
 standards.  Once such a violation has been determined, the land disturber is
ordered to prepare an erosion and sediment control  plan, to have it approved
 by the district, and to  implement the plan.   Either the district or permit-
 issuing authority may initiate a  court action for an injunction or other
appropriate relief to enforce the order.

     County or local  enforcing agencies in Michigan are responsible for en-
forcing the state land use regulations.   Persons who fail  to get a required
permit for earth-disturbing activities are guilty of a misdemeanor.   County
or local  enforcing agencies notify the Water Resources Commission of all
violations of the state regulations or local  land use ordinances.  If the
                                  p-IV-B-8

-------
commission finds a violation has occurred, it proposes an appropriate correc-
tion agreement to the violator.  If the violator does not agree to the pro-
posal, an order requiring correction, enforceable by the court, is issued by
the commission.  The county or local enforcing agency may also issue cease
and desist orders and revoke permits upon finding violations of regulations
or approved soil erosion and sedimentation control  plans.

     The statutes giving the Ohio Division of Soil  and Water Districts author-
ity to adopt rules for agricultural pollution and urban sediment pollution
abatement do not provide enforcement powers.   If, however, the conservation
districts receive complaints on violations of the standards, they will attempt
to find a solution through voluntary cooperation.

     Soil and water conservation districts are involved in enforcement in all
these states except Michigan.  However, soil  erosion control in Illinois and
Ohio is voluntary; there are no means of enforcing  compliance with the stan-
dards, except through cooperative efforts.  Only in Iowa and South Dakota do
soil and water conservation districts meet the regulatory agency requirements
of a Section 208 areawide water quality management  plan from an enforcement
standpoint.

CONCLUSIONS

     All 13 North Central states have soil and water conservation districts
or similar local governmental units.  Districts in  nine states either possess
or are involved with some type of nonpoint source pollution regulatory power.
Adopting land use regulations in the five states with permissive regulatory
powers is a difficult process and involves stiff requirements for approval on
a referendum.  Enabling legislation in these states permits districts to in-
clude adequate conservation practices and measures  in the land use regula-
tions for controlling soil erosion on agricultural  land.  None of the permis-
sive regulatory states specify who is to administer land use regulations or
how they are to be administered.  All the states except Nebraska provide for
enforcement powers.

     State administrative agencies in states with more mandatory regulatory
powers — Illinois, Iowa, Michigan, Ohio and South Dakota—have more authority
in the nonpoint source pollution control process than do state agencies in
the states with permissive regulatory powers.  Conservation practices and
methods are based on soil loss limits, conservation standards and land-
disturbing activities, and they apply to both agricultural and non-
agricultural uses of land.  Only in Illinois, Iowa  and South Dakota are soil
and water conservation districts responsible for administering land use regu-
lations.  Permits are used for land disturbances in Michigan and South Dakota.
Agricultural land in Illinois, Iowa and South Dakota is deemed to comply with
the regulations unless a complaint is made; the districts then provide for
enforcement.  Counties and a state agency provide enforcement in Michigan,
and there are no enforcement provisions in Ohio for either agricultural or
urban pollution control rules.  Only districts in Illinois, Iowa and South
Dakota have sufficient powers to be considered adequate management agencies
for implementing the Section 208 areawide water quality management plans.
                                   P-IV-B-9

-------
 REFERENCES
 Davey,  W.B.  (1977).   Conservation  Districts  and  208 Water Quality Management.
    National  Association  of  Conservation  Districts, Washington.
 Holmes,  B.H.  (1979).   Institutional  Basis  for  Control of Honpoint Source
    Pollution  Under  the Clean .Water Act--Hith Emphasis on Agricultural
    Nonpoint  Sources.   U.S.  Environmental Protection Agency WH-554, Washington.
 U.S.  Environmental  Protection  Agency (1973).   Methods and Practices for
    Controlling  Water  Pollution from  AgriculturalNonpoint Sources.  EPA-430/9-
    73-015, Washington.
 U.S.  Environmental  Protection  Agency (1978).   Alternative Policies for
    Controlling  Nonpoint  Agriculture  Sources  of Water Pollution.  EPA-500/5-
    78-005, Washington.
 Federal  and State Statutes  and Administrative  Regulations
 United  States Code, Title 33,  Sec. 1238  (1982).
 Illinois Annotated  Statutes, Chap. 5, Sees.  128  to 130  (Smith-Hurd 1975).
    (permissive)
 Illinois Annotated  Statutes, Chap. 5, Sees.  111(8), 138.3 to 138.9 (Smith-
    Hurd  Cumulative  Supplement  1984-1985).  (mandatory)
 Illinois Department of Agriculture.   Illinois  Erosion and Sediment Guidelines
    (April  18, 1980).
 Iowa  Code  Annotated,  Sees.  467A.42 to 267A.51  (West 1971 and Cumulative
    Supplement 1984-1985).
 Kentucky Revised Statutes,  Sees. 262.350 to  262.520 (1981 and Cumulative
    Supplement 1984-1985).
 Michigan Compiled Laws Annotated,  Sees. 282.104  to 282.117 (1979 and
    Cumulative Supplement 1984-1985).
 Michigan Administrative  Code,  Rules  323.1701 to  323.1714 (1979).
 Nebraska Revised Statutes,  Sees. 2-3244 to 2-3250 (Reissue 1983).
 North Dakota Century  Code,  Sees. 4-22-27 to 4-22-39 (1975).
 Ohio  Revised Code, Sec.  1515.30  (Page Supplement 1983).
 Ohio Administrative Code, Chapts.  1501:15-1, 1501:15-5.
 South Dakota Codified Laws  Annotated, Sees. 38-8A-1  to 38-8A-21  (1977 and
   Supplement 1984).
 South Dakota State Conservation Commission and South Dakota  Department of
   Agriculture.  Suggested  Guidelines for Local  Erosion and  Sediment Control
   Programs (1977).
Wisconsin Statutes,  Sec.  92.11  (1981-1982).
                                  P-IV-B-10

-------
                    EMERGING  INSTITUTIONS FOR NFS CONTROL
                                     By
                               Glenn E. Stout
                           Water Resources Center
                           University of Illinois
                        2535 Hydrosystems Laboratory
                           208 North Romine Street
                             Urbana, IL   61801


     The management of our water and land resources require an
interdisciplinary approach involving the many users of the resources.  For
fifty years, agriculture has been attempting to reduce nonpoint pollution,
NFS, by various programs which had some success in the fifties, but, due to
new technologies, the grass waterways approach has disappeared.  New
cultivation techniques are helping to reduce the NPS, but the users of the
land are often unappreciative of the impact of the loss of soil and its
consequences throughout the watershed and the subsequent deposition in lakes
and deltas.  We need a certain amount of erosion in order to maintain the
coastal estuaries and delta regions.  But, an excess amount is unnecessary for
the natural ecosystem.  The city dweller and the rural land manager has not
been aware of the consequences of NPS as it affects the rest of the land
resources, lakers and the ecosystem.  Soil erosion fills in drainage ditches
and causes lakes to lose their storage capacity and subsequent transformation
into a wetland.

     In the late 70s, residents of Mattoon approached the Water Resources
Center for assistance in restoring their man-made lake which was rapidly
filling with sedimentation.  It was rather obvious that the problem was on the
watershed and that there was a need for the lake owners to talk and
communicate with the farmers in the area.  This was a very difficult task
since it involved a municipality, residents around the lake as well as the
users of the land on the watershed.  In fact, the city officials were
reluctant to take any action because of the possible incresed taxes that might
be invoked upon their city.  Finally, after numerous meetings and publicity in
the area, the region developed a concensus that there was a need to work
together.

     The city of Mattoon, who owned the lake, changed its attitude and began
to support erosion control program on the Lake Paradise watershed.  Farmers in
the area received free water from the city, but, at the same time, were not
aware that their soil was creating a water quality and supply problem for
future users.  Everyone was finally awakened to the issue and now there is a
general concensus in the area that erosion is a problem for everyone.  In
order to maintain a high-quality water supply, the problem has to be resolved
at the source.

     The Water Resources Center has held three meetings on lake and watershed
management (1, 2).  Last fall, 200 people gathered in Springfield for a
meeting which involved the Association of Soil and Water Conservation
Districts, the EPA, owners and managers of lakes and scientists.  As a result
                                  P-IV-C-l

-------
 of  this meeting  and other efforts in the state,  there  have been  several new
 cooperative  programs developed involving the  integration of lake and watershed
 management.

      Several years  ago,  David Daily, a student of  soils at the Southern
 Illinois  University at Edwardsville, undertook a personal project to determine
 the soil  losses  on  the watershed of 4,000 acres  for Holiday Lake, a nearby
 private lake,  which was  experiencing a heavy  siltation (3).  He went to the
 management of  the lake and pointed out to them that the source of the problem
 was the watershed.   The  owners were receptive to his concerns, since they
 would have to  spend $45,000 per year for dredging.  Mr. Daily then explained
 the potential  solutions  to the landlords on the  watershed.  ASCS was
 approached for matching  funds and the local Madison County Soil  and Water
 Conservation District agreed to administer a  sediment  reduction  plan
 regardless of  the ASCS backing.   As a result,  the  Holiday Lake Homeowners
 Association  are  providing funds to combat the erosion  on critical portions of
 the watershed.   This effort was initiated outside  of government  and is an
 excellent example of how people can work together  in forming a local
 institution  to solve local  problems.

      In October  at  the North American Lake Management  Association meeting in
 Geneva, several  papers are being presented which will  describe additional
 recent examples  of  cooperative local  efforts  involving the Association of Soil
 and Water Conservation Districts,  the Illinois EPA and the private sector.

      The  big issue  in solving local  problems  is  the compatibility of the
 individuals--everyone has to  be unselfish and interested in the welfare'of the
 entire community.   Although there will  be personal biases, these must be
 minimized in order  to create an atmosphere whereby an  institutional
 arrangement  is established to solve local  problems.  Universities and state
 agencies  can provide technical  assistance, but they should not be involved in
 the legislation  and demands of extravegant plans to resolve these problems.
 Many of the local erosion NFS programs  could  easily be solved by small units
 of  government  or individuals  working  together to protect the environment for
 the future of  mankind.
                                 REFERENCES
Water Resources Center.  (1980).  Proceedings of a conference on restoring
man-made lakes in Illinois - February 19, 1980.

Water Resources Center.  (1981).  Proceedings of a round table on reclaiming
and managing lakes inIllinois.

Water Resources Center.  (1984).  Proceedings of the Illinois Conference on
Lake and Watershed Management - November 8-9, 19B4.
                                 P-IV-C-2

-------
Technical Reports

-------
        APPLICATIONS OF GEOGRAPHICAL INFORMATION  SYSTEM  AND  HYDROLOGIC
              MODELING TO AN AGRICULTURAL WATERSHED IN ILLINOIS

                           Ming T.  Lee,  Hydrologist
                                     and
                     Rodolfo Camacho, Research Assistant
                         Illinois State Water Survey
                           Champaign, Illinois, USA


                                   ABSTRACT

The state-of-the-art appraisals of  water quality  impacts  induced  by  non-point
sources are difficult due to many factors.   Lack  of  existing research data
base, the diffusive characteristics of non-point  source  pollutants,  and the
technical difficulties relate the instream water  quality  to  the topographic
features, soil characteristics, land use, land cover,  soil moisture  condi-
tions, and the management practices in the watersheds.   In order  to  overcome
these difficulties, the comprehensive monitoring  and evaluation program was
established in Highland Silver Lake Watershed.

The basic techniques may be divided into three categories:   1)  field monitor-
ing, 2) geographical information system, and 3) watershed modeling.  These
three components are mutually dependent. The field  monitoring  is described
in an accompanying paper.  The geographical  information  system  consists of
two basic types of data:   cartographic coordinates which identify the points,
lines, and polygons and attributes  data which describe the characteristics of
these features.  The computer oriented GIS  performs  manipulation  and analysis
which include map overlay,  buffer generation, dissolve boundaries (regroup
and reclassify), tabular analysis and network analysis.   The watershed model-
ing utilizes the data from field and GIS to  evaluate the effects  of  watershed
management practices.  The MULTSED  and AGNPS models  were used.  Results of
the modeling effort are discussed.


Keywords:  Nonpoint source pollution, geographical information  system,
           watershed modeling, Best Management Practices, agricultural
           watershed, sediment load, runoff, water quality.
                                  T-I-A-l

-------
 Introduction

 Watershed modeling  is a tool for hydrologic system synthesis, prediction, and
 optimization.  Hydrologic modeling requires simplification or abstraction.
 There are two ways  to pursue this abstraction:  lumped or distributed systems
 (Chow,  1964; Haan et al., 1982).  In lumped systems, non-uniform parameters
 such as rainfall, soils, land use, and topographic characteristics are aver-
 aged to obtain representative values for the entire drainage basin.  On the
 other hand, a distributed system has the built-in capability of assembling
 the non-uniform parameters and simulating the interaction of these parame-
 ters.   Thus, a distributed system should be more flexible for describing
 spatial-temporal dynamics of overland and channel flows.  However, this type
 of model requires more detailed input data.

 One way to overcome these difficulties is to use a geographical information
 system  (GIS).  A brief description of 6IS is as follows.

 Geographical Information System

 In general, a geographical information system consists of two basic types of
 data:   cartographic coordinates which identify the points, lines, and poly-
 gons; and attribute data, which describe the characteristics of these fea-
 tures.   The computer-oriented GIS such as ARC/INFO (ESRI, 1984) can perform
 sophisticated manipulation and analyses which include map overlay, buffer
 generation, dissolve (regroup and reclassify), tabular analysis, and network
 analysis.  The system contains a series of computer mapping and display
 capabilities for generating high quality cartographic displays.  The user can
 specify size and scale to produce the desired map outputs.

 Highland SiTver Lake Project:  An Example

 The Highland Silver Lake Project is one of thirteen experimental projects for
 the National Rural  Clean Water Program in the United States. The project ha^
 been in  operation since September 1980.  The project consists of field data
 collection and evaluation of different land management practices applied in
 the watershed.  Since land management practices consist of different tillage
 practices and different crop sequences on different soil  and topographical
 conditions, the collected field data alone will not be enough to determine
 the effects of management practices in the entire project area.  Therefore,
 utilization of a watershed model  is needed to evaluate the impacts of land
 management practices on water quality and on stream and reservoir sedimenta-
 tion, as well as the costs and benefits of the program.

 Data Base of Highland Silver Lake Watershed

 In the  Highland Silver Lake Project, the watershed data base consists of a
 hydrographic map (stream networks and water bodies), soil  maps, land use
 data, and slope map.  The hydrographic map was digitized from a standard U.S,
 Geological  Survey 7-1/2 minute topographic map with the recent aerial  photor
 graphs as supplemental  data sources.  The soil  map was digitized from pub-
 lished county soil  survey maps.   The land use data were obtained from recent
aerial photographs.   The slope map was developed on the basis of topographic
maps published by the U.S.  Geological  Survey.   Watershed and sub-basin boun-
daries and locations of monitoring stations were also coded in the GIS.

                                T-I-A-2

-------
Agricultural Non-Point Source Pollution Model  (AGNPS)

This model was developed by the North Central  Soil  Conservation  Research
Laboratory, Agricultural Research Service,  USDA (Onstad  and  Young,  1983).
AGNPS is a single event, deterministic, and distributed  model.   It  configures
the watershed into small square subareas called "cells."  It is  at  this level
that all characteristics are established and calculations  made.

Figure 1 shows a subwatershed of the Highland Silver Lake  watershed. The
model is composed of the following parts:  (1)  runoff component, (2) channel
flow, (3) erosion, (4) sediment transport,  (5)  nutrient  transport,  and
(6) feedlot.

The basic output from AGNPS includes a hydrology component which estimates
both volume and peak runoff; a sediment component which  estimates upland
erosion, gully erosion, and sediment yield; and a nutrients  component which
estimates nitrogen and phosphorus yields in terms of concentration  and load.
The output can be examined on a cell basis  or for an entire  watershed. Sedi-
ment analyses are broken down into particle-size classes.

The second model is the Multiple Watershed  Storm Water and Sediment Runoff
Simulation Model (MULTSED) (Simons et al.,  1978 and 1981).

MULTSED Model
This model is also a single event, distributed,  deterministic model.  The
model contains two basic components:   a hydrologic and hydraulic  routing
component and a sediment component.  The calculations  are  conducted in three
basic units:  (1) two-plane and single channel,  so called  Wooding-plane sub-
watershed as shown in figure 2 (Wooding, 1965);  (2) channel  unit;  and (3)  a
single-plane unit. The number and size of these  units  for  a  specific  water-
shed can be selected by the users.

Use of GIS for Input Data Preparation

The two selected models that have been described represent two  types  of model
configurations. AGNPS represents the  grid-cell system.  MULTSED represents
the irregular configuration which reflects the hydrologic  units such  as
Wooding-plane, single overland flow,  and single  channel.

The input data for the AGNPS model consist of:   (1) cell identification
number; (2) receiving cell identification number;  (3)  SCS  curve number;
(4) land slope in percent; (5) land slope shape  factor;  (6)  field slope
length in feet; (7) channel slope in  percent;  (8)  channel  side  slope  in
percent; (9) Manning's roughness coefficient for the channel;  (10) soil
credibility factor; (11) cropping factor; (12) practice factor; (13)  surface
condition constant; (14) aspect; (15) soil texture number; (16) fertilization
level; (17) availability factor; (18) point source level;  (19)  gully  source
level; (20) chemical oxygen demand factor; and  (21) impoundment factor.

In order to compile these input data  from the GIS data base, the  first step
is to construct a "template" which delineates the watershed  into  sub-units.
In the case of the AGNPS model, the grid-cell system is the  template. The

                                  T-I-A-3

-------
Figure 1.  An example of grid-cell system of AGNPS model
                       T-I-A-4

-------
                                   Line E
(a) Original Subwatershed Topographic Map
     (b) Wooding Plane Representation
Figure  2.   Watershed configuration of  MULTSED model




                       T-I-A-5

-------
 computer  software  is able to retrieve the attribute data and the sub-units
 from the  data base.  However, few items in the input data require field
 inspection to acquire the necessary data.  Similarly, the data can be re-
 trieved for the MULTSED model.

 Results

 AGNS was  run for field site 5 of the Highland Silver Lake watershed using a
 10-acre grid system as shown in figure 1.  A 4.4-in. rainfall storm was
 considered which represents an extreme storm for this watershed in the last
 three years.  Table 1 shows the results given by the AGNPS and MULTSED model
 for  the same event.

     Table 1.  AGNPS and MULTSED Model Results for a 4.4-in. Rainfall Event
             in Field  Site 5 of  the  Highland  Silver  Lake Watershed

                                           AGNPS          MULTSED

     Runoff volume (in.)                    2.8             3.6
     Runoff ratio  (%)                      64.0            82.0
     Peak runoff rate (cfs)               382.0           283.0
     Sediment yield (tons/acre)             0.327           0.089

 The  primary results indicated that runoff volumes and runoff ratio show
 higher values by using MULTSED model, but peak runoff rate and sediment yield
 show higher values by using AGNPS model.  Further verification with field
 observation is being conducted.

 Three events during 1982 and 1983 were selected for the calibration ofthe
 MULTSED model.  The events were selected so that they fairly represent the
 average characteristics of the events for these years.  The events considered
 were:  1) May 31, 1982, with a rainfall of 0.88 in.  and a duration of 4
 hours; 2) June 4, 1982, with a rainfall of 0.4 in. and a duration of 11
 hours; and 3) March 18, 1983, with a rainfall of 1.34 in.  and a duration of
 16.5 hours.

 The  parameters used in the calibration were:   1) the hydraulic conductivity,
 2) initial soil moisture, 3) rainfall splash detachment coefficient, and 4)
 roughness coefficients.  These parameters were adjusted by doing sensitivity
 analysis  on their value ranges until the simulated and measured runoff and
 sediment  concentration hydrographs were matched.  Other parameters such as
 ground and canopy cover, porosity, average suction head, final soil  moisture,
 interception by ground cover or canopy, and soil  temperature were kept con-
 stant.  Table 2 shows  the results of the three selected events.  A selected
 runoff and sediment hydrographs are shown in figure  4.

 Conclusions

 1.  Application of the MULTSED model for a portion of the  Highland Silver
 Lake watershed has been performed.  Satisfactory simulation of runoff and
 sediment hydrographs was obtained.  Very little is known on the relationships
 between the detachment coefficients and the land use and watershed management
 practices in the MULTSED model.   An attempt to find  these  relations  will  be
done when the entire basin is studied.

                                  T-I-A-6

-------
              Table 2.   Runoff  and  Sediment  Yield  for  Field  Site 5 and Gaging
               Station 3 of Highland Silver Lake Predicted by MULTSED Model
  Event

Subarea A
May 31, 1982
June 4, 1982
March 18, 1983
          Measured  Computed         Runoff  Measured  Computed
Rainfall   runoff    runoff   Error  ratio   sediment  sediment  Error
 (1n.)     (in.)     (in.)     (%)    (%)      (tons)     (tons)     (%}
 0.88
 0.40
 1.34
Subarea B
May 31, 1982     0.88
June 4, 1982     0.40
March 18, 1983   1.34
0.198
0.093
0.352
           0.213
           0.053
           0.550
0.204
0.097
0.355
          0.222
          0.053
          0.521
3.04   22.5
3.85   23.3
0.64   26.3
          3.99
          0.06
         -5.2
       24.2
       13.3
       41.0
 N.A.
 0.55
 2.44
38.32
 4.60
80.37
 1.15
 0.48
 2.23
27.10
 4.61
80.58
-13.5
 -8.7
-27.0
  2.1
  0.25
N.A. = not available
                                   T-I-A-7

-------
            LEGEND
    SW  = Subwatershed Unit
    PL  = Plane Unit
    CH  = Channel Unit
     A  = Drainage Area of FS5 (SW1)
     B  = Drainage Area of SyV2, PL1, PL2
GS3
Figure  3.   Gaging  Station 3  Sub-watershed  of Highland  Silver Lake,
            Highland,  Illinois
                              T-I-A-8

-------
     024
<£ °
|i
<  0.5
oc
   150
". 125
$ 100
K
<

I  75
   50
   25
                     Time, hours
                                         2000
                                       a

                                       ^ 1500
                                       o

                                       z
                                       o

                                       <

                                       £ 1000
                                       LU
                                       o

                                       O
                                       o
                                      o
                                      LU
                                      I/)
5         10
 TIME, hours
15
                                          500
                                                                             GS3
                                                       5          10
                                                        TIME, hours
                                                                           15
                                                    •    Measured Concentration

                                                   	Measured

                                                   	 Predicted
                                                   FSB   Field Site No. 5

                                                   GS3   Gaging Station No. 3
               5         10

                TIME, hours
  Figure 4.   Runoff and Sediment  Hydrographs of March  18,  1983 of
               Highland  Silver Lake  Watershed, Highland, Illinois
                                 T-I-A-9

-------
2.  A methodology illustrates that a geographical information system can
provide an efficient way to compile necessary input data for complex dis-
tributed hydro!ogic models.

3.  The development of the data base in a geographical  information system  is
independent of a specific model.  The data can be retrieved from GIS to meet
the requirement of different geometric configurations of watershed models.

4.  The sensitivities of various geometric delineations and hydrologic proc-
esses in a watershed can be evaluated by utilizing a geographic information
system.
References

Chow, V. T. (ed.)  Handbook of Applied Hydrology. McGraw-Hill  Book  Company*
     1964.

Environmental System Research Institute, ARC/INFO User Manual, Version 2.3,
     Redlands, California, 1984.

Haan, C. T., H. P. Johnson, and D. L. Brakensiek, Hydrologic Modeling of
     Small Watersheds, American Society of Agricultural  Engineers,  Monograph
     Series, 1982.

Onstad, C. A., and R. A. Young, A Procedure for Prioritizing Water  Quality
     Problem Areas, Paper No. 83-2156, presented at the 1983 Summer Meeting,
     American Society of Agricultural Engineers, held at Montana State Uni-
     versity, Bozeman, Montana, June 26-29, 1983.

Simons, D. B., R. M. Li, and B. E. Spronk, Storm Water and Sediment Runoff
     Simulation for a System of Multiple Watersheds - Vol. I.  Water Routing
     and Yields, CER77-78DBS-RML-BES47, Colorado State University,  April
     1978.

Simons, D. B., R. M. Li, W. T. Fullerton, and  T. R. Grindeland,  Storm Water
     and Sediment Runoff Simulation for a System of Multiple Watersheds -
     Vol. II. Sediment Routing and Yield, Colorado State University,  October
     1981.

Wooding, R. A., A Hydraulic Model  for the Catchment-Stream Problems^
     I. Kinematic Wave Theory, Journal of Hydrology, Vol.  3, No.  3/4,
     pp. 254-267, 1965.
                                T-I-A-10

-------
        Delineating Sources of Agricultural Nonpoint Source Pollution
                W.T. Dickinson*,  R.P.  Rudra*,  and G.J.  Wall»»
*  Professor and Assistant Professor, respectively,  School of Engineering,




   University of Guelph, Guelph,  Ontario,  Canada,  N1G 2W1




** Regional Director, Environment Canada,  55 St.  Clair Avenue East,  Toronto,




   Ontario, Canada, M4T 1M2
                                T-I-B-1

-------
 INTRODUCTION




      Fluvial sediment  and  associated nutrients and chemicals are now



 acknowledged to  be  a major cause of contamination in river and laKe systems



 throughout North America  (7,  8).  A prime source of these contaminants is




 sheet and rill erosion from agricultural cropland (19).  Such soil erosion is



 widely distributed  over the landscape, but the rates at which soil erodes and




 yields sediment  to  streams can be highly variable from field to field and from




 watershed to watershed (15).  Therefore, although soil erosion has been termed




 a nonpoint source of pollution, most basins have more or less specific




 locations which  yield  much of the sediment and associated contaminants.  The




 design and selection of cost-effective non-point source pollution control




 strategies requires delineation of key source areas.



      Recent  erosion/sedimentation research studies at Guelph have led to the



 development  of a method for estimating sources and rates of sediment yield in



 small agricultural  watersheds (2, 6).  Examples of the application of the




 method to two agricultural watersheds in Southern Ontario are presented in




 this  paper,  to reveal  the  utility of watershed modelling for delineating soil



 erosion and  sediment yield source areas.



 LITERATURE REVIEW



      The notion  of  sediment contributing areas, i.e. that prime sources of




 sediment in agricultural regions constitute an area less than and in some



 cases much less  than the total watershed area, has received considerable



attention in recent years  (3»  13» 1b»  17).   Much in the erosion/sedimentation



 literature, although not specifically focussed on the  definition of sediment



 sources, provides useful background for the location of significant sediment



 sources.  Some field-scale models provide estimates of annual erosion (22),



and others predict annual sediment loads directly (9,  13»  22).    Another



 category of models predicts annual amounts of erosion  and sediment on the




                                 T-I-B-2

-------
 basis of the Universal Soil Loss Equation (23) and a functional relationship



 for watershed delivery ratio or field transport factors (10, 14, 16, 17, 1$,



 21).




     Single storm event and continuous erosion/sediment/chemical models have




 also been developed for both the field scale (11) and the watershed scale (1).




 This group is more descriptive of the physical and chemical processes involved




 and can be used to delineate key sources.  However, the large volume of input




 data required for these models, some of which is not always readily available,



 is often a constraint on their use.




     Therefore, although there is now an extensive literature on soil erosion



 and fluvial sedimentation which provides useful background and detailed




 distributed models, there has not been a method developed with the prime




 purpose of delineating key source areas.  Further, although there has been



 considerable evidence that stream sediment loads are usually highly variable



 throughout the year (2, 5, 12, 19, 20), the variation of seasonal sediment



 sources has not been addressed.  GAMES - the Guelph model for evaluating the



 impact of Agricultural Management systems on Erosion and Sedimentation - has



 been developed for the purpose of predicting on a seasonal basis the magnitude




 and location of soil loss by water erosion and subsequent sediment yield



METHODOLOGY



     Several criteria have been kept in mind during the development of GAMES.




The model was to take account of seasonal variations in soil loss and sediment



 transport mechanisms, but was not to become encumbered with the data



requirements of an event on continuous model.  The basic computational land



 units were to be field size, with the intent of applying the model to small



agricultural watersheds.  Although a predictive capacity was deemed to be



desirable, the prime purpose of the model was to distribute sediment yields






                                 T-I-B-3

-------
 throughout a watershed on a quantitative and objective  basis.

      The resulting model, GAMES,  has been suitably constructed  for  the

 computation of seasonal potential soil loss  (the soil loss  component) and

 seasonal sediment yield (the delivery component) for field-size cells in small

 agricultural watersheds.   The soil loss component is based  on the Universal
                                                                    *
 Soil  Loss Equation (23) has been  modified for seasonal  application.  That is,

           ^3 - RsCsKaLS.P3

 where AS = seasonal soil  loss per unit area,  Rs  = seasonal  rainfall factor,

 Cs  s  seasonal land use or cropping factor, Ks =  seasonal soil credibility

 factor,  LS = slope gradient and slope  length  factor, and P  = seasonal

 erosion  control practice  factor.   The  sediment yielded  by each  field unit to

 downstream field and  stream units is determined  from the product of the

 seasonal potential soil loss and  the seasonal delivery  ratio for the unit.

 The seasonal delivery ratio,  or percentage of  the potential soil loss which

 can be expected to be delivered downstream in a  given season, is determined

 from  the  expression,
          DR3 =a
 1
 -   .  S .  Hc<
L.ns
where 0 £ DRS <. 1J n3 = seasonal surface roughness coefficient, as indexed

by Manning's n; S = cell slopej Hcs = seasonal hydrologic coefficient, an

index regarding the probability and depth of overland flow in a season; Lg s

seasonal length of overland flow path; and a, 3 = calibrated parameters.

     Application of the soil loss and delivery components initially involves

the development of a composite overlay of land use, soils, and land slope to

divide the watershed area into land units each of which is characterized by a

single land use, a single soil type, and a single slope class.  Then values of

the various variables in the USLE (23) and DR (2) expressions are determined

                                T-I-B-4

-------
and assigned to each unit.  Determination of the a and 3 parameters for a



watershed involves an optimization routine to compare the accumulated water-



shed sediment load calculated by means of GAMES for assigned values of and a




and 3 with a measured sediment load or one estimated by another means.  Output



for a calibrated application is generated on a unit by unit basis for each




season considered, and includes potential soil loss, potential soil 3,033 per




hectare, delivery ratio to the adjacent downstream unit, delivery ratio to the



nearest stream, sediment yielded to the stream, and sediment yielded per hectare.



     With the selection of soil loss and sediment yield tolerance levelsf J.t




is possible to extend the application of GAMES into the area of problem



identification (4).  The framework shown in Table 1 has been suggested on the



basis of distinct definitions for erosion and sediment problems.  An erosion




problem is considered to be fundamentally a source problem, the loss of soil




creating hardships such as decreases in agricultural productivity, loss of



tillable land, damage to buildings, roads, equipment etc.  A sediment problem




involves downstream situations e.g. siltation of streams and reservoirs,



pollution of fish spawning areas, increased water treatment costs.  Areas of a



watershed which fall into problem category I may be expected to exhibit both




high soil erosion rates and high sediment yield rates.  Areas in category II



have high erosion rates, but contribute relatively little soil to downstream



locations.  Category III includes those areas which have relatively low soil




loss rates, but from which most of the eroded soil moves into the stream



channels contributing significantly to sediment problems.  Category IV is the



"no problem" category.



     For the sake of illustrating the utility of a model for delineating soil



loss and sediment yield source areas, GAMES has been applied to two small



agricultural watersheds located in the Thames River Basin of Southern Ontario.





                                 T-I-B-5

-------
Table 1:  Framework for soil erosion and sediment
          problem identification
Problem
Category
I
II
III
IV
Erosion*
Problem
X
X
.
"
Sediment**
Problem
X
-
X
"
*  Agricultural productivity is adversely affected

** Sediment yield and associated nutrient contribu-
   tion to the stream is excessive.
                 T-I-B-6

-------
The Stratford/Avon Watershed is comprised of 434 hectares of rolling upland




soils, cropped predominantly in continuous corn with fall ploughing and few



if any conservation practices.  The Big Creek Watershed is 2141 hectares in



extent on quite flat lowland soils.  It is almost totally in row crops (e.g.



continuous corn, beans, vegetables) which are fall ploughed.



SAMPLE RESULTS




     Sample output maps from the GAMES application are presented in Figures 1




through 6.  Figures 1 and 2 reveal the spatial pattern of spring potential




soil loss quantities for the upland and lowland watersheds, respectively.  The




estimated average spring gross erosion per unit of watershed area is not large




for either basin, being 2.5 tonnes per hectare for the lowland basin and 3.5



tonnes per hectare for the upland basin.  However, the spring erosion



estimates are quite variable in the upland Stratford/Avon Watershed, ranging




from essentially zero to 25 tonnes per hectare.  The range is much smaller




across the lowland Big Creek Watershed, varying only from 0 to 3 tonnes per



hectare.  It is also evident from Figure 1 that a major portion of the eroded




soil in the upland watershed moves within a small percentage of the basin.



For example, 92 percent of the spring sheet and rill erosion volume is



estimated to occur in 4? percent of the watershed area.  The areal



distribution of soil loss is more uniform in the lowland basin (Figure 2).



About 56 percent of the landscape contributes 64 percent of the sheet and rill



erosion.



     Sample spatial distributions of sediment yield from the upland and




lowland agricultural basins are presented in Figures 3 and 4.  Similar to the



erosion picture presented above, but more pronounced, the great majority of



the spring sediment load leaving the upland watershed is estimated to emanate



from a small percentage of the watershed area.  For example, 83 percent of the



load is generated in 14 percent of the basin.  Indeed, such a watershed is




                                T-I-B-7

-------
  Tonnes Per Hectare
Figure 1.   Spatial variation of  soil erosion on upland watershed for
          spring conditions,

                          T-I-B-8

-------
                            Tonnes Per Hectare
                                           Q<2.5
Figure 2.
                       so11 eroslon
                      T-I-B-9

-------
Tonnes Per Hectare
  Figure 3.  Spatial variation of sediment yield on upland watershed for
           spring conditions.
                           T-I-B-10

-------
                                  Tonnes Per Hectare
Figure 4.   Spatial variation of sediment yield on lowland watershed for
          spring conditions.
                            T-I-B-11

-------
    Problem Type
Figure 5.  Spatial distribution of soil erosion and/or fluvial sediment
          areas on upland watershed for spring conditions.
                               T-I-B-12

-------
                                       Problem Type
Figure 6.  Spatial distribution of soil  erosion and/or fluvial  sediment
          areas  on lowland watershed for spring conditions.

                               T-I-B-13

-------
 characterized  by rather distinct  sediment  sources.  On the other hand, the




 lowland watershed is seen to generate  its  sediment load rather uniformly



 across  the  basin,  and sediment  yield hotspots are not evident (Figure 4).




      Figures 5 and 6 present maps of the four soil erosion and sediment




 problem categories on the basis of a spring erosion tolerance level of 2.5




 tonnes  per  hectare and a spring sediment yield tolerance level of 1.0 tonnes




 per hectare.   A  quantitative summary relating to these figures is given in



 Table 2.



      The sample  results regarding erosion  and sedimentation problem




 identification reinforce the results discussed previously, but also reveal



 additionalsignificant points.  It is  clear from Table 2 that 4? percent of




 the upland  watershed and 75  percent of the lowland watershed contribute




 significantly  to erosion and/or sediment problems.  In and upland



 Stratford/Avon watershed 14  percent of the basin exhibits almost half of the



 soil  loss and  is the  source  of  more than three quarters of the sediment load.




 Only  4? percent  of the  basin is estimated  to have spring erosion rates greater



 than  the defined tolerance level, accounting for 92 percent of the total basin




 soil  loss; while only  14 percent  of the basin is estimated to serve as a prime




 sediment source, contributing 63  percent of the downstream loads.  More than



 half  of the landscape  of this basin exhibits potential spring soil loss and



 sediment yield less  than the  defined erosion and sediment yield tolerance



rates.



      The lowland Big Creek watershed, typical of many level lowland watersheds



in Southern Ontario, has more than half of its area in erosion problem



categories.  Fifty-six percent of the basin has spring erosion rates greater



than  the defined tolerance level.   About one third of this basin is estimated
                                 T-I-B-14

-------
Table 2.  Quantitative identification of soil erosion and sediment
          problems in selected agricultural watersheds
Problem
Category
I


II


III


IV


Comparative i
Percentages
% of Area
% of Soil Loss
51 of Sediment Load
% of Area
% of Soil Loss
% of Sediment Load
% of Area
% of Soil Loss
% of Sediment Load
% of Area
% of Soil Loss
% of Sediment Load
Stratford /Avon
Watershed
14
45
83
33
47
15
0
0
0
53
8
2
Big CreeK
Watershed
12
13
18
44
51
34
19
17
30
25
19
18
                          T-I-B-15

-------
 to  provide  a prime  source of  sediments  contributing to half of the downstream




 loads.   It  should be noted  that  19  percent of the basin exhibits potential



 soil  loss less  than the  defined  erosion rate tolerance.  Nonetheless, the same



 area  contributes about one  third of the downstream sediment load at sediment




 yield rates above the defined sediment  yield rate limit.



 CONCLUSIONS




      A watershed soil erosion and fluvial sedimentation model, GAMES, has been




 shown to provide a  very  useful tool not only for evaluating the seasonal and




 spatial variability of field  potential  soil loss and sediment yield in an



 agricultural watershed but  also  for delineating spatially the distribution of




 erosion and sediment problems.   For selected soil loss and sediment yield



 tolerances, the approach affords in map and summary statistics form



 information regarding the location and extent of areas contributing to (i)



 both  erosion and stream  sediment problems, (ii) only erosion problems,' (iii)



 only  sediment problems,  and (iv) no problems.  Such description and




 delineation of agricultural nonpoint sources is proving to be of great




 assistance  in the planning and targetting of nonpoint pollution control



 programs in rural areas.



 ACKNOWLEDGEMENTS



      Sincere appreciation is  expressed to all members of the very large team




 of researchers associated with the various phases of this project.   And in



addition to the funds provided by Environment Canada through the Department of




Supply and Services, the financial support of the Natural Sciences  and



Engineering Research Council of Canada,  the  Ontario  Ministry of Agriculture



and Food, the Thames River Implementation Committee,  and the Stratford/Avon



River Environment Management Program is  gratefully acknowledged.
                                T-I-B-16

-------
REFERENCES CITED




1.  Beasley, D.B., L.F. Huggins and E.J. Monke.  1980. ANSWERS: A model for



     watershed planning.  Transactions of the ASAE, 23:938-944.




2.  ClarK, D.J.   1981.  An expression for determining sediment delivery ratio.



     M.S. Thesis.  Univ. of Guelph, Guelph, Ontario, Canada.




3.  DicKinson, W.T. and G.J. Wall.  1977*  Temporal and spatial patterns in



     erosion and fluvial processes.  Proc. 5th Guelph Symp. on Geomorphology;



     "Research in Fluvial Geomorphology."  Geo Abstracts Ltd., East Anglia.



4.  Dickinson, W.T. and R. Pall.  1982.  Identification and control of soil




     erosion and fluvial sedimentation in agricultural areas of the Canadian




     Great Lakes Basin.  Final Rep. to Supply and Services, Canada.  School




     of Engr., Univ. of Guelph, Guelph, Ontario, Canada.




5.  Dickinson, W.T., A. Scott and G.J. Wall.   1975.  Fluvial sedimentation in



     Southern Ontario.  Can. J. Earth Sci., 12(11):1813-1819.



6.  Dickinson, W.T., R. Pall and G.J. Wall.  1984.   GAMES - A model for identi-




     fying sources and amount of soil erosion and fluvial sediment.  Can.



     Hydrol. Symp., Quebec City, P.Q., Canada.



7.  Duttweiler, D.W. and H.P. Nicholson.  1983-  Environmental problems and




     issues of agricultural nonpoint source pollution.  From "Agricultural



     Management and Water Quality", edited oy F.W.  Schaller and G.W. Bailey.



     Iowa State Univ. Press, Ames, Iowa, 3-16.



8.  International Joint Commission.  1980.  Annual  report of the International



     Joint Commission.   I.J.C., Windsor, Ontario.



9.  Jansen, J.M.L. and R.B. Painter.  1974.  Predicting sediment yield from



     climate and topography.  J. of Hydrol.,  21:371-386.



10. Kling, C.P. and C.W. Olson.  1974.  The sediment transport computer model.



     Cornell Agri. Mimeo 74-11, Dept. of Agro., Cornell Univ., Ithaca, N.Y.






                                T-I-B-17

-------
11. Knisel, W.D. Jr., Ed.   1981.  CREAMSJ A field scale model for chemical,




     runoff and erosion from agricultural management systems.  U.S. Dept. of



     Agri., Cons. Res. Rep. No.26.



12. Mildner, W.F. and R.C.  Boyce.  1979.  Monthly variation in soil loss and



     sediment yield.  ASAE  Paper No.79-2528.  Amer. Soc. Agri. Engr.,




     St. Joseph, Ml.



13« McPherson, H.J.  1975.  Sediment yield from intermediate-sized stream




     basins in Southern Alberta.  J. of Hydrol., 25:243-257.




14. Onstad, C.A. and C.R. Foster.  1975.  Erosion modelling on a watershed.




     Transactions of the ASAE 20:89-95.



15. Pall, R., W.T. Dickinson and R. McGirr.  1982.  Climatic and soil



     conditions significant to soil erosion and sedimentation in Southern



     Ontario.  NAR-ASAE Paper No. 82-201.   Amer. Soc. of Agri. Engrs., St.



     Joseph, MI.



16. Piest, R.F.  1965.  The role of the large storm as a sediment contribution.




     Proc. 1963 Federal Inter-Agency Conference on Sedimentation.  U.S.Dept.




     of Agri., Misc. Publ.  970:97-108.




17. Renfro, G.W.  1972.  Use of erosion equations and sediment delivery ratios



     for predicting sediment yield.   Sediment-Yield Workshop.  U.S.  Dept. of



     Agri. Sedimentation Lab.,  Oxford,  Miss. 33-45.




18. Robinson,  A. R.   1977.   Relationship between soil erosion and sediment



     delivery.  IAHS Symp. on Erosion and  Solid Matter Transport in Inland



     Waters.   IAHS Publ.  No.122.




19. van Vliet, L.J.P.,  G.J.  Wall and W.T.  Dickinson.  1978.   Erosion losses



     and sediment delivery ratios for agricultural watersheds.   I.J.C.  Tech.



     Rep. re:  Great Lakes  Pollution from Land Use Activities, Windsor,



     Ontario,  Canada.








                                T-I-B-18

-------
20. Wall, G.J., W.T. Dickinson and L.J.P. van Vliet.  1979.  Agricultural



     sources of fluvial suspended sediment.  Prog. Wat. Tech., 11(6):U8l-M99.



21. Williams, J.R.  1972.  Sediment yield production with Universal Soil Loss




     Equation using runoff energy factor.  Sediment-Yield Workshop, U.S.



     Dept. of Agri. Sedimentation Lab., Oxford, Miss., 244-252.




22. Williams, J.R. and H.D. Berndt.  1976.  Sediment yield prediction based on




     watershed hydrology.  ASAE Paper No. 76-2535.  Amer. Soc. Agrl. Engr.,



     St. Joseph, Ml.



23. Wischmeier, W.H. and D.D. Smith.  1965.  Predicting rainfall-erosion




     losses from cropland east of the Rockey Mountains.  Agri. Res. Ser., U.S.



     Dept. of Agri. Handbook No. 262.
                                 T-I-B-19

-------
            COUPLING NOMPQINT POLLUTION ME. WATER QUALITY MODELS:
              AN EXAMPLE FOR THE GJJEEJJ. BAY - FjQX. RIVER WATERSHED
    Thoaas M.  Heldtke, Associate Professor,  Department  of Civil
    Engineering, Wayne State University,  Detroit, MI 48202,  Martin T.
    Auer, Associate Professor, Department of Civil  Engineering, Michigan
    Technological University, Houghton,  MI 49931,  Raymond P.  Canale,
    Professor, Department of Civil  Engineering, University of Michigan,
    Ann Arbor, MI 48109, and Theodore A.D.  Slawecki,  Computer  Engineer,
    Limno-Tech, Inc.,  Ann  Arbor, MI   48103
                                 ABSTRACT

Green Bay is a major gulf located in the northwest corner of Lake Michigan
which has been cited as one of the major water quality problem areas in the
Great Lakes.  Historically, interest in water quality degradation has been
limited  to the lower Pox River  and extreme southern Green Bay, the site of
municipal  and industrial pollutant discharges.  Recent surveys indicate
dissolved oxygen  depletion far out into  the  bay,  beyond  the immediate
Influence of the Fox River's  organic  loading.  Attention is now focused uppn
nonpoint  sources  of phosphorus as  stimulants for primary  production  and the
ultimate  cause of oxygen depletion problems.  The Fox  River contributes 78$ of
the annual tributary load  of  total phosphorus to Green Bay.  Of this amount,
only 3% is attributable to the Green Bay Metropolitan  Sewerage Authority, the
region's  largest  point  source discharger of  phosphorus.

Nonpoint  source loads of total phosphorus are estimated using a microcomputer
model which  considers  land use, soil  texture,  and  the phosphorus loading
associated  with each location in the land use/soil texture matrix (UALs,  unit
area loads,  kgP/ha'year).  Land use classifications and soil  textures are
developed specifically  for the  Fox River watershed and UALs are derived from
recent studies  of nonpoint source loadings (PLUARG,  208 Areawide Water Quality
Management Planning Program).  Annual  total  phosphorus loads to Green Bay from
the Fox River are calculated for  baseline  (= existing) conditions  and six
additional  land use scenarios as the sum of  the products of the land use/soil
texture and unit area load matrices for each component sub-basin in the Fox
River watershed.

Load estimates  are input to a steady-state mass balance phosphorus model which
calculates the phosphorus  concentration along the major  longitudinal  axis of
Green Bay.   Trophic levels  corresponding  to points along the phosphorus
gradient  are  established using a trophic index.   A  marked gradient in trophic
conditions  exists along the bay which is sensitive to variations in land use
practices impacting nonpoint  phosphorus loads.   The hypothetical scenarios
help demonstrate the utility and of  interactive,  coupled  nonpoint  pollution
and water quality microcomputer  models.

Keywords: Nonpoint source pollution model, water quality model,  abatement of
         pollution from  agriculture,  phosphorus,  eutrophication,  Green Bay,
         Lake  Michigan, Great Lakes

                                 T-I-C-1

-------
INTRODUCTION

Modeling approach

     Microcomputer models for the estimating of nonpoint source pollutant
loadings and receiving water quality response  may be  coupled and used to
examine the impact of land use activities on  water quality.  This manuscript
focuses on nonpoint source loads of phosphorus  to  Green Bay, Lake Michigan and
their  relation to  the trophic  state gradient  within  that system.   Seven
hypothetical land use managment scenarios are described to illustrate the
application of interactive planning and management tool.   The methodology is
versatile and may be modified to address other substances and their impact
through  construction of pollutant-specific unit area  load tables  and
quantification of  sources and sinks for the parameter of  interest.

Green Bay and the  Fox River

     Green Bay is  a  gulf approximately 160 km  long and 22 km wide, located in
the northwest corner of Lake Michigan  (Figure 1).  The bay has a  mean depth of
15.8 m, a volume of 67 km^ and a residence time of 6 years (Mortimer,  1978).
Historically,  industrial  and  municipal  discharges of  oxygen-demanding
substances and plant nutrients have led to severe dissolved oxygen depletion
during summer in the lower Fox River  and extreme southern Green Bay. Recent
evidence of  hypolimnetic oxygen depletion  over much greater  portions of the
bay  has  resulted in  the current interest in  nonpoint  sources  of  algal
nutrients,  specifically phosphorus. The Fox River is  the single  greatest
hydrologic and pollutant  source for Green Bay,  contributing ^5%  of the annual
tributary  flow and 78$  of the annual  tributary  total phosphorus  load
(Roznowski  and Auer,  1984). Strong  longitudinal gradients  in trophic status
are set up in response to the Fox River phosphorus loading — hypereutrophic
conditions exist  in the  southernmost region near the river's  mouth,  while
oligotrophic conditions prevail in the northern reaches near the  junction with
Lake Michigan.
NONPOINT SOURCE MICROCOMPUTER MODEL

Model Structure

     The microcomputer model  used  to  assess average annual total phosphorus
loadings from nonpoint sources within a given drainage system is  based on
three critical assumptions:

     1 ]   The  total  phosphorus  load  generated by surface runoff from a given
         site may be represented by a unit area  load  (UAL) —  the mass of
         phosphorus contributed per unit area of land per unit time.

     2]   Under identical climatological conditions, the UAL for a given area
         is  principally  determined by the predominant land use  and  soil
         texture within that  area.

     3]   UALs are derived to  reflect nonpoint source loadings for an  "average"
         year of wetness  (total precipitation).
                                 T-I-C-2

-------
                                                           Lake
                                                    K     Michigan
         rLake
        f Winnebago
 I
N
I
FIGURE 1.  Location of Green Bay and major morphological features.
                         T-I-C-3

-------
     To evaluate the annual nonpoint source total phosphorus loading, the area
 under study is divided into a set of sub-basins or sub-watersheds representing
 the  major  hydrologic units within the system.   Each  sub-basin is then further
 divided  into  a set of  cells which represent the  area comprised of a specific
 land use/soil texture combination.   The area of each  cell is multiplied by its
 appropriate UAL and the products summed to arrive  at  the  total annual nonpoint
 source load  for any sub-basin.   The cumulative load for the entire drainage
 system is simply  obtained as the sum of the  contributions  for each  of the
 major sub-basins as indicated by  Equation  1 below:
      n         n
TP =  y   TP.  =  y
      L,     i    Li
     i=l      i-1
                                   m
UAL.
                                       c=l
                                                                     (1)
where      TP:  annual nonpoint  source load for the entire
                drainage system

          TPi:  annual nonpoint  source load from sub-basin  i
                (i = 1, 2,  ...,  n)

       UALj k:  annual nonpoint  source total phosphorus load
                per unit area of land use j and soil texture k
                (j = 1, 2,  ....  m; k = 1, 2, ..., 1)

       *i i k:  area °f land  use j and soil texture k within
         'J>    sub-basin i
     The microcomputer model allows  a user to quickly conduct a preliminary
macroscopic nonpoint pollution  loading analysis on any drainage area of
interest.   The accuracy or goodness of the analysis is a direct function of
the user's ability to accurately represent the land use, soil  texture and
hydrologic characteristics  of the system under study.

     The model is comprised  of two  major sub-programs:  (1) SETUP,  which
prompts the user to define the land use and soil texture  matrix for each sub-
basin in the drainage system and to provide the UAL matrix for the  pollutant
of interest, and  (2) LOAD, which integrates the data provided to SETUP with
new information on land use changes or pollutant controls.  Output  from LOAD
includes tabular and graphical displays of nonpoint source pollutant loadings
for each major sub-basin and  for any set of land use management/pollutant
control scenarios  under consideration.

Application to Fox River/Green Bay

Land Use/Soil Texture Description

     The  nonpoint  source  pollution model  is applied in  a preliminary
assessment of annual total  phosphorus loads  to  Green  Bay from the  Fox  River
drainage basin.  Information on land use and soil classifications within the
area were obtained from the  Fox Valley Water  Quality Planning  Agency (FVWQPA)
and Soil Conservation Service  offices for the five counties comprising the Fox
River watershed.  For the purposes of this demonstration, soil classifications
were grouped into three major soil texture categories ~ coarse,  medium,  and
fine;  eight land use  categories were considered in the  analysis:

                                  T-I-C-4

-------
          1]  Residential                  5] High Tillage Cropland
          2]  Commercial                   6] Low Tillage Cropland
          3]  Industrial/Transportation     7] Pasture
          4]  Institutional/Government      8] Woodland/Residual

     The Fox River watershed was divided into three sub-basins (Lower Fox
River, Lake Winnebago, and River-Lakes) as shown in Figure 2.  Acreages for
each soil association in a given sub-basin were calculated from county soil
survey maps.  Soils in the River-Lakes (97$ fine) and Lake Winnebago (99$
fine) sub-basins are assumed to be comprised of 100$  fine textured soils.  The
Lower Fox River sub-basin contains 35$ coarse soils,  20$ medium soils,  and 15$
fine soils.

     Land use areas within the three sub-basins are dominated by cropland  (64-
69$), residual  (17-25$), and  residential (4-8$).   The residual  category
includes woodlands, wetlands, and grasslands.  Because UALs for any areal cell
are dependent on the predominant  land use and soil texture,  it is necessary to
apportion land  use areas to the three soil texture classifications noted
above.  For this demonstration, the acreage  of each land use corresponding to
a given  soil  texture category is approximated using a simple weighting ratio
of soil  texture area to total  sub-basin area. The acreage within each sub-
basin corresponding to a  specific soil texture and land use is defined by the
resulting land use/soil texture matrix.

High Tillage  vs. Low Tillage Cropland

     The  land use  analysis  conducted  for  this   demonstration did  not
distinguish between low tillage and high tillage cropland areas.  High  tillage
crop management  practices are  defined here as conventional or deep  tillage
methods  consisting of complete soil inversion following crop harvest.   Farming
areas utilizing  conservative tillage management systems,  including   minimum
tillage  (e.g., noninversion soil conditioning with chisel plows) and no till
methods,  have been grouped together and defined here as low tillage cropland.
For the purposes of baseline calculations, it is assumed that the high  tillage
and low  tillage cropland areas are equally represented  in the three sub-
basins .

Unit Area Load Matrix

     Unit area  loads used to estimate nonpoint source  total phosphorus
contributions within the Fox River watershed were derived from storm event
monitoring data and literature values reflecting similar climatological, land
use and  soil  texture conditions.   Among the major  information sources utilized
in developing UALs  were:  a] surface washoff measurements  from  21  test
watersheds in northern  Virginia (NVPDC/VPI,  1978), b] monitoring of  12 test
watersheds in metropolitan Washington,  DC (NVPDC,  1983)f  c]  data from  11 test
watersheds in the Chesapeake Bay drainage basin (Hartigan et al., 1983), d]
pilot watershed measurements in the  Great Lakes basin (PLUARG, 1978), e]
methodologies of the  water  quality  assessment procedure  for  toxic  and
conventional pollutants (Tetra Tech, 1982), and  f] simulations of water
quality  and  hydrology  for the Occoquan Reservoir,  Virginia (Office of
Comprehensive Planning, Fairfax County,  Virginia,  1982).
     The UALs applied in this demonstration for the Fox River and Green Bay

                                  T-I-C-5

-------
       FOX RIVER WATERSHED
              SUB-BASINS
 RIVER LAKES
 SUB-BASIN
                          LOWER FOX RIVER
                          SUB-BASIN
                      LAKE WINNEBAGO
                      SUB-BASIN
FIGURE 2. Component sub-basins of the Fox River watershed.
              T-I-C-6

-------
are presented in Table 1.  Total  phosphorus contributions  to the Fox River and
Green Bay from interf low/basef low water are not included in the UAL values
presented in Table  1; these loads are treated as constant at a rate of 0.018
kg/ha*year over  the  entire watershed.   The UALs  applied in this demonstration
are intended to represent annual total phosphorus  contributions from surface
runoff events within the Fox River  watershed during an "average" year of
wetness or total precipitation.  The current analysis has not attempted to
quantify and compare climatological and physiographic  factors  among the areas
from which UAL information was compiled.
Table 1.   Total Phosphorus UAL Matrix  (kg/ha*year)
       Land Use
                                      Coarse
Soil Texture
   Medium
Fine
Residential
Commerical
Industrial/Transportation
Institutional/Government
High Tillage Cropland
Low Tillage Cropland
Pasture
Woodland/Residual
0.06
0.28
0.18
0.09
0.55
0.22
0.06
—
0.13
0.28
0.37
0.18
0.74
0.29
0.07
0.02
0.18
0.28
0.55
0.28
0.92
0.37
0.09
O.OM
Average annual  total phosphorus loads are  calculated for each sub-basin as the
product of the  land use/soil texture matrix and the unit area load matrix.

Land Use Scenarios and Loading Estimates

     The nonpoint source microcomputer model  is used  to generate estimates of
average annual  total phosphorus loads to the  Fox River  for seven hypothetical
land use  scenarios.   These test cases are not intended to represent expected
land use trends within the drainage  area; rather,  they demonstrate  the  value
of the model in obtaining a rapid,  macroscopic estimate of  nonpoint loadings
under a diversity of land use conditions.  In addition, the test cases show
how  the  model may  be  used  to  examine  potential  load  reductions  and
concommitant water quality improvements resulting from implementation of
remedial action within given portions of the watershed.

     The seven  scenarios include a baseline (=existing) condition, a "best-
case" condition (all basin acreage in woodland), a "worst-case" condition (all
basin acreage in high tillage), and three intermediate scenarios  (baseline
conditions plus combinations of high  and low tillage).  The estimated nonpoint
source total  phosphorus loadings to  Green Bay from  the Fox River are presented
in Table 2.

Model Output  and Analysis

     An examination  of model-estimated loadings  points to the importance of
the  cropland  contribution  to the  overall  load  — an  expected result
considering  the high UAL assigned to cropland areas and the high acreage of
cropland present  within eachof  the Fox River sub-basins.   The average annual
total phosphorus  load  for  Scenario 1 — baseline conditions — is estimated at

                                   T-I-C-7

-------
Table 2.  Land Use Scenarios and Annual Phosphorus Loads
Scenario
 1. Baseline (includes 50$  High Till)
 2. Woodland
 3. High Till Cropland
 4. Baseline/100$ High Till
 5. Baseline/75* High Till
 6. Baseline725$ High Till
 7. Baseline/100$ Lo Till

 « Watershed total;  includes baseflow
                                               Total Phosphorus Load
                                              (kgP/year)§     (kgP/day)«
                                               861,725
                                                91,993
                                              1,675,647
                                              1,186,948
                                              1,024,338
                                               699,114
                                               536,499
2361
 252
4591
3252
2806
1915
1470
861,725  kgP (2361  kgP/day).   Almost 90$ of  that total is attributable to
runoff contributions fom cropland.  The influence of cropland on nonpoint
loadings is  further evidenced by  the significant  increases and decreases in
the  expected  loads which  accompany  changes in  the percent of  cropland
distributed between the high tillage and low  tillage  categories.   The water
quality implications of these hypothetical land use scenarios are now examined
by coupling output from the nonpoint source  model to a microcomputer-adapted
trophic response model.
WATER QUALITY MICROCOMPUTER MODEL

Model Structure and  Baaelina Conditions

    The water quality microcomputer model is a steady-state,  12-cell mass
balance model for phosphorus;  details of model  development,  calibration, and
verification are provided elsewhere (Auer and Canale,  1985).   The model
considers sources (tributary loads)  and  sinks (net settling,  mass  transport)
for total phosphorus in Green Bay and calculates the steady-state  phosphorus
profile for each of 12 model cells  oriented along the major (NE/SW) axis of
Green Bay.   A comparison of model  output with summer average  total  phosphorus
concentrations for  1982 is provided  as Figure 3.  Trophic conditions for each
model cell  are  evaluated  using the methodology  of Chapra and Dobson (1981); a
total phosphorus concentration <11.0 ugP/L Indicates oligotrophy,   11.0 - 21.7
ugP/L — mesotrophy,  and >21.7 ugP/L — eutrophy.  The gradient in trophic
state corresponding  to  the phosphorus levels in Figure 3  (Baseline  + 50$ High
Till)  is presented as Figure  4.   Under baseline  conditions,  the four
southernmost model cells (approximately 20 km) are eutrophic, the mid-bay
region is mesotrophic  and the outer reaches are oligotrophic.  It should be
noted that significant  oxygen  depletion  occurs in  the  hypo limnetic waters of
the mid-bay  region — an area currently classified as mesotrophic.
Response to
                      Scenarios
     The microcomputer nonpoint source loading model estimates a baseline
annual total phosphorus loading of 2361 kgP/day; this value compares well with
the five-year average (2174 kgP/day, range 1240-2716 kgP/day) calculated from
monitoring data and  discharge permit reports by Roznowski and  Auer (1984).
                                 T-I-C-8

-------
        wo-
                           Green Bay Summer Average -1982
                                    Data-mean*S.D.

                                — Mode! Output
                   20       40       60       80       100

                 DISTANCE FROM FOX RIVER MOUTH (km)
FIGURE 3.  Steady-state model output compared with measured summer average total phosphorus con-
          centration at 12 stations  in Green Bay; bars (standard deviation, n = 18) indicate temporal
          variation at each station.
                                 T-I-C-9

-------
                  Baseline
  Scenario I
                              TROPHIC  STATE
                                 OLIGOTROPHIC
                                 MESOTROPHIC
                                 EUTROPHIC
                Woodlands
             100% HiTill
  Scenario 2
Scenario 3
            Base* 100% Hi
          Base •»-75% Hi
  Scenario 4
Scenario 5
             Base-i-25% Hi
          Base+100 %Lo
  Scenario 6
Scenario 7
FIGURE 4. Trophic status for Green Bay under baseline conditions and 6 hypothetical land use scenarios.
                         T-I-C-1C

-------
The microcomputer water  quality model  calculates  summer average  total
phosphorus  concentrations based on the  summer average  tributary  total
phosphorus  load. Scenario-specific summer average loads for  the Fox River are
calculated according to Equation  2:


                        W =  '"seen / wbase>  * wsum                  <2>

where:

    W  =  scenario-specific  summer  average total phosphorus load (kgP/day)

 Wscen  =  scenario-specific  annual  total phosphorus load (kgP/day)

       =  baseline annual total phosphorus load (kgP/day)

  Wgum  =  baseline summer average total phosphorus load (kgP/day)


The baseline (1982) summer  average total  phosphorus loading  to Green Bay from
the Fox River is  959 kgP/day  (Auer and Canale,  1985) and the baseline annual
total  phosphorus loading to Green Bay  (Table 2, Scenario 1) is 2361 kgP/day --
thus the  summer load accounts  for  approximately 41$ of the annual  load.

    Illustrations of  the gradient in trophic conditions for each land use
scenario are presented in Figure 4.   The most dramatic change in  trophic
status is  that for  the hypothetical  "best case" condition (Scenario 2)
corresponding to 100  percent woodland where  oligotrophic conditions  prevail
over the  entire length  of the  bay.  The load yields a summer  average Fox River
total  phosphorus concentration of  11.3 ugP/L,  only  slightly above the limit
for oligotrophy, and approaching the Lake Michigan boundary condition of 8
ugP/L.  Output from this scenario may reflect conditions prior  to cultural
development  in the basin.   Only  slightly  less  dramatic is  the  hypothetical
"worst  case"  condition  (Scenario  3), where  the entire  basin  is characterized
by high tillage cropland.  In this case,  eutrophic conditions persist for
approximately 65 km out the bay  and overlay  the  regions most susceptible to
hypolimnetic dissolved  oxygen depletion.  These results indicate that a wide
range in  water quality conditions is  possible for Green Bay --  and that the
current gradient in trophic state  is strongly related to land use practices.

    The  "worst-" and  "best case"  scenarios demonstrate  the sensitivity of the
system to  land use practice, but are not realistic  or viable management
options.  Scenarios 4-7 examine  the impact of variation in high tillage/low
tillage  ratios under  the baseline  scheme  for land use.    The output for
Scenario  5  (Baseline +  75$ High Till)  is not  appreciably different from that
of the baseline condition.  Expansion of high tillage practices to baseline
plus  100$  high  till   (Scenario 4)  extends  the  region   of  eutrophy  by
approximately 10  km.   Similarly, reduction of high tillage practices to
baseline plus 25$  high till (Scenario  6) or baseline  plus 100$ low till
(Scenario 7) reduces  the region of eutrophy  to the southernmost 10-15 km of
the bay.  Implementation of 100$  low  tillage  practices offers the additional
advantage of extending the oligotrophic zone into the mid-bay region where
oxygen depletion  has been reported  under baseline  conditions.  Specific
water  quality impacts may  be  evaluated for each  scenario  through  the
application of indices  relating  phosphorus and  transparency or  chlorophyll.

                                 T-I-C-11

-------
CONCLUSIONS

     Microcomputer models  for  nonpoint source phosphorus loads  and water
quality are coupled to examine the impact  of  variation in land use practices
on  trophic state in  Green Bay.   Trophic  response to various  land  use
management scenarios is dramatic because of the  large contribution O8555) by
cropland to the annual Fox River nonpoint total phosphorus load. Model output
indicates that land  use management,  especially low tillage practices, offer an
opportunity for  improved water  quality in the  bay.  The  utility  of  this
interactive planning and management tool has  been  demonstrated for several
hypothetical land use strategies in the Fox River/Green Bay system.
REFERENCES

Auer, M.T. and R.P. Canale. 1985.  A phosphorus budget for Green Bay.  Report
   to U.S. Environmental Protection Agency, ERL-Duluth, MN (In Preparation).

Chapra,  S.C.  and H.F.H. Dobson.  1981.  Quantification of  the lake trophic
   typologies of Naumann (surface quality) and  Thienemann (oxygen) with
   special reference to the Great Lakes.  J.  Great  Lakes  Res., 7(2): 182-193.

Fairfax County Office of Comprehensive  Planning. 1982. Occoquan Basin Study.
   Fairfax County, Virginia, 162 pp.

Hartigan,  J.P., Quasebarth, T.F., and E.  Souther land.  1983. Calibration  of  NFS
   model loading  factors.  J. Env. Eng.  Div., ASCE,  109(6):  1259-1272.

Mortimer, C.H. 1978. Water movement, mixing, and transport  in Green Bay. p.
   10-56, In Green Bay Workshop Proceedings, Univ. Wisconsin Sea Grant Pub.
   No.HIS-SG-78-234, Madison, WI

Northern Virginia Planning District Commission  and the Virginia Polytechnic
   Institute  and State University. 1978. Occoquan/Four Mile Run Nonpoint
   Source Correlation Study:  Final Report.  Metropolitan Washington Water
   Resources  Planning Board, Washington, DC.

Northern Virginia Planning District Commission.  19&3. Washington  Metropolitan
   Area Urban Runoff Demonstration Project:  Final Report.  Metropolitan
   Washington Council  of Governments, Washington, DC.

Pollution from Land Use  Activities Reference Group. 1978.  Environmental
   Management Strategy for the  Great  Lakes System.  Final Report to  the
   International  Joint  Commission, 115 pp.

Roznowski, D.M. and M.T. Auer. 1984. Tributary loadings to Green Bay: A mass
   balance approach.   Report  to U.S. Environmental Protection Agency, ERL-
   Duluth, MN, 74pp.

Tetra Tech, Inc. 1982. Water Quality Assessment: A Screening Procedure  for
   Toxic and Conventional  Pollutants — Part  1.  U.S. Environmental Protection
   Agency,  Report No. EPA-600/6-82-004A,  570 pp.
                                 T-I-C-12

-------
               MODELING  VERTICAL  FLUX  OF  PESTICIDES  WITH  CREAMS

  R. A.  Leonard, Soil  Scientist;  W.  G.  Knisel,  Research Hydraulic  Engineer;
                       D. A.  Still, Agricultural  Engineer
                       U.  S.  Department of  Agriculture
                         Agricultural Research  Service
                   Southeast  Watershed  Research  Laboratory
                                Tifton,  Georgia

                  A. W.  Johnson,  Research Plant  Pathologist
                       U.  S.  Department of  Agriculture
                         Agricultural Research  Service
                  Nematodes,  Weeds,  and Crops  Research Unit
                                Tifton,  Georgia


                                   ABSTRACT

Mathematical  models to  assess  nonpoint  source  pollution  and  evaluate  the
effects  of management  practices are  needed to respond to  water  quality legis-
lation.   The  USDA model,  Giemicals,   Runoff,  and ^rosion  from  Agricultural
Management ^ystems (CREAMS),  which has  "Keen used  successfully  in modeling pes-
Ticides  in surface runoff  from  field-sized  areas  was modified  to simulate pes-
ticide flux  and persistence  in  the soil plant  root zone.   The  modified  version
of  CREAMS  uses the daily  rainfall option,  and the  soil  erosion  submodels  in
the  original  CREAMS.    The  modified  version  considers  pesticide  adsorption
characteristics, degradation  rates,  and soil characteristics by  horizon.   Pes-
ticide distribution in the root zone is computed with time and  depth,  as well
as  the  quantities  leached below  the root  zone.   Pesticide concentrations  in
water and  sediment  fractions of  surface  runoff are  also  computed.   Up  to  10
different  pesticides  and  multiple  applications  of each can  be  considered
simultaneously.  Preliminary  tests comparing measured persistence  and movement
of the nematicide fenamiphos  (Ethyl  3-methyl-4-(methylthio)phenyl(l-methyleth-
yl)phosphoramidate) with simulation  results appears  promising.

Utility  of the model in  long-term simulation was  demonstrated.   Application  of
the  model  in  comparing  relative effects  of  management,  climate, pesticide
properties,  and  soil   characteristics  on  potential movement  of  pesticides  to
groundwater  is envisioned.

Keywords:  Groundwater pollution, water  quality,  pesticide leaching,  runoff,
           erosion
                                    T-I-D-l

-------
 INTRODUCTION

 Mathematical models  to  assess  nonpoint  source  pollution  and  evaluate  the
 effect  of management practices  are  needed  to adequately respond to  the water
 quality legislation implemented since  about 1970.  Nonpoint  source  pollution
 from  agricultural  areas  must be assessed,  specific problems  identified,  and
 alternative practices  developed to reduce or minimize  problems.   Sediment and
 chemical  losses  from surface runoff and chemical  losses  from the  root  zone by
 percolation must be addressed.   Monitoring  every field or farm to  measure pol-
 lutant  movement  is impossible,  and landowners and  farm operators  need  to know
 benefits  and costs before they  implement alternative practices.

 Although  models  are available for  surface  runoff  and  chemical  transport,  the
 burden  is upon the model  user to select from several  potential  models  the one
 that  will  best  represent the conditions,  practices,  and  desired  results  for
 specific  problems.  The  U.  S.  Environmental  Protection Agency has  sponsored
 development of  several  models;  e.g.  PRT (Crawford  and  Donigian, 1973),  ARM
 (Donigian and Crawford,  1976a), NPS  (Donigian  and  Crawford,  1976b),  culmina-
 ting  in HSPF (Johanson et  al.  1980) which  is  a comprehensive model with  the
 capability of continuously simulating the dynamics  of  river  basins and system
 response   to implementing  basin-scale  water  quality  management  decisions.
 Other  field-scale  pesticide  models  were described  by Haith  and Tubbs (1981)
 and Steenhuis and  Walter  (1980).   In 1978, the  U. S. Department  of  Agricul-
 ture, Agricultural  Research Service,  began  a national  project to develop math-
 ematical  models  for evaluating nonpoint source  pollution.   The first  develop-
 ment was  the CREAMS model,  published  in 1980 (Knisel,  1980).

 With  reports of widespread  contamination   of  groundwater  (Pye  and  Patrick,
 1983),  concerns   have  mounted for  the  protection  of  this valuable  resource.
 Agriculture may  be  an  important contributor of organic contaminants  such  as
 pesticides (Cohen et al. 1984;  Marti  et al. 1985).   In most of the  reported
 incidences  of  groundwater  contamination  by  pesticides,  contamination  has
 occurred  because of combinations of several conditions  and prerequisites  con-
 ducive  to pesticide transport.   These may have  been heavy applications  of per-
 sistent and mobile compounds  applied  to soils with little adsorptive  capacity,
 spills, shallow  water  tables, and high  rainfall  with  annual  aquifer  recharge.
 As with surface  water  problems,  models  are   needed  to  assess  the potential  for
 transport to groundwater  and evaluate corrective measures.   The PESTANS model
 (Enfield  et al.  1982) was developed  to  provide  a one-dimensional projection of
 vertical  chemical  movement through the  unsaturated zone.  The Pesticide  Root
 Zone  Model  (PRZM)  by  Carsel  et al.  (1984)  was  developed  for evaluating  the
movement  of pesticides  within the root  zone and the  lower  unsaturated zones.
 This  model  incorporates  a daily rainfall/runoff method  similiar  to  that  in
 CREAMS  and a numerical  solution  for  the chemical  transport  equation.  The PRZM
model  is  currently undergoing extensive  verification and  testing.

 The USDA  CREAMS model is  widely  used  (Knisel and Svetlosanov,  1982; U.  S.  Soil
 Conservation Service,  1985)  and has  received  rigorous testing and  evaluation
 by the  developers  and other scientists  (Leonard  and Wauchope,  1980; Lorber  and
Mulkey, 1982; Nutter et  al.  1984).   Although   CREAMS  routed plant  nutrients
 through the root  zone,  pesticides were  not;  leaching only below the surface 10
mm zone was  considered.
                                    T-I-D-2

-------
Users of CREAMS  have expressed considerable interest in an  addition  to  CREAMS
for pesticide  leaching  in the root zone.   This  paper describes such  an  addi-
tion presently called CREAMS  1.7R.  This  addition  is  in  keeping with  the  orig-
inal CREAMS structure and  simplicity.   A  more  comprehensive  model,  CREAMS2,  is
currently  being  tested  and will  also  consider pesticide leaching  through  the
sol urn (Leonard and Ferreira,  1985).

THE MODIFIED CREAMS  MODEL

The  original  CREAMS model consists  of  three major components:   hydrology,
erosion/sedimentation,  and chemistry.   The hydrology component  estimates  run-
off volume and  peak rates, infiltration,  evapotranspiration,  soil water  con-
tent, and  percolation on  a daily basis.    If  detailed  precipitation   data  are
available,  infiltration may be estimated  at  histogram  breakpoints.   The  ero-
sion component  estimates erosion and  sediment yield, including  particle-size
distribution at the  edge  of the field.  Storm  loads  and  average concentrations
of  sediment-associated  and dissolved  chemicals  in the  runoff are also  esti-
mated as well as nitrate-nitrogen leached  from the root  zone.

The daily  hydrology model of  CREAMS  accounts  for soil  water  by dividing  the
root zone  into seven soil  layers  and  estimating flow through  each  layer  using
a storage-routing technique.   CREAMS 1.7R  uses this  daily  hydrology option  ex-
clusively  and routes pesticides through the same  seven  layers  using pass-files
of water content and flux  generated from  the hydrology model.   Plant  nutrients
are not  simulated  in CREAMS  1.7R  to  simplify input  requirements.   Also,  the
streamlined  Soil  Conservation  Service version of  pesticide  input files  and
their crop  rotation  aspect (U. S. Soil Conservation Service,  1985) was  incor-
porated into CREAMS  1.7R  for  user convenience. For  long-term  simulation  runs,
the irrigation option of DelVecchio et al . (1983) was  added  so that  adequate
soil moisture was maintained  for  crop  growth without  having  to manually  speci-
fy irrigation days and  amounts.

A schematic of the  pesticide  submodel  in CREAMS with the CREAMS  1.7R  addition
is given  in Fig. 1.  Pseudo  first-order functions  are  assumed for  pesticide
dissipation from  foliage and  soil  after  pesticide application.  Pesticide  in
the runoff  active soil  layer  is partitioned between  the  solution phase and  the
soil phase  by the following relationship:

                           (CWQ) + (CSM) = A Cp

and
where  Cw is  pesticide  concentration in  runoff  water;  Q  is  the  volume  of
water per unit  volume of surface  active  layer;  Cs is pesticide concentration
in soil equilibrating with  runoff;  M  is mass  of  soil  per  unit  volume of  active
surface layer; A  is  an  extraction  ratio  specifying the soil to  water  ratio  in
the  extraction  zone; Cn  is the gravimetric  concentration  of pesticide  resi-
due  in  the  soil;  and  Kj  is the  coefficient for  partitioning  the  pesticide
between sediment  and water  phases.   Before  runoff  computation, a  portion  of
the pesticide mass is moved  below  the surface 10 mm  by the  function:
                                   T-I-D-3

-------
                                             PESTICIDE
                                             APPLICATION
                     FRACTION  ON SOIL
                           ADD
                     PREVIOUS RESIDUES
                         COMPUTE
                      CONCENTRATION
                       OF  RESIDUE
                        ADJUST  FOR
                    DOWNWARD MOVEMENT
                         COMPUTE
                        AVAILABLE
                         RESIDUE
                        FOR STORM
                                             RAINFALL,
                                          RUNOFF, SEDIMENT
                                          (HYDROLOGY AND
                                          EROSION  MODELS)
FRACTION  ON  FOLIAGE
       ADD
 PREVIOUS   RESIDUES
     COMPUTE
      MASS
    OF  RESIDUE
     WASHOFF
     FRACTION
  CONCENTRATIONS
     IN  WATER
 AND SEDIMENT  AND
    TOTAL  MASS
Fig.  1.   Schematic of CREAMS 1.7R model
                                         T-I-D-4

-------
                       Z  = Z0e
F    RF-RO-S     "1
[^ UKd (1-p) + p  J
where Z  is  the pesticide mass remaining  in  the surface layer  after  infiltra-
tion; Z0,  the  initial  pesticide mass;  RF,  rainfall;  RO,  runoff; S,  initial
abstraction;  D,  soil  particle density; and p, soil porosity.   In  routing  pes-
ticide through the root zone, the  mass (Z0-Z) is  added  to the adjacent  soil
layer below,  along with the  volume  of water percolate.  Pesticide  concentra-
tions in the  water  and soil phases  of this  hypothetical  reservoir are  then
computed using the relationship  Cs  =  K(jCw.   In the model, Kj  is calculated
from K
-------
 Four replicate soil  cores were removed from each plot to depths  of  1  m.   Each
 core was subdivided into 0-10,  10-20, 20-30, 30-40, 40-50,  50-75,  and 75-100
 cm  depth intervals  for pesticide  analysis.   Soil   samples  were  removed  at  5
 time intervals, spanning  a period of 1-40 days after  pesticide application.
 Groundwater wells  were sampled  weekly  after  first  determining water  table
 depth and evacuating to  remove  stagnant water  in  the well  casing.   Rainfall
 and  other meteorlogical data were available from a station  located adjacent to
 the  study area.

 Several  pesticides  were included in this study,  however, only data  on fenami-
 phos in 1983  will  be  presented  here.  Fenamiphos  (Ethyl  3-methyl-4-(methyl-
 thio)phenyl(l-methylethyl)phosphoramidate)   is   a systemic  nematicide  giving
 broad spectrum nematode control  for a variety  of crops.   A  granular  formula-
 tion of  fenamiphos  was  applied in 1983 at  a  rate of 6 kg/ha  active  ingredient
 in  30-cm bands- in  each  row  and  incorporated  7.5  cm deep.    The 6  kg/ha  is
 broadcast equivalent,  i.e.  6 kg/ha if  broadcast uniformly,  however,  since only
 one  third of  the soil  was  treated (90-cm row spacing),  actual mass applied was
 2 kg/ha  on  an  areal  basis.

 Soil  samples  (Ca. 300  g)  were air dried,  thoroughly mixed  and stored under re-
 frigeration pending  analysis. Samples were extracted by shaking  for two hours
 with 50% chloroethanol.  After extraction,  fenamiphos was  converted  to fenami-
 phos sulfone  by oxidation  with  magnesium sulfate and potassium  permanganate,
 partitioned into methylene  chloride, and  determined  by  gas chromatography
 using a  thermionic  detector operated  in the nitrogen mode.   Recoveries by this
 method were >95% with  detection  limits of 0.005 ug/g soil.

 Fenamiphos  concentrations  found  in  the root zone are plotted  in Fig.  2(a) with
 respect  to  time and depth.  Points plotted are at  midpoints  of  each  sampling
 depth interval and  are  averages  of four replicates.   Movement  of  fenamiphos
 below the incorporation  zone with time  after  application  is  indicated;  how-
 ever, concentrations rarely exceeded 0.005  ug/g below 50 cm.   Apparently, most
 of the fenamiphos had dissipated  by degradation in  34 days.

 Using appropriate soils data and measured  rainfall/irrigation  inputs,  fenami-
 phos  degradation and movement was  simulated  as  shown in Fig.  2(b).   However,
 no model  calibration to the  specific  site  was  done.   A degradation  constant
 based on  a  14-day half-life was  estimated from  previous  work  of Johnson et al.
 (1982).   A  KOC value of 160 was  assigned  by averaging measurements made in our
 laboratory  on  20 Coastal  Plain soils.   Incorporation  depth was set at 7.5 cm
 and  all  other parameter values  estimated  as recommended  (Knisel, 1980).  In
 the  model runs,  a root  zone of 46 cm was assumed and divided  into approximate
 increments  of 0-1,  1-7,  7-15, 15-22,   22-30, 30-38, and 38-45 cm.    Computed
 concentrations are plotted  in  Fig.  2(b) and are  at  the  midpoint  of  each  depth
 increment with respect  to  time.   Comparing  Fig. 2(a) and Fig.  2(b),  a reason-
 able  correspondence between measured and  simulated  results  is measured consid-
ering that  depth increments and  times  were  not  identical.   Model   output  dates
were  chosen to correspond  to rainfall events,  whereas  actual sampling  dates
were  somewhat  arbitrary.   The estimated 14-day  half-life provided good  repre-
 sentation of  observed  fenamiphos persistence.  Based on field-collected  data,
fenamiphos  appeared to  be more mobile  than  simulated by the model.  Note con-
centration maxima at lower  depths  and  broadening of  the  concentration  profiles
 in the  observed vs  simulated.   However,  several variance  factors  should be

                                   T-i-D-6

-------
           PESTICIDE  IN  SOIL  (pg/g)
           1234
PESTICIDE   IN  SOIL (pg/g)
1234
 60
                                              50
                                              60 -
                                                                     (b)

                                                             FENAMIPHOS,  SIMULATED

                                                              DAYS  AFTER APPLICATION
                                                                 	O	  3
                                                                 	a	  6
                                                                 	O	  14
                                                                 	•	  21
                                                                 	O	  40
Fig.  2.   Comparison of measured  and simulated fenamiphos concentration  in
           Bonifay sand
                                       T-I-D-7

-------
 considered.   Numerical dispersion may  be created by the  sampling  and simula-
 tion  depth  increments chosen.  The model assumes  a  constant  7.5  cm incorpora-
 tion  depth;  however,  in the field,  the incorporation implement cannot be regu-
 lated exactly  and  the depth  of incorporation  may  have  varied  considerably.
 Establishment  of row beds  at  planting  and  soil subsidence with  rainfall  also
 creates  surface roughness  and distorts  the  surface reference.

 In  model application,  a  soil organic  matter  profile of  1.0,  1.0,  0.7,  0.5,
 0.3,  0.2,  and  0.1%  was assumed for  the seven depth increments.   Organic matter
 contents were  not available on the  specific site,  only an approximate value of
£ 1%  was available  for the surface  horizon.  Lower organic matter content than
 assumed  would  have  contributed to an  underprediction of  pesticide movement.

 During the model simulation period  depicted in Fig.  2(b), simulated loss below
 the 46 cm  root zone was 10.2  g/ha.   Computed  percolation below  the  root  zone
 during this  period  was 9.6 cm giving  an average fenamiphos concentration of
 0.011 mg/1  in  the  percolate.  At saturation,  this solution  concentration  plus
 that  adsorbed  at equilibrium by  the  soil  organic  matter would give  a concen-
 tration  in  soil of  0.007 ug/g on a dry weight basis.   The practical  limit of
 detection  for  fenamiphos  by actual  analysis was  0.005  ug/g,  so  it  is obvious
 that  comparisons are  made  in a range  of uncertainty  in  the actual data.  Using
 Fig.  2(a),  an interpolated  fenamiphos  concentration  at the  45-cm  depth  of
 0.082 ug/g produces 0.2 mg/1  in  the  soil  solution  at saturation.   At this  con-
 centration,  if 3.7 cm of   water  percolated through this plane  (predicted  by
 model simulation),  74.2 g/ha fenamiphos could have  been  transported.   This is
 considerably greater  than  the 10.2 g/ha simulated.   However,  within the'limits
 of detection no fenamiphos was transported  below the 55  cm depth.   Also,  none
was detected  in sampling  wells or tile  outflow.    For  the  entire  simulation
 period,  however, about 10  g/ha could  have been leached  and would  not have  been
detected if  concentrations remained below 0.005  ug/g soil.  Clearly, we are in
 a concentration range  that does not give  a complete test  of  the model,  and
 illustrates  the difficulty in obtaining field  test data  for pesticides ranging
 from  the nonadsorbed  to moderately  adsorbed  (such  as  fenamiphos).    For  this
 and other  reasons,  models  are best used  to make relative comparisons  and  not
 absolute predictions.

An indication  of model  sensitivity to pesticide  Koc  is  provided for two hypo-
thetical pesticides in Fig. 3.  Model runs were performed  using  identical  in-
puts  as  in simulations  depicted  in Fig.  2(b) except  that  in test  A,  Koc =  500
and in test  B,  Koc  =  5 were assumed.  A degradation  half-life of  50 days  for
both  and identical  application  rates of 1.7  kg/ha  were  also  assumed.   Com-
paring these simulations demonstrates that  significant movement of  very weakly
 adsorbed pesticides would  have  occurred below  the root  zone  under  rainfall
patterns observed in  1983.  Of the total amount  applied, about  31%  of pesti-
cide  B moved below  the  root zone  compared to 0.01% of pesticide A.

 In the fenamiphos  simulation  for 1983  (Fig.  2[b]), 0.74% of  the  application
was estimated   to  move below the  46-cm root zone  during  the  entire year.
Annual losses  will, however,  be highly dependent on rainfall occurring within
a few weeks  after application.   Field experiments are usually  limited  to  1  to
3 years  and  cannot  assess the  year-to-year  variation  expected in  leaching
losses.  Using daily  rainfall  records for the  period 1958-77 at  Tifton, 6A,  a
20-year  simulation  of  fenamiphos  leaching  at the experimental site  was  per-

                                  T-i-D-8

-------
          0.25
PESTICIDE  IN SOIL  (>ig/g>
 0.50     0.75	1.00
                                      1.25
                                             1.50
  4Q
  50
  60
          PESTICIDE  A. SIMULATED

          DAYS AFTER APPLICATION


             	O	 14
             	•	 21
             	Q	 40

               K-v • 500
                                                     PESTICIDE  IN  SOIL (>ig/g)
                                                  0.25	 0.50    0.75	1.00
                                                                                        1.25
                                                  DAYS AFTER APPLICATION
                                                           -  3
                                                           -  6
                                                       -O	  14
                                                       —	  21
                                                       -Q	  40

                                                       Knr • 5
                                                    60
Fig.  3.   Comparison  of mobility in  Bonifay sand of  two hypothetical  pesticides
           differing only  in K   .   Rainfall  and  other  inputs same as  used  in
simulation shown
                 QC.
               if  Fig.
                                        2.
                                        T-I-D-9

-------
formed,  and  results are presented in Fig.  4.   Annual losses ranged  from  near
zero  to  about 5%.  The 5%  loss  occurred  in 1973 when 20.3 cm of  rainfall  oc-
curred  within a  14-day period  after application.   These  results  demonstrate
the  value of  long-term simulations in assessing  expected leaching  below  the
plant root zone  and  serves  as  a  useful  tool  in  extending  field observations.

SUMMARY  AND  CONCLUSIONS

Based on  preliminary evaluations,  a modification of  CREAMS which  provides  sim-
ulation  of pesticide concentrations  in  the  root zone, and masses  leached  below
the root  zone appears  promising.   Testing  and  verification efforts  will  con-
tinue.  Model  utility in  long-term simulations  was demonstrated.   Model  appli-
cations  in evaluating relative effects  of management decisions, differing  soil
and pesticide properties,  and  weather patterns  on potential pesticide  loading
to groundwater are  envisioned.

REFERENCES
Carsel,  R.  F.,  C.  N, Smith,  L.  A.  Mulkey, J.  D.  Dean,  and P. Jowise.   1984.
Users Manual  for  the Pesticide Root Zone Model  (PRZM) Release 1.   U. S.   En-
viron. Prot.  Agency.  Athens,  6A.
Cohen, S.  Z., R.  F.  Carsel, S. M. Creeger, and 6.  G.  Enfield.  1984.   Poten-
tial  for pesticide  contamination  of groundwater  resulting from  agricultural
uses.  In:  R. F. Krueger  and  J.  N.  Seiber  (eds.).   Treatment  and  Dispersal  of
Pesticide Wastes.  Amer. Chem. Soc.  Washington, DC.   pp.  297-325.
Crawford,  N.  H.  and  A.  S. Donigian.   1973.    Pesticide  transport and  runoff
model for agricultural lands.  EPA-660/2-74-013.   U. S. Environ. Prot. Agency,
Washington, DC.
DelVecchio, J.  R.,  W. G.  Knisel,  and  V.
irrigation  on pollutant  loads.    Proc.
Amer. Soc. Civil Eng., Jackson, WY.  pp.
                                Crawford.
                                                         1983.   The impact  of
                                                          Drainage  Conference,
Donigian, A. S., Jr.  and  N.  H.
trients  on  agricultural  lands.
Technology Series.   Office of
Agency, Washington, DC.   317 p.
Donigian, A. S., Jr., and  N. H.
from  the  land surface.   U.  S.
3-76-083.
Enfield, G. G., R. F. Carsel, S.  Z.
Approximating pollutant transport to
Haith, D. A. and L. J.  Tubbs.   1981.
ricultural nonpoint source pollution
Univ., Ithaca, NY.
           A.  Ferreira.
           Irrigation  and
           113-123.
            1976a.   Modeling pesticides and nu-
   EPA-600/2-76-043.   Environmental  Protection
Research and Development,  U.  S.  Environ. Prot.
                                Crawford.   1976b.  Modeling  nonpoint  pollution
                                Environ.  Prot. Agency,  Athens,  GA.    EPA-600/
                                     Cohen,  T.  Phan, and D. M. Walters.  1982.
                                     groundwater.   Ground  Water  20:711-721.
                                       Operational  methods for analysis of  ag-
                                     .  Search:  Agriculture 16:20 pp.  Cornell
                                       H.  Davis.
                                                  1980.  User's
                                                U.  S.  Environ.
                                 Manual  for the
                                 Prot.   Agency,
Johanson, R. C., J. C.  Imhoff,  and  H
Hydroldgic  Simulation  Program-FORTRAN (HSPF).
Athens, GA.  EPA-600/9/80-015.
Johnson, A.  W.,  W. A.  Rohde,  and  W.  C. Wright.   1982.   Soil distribution  of
phenamiphos  applied  by overhead  sprinkler irrigation  to control  Meloidogyne
incognita on vegetables.  Plant Disease 66:489-491.
Knisel, W. G. (ed.).   1980.   CREAMS:   A Field-Scale Model for Chemicals,  Run-
off, and Erosion from  Agricultural  Management  Systems.   U. S.  Dept.  of  Agric.,
Sci. and Educ. Adm., Conserv. Research Rep. No.  26.
                                  T-I-D-10

-------
UJ

8
i-
                     S
                     !•
                     z
                     o
                     §2
                     O I
                     t-


                     So
                             FENAMIPHOS INCORPORATED IN
                             BONIFAY SAND  FOR  NEMATODE
                                   CONTROL
                           2  4   6   8  10  12   14  16  18  20
                           YEAR IN  20-YR. SIMULATION PERIOD
                                      1958 - 77
Fig. 4.   Results  of 20-year  simulation  of fenamiphos leaching below  46  cm root
          zone  in  Bonifay  sand using  rainfall record. 1958-77, at Tifton, GA
                                      T-I-D-ll

-------
Knisel,  W.  G.  and Svetlosanov, V. (eds.).   1982.   European and  United  States
Case Studies  in  Application of the CREAMS Model.   International  Institute  for
Applied  Systems  Analysis,  Laxenburg,  Austria.   Collaborative Paper  CP-82-S11.
Leonard,  R.  A.  and V. A. Ferreira.   1985.   CREAMS2.   The  chemistry  submodel.
Proc. Natural Resource  Modeling Symposium.  U. S.  Dept.of  Agric.,  Agric.  Res.
Service.  In Press.
Leonard,  R. A. and  R. D.  Wauchope.   1980.  The pesticide submodel, ch.  5,  pp.
88-112.   In:   Vol.  I.  Model  documentation, CREAMS:   A Field Scale  Model  for
Chemicals,  Runoff,  and  Erosion  from Agricultural  Management  Systems.W.  G.
Knisel (ed.).  U. S.  Dept.  of  Agric.,  Conservation  Res.  Rep. 26.  643 pp.
Lorber,  M.  N. and L.  A.  Mulkey.   1982.  An evaluation of three  pesticide  run-
off loading models.   J.  Environ.  Qua!.  11:519-529.
Marti, L. R., J.  De Kanel,  and R. C.  Dougherty.  1985.  Screening  for  organic
contamination .of groundwater:   Ethylene dibromide  in  Georgia irrigation  wells.
Environ.  Sci. Tech.  jn  Press.
Nutter,  W.  L., T. Tkaus, P. B.  Bush, and D.  G.  Neary.   1984.   Simulation  of
herbicide concentrations in stormflow from forested watersheds.   Water  Re-
sources  Bulletin 26.  In  Press.
Pye, V.  I.  and  R.  Patrick"!1983.    Ground  water  contamination  in  the  United
States.   Science 221:713-718.
Steenhuis, T. S.  and  M.  R.  Walter.   1980.  Closed  form  solution  for  pesticide
loss in  runoff water.  Trans.  ASAE 23:615-620, 628.
U. S.  Department of Agriculture.  1972.   Soil  Conservation  Service, SCS  Na-
tional Engineering  Handbook, Sec. 4,  Hydrology,  548 p.
U. S.  Soil  Conservation Service.  1985.   An  SCS  user's guide for  CREAMS:  A
Field Scale Model for Chemicals,  Runoff,  and Erosion  from Agricultural'Manage-
ment Systems.  Tech.  Release No.  72.   In  Press.
                                   T-I-D-12

-------
          NON-POINT POLLUTION ARATFMFNT  IN TAMPA BAY

        Bernard E. Ross, Professor and hark Ross Research Assistant
                Civil Engineering and Mechanics Department   •
                       University of South Florida
                          Tampa, Florida, USA

                                ABSTRACT

This paper presents the most recent evaluation of the effect of pollutlonal
loadings upon Tampa Bay.

In the years between 1970 and 1984, the Tampa Bay Region has been the
subject of A major pollution abatement programs.  In each case urban runoff
has been evaluated. In three of the programs all point and non-point sources
were evaluated.. Loadings of pollution parameters as determined for the year
2000 in all of the programs displayed a remarkable similarity. The
projections  were prepared by  divers groups of professionals over a long time
span using totally different techniques. Thus loadings determined for the
year 2000 for the wet and the dry seasons can be accepted as good estimates.

This report presents the methods used to evaluate the effect of the pollutants
upon the water quality of Tampa Bay.

A data file of water quality parameters dating from 1969 through 1983 were
subjected to multiple regression analysis.

An improved water quality model was developed calibrated and verified to
more closely represent the nutrient storages and flows In Tampa Bay. In
particular the effect of phytoplankton growth on light distributions  in the bay
were addressed. Field measurements were conducted to determine pertinent
rate coefficients for the model. Benthic processes were found to be dominate
in Hillsborough Bay and some portions of Upper Old Tampa Bay.

Keywords: Tampa Bay, Nonpoint pollution.point sources pollution, Statistical
          study, Ecological models.water quality, hydraulic models, benthic
          processes, eutrophcation
                                T-I-E-l

-------
 INTRODUCTION:

 Tampa Bay is a shallow well-mixed bifurcated estuary on the west central
 coast of Florida-see Figure 1.  It is beset by all of the problems attending the
 rapid development of the area. The University of South Florida (USF), Center
 for Mathematical Models has been involved in numerous studies relating to the
 water quality of Tampa Bay, dating from 1971 through the present. The 1984
 Wasteload Allocation Program (WLA) sponsored by the Florida Department of
 Environmental Regulation (FDER) is the occasion of this paper.

 From the first major study of a portion of Tampa Bay conducted by the Federal
 Water Pollution Control Administration IFWPCA 1969], to the present
 studies, the fundamental problems of the bay remain. The consistent
 problems are
            A  Excessive nutrient levels which encourage algae growth,
            B.  Excessive benthic oxygen demand which produces oxygen
               depletion in the benthos and oxygen stratification in the
               water column.
            C.  The gradual disappearance of sea grasses.

 The WLA program was designed to address each of these problems in
 attempting to allocate whatever  resources the bay had to assimilate
 non-point and point source loadings. A program was devised to review all
 existing data, photographic and archival.  Field tests were conducted to
 provide rates for the mathematical model,  calibration and verification data
 were obtained.  The existing data and the new data were subjected to
 statistical analysis.

This paper concentrates on the tasks performed by USF.

PERTINENT FACTS REGARDING TAMPA BAY

Tampa Bay is approximately 35 miles long (67.3 km), and 12 miles (15.5 km)
wide at the widest point. The bottom area is 8.6 billion square feet (1.14
xlO7 km2).  The volume at an average high tide is 116 billion cubic feet (5.6
xlO9 m3).  The normal tidal prism is 13.5 billion cubic feet (6.5 x 108 m3).
Tides are semi diurnal.  Point source input to the bay is approximately 26.6
mi 11 ion cubic feet per day  (1.27 x 106 m3).  Wet weather nonpolnt source
input is approximately 260 million cubic feet per day (1.27 x 109 m3).
Pollutants enter Tampa Bay at the 39 locations shown in Figure 2. Pollutant
loads for this study were arrived at by a statistical separation of low and

                               T-I-E-2

-------
      Figure 2
   Nutrient Discharge
       Points
T-I-E-3

-------
high flows by durations, COM ( 1 984).  Runoff from ungaged areas were
determined by use of U. S. Geological Survey empirical relations developed in
the region ,USGS (1983). Parameter concentrations were determined by use
of empirical relations from the same source. Yearly total loads determined in
this study were within a few pounds of those determined in 1971 - for the
year 2000, TBRPC ( 1 973).  Agreement of the year 2000 estimates for the
1976 determination and the present prediction was also within a total of
2200 poundsJBRPC ( 1 976).  Natural phosphorus loadings are higher for the
Tampa Bay region since this is an area of natural deposits and mining.

STATISTICAL ANALYSIS RESULTS

Samples of waters in Hillsborough Bay subjected to bioassay always indicate
a nitrogen limitation to growth .City of Tampa ( 1 983).  Statistical analysis
of the 1 980- 1 98 1 data shows no correlation of chlorophyl a and nitrogen  in
Hillsborough Bay but a  positive correlation in other parts of the bay.  Strong
correlation existed between BODu and light extinction everwhere. A
correlation between chlorophyl a and ortho phosphorus existed everywhere.
The latter finding was surprising inasmuch as phosphorus concentrations are
high in Tampa Bay .

In brief, the statistics showed that during the wet season planktonic growth
for the most part is limited by some factor other than nitrogen or phosphorus,
possibly light. The relation between  chlorophyl  a and phosphorus is the
result of the spatial proximity of the two since concentrations of phosphorus
were hundreds of times the half saturation concentrations required for
chlorophyl a.

THE HYDRAULIC MODEL USED IN  THE  STUDY

The hydraulic model used in the study is the one developed at USF in 1970.
The model comprises the vertically integrated equations of motion and
continuity in two dimensions, Ross( 1 984).  The basic equations used are
given:
                                  _gDati + x -
     dt    D dx  D  dy        p  dx    dx       D2
                                     * Y -
    dt    Ddy  D dx       p dy    dy         D2

                                T-I-E-4

-------
                                 ^i*t~«««j««^"" \jf        dK"j
                                 '•;—tcs-^ir     .rJT.:::

                                 • r-t^r^I^rtlii!.1.1.* .•£.•::**•<
T-I-E-5

-------
    Figure 5

Net Velocities

    USF
                                                                               mnpn ant.  n.
                                                                               0-3D02HI-ND PS
                                                                               IUIIII
                                                                               HMD
                                                                               oau
                                                                               txua
                                                                               (Licnr
                                                                               isniLi
                                                                                 OIL
                                                                               anus
                                                                                 cic
                                                                                 OIF
                                                                                 DIN
                  lty P«r«wtcr« m Mm U.S f.  Eo.logiool Bay Had.I

                  n) I lions uF povndE in ttn toy

                  eofc*ntraHon In th» boy
                           T-I-E-6

-------
        OISS.
        Basic Equation:
                         PG
                         f 1KC
  Parameter


  ALGAE(A)

  CBOD(C)


  ORGP(P1)          Kp


  ORTHO-P(P2)       0

  ORGN (N1)         KN


  AMMONIA-N (N2)    f2


  NITRATE-N (Nj)    0
                         Table  l Ecological nodel Equations  \

                         Decay Rate(K)          Source/Slnk(R)
                    K0
0


RC
                                                           - apUA
                                                     onpA
                                                                    Sp,
                    "  flKCC
Terns:

PG   -
P
oc
ON
aOG
°OR
°ON
• grazing rate
« respiration rate • 0.1 p'
- algae settling rate
- local specific growth rate (see below)
» deoxygenation rate
- 0X/(0X+0.01)
- settling rate of CBOD
- reaction rate of Organic P

- settling rate of Organic P

» ratio of phosphorus to biotnass
» benthic source rate of Ortho-P

- hydrolysis  rate of Organic N

• settling rate of Organic N

- ratio of nitrogen to bioraass
• nitrification rata

- Ox/(Ox+0.5)
- settling rate of ammonia

• 0.9N,/(0.9N2+0.1N3)
• bentnic source rate of ammonia

» benthic source rate of nitrate

• reaeration  rate
- DO  saturation
» photosynthic oxygen production  rate
» oxygen consumption rate  by respiration
• oxygen uptake per unit nitrification  »  4,
                                                   Supplementary Equations:
   U    "

   where:

   M1
   KN
   Kp
   L
                                                             iNj+Nj+XjjyvrjT^
   a
   b
   IA
                                                   OCL
 maximum specific growth rate
 half saturation value for inorganic V
 half-saturation value for inorganic P
 f%(exp(-ab) - exp(-b))
 2?718...
 photoperlod (fraction of day)
 extinction coefficient » fa£ + 0.0S2A
 algae-free secchi depth
 exp (-KeD)
 IA/IS
 avg light intensity over daylight hours
 saturation light Intensity

 11.6101 TJ°-969D-l-673+KowS

 wind coefficient for reaeratlon
 wind speed
 chlorophyll-a concentration • oc^A
• ratio of chTorophyll-£ to bioraass
 57
                                         T-I-E-7

-------
  Table 2  Calibration Constanta
              I                            B.S7U IkacUntiM
           §>Ty "• AVttM^B %Amfft            3cv.J
                             •
 CM. a-iq «OgBl irioKUB !•_)                A.U Mr  dH/og Ua
                                          O.D994&V
 Ai|«nic MBwboBM decmr tK f              9.005 day*
        UllJinan bfAMdyclB xat* (R )        fl.fl«» J«y~
                                 w
                  l«» «te  OLJ    ..        9.3 W1
iCMBMttllfli SKS *rj                    S.3'
 Ocg. BMutaam flDct±iA) tam  i«  }          o.ttli 2r/tUf
                                          p.oJ
                        *J                D.4 W»
         la alv*l MXBKDV (Hjjl              P.wi* i
        u co ilcQ ^riir**  bi_)            O.ffli i
 a- paaveilca r«r alfal gcoMtft              1«C m«
 M imMHini QV Jdgal reontotiim           3.0 09 Qytaf Itoai
                           T-I-E-8

-------
    dt    dx   dy

These equations expressed in finite difference form are solved explicitly on a
grid representing Tampa Bay. The resulting model was calibrated once in
1970 by the use of sodium f luoroscene dye which was released on target and
photographed repeatedly from the air.  The model has been verified in seven
major studies including the study of this report, Ross (1984).  Verifications
in elude tide heights and velocities. A graphic comparison of net transports
in Tampa Bay is given in Figures 3 and 4. Figure 3 is the results from the
US6S model of Tampa Bay and Figure 4 is the results from the USF model with
the same input data ,Goodwin and Ross( 1984). Comparison of the results from
these two very different models indicates that major circulations do exist in
Tampa Bay.  Net velocities in Tampa bay are shown in Figure 5 ,Ross( 1973).
It has been shown that suspended material tends to congregate in velocity
gyres, Ross(l985).

For this study the driving functions for the model were obtained by  fitting the
observed tide data with an astronomical harmonic series, this allowed
extensions beyond the observed period.   Hydraulic calculations were
performed for each one minute period for a real time of 14.4 days (a repeating
tidal period). Transports, velocities, depths,and accumulated transports were
recorded for each 10 minute interval. These data are the input for many other
auxllliary models including a new Ecological model  developed for this study.

THE ECOLOGICAL MODEL USED IN THIS STUDY

The ecology of Tampa Bay  is shown in Figure 6.  Instead of the open system
suggested by Lewis (I960), a semi-closed system Including the recycling
from the carnivores is the basis for the new model.

The model is based upon the vertically Integrated equations of conservation
of mass including sources and rate processes.  The basic equation and the
applicable rates and parameters are shown in Tables I  and 2.
An example of the output of the program is given in Figure 7.  The form of the
output shown is of invaluable aid in the calibration and verification process.
The entire cycle is visible on one page. The rates used in the model  were set
at default rates as given by Ztson( 1978). A box model created to summarize
the ecologic system (part of the main program) is useful in the calibration
process.  Average values of the concentrations were used to adjust
approximate rate coefficients. This can  be done in minutes.  A long run
shows that other changes may be necessary.
                              T-I-E-9

-------
 During calibration and verification, time varying hydraulic conditions were
 utilized in the hydraulic model as discussed. The ecotogic mode) was
 initialized close to fact then time varying loads are introduced and the model
 run for a 40 day period.  For discharges from the 39 sources, independent
 tests discovered the dynamic equilibrium time of 40 days  for Tampa Bay.

 PARTIAL RESULTS OF THE STUDY

 Some of the important results of study are summarized in six figures.  Figure
 8 and Figure 9 show year 2000 with nonpoint source pollution only -  present
 bottom conditions. Figures  10 and 11 show nonpoint sources with best
 practical  treatment- present bottom conditions.  Figures 12 and  13 show the
 results for nonpoint sources with BMP -and benthic uptake reduced by SOX.

 CONCLUSIONS :

 The six figures from 8 through 13 show that  reducing nonpoint sources by 30%
 by the use of the Best Management Practice only slightly reduces  the spatial
 extent of  poor water quality, however, slightly reducing the benthic uptake
 of nutrients  greatly improves water qual i ty.

 The computer simulations also verified that  the phytoplankton in  Hillsborough
 Bay are nitrogen limited and that they grow to a density at which they become
 light limited. Self shading increases the oxygen  demand and detritus
 settling. The uptake of nutrients from the bottom contributes to the algae
 growth  tending to fix the location of the high density.  The process spreads
 slowly  with  increased deposition. This is the classic pattern of
 eutrophication. While the model does not calculate the vertical distribution
 of dissolved oxygen, it is safe to assume that in areas of high benthic demand
 , anoxic conditions exist in the lower water column much of the time even
 though the average concentrations throughout the water column may meet
minimum standards.
                               T-I-E-10

-------
         - BtMlM •/
   »ltod*l!  TMJ> MM «ltl ••
   CMtrvl* ••* •» n -
«•!•» DO •* ttifk flow
         Figure 8
      Minimum  D.O.

        Yow2000
     Nonpolnt Sources Only
        No BMP
              *f
iM*n TM* mt
OMtr»I» «n4 M n
      Figure 9
      Year 2000
  Average Chlorophyl a
  Noo point Sources Only
     NO BMP
                       T-I-E-11

-------
                     - Ceeolta ol
            tat CsSelt Year »CCO tdtb
            Orbea ecd Aoricultcral t»»
               ej fO - nalca £3 at
                   Fiqure 10
                   Year 2000
              Nonpoint Sources Only
                With BMP
              Minimum DO
                                                                         Baetdta of
                                          Orteo end AQtltttlterol
                                          r s4 K, pg .
                                          Dleb Pica
                   o e»
       Figure  '11

  Average CMoro.o
      Yocr2000
NOT point eourcoo only
   WithBHP
- Catele* e» Ittyt
 sees o* nj. c.i
c^li-iii:
V •'	d
                   Figure  12
               Minimum   DO   •
              Nonpoint Sources only

               With Benthf c demand
                 Reduced
 Day
                                                                                  o*
                                                                               to M
                                                                      CaAnod Bcatble
                                                                          « a« Qlfib Pica
      Figure  13

  Average Chloro. a
  Nonpoint Sources Only
  With BonthfcOomend
    Reduced
                                 T-I-E-12

-------
REFERENCES

1.  Rates, Constants, and Kinetics Formulations in Surface water Quality
    Models.  Zison, S.W., W.B. Mills, D. Deimer, C.W.  Chen of Tetra
    Tech, Inc. for USEPA (ORD) EPA-600/3-78-105; December 1978.

2.  "Results of Phytoplankton Bioassay Experiments".   City of Tampa
    Department of Sanitary Sewers, June 1983.

3.  "Tributary Streamflows and Pollutant Loadings Delivered to Tampa
     Bay".  Camp, Dresser & McKee, Inc.  January 1981.

1.  "A New Geographies Description of the Boundaries and Subdivisions of
     Tampa Bay".  Lewis, R.R. Ill and R.L. Whitman.  BASIS Symposium,
     May, 1982.

5.   "Courtney Campbell Causeway Tidal Flushing Study".   TBRPC with B.E.
     Ross and M.W. Anderson for FOOT.  69pp.  February,  1983.

6.   "Problems and Management of Water Quality in Hillsborough Bay,
      Florida".  Hills. Bay Technical Assistance Project, SE Region
      FWPCA, December, 1969.

7.   "Water Quality Characteristics of Urban Runoff and Estimates of
      Annual Loads in The Tampa Bay Area, Florida 1975-80. USGS, 1981.

8.   "Area Wide Water Quality Management Plan". TBRPC, 1976.

9.   "Water Quality Management Plan Tampa Bay Basin".  TBRPC, 1973.

10.  "Waste Load Allocation". Vol. I - Hydraulic Model,  B.E. Ross, 1981.

11.  "Waste Load Allocation". Vol. Ill - Model Validation, B.E.  Ross,
      1981.

12.  "Circulation Patterns in Tampa Bay".  B.E. Ross,  1985.
                                 T-I-E-13

-------
       POLLUTION ABATEMENT IN THE RUHR RIVER BASIN,
                          WEST GERMANY
                   D.R. Albrecht and K.R.  Imhoff
1.  The Ruhr System
Nonpoint pollution abatement has been an integrated  part of the
comprehensive water management in the  Ruhr River basin  since
72 years.  Innovative approaches to water management in this area
developed over seven decades ago as increasing  pressure on a  very
limited supply of water by the massive  urban-industrial complex in
the Ruhr  Industrial Area  required  decisive action to resolve severe
water related problems. Special  legislation created two organizations
which were empowered, respectively, to care for water quantity and
water quality. A joint  directorship  of the two was formed  47 years  ago,
thus institutionalizing  the fact that management of water quantity and
water quality are inextricably intertwined.  The Ruhr River  water
management associations are characterized by compulsory membership
for all major  water users and  suppliers, i.e.  communities, industries
and water works. A  comprehensive water management in the  Ruhr
catchment is  all  the more  necessary because approximately one-third
of the average seasonal low flow is exported as drinking water to other
river  basins  in the industrialized area (Fig. 1).

Northrhine-Westphalia  is one of the most industrialized states in West
Germany.  In the industrial district - known under the name  "Ruhr District" -
water consumption and  the volume of wastewater are seven times  as high per
km2 as in the average of  West Germany.  The principal  source  of  water
supply in the district  is the Ruhr  River  which at present has to meet
70 % of the whole demand. The  Ruhr River has a drainage area of
4,488 km2. It flows at the southern border in parallel to the large  urban-
industrialized district from east to  west.  The long term average precipitation
varies within the river basin  from  800 to 1,400 mm per year. Due to seasonal
factors, the river flow  varies between a minimum of  4 m3/s  and a maximum
of 2,200 m3/s. The present permanent export rate of drinking  water
abstracted from  the  Ruhr basin is  12 m3/s.  Thus, in dry periods a

                              T-I-F-1

-------
    To supply cities and industries %
        annually 410 million m3
   are pumped out  of the Ruhr Valley
    320 Mill  m3 to the Emscher
      82 Mill  m3 to the Lippe
      6 Mill  m3 to the Wupper
  and 2 Mill  m3 to the Ems Basin
c±i Water Works
     Fig.  1:  Water export from the Ruhr  to adjacent river basins
considerable water deficit must be compensated for by  the  operation of a
system  of  reservoirs. In total  a volume of 471 million m3  of storage capacity
exists to augment low flow.

Besides  providing sufficient  water for  water supply the Ruhr River also
has the  task of carrying the wastewater of about 2 million  people  living
in its drainage area  and of various industries located  in  the basin. Thus,
effective wastewater  treatment  is necessary.  The Ruhr River Association
operates a system of 118 treatment works, 40 of  which are  vested with
tertiary  treatment by polishing lagoons.

One of  the first  tasks of the association was to assist the municipalities
to design and construct sewer  systems, in order to improve  their hygenic
conditions. At this stage the former  nonpoint pollution of communities
was changed into point  pollution for  the major rivers.  In many cases
smaller  tributaries of the Ruhr were integrated into the sewerage  system.
The following period  was characterized by the increasing control of point

                               T-I-F-2

-------
pollution  and increasing degrees of  wastewater treatment.  At
present, the major load of point pollution is controlled by  treatment
efficiencies  of more than 90  % in respect to BOD. In this situation  the re-
maining  nonpoint  pollution has  become the largest  residual fraction of
organic load to  rivers. Sewage and unpolluted water of creeks are
presently separated and many tributaries in the cities  are  now "renaturized".

2. Purification Lakes
According to the innovative  design of the  first director of the Ruhr  River
Association,  Dr. Karl  Imhoff,  a  system of  five river impoundments  has
been constructed in addition  to the system  of  treatment works to compensate
for residual pollution from treatment plants and nonpoint pollution  above
major water works  intakes [1],  (Fig.  2). Both concepts, polishing lagoons
and river impoundments provide facilities for sedimentation and extended
biological activities. River water polishing  by impounded lakes has  proven
to be a feasible measure to improve  river water quality and to control
nonpoint pollution. The always available large  amount of fresh water  has
compensated for the considerable oxygen demand of storm  water runoff at
any time. In contrast  to other German rivers,  a fish kill was never observed
in the impounded  Ruhr after heavy rainfalls.
                                                      Dortmund
                       0  A  8  12 16  20km
Impounded
Lake

Hengstey
Harkort
Kemnade
Balden ey
Kettwig
year of
impounding

1929
1930
1979
1932
1949
average
depth
(m )
1,90
2,30
2,40
3,25
2,60
surface

(km2)
1,70
1,38
1,25
2,60
0,55
capacity in
the year of
impounding
(mio. m3)
3,3
3,2
3,0
8,7
M
                                                     N
      Fig.  2:  River impoundments in  the  Ruhr valley
                             T-I-F-3

-------
 The selfpurification effect in river impoundments is  mainly provided
 by sedimentation. In Lake Hengstey  100,000  m3  of sediment have
 accumulated and in  Lake Harkort some 350,000 m3.  During 50 years
 the volume of Lake  Baldeney has diminished  by  1.5  x 106 m3.  Subsidence
 caused by coal  mining until  1970 is estimated to be  of the same order.
 Consequently,  some 3 x 106 m3  have  been accumulated,  resulting  in a
 sedimentation  rate of 60,000 m3  per year  [2]. This  rate depends  on the
 flow characteristic of the river and on: size and morphology of the
 impoundment.  For example,  in Lake Baldeney maximum sediment accumulation
 in terms of m3 per day  is observed at flows  of the order of three times
 long term average flow, despite the then  considerable turbulence.  At
 lower flows circumstances for sedimentation are more favourable, but the
 input of settleable solids is  then limited.  At flows equal  to ten times long
 term average, deposits start to  be mobilized, and only flows exceeding
 15 times average flow flush deposits  partly into downstream reaches.
 Table  1 indicates analytical data of the sediments in Ruhr lakes.
    Table 1:  Average composition of sediments  in the Ruhr  lakes
Parameter
Dry solids
volatile solids
Fe
Cu
Zn
Ni
Cr
Pb
Cd
POn
N total
Dimension
%
% of dry sol.
mg/kg dry sol.
mg/kg dry sol.
mg/kg dry sol.
mg/kg dry sol.
mg/kg dry sol.
mg/kg dry sol.
mg/kg dry sol.
mg/kg dry sol.
mg/kg dry sol.
Pur
Hengstey
39.4
12.3
69,200
1,240
4,030
261
440
450
18
14,400
3,750
ification La
Harkort
45.8
11.0
59,900
830
3,120
250
280
480
29
11,110
2,840
ke
•
Baldeney
45.6
14.0
50,100
730
3,540
312
400
525
36
17,700
6,820
A detailed  research program on the influence of the largest Ruhr  River
impoundment,  Lake Baldeney, on the river water quality was performed
from 1979 throughout 1982 [3]. Some  results will be described below:
                              T-I-F-4

-------
On average, 30,000 tons  of suspended solids are retained
annually in  Lake Baldeney, which is 20 % to  50 I of the input.

The average difference between  BOD5-input and output is
2,200 tons per year, being equivalent to a purification capacity
of 100,000 population equivalents.  20  % to 30 % of  the BOD5
input  is  retained.  Fig.  3 represents the decrease of BOD5-load
in dependency of  flow, distinguished  for winter and summer
half years.
               10   20      50    100    200     500  1000
                              Flow (m3/s)
      Fig. 3:   Impounded  Lake Baldeney: percentual decrease
               of  BODs-load in dependency  of flow
By evaluating 284 influent and effluent samples  a detailed
phosphorus balance  was performed. Fig. 4 indicates the
retention of the P-load in dependency  of flow.  At summer
low flows 22 %  or  70  tons of dissolved phosphorus are retained
by algal development and subsequent sedimentation of phyto-
plankton or grazing zooplankton.  During the winter half year
16  % or 40 tons of dissolved phosphorus are eliminated.  In
addition,  170  tons of phosphorus  is settled per  year with solid
particles.
                        T-I-F-5

-------
                 50
               100
            Flow (m3/s)
150
     Fig. 4:
Decrease of dissolved phosphorus in the
impounded  Lake Baldeney
Due to geochemical reasons and  because of a considerable
number of metal finishing factories in the  Ruhr basin, the
heavy  metal content of the sediments is  ten times as high as usual.
The average elimination rates for heavy  metals in Lake Baldeney
are as follows:
Element
 Fe   Pb   Cr   Cu   Cd   Zn    Ni
Elimination
rate %
 50   46    38    33    33    31   10
A separate investigation on origin and fate of heavy metals
in the Ruhr basin has indicated that a fraction  of  41  % of
the total heavy  metals in Ruhr water  is of geochemical origin,
i.e.  from  nonpoint  sources (Fig.  5) [4],
                        T-I-F-6

-------
                                         measured in one year without floods
                                            I nput     _ Output+ Retention
Percentage of the
catchment area
                             ;j Discharge at',';
                             '' river mouth \'<
                                  283 t/a!;
                                                               !; Retention in !;
                                                               ;>sediment and!'
                                                               :', subsoil    |.
                                                               iXiXoxo.1:1:0,
                                                               '/Exported with •
                                                               'drinking water!
                                                               '.over the
                                                               ; watershed
                                                                    50 t/a
                                                                 iX.ov.'ViV.':'
                                                              Sum: 483 t/a
Sum: 483t/a
H partly estimated
          Fig. 5:  Balance of heavy metal loads in  the
                   Ruhr River basin
       Bacterial  counts are reduced  in the  lake by agglomeration
       at particles and sedimentation to the order of 90 %.

       Nitrification  occurs with considerable lower rates in  the lake
       as in free flowing  reaches of the river.  Oxygen consumption
       by nitrification processes is therefore  not  important in the
       impounded reaches  of the Ruhr [5].

       A potential secondary pollution is created  in  the  impoundments
       by mass algal  development in summer.  Consequently, oxygen
       depletions may  result and  have to be compensated for by instream
       aeration [6].

In order to keep sufficient conditions for boating  and other recreational
uses a minimum depth between 2 and 2.5 m is required. Sediment deposits
in the Ruhr impoundments are observed annually  by echo  sounding
monitoring.  By hydraulic dredging  1.1  million m3  of sediments were
removed from Lake  Baldeney in 1983/84. The  total cost including
sedimentation lagoons and  land use was 11  million  DM. As  supernatant water
from the lagoons  flows back into the river  (Fig.  6)  its quality has been analyzed,
                               T-I-F-7

-------
   Essen-Bredeney
                                                boat harbour
                                                Heisingen
                     sediment removal
                     tills2.3m depth
  Essen-Werden
                     areas of sediment deposit
                     fixed sludge pipeline
                                                                ditches for
                                                                supernatant
                                                                water
                                                           Essen-Kupferdreh
                                      1 km
      Fig.  6:  Sediment removal from Lake Baldeney

The following results were obtained:

       BOD5                       8.4    mg/l
       COD                       23     mg/l
       NH4                        8     mg/l
       Fe, Mn                    <0.1    mg/l
       Cu, Zn, Ni, Cr,  Pb,  Cd   <0.01   mg/l

Half a year after finishing sludge dredging  wild  herbs, bushes, and
trees have grown already without any  human aid to a height of half
a meter on the  surface of the sediment  lagoons.  Some rare species  are
among them. According to  German legislation a detailed plan  for land-
scaping of the lagoons had to be  designed before starting the measure,
but it is doubtful if this will be performed in future, as a new probably
also valuable  "substitute" habitate develops.
3.  Stormwater Treatment
After completion of the system of biological wastewater treatment plants
and purification  lakes stormwater  of the mostly combined sewerage systems

                               T-I-F-8

-------
remained as the now major pollution to rivers. According to a basic study
for the German Federal  Government,  the load of organic pollution from storm
overflows without any treatment represents on average 12 % of the per capita
organic load of domestic wastewater,  which is shown as 0.12  pollution units
(pu) per capita in Fig.  7.

Although the accurate figure is strongly dependent on the  particular
circumstances of a drainage area,  i.e.  topography, size,  urban structure,
and others, 0.12 pu/capita give a general idea  of  the  relative order of
magnitude of the waste  load resulting from storm water.  Fig. 7 indicates
schematically the waste  load to rivers which can be reduced by 50 % if in
accordance to the German guideline A 128 [7] storm water  flow is retained
by storage and later is  discharged to the municipal treatment plants,  where
up to twice the dry weather flow  is treated  biologically.  The dimensioning
rules  of this guideline consider especially the storage  of the  highly polluted
first flushing  flood  and  ensure the transport of up to 90 % of the storm  water
flow  into the treatment  plant at critical  conditions  in the  receiving water by
calculating a critical storm water flow rate for any particular case.  In
dependency of the ratio between average low flow  of the  receiving water
and the total wastewater flow  including  the storm  water the critical .rainfall
rate  is  between 7 and 15  l/(s  • ha).  It is assumed  that herewith 75 %  of  the
annual organic load from storm water can  be removed.
waterwork
                             sewerage system
                             for 5,000-1 0,000
                             inhabitants
                          untreated
                          stormwater
                          0.12 pu/capita
                          wastewater
                          treatment plant
                                   stormwater retention and
                                         biological treatment
                                         0.03 pu/capita
                  0.05 pu/capita
      Fig.  7:   The new  storm water treatment concept results  in
               more than 50 % load reduction

                              T-I-F-9

-------
 The German Federal Law on  Wastewater Charges includes a charge for
 storm water discharges in  addition to the charge for  dry weather
 effluent of treatment  works.  According to State Water Laws the storm
 water charge can be  reduced or even cancelled if storm water retention
 tanks of appropriate  size are existent. Thus,  a double incentive is
 given to construct storm water  retention facilities from the financial and
 the water  pollution  control point of view.  A new  100  million DM program
 to fight this part of pollution has therefore been approved  for the Ruhr
 basin.

 Some of the stringent State requirements were  incentive to  the development
 of sophisticated models to determine the pollution load from storm overflows.
 It seems,  however,  that  this ambitious approach has  failed because of the
 most  complex interrelationships.  As already  the accurate determination  of
 the discharged quantities from the particular systems is a very difficult
 approach,  the Ruhrverband has decided to develop  a practical hydraulic
 model.  Recent discussions indicate the tendency of changing sophisticated
 State regulations on  storm water charges to more simple calculations.

 4.  Particular  Strategies
 To avoid detrimental effects for the  Ruhr  water works from accidences
 on highways a detailed field  study has been performed  in the early  sixties [8].
 It resulted in design  recommendations for  retention tanks to be constructed
 at highways in the Ruhr valley.  The effluent from highways thus passes
 in any case a tank where oil  and other hazardous  substances can be retained and
 abstracted before reaching the Ruhr  River.  The structures are simple and
 not equipped  with automatical devices.  They are designed for a storm water
 flow  rate of 100 l/(s • ha) and a  retention time  of  10  minutes. The hydraulic
 surface loading is 12  m/h.  The tanks  are operated by highway maintenance
 services and observed by personnel  of the Ruhr River  Association.

 A  new French  investigation has  indicated that  from double laned  roads
 230 to tOO  kg COD/(km • a) result. For lead 0.9 to 1.3 kg/(km •  a) were
 observed.  The pollution load  from roads is  comparable to  that of domestic
 wastewater related to  the surface area. Only few  heavy rainfalls  may flush
 30 % of the annual load into the environment [9].

Other strategies have to  be applied in the case of major agricultural districts.

                              T-I-F-10

-------
It is an aim to educate farmers to save artificial  fertilizers,  to apply
manure properly,  and to avoid  erosion for example by appropriate
ploughing in hilly areas, especially in the catchments of  reservoirs,
which are even more sensitive to eutrophication as river  impoundments.

5.  Summary and  Conclusions
In the past,  nonpoint pollution  has been transferred into point pollution
by the construction of sewerage systems. Consequently,  this point
pollution was more and more decreased over  the  decades. The greater
the progress in this area of water pollution control  the larger the
percentage of residual  nonpoint  pollution  as a part of the total river  load.
For the Ruhr basin a  system of river impoundments has  proven  to  be a
feasible measure for river water polishing and  nonpoint pollution reduction.
As impoundments on the other hand may create disadvantages by forming
secondary pollution from eutrophication processes, the task  remains to
find  an optimum between impounded and free flowing reaches in  a river
system.

A new storm water treatment concept will lead  to further reduction of
residual pollution  to rivers.
                              T-I-F-11

-------
References:
[13
Imhoff, K,
Die Stauseen im Ruhrtal.  Wasserkraft und
Wasserwirtschaft, Vol. 26 (1931),  pp 85-89.
[2]
Imhoff, K.R,
The Design  and Operation of the Purification
Lakes in the Ruhr Valley. Conference Paper No.8,
Annual Conference of  The Institute of Water
Pollution Control, September 5  -  8,  1983, England,
Edwards  The Printers  Ltd., Coventry, England.
[3]      Koppe, P.
         et al.:
                Untersuchungen iiber den  Einfluft der Stau-
                haltung eines Flusses auf die Wasserbeschaffen-
                heit.  Forschungsbericht BMFT -  FB 02 WT  854,
                December 1983.
[4]      Imhoff, K.R.,    Heavy Metals in the Ruhr River and their Budget
         Koppe, P. and   in  the Catchment  Area.  Progress in Water
         Dietz, F.:       Technology, Vol.  12,  pp. 735 - 749.


[5]      Klopp,  R. and   Abhangigkeit der  Ammoniumkonzentration
         Koppe, P.:      des Ruhrwassers  von  den hydrologischen
                         und meteorologischen  Bedingungen. Forum
                         Stadtehygiene (in press).
[6]



[7]
Imhoff,  K.R.    Instream Aeration  in the Ruhr River.  Progress
and Albrecht,D. :in Water Technology, 1978,  Vol.  10, p. 277.
Abwassertech-
nische Vereini-
gung  (ATV):
Arbeitsblatt A 128: Richtlinien  fur die Bemessung
und Gestaltung von Regenentlastungen in Misch-
wasserkanalen, 1977.
[8]
Imhoff,  K.R.:
Olabscheidebecken fur FernstraBen im  Ruhrtal.
Gas- und Wasserfach, Vol. 108 (1967),  pp.
43 - 45.
[9]     Balades,  J.D.    Chronic Pollution of Intercity Motorway  Runoff
        et al.:           Waters. Wat. Sci. Tech., Vol 17 (1984), pp. 1165-1174,
                             T-I-F-12

-------
                 USING LINEAR OPTIMIZATION FOR URBAN NONPOINT
                   SOURCE POLLUTION WATER QUALITY MANAGEMENT

                               J. G. Garland III
             Master of Science Civil Engineering Degree Candidate
                                 F. S. Tirsch
                   Assistant Professor of Civil Engineering
                                C. A. Markowski
    Assistant Professor of Management Information Systems/Decision Science
                                  C. C. Churn
                   Assistant Professor of Civil Engineering
                           Old Dominion University
                            Norfolk, Virginia, USA
                                   ABSTRACT

Urban nonpoint source pollution is a principal cause of water quality
problems in the United States.  This pollution can be reduced by implementing
management practices.  One factor influencing management practice selection
is the spending level at which diminishing marginal returns occur.

This study demonstrates diminishing marginal returns for spending to reduce
nonpoint source pollution in Hampton, Virginia. A linear optimization program
selected the best combination of management practices to minimize the cost of
nonpoint source pollution control. Management practice options were grass
swale roadways, porous pavers, detention basins, ponds and fertilizer
management.  Pollutants evaluated were total nitrogen, total phosphorus,
suspended solids, fecal coliforms and five-day biochemical oxygen demand.

Additionally, a sensitivity analysis was conducted for total phosphorus to
determine how sensitive overall nonpoint source pollution control costs were
to key management practice data used in the linear optimization problem.
Part of the sensitivity analysis examined the impact of using more site-
specific cost data.

Significant diminishing marginal returns occurred above 40.0 percent removal
for nitrogen and 50.0 percent removal for phosphorus, suspended solids and
fecal coliforms. Overall pollution reduction cost was most sensitive to the
pollution removal data for grass swale roadways and ponds.  Using more site
specific cost data in the sensitivity analysis lowered the spending level of
diminishing marginal returns.

Keywords;  Urban runoff pollution, nonpoint source pollution, linear
           optimization, water quality models
                                   T-I-G-1

-------
 INTRODUCTION

         In a recent United States Environmental Protection Agency report to
 Congress on nonpoint source (NFS)  pollution,  six of ten Agency regions
 specified NFS pollution as the principal cause of water quality problems
 (USEPA,1984). This report also listed urban NFS pollution as one of the most
 cxxnmonly cited problems.   Urban NFS pollution presents a special challenge to
 water quality managers,  because the many available pollution removal and
 treatment options  {management  practices) have cost and pollution removal
 characteristics which vary with land-use.  This complicates selecting optimum
 treatment practices and spending  levels. The  focus of this study is on a
 factor which influences decisions on spending levels — diminishing marginal
 returns.  Diminishing marginal  returns occur at the point where less pollution
 removal  occurs per dollar spent.
         This study's objectives were to find  out if diminishing marginal
 returns  would occur for spending  on management practices in Hampton, Virginia
 and to evaluate how sensitive  the point of diminishing marginal returns was
 to the management  practice and land-use data  in the problem.  The sensitivity
 of the results to  the data was evaluated to determine the importance of data
 accuracy.   This information is essential to evaluate the reliability of the
 results  and the effect of data assumptions.

 METHODOLOGY

         To achieve the diminishing marginal return objective, a linear
 optimization problem was  created  for each pollutant.  These problems included
 the cost and pollution removal characteristics of every management practice
 in every land-use  area. The data  for the allowable pollution was set at the
 desired  level  and  the problem was  solved.  The problem solution contained the
 overall  cost of achieving the pollution level and a cost breakout, i.e., a
 list of  the management practice area recommended in each land-use that
 contributed to the overall  cost.   The allowable pollution was made more
 stringent  in 10.0% increments from 10.0% to 60.0% pollution reduction.  This
 systematic adjustment provided cost and management practice data for a range
 of pollution  removal  which was used to analyze diminishing marginal returns.
 These results are  reported in a graphic presentation of percent pollution
 removal and cost.   The results are also presented in the form of cost per
 percentage pollution  removal.
        The sensitivity of the problem to the data was evaluated using a
 single pollutant.   The approach was to change one data element and solve the
 linear optimization problem again at each of the pollution removal
 increments. The difference between the new (modified data) cost and the
 previous cost was  the result of the data element change.   These differences
 are reported  for 50.0% pollution removal in the results.  Another form of
 sensitivity evaluated was the sensitivity of the point of diminishing
marginal returns to the level of detail  of the land-use data.  In this
analysis a  comparison was made between two linear optimization problems for
the same city subbasin.  A consolidated problem divided the subbasin into
 8 areas and a parcel  problem divided the subbasin into 33 smaller areas.   The
affect on the level of spending at which diminishing marginal  returns occurs
is shown.
        Linear optimization has been used as a water quality management tool
in the past. Anderson and Day (1968) minimized the regional operating costs

                                   T-I-G-2

-------
of conventional point source waste treatment with linear optimization.
Revelle et al. (1968) compared linear programming results to dynamic
programming to show both techniques yielded essentially the same result.
Smith and Morris (1969) optimized water-use goals and management options
using point sources as the modified management practice and dissolved oxygen
as the water quality constraint.  Watershed Handbook (USEPA,1981) illustrated
how to use linear programming to minimize the cost of reducing point and NPS
pollution flowing into a water body.

Linear optimization model

        The linear optimization approach was to minimize the cost of NPS
pollution control subject to land-use and water quality constraints.  The
land-use constraints required the sum of all the management practices and the
land where there was no recommended management practice to equal the total
area in each land-use category.  Water quality constraints set limits on the
NPS pollution permitted in each land-use area.  Without water quality
limitations, no management practices were necessary and all the urban NPS
pollution reached the receiving water.  Making the water quality constraints
more stringent, i.e., requiring larger reductions in pollutants, caused the
program to select a land-use area and management practice for implementation
to reduce pollution.
        The management practice options were grass swale roadways (GSR),
porous pavers (PP) — a composite management practice including porous
pavement and modular pavement, detention basins (DB), ponds (PO) and
fertilizer management education programs (EM).  The land-use categories were
commercial strip, heavy industry, light industry, high density residential,
multi-family residential, central business district, low density residential
and open land.
        The objective function (Equation 1) minimized the total cost (C) of
implementing area in each of the management practice options (Xij) with
individual costs (Cij).  The management practice options were represented by
subscript j, and the land-use categories were represented by subscript i.

                              m    n
Equation 1.     Minimize C =  Y"   V"   C. .X.  .
                             i=L  T^l   1-'  -1
        The land category constraints (Equation 2) required the sum of the
management practices in each land-use area i to equal the area in i (Ai).
Equation 2.                 )    X.. = A.    for i = l,2,...,m
                           3=0   ^    1

        Each hectare (ha) of management practice Xij was assigned a fixed
amount of NPS pollution that would be generated yearly (Lij).  The sum of
LijXij for each land-use area was a water quality constraint equation
(Equation 3).  The right side of the water quality constraint (w.q.)  was set
at fixed increments.  Non-negativity constraints required Xij be greater than
zero (X. . - 0) because negative area was not feasible. A detailed formulation
of the problem is described by Garland (1985).

                                  T-I-G-3

-------
                             n
 Equations.                 T   L. .X..-w.q.   for i = l,2,...,m
         The result of  these equations was a  linear programming problem with
 40 variables and 16 constraints  for each pollutant.

 Management practice costs

         The assigned cost of each management practice  ($/ha) was based on the
 management practice cost and the different characteristics of each land-use
 area (Table 1).   Land-use influenced cost because most management practices
 could only be applied  to a  fraction of the total land-use area.  Other land-
 use factors which affected  cost  were length  of roadway, percentage of
 impervious area  and dwelling unit density.   These factors influenced GSR, PP
 and FM costs respectively.
         GSR for  example, can only be applied to roadways in certain land-use
 areas because they require  wider right-of-ways and are not feasible on main
 highways with a  high number of access roads.  This led to the assumption that
 GSR were not applicable  in  commercial strip  and central business districts,
 and they were only feasible on 20%  of industrial land. GSR were not
 considered a management  practice option  for  open land, because limited
 roadways would only provide very limited overall pollution reduction.  The
 cost ($/meter) for GSR was  obtained from the Hampton Roads Water Quality
 Management Agency (1978).   The length of roadway in each hectare (m/ha)'for
 each land-use was multiplied by  the cost ($/m) and feasible area (e.g., 0.20
 for industrial land) to  calculate the GSR cost per hectare.

 Management practice effectiveness

        Both the  rate of pollution generation and the ability of the
 management practice to remove the pollutant  for each land-use area were
 needed to  create the water  quality  constraint.  Pollution loading data was
 obtained from a  1978 STORM  (Storage Treatment Overflow Runoff Model)  analysis
 of  the Hampton area conducted by the Hampton Roads Water Quality Agency
 (1978).
        Management practice effectiveness for this linear programming format
 was expressed in units of kg pollution loading/ha.  Literature values
 provided figures for how effective each management practice was,  but the
 literature did not detail how extensively the management practice could be
 applied in each  land-use hectare.  For example, porous pavers (PP)  are 100%
 effective  at  removing surface runoff in a parking lot by infiltration if
 properly designed. However,  PP are not 100% effective at removing NPS
pollution on  a hectare of any land-use category because they can not be
applied to the entire area. The PP treatable percentage of each land-use
 category had  to be estimated based on the impervious area of the land-use
categories.   The effectiveness data is displayed in Table 2.   For example,  in
Table  2 the pond  (PO) management practice in commercial strip land-use will
remove  50% of the total P,  SS and FC;  40% of the total N and 20%  of the BODj.
        Both  PO and DB were assumed to be applicable to all of the area in
every  land category.  Pollution removal effectiveness was based on the
removal rate of three PO systems in the local area  examined by Anderson and

                                  T-l-G-4

-------
                                   TABLE 1

       MANAGEMENT PRACTICE COST SUMMARY IN 1984 DOLLARS PER HECTARE


Land-Use

Commercial
Strip
Heavy
Industry
Light
Industry
High Density
Residential
Multi-Family
Residential
Low density 5
Residential
Central Bus.
District
Open Land 2
SOURCES:
a
Areaa
(ha)
811

333

249

700

109

,760

67

,663
Hampton
Management Practice
($)
Grass Porous
Swales Pavers
44,702

911 38,935

911 27,399

911 38,935

319 12,256

190

47,587

• • • • • •
h
Pond0
11,577

4,787

4,532

22,519

22,373

15,875

75,582

1,820
Roads Water Quality Management
h
Basin
36,964

15,284

14,472

71,894

71,427

50,679

241,312

5,812
Plan (4)
Fertilizer
Management
• • •

• • •

• • •

141

469

32

• • •

• • •
and
Tourbier and Westmacott (10).
        NOTE: Costs include equipment, material, installation, capital and
maintenance factors. For a complete discussion of cost assumptions see
Garland (3).  All cost data converted to January 1984 dollars using the
Department of Commerce Composite Cost Index. Infeasible combinations "...".
        aThe city-wide land-use data was from Hampton, Virginia. Hampton is
a 10,753, ha city located on a peninsula.                  ^
         Volume.,determined by inflow to area ratio of 53 m /inflow ha for
ponds and 170 m /inflow for detention basins. Price increased by a real
estate factor in each land-use category.
                                  T-I-G-5

-------
                               TABLE 2

        MANAGEMENT PRACTICE POLLUTION REMOVAL EFFECTIVENESS
(Percent)
.Land-
Use Poll.
Comm.
Strip

Heavy
Ind.


Light
Ind.


High
Density
Resid.


Total P
Total N
BOD.
ss-
FC
Total P
Total N
BOD.
SS
FC
Total P
Total N
BOD.
SS
FC
Total P
Total N
BOD.
SS
FC
Mgt.
GS PP

• *

10
13
15
15
15
10
13
15
15
15
50
65
73
77
77
62
62
62
62
62
54
54
54
54
54
38
38
38
38
38
54
54
54
54
54
Practice3
PO DB FM
50
40
20
50
50
50
40
20
50
50
50
40
20
50
50
50
40
20
50
50
80 ..
57 ..
63 ..
70 ..
70 ..
80 ..
57 ..
63 ..
70 ..
70 ..
80 ..
57 ..
63 ..
70 ..
70 ..
80 10
57 10
63 ..
70 ..
70 ..
Land-
Use
Multi-
Family
Resid.

Low
Density
Resid.


Cent.
Bus.
Dist.

Open
Land


Poll.
Total P
Total N
BOD
SS
FC
Total P
Total N
BOD
SS b
FC
Total P
Total N
BOD
SS
FC
Total P
Total N
BOD
SS
FC
Mgt.
GS PP
50
65
73
77
77
50
65
73
77
77



* •
* *
• *
. .
* •
17
17
17
17
17

• •
• *
66
66
66
66
66
* *
• *

Practice
PO DB FM
50 80
40 57
20 63
50 70
50 70
50 80
40 57
20 63
50 70
50 70
50 80
40 57
20 63
50 70
50 70
50 80
40 57
20 63
50 70
50 70
10
10
* •
10
10
. .
* *
• *


• •
• *

        SOURCE:  Anderson et al. (1982).

        NOTE: For a complete discussion of effectiveness
assumptions see Garland (1985).  Infeasible combinations "..".

         GS — grass roadway swales, PP — porous pavers,  PO —
ponds, DB — detention basins and FM — fertilizer management.
                             T-I-G-6

-------
coworkers  (1982).  These researchers showed PO needed a 53 m /inflow
hectare ratio in order to achieve a 50% SS reduction. A 50% removaKrate for
SS for DB corresponded to a storage volume to inflow ratio of 170 m /inflow
hectare.  FC were assumed to behave as particles, because most bacteria are
attached to suspended solids.
        FM programs were assumed to be applicable only in the residential
areas since this management practice is directed toward the reduction of P
and N pollution caused by excessive fertilizer application practices.  The
effectiveness assigned to this management practice was a 10.0% reduction in
the total N and total P runoff.

RESULTS

       The plots of the city-wide data for cost and percentage of pollution
removal are shown in Figures 1, 2 and 3. A visual inspection of the plotted
data revealed linear segments in the relationships between cost and
percentage pollution removal.  For BODj., the linear range extended from 10.0%
to 60.0%. Total P, SS and FC appeared linear between 10.0% and 50.0%, and
linear at a new slope between 50.0% and 60.0%.  Total N appeared linear from
10.0% to 40.0% and then linear at a new slope above 40.0%.
        A best fit line was developed with linear regression and the slope
used to calculate the cost per percentage pollution removed ($/!%). This data
is shown in Table 3.  There was a perfect correlation of the data (one) when
the points being measured represented costs from the same set of recommended
management practices. When the correlation dropped slightly, e.g., to 0.9999,
this indicated a region where a new management practice was added to the
treatment set or one of the management practices was no longer recommended.
        No feasible solution was possible above 57.0% removal for total N
because none of the management practice options would remove more than 57.0%
of the pollution load for industrial and open land.
        BOD- was the most expensive pollutant to remove in the 10.0% to 50.0%
removal range, costing about 1.5 times the removal cost of the other
pollutants. BOD5 cost more to remove due to the low effectiveness of PO.
Because PO were only capable of removing 20.0% of the initial BOD<-, the DB
management practice had to be used in some land categories for removal rates
starting at 20.0%. DB cost between three and four times as much per hectare
as PO creating the high cost for BOD^ pollution removal.
        Above 40.0%, total N removal costs also increased due to the switch
in recommended management practices to DB instead of PO and GSR in industrial
areas. For SS, FC and total P, DB were not selected until above 50.0%
pollution removal.
        The extremely high cost of total P removal above 50.0% was caused by
the requirement to implement DB in residential areas while all the other
pollutants were still removed by GSR management practice.  DB cost from 70 to
270 times the cost of GSR per hectare in residential areas.
        As Figure 4 shows, linear segments also occurred in the subbasin
analysis by parcels and by consolidated land-use areas. The first segment was
from 10.0% to 50.0% pollution removal.  In this range, the parcel cost per
percentage pollution removed was $47,825/1% removed and the consolidated cost
was $48,679/1% removed. The second segment was 50.0% to 60.0% pollution
removal, where parcel analysis cost was $2,097,985/1% removed and the
consolidated cost was $1,231,235/1% removed.
        The subbasin analysis showed that more detailed examination lowered

                                  T-i-G-7

-------
M

O
I
oo
                 60.0-

                 500-

                 40.0-
        Percent
        Pollution  30.0-
        Reduction
                 20.0-

                 10.0-

                   0.0
                     0
10
                        BODc
                      *SS "
20
     30

10~5  Dollars
      50
      60
        FIGURE 1. CITY-WIDE POLLUTION REDUCTION COSTS FOR BOD5
                 AND SS IN 1984 DOLLARS.
                                                  60.0-1

                                                  50.0-

                                                  40.0-
                                        Percent
                                        Pollution  30.0-
                                        Reduction
                                                  20.0-

                                                  10.0-

                                                   0.0
                           0
30
60
    ~90~
10~6  Dollars
120
150
180
                                               FIGURE 2. CITY-WIDE POLLUTION REDUCTION COSTS FOR TOTAL
                                                        P IN 1984 DOLLARS.
                  60.0-

                  50.0-

                  40.0-
        Percent
        Pollution  30.0-
        Reduction
                  200-

                  10.0-

                   0.0
                            10
                      » Total N
                        FC
       20
       30

  10~6 Dollars
40
50
60
         FIGURE 3  CITY-WIDE POLLUTION REDUCTION COSTS FOR FC
                  AND TOTAL N IN 1984 DOLLARS.
                                                  60.0-j

                                                  50.0-

                                                  40.0-
                                        Percent
                                        Pollution  30.0-
                                        Reduction
                                                  20.0

                                                  10

                                                   0.0
                                                                  Total  P Subbasin
                                                                  • Consolidated
                                                                  A Parcels
                                                                                                            12
                                                                                       16
                                                                         20
                                                                         24
                                                                                                       10~6 Dollars
                                               FIGURE 4. SUBBASIN AND PARCEL ANALYSIS POLLUTION
                                                        REDUCTION COSTS FOR TOTAL P IN 1984 DOLLARS

-------
                             TABLE 3

                     CITY-WIDE DATA ANALYSIS
             Pollutant
            Pollution
            Removal
            Range
Cost/Percent
Pollution
Removed
(1984 Dollars)
BOD,-
SS

PC

Total N

Total P

10-60
10-50
50-60
10-50
50-60
10-40
40-60
10-50
50-60
992,603
427,460
1,996,428
429,938
2,006,944
526,510
2,364,271
592,273
12,735,077
                              TABLE 4
            SENSITIVITY ANALYSIS SUMMARY FOR 50% TOTAL P
                 REMOVAL AND CITY-WIDE LAND-USE DATA
Mgt. Practice/
Variable       Change to all Land-use Areas
                                        Total Cost
                                        % Increase
Grass Swale
  Roadways
Ponds
Ponds
Open Land

Porous Pavers
Grass Swale
  Roadways
Open Land

Fertilizer
  Management
Lower effectiveness 10%

Lower effectiveness 10%
Increase cost 10%
Raise area 20% and decrease
other areas fractionally
Increase cost 10%
Increase cost 10%

Raise area 10% and decrease
other areas fractionally
Raise cost 10.0%
Lower effectiveness 10.0%
                 154.38

                  19.56
                  13.57
                   1.45

                   0.86
                   0.62

                   0.53

                   0.00
                   0.00
                             T-I-G-9

-------
 costs  at lew pollution removal  levels and raised costs above 50% pollution
 removal.  This effect was caused by using 33 different prices (one for each
 parcel for DO and DB)  rather than 8 different prices (one average figure for
 PO and DB for each  land-use category).  Below 50.0% pollution removal in the
 parcel analysis,  some  PO and DB parcel prices were inexpensive enough to
 cause  their  selection  while they were not selected in the consolidated
 problem which used  the average  figures.  Above 50.0% pollution removal, all
 of the inexpensive  PO  and DB parcel area had been used.  Additionally, there
 were no other management practice options which could achieve the desired
 pollution removal.  The result was large overall cost increases in the parcel
 problem because more expensive  PO and DB area was recommended.
        The  parcel  approach had the additional advantage of providing more
 information  on where to implement the recommended management practices.  For
 example,  the consolidated problem recommended 9.9 ha of GSR in low density
 residential  area, while the parcel problem specified GSR in low density
 residential  area  in the following parcels: 0.0 ha in parcel A, 3.2 ha in
 parcel B,  0.8 ha  in parcel C, 5.6 ha in parcel D, 0.3 ha in parcel E.  The
 disadvantage of the parcel problem was the problem size — 165 possible
 management practice and land-use combinations and 66 constraints.
        The  cost  sensitivity analysis results are shown in Table 4.  At the
 total  P removal levels examined, the greatest overall cost change was caused
 by changing  PO cost. FM was not selected at 50.0% pollution removal, so the
 change in FM cost had  no impact on overall cost. Below 50.0% pollution
 removal,  a 10.0%  increase in FM cost created a change in overall cost of
 approximately 1%, e.g., at 10.0% pollution removal a 0.65% overall cost
 change occurred as  the result of a 10.0% increase in FM cost.  Increasing PO
 costs  had a  greater effect than other cost sensitivity changes because the PO
 fraction  of  the overall cost was greater.  For example, the area recommended
 for PO in all the land-use categories (4,106 ha) accounts for 62.74% of total
 costs  at  50.0% pollution removal compared to 6.20% of the overall cost for
 the GSR area (6,444 ha).
        Decreasing  management practice effectiveness caused an increase in
 pollution removal costs due to the increase in management practice area which
 had to be implemented  to achieve the desired pollution reduction.  In some
 cases  the decrease  in management practice effectiveness made it necessary to
 switch practices.   The increases in overall cost caused by the drop in the
management practice effectiveness of PO and GSR (19.56% and 154.38%
 respectively) at  the 50.0% pollution removal level was caused by a switch in
management practices. For example, at the 50.0% pollution removal level PO
 capable of removing 50.0% of the pollution could be implemented on all the
 land-use  area and achieve the desired pollution reduction.  When PO
effectiveness is  reduced to 45.0%, another more effective and more expensive
management practice (DB) must be selected in order to achieve 50.0% pollution
removal.  When DB was  selected to augment PO or GSR at the 50.0% pollution
 level  the cost increased greatly for each percentage of pollution removal.
        An additional  sensitivity analysis raised the effectiveness of GSR in
residential areas to 80.0% for total P.   The primary reason for this analysis
was to simulate the lack of diminishing marginal returns demonstrated by BOD5
with another pollutant.  One of the main differences between BOD^ and the
other pollutants was GSR removal effectiveness.  GSR are 73.0% effective for
removing BOD^ in residential areas.  Improving the GSR effectiveness to 80.0%
for total P eliminated the need to recommend any other management practices
in the residential  land categories.  This switch to one management practice

                                  T-I-G-10

-------
for the residential areas did cause a dampening of diminishing marginal
returns between 40.0% and 50.0% pollution removal.  The results were similar
to those shown for BOD^.
        The fraction or the open land land-use area was examined to see how
sensitive the results were to errors in land-use data collection caused by
using zoning data. An open land area data error would occur if a 10 ha area
zoned commercial strip was in fact 8 ha of commercial strip and 2 ha of open
land.
        Increasing open land area increased removal cost in the 10.0% to
50.0% pollution removal range and decreased removal cost above 50.0%. This
occurred because the average cost of treatment of open land below 50.0%
pollution removal exeeded the average cost of treatment in other land-use
areas.  However, above 50.0% removal DB and PO, the recommended practices on
open land, were less expensive than the management practices recommended on
other land-use areas.

CONCLUSIONS

        Significant diminishing marginal returns did occur for all pollutants
except BOD5 at spending levels between $20 and $28 million.  This spending
level is too high to effect Hampton's expected NFS pollution control program,
because it represents approximately 25% of the city's total expenditures.
        Pollution removal costs were the sum of the costs of the management
practices in each land-use category.  As additional acres of a set of
management practices were added to meet the more rigid water quality
criteria, a linear relationship was formed between cost and pollution
reduction. Diminishing marginal returns occurred when the cost of removing an
additional percentage of pollution increased.
        A new set of management practices was selected when the old set would
no longer achieve the desired amount of pollution removal.  This occurred at
40.0% pollution removal for total nitrogen and 50.0% removal for total
phosphorus, suspended solids and fecal coliforms.  New management practices
were introduced causing a change in the cost and pollution removal
relationship.  Each new set of management practices was capable of removing
less pollution per dollar spent than the previous management practice set
thereby creating diminishing marginal returns.
        The five-day biochemical oxygen demand cost per percentage removal
remained approximately the same from 10.0% to 60.0% pollution removal levels.
A sensitivity analysis showed this situation was caused primarily by the
assumptions making grass swale roadways inexpensive and highly efficient at
removing biochemical oxygen demand through 73.0% removal levels.
        In general, pollution removal effectiveness was the most sensitive
input variable, then management practice cost and then the percentage of open
land.  The point of diminishing marginal returns was also affected by the
level of detail of the analysis. The parcel analysis showed more site
specific management practice costs reduced the spending level of diminishing
marginal returns. By using more detailed cost data for ponds and detention
basins, this analysis showed the spending level needed to encounter
diminishing marginal returns for total phosphorus in the subbasin decreased
by $201,436.
        While this study focused on spending levels for treating urban
nonpoint source pollution, linear programming is a potentially valuable tool
for selecting the optimum combination of acreage for application in each

                                  T-I-G-11

-------
management practice option. The optimum set of management practices under the
assumptions of the study were similar for all pollutants. Ponds and detention
basins were recommended for commercial  strip.  Porous pavers were recarmended
for central business districts.  For industrial areas, grass swale roadways
were recommended below 10.0% pollution  removal. Between 10.0% and 50.0%
pollution removal, a combination of ponds and grass swale roadways was
recommended.  Where five-day biochemical oxygen demand was the pollutant,
detention basins were  recommended instead of ponds in the 10.0% to 50.0%
pollution removal range to compensate for poor biochemical oxygen demand
removal by ponds. A combination of ponds and detention basins was recommended
for all pollutants above 50.0% pollution reduction. The management practice
recommendation below 50.0% removal for  residential areas was grass swale  .
roadways. Above 50.0%  pollution reduction, a combination of detention basins
and grass swale roadways was recommended. Open land treatment was by ponds
below 50.0% pollution  removal and a combination of ponds and detention basins
above 50.0% pollution  removal.

REFERENCES

Anderson, G. F., Neilson, B.J.; and Campbell, D.H. (1982). Management Practice
Evaluation for Urban Areas in Hampton Roads Vicinity. Gloucester Point,
Virginia: Virginia Institute of Marine  Science.
Anderson, M. W. and Day, H.J. (1968). Regional management of water quality —
a systems approach. J. Water Poll. Cont. Fed.,40, 1679-1687.
Garland, J. G. (1985). "Selecting optimum urban nonpoint source pollution
management practices using linear programming computer modeling." M.S. 'thesis,
Old Dominion University, Virginia.
Hampton Roads Water Quality Agency (1978). Hampton Roads Water Quality
Management Plan. Virginia Beach, Virginia.
Northern Virginia Planning District Commission (1979). Guidebook for Screening
Urban Nonpoint Pollution Management Strategies. Final report prepared for
Metropolitan Washington Council of Governments, Wash., D.C..
Revelle, C. S. (1968). Linear programming applied to water quality management.
Water Resource Research,4, 1-9.
Smith, E. T. and Morris, A. R. (1969). Systems analysis for optimal water
quality manaaement. J. Water Pol1. Cont. Fed..,41, 1635-1646.
Tourbier, J. T. and Westmacott, R. (1981). Water Resources Protection
Technology; A Handbook of Measures to Protect Water Resources in Land
Development. Urban Land Institute, Wash., D.C..
U.S. Environmental Protection Agency (1984). Report to Congress; Nonpoint
Source Pollution in the U.S.. U.S. Government Printing Office,  436-672/879.
U.S. Environmental Protection Agency (1981). Watershed Handbook
EPA-905/9-84-002.
                                  T-I-G-12

-------
                  DESIGN OF URBAN DETENTION BASINS FOR
                              J Bryan Ellis
                     Urban Pollution Research Centre
                          Middlesex Polytechnic
                      Queens,. Enfield, EN3 ASF, UK.
                                ABSTRACT

Engineering effort has been traditionally placed on improvements in the
hydraulic design and management of urban storm drainage detention
facilities with quality control considerations of stormflow performance
receiving relatively little attention.  Nevertheless/ the pollutional
impact of such routed flood discharges on receiving streams in urban
catchments can be significant.  The paper examines pollutant
settleability and removal efficiencies of flood detention basins and
considers the effects of particle size, pollutant form and timing on trap
efficiency.

It is shown that timing and partitioning of pollutants during stormflow
events are not amenable to the quality control function of detention
basins as best management engineering practice.  The significance of
particle size and settling velocities on pollutant removal within storage
basins are demonstrated through sedimentation column tests and
application of trap efficiency formulae.  Efficiencies of between 18 to
62% were found for various pollutants although these values were
considerably reduced if adjustments are made to allow for turbulent flow
routing through the basin.
Ke^words^  Urban runoff pollution, pollution retention, trap efficiency,
particle size and contaminant controls, detention basins.
                                  T-I.-H-1

-------
NOMENCLATURE

V     =   particle setting velocity

L.    =   basin  length
 D

v     =   mean flow velocity  in basin

y     =   basin  depth

n     =   manning roughness coefficient for basin floor

g     =   gravitational constant

v     =   shear  velocity
n    =   efficiency function
INTRODUCTION

It is well recognised that urbanisation causes permanent changes in the
storm runoff regime and  leads to increases in both the magnitude and
frequency of peak flows discharged to receiving streams.  It is equally
clear that urban developments lead to a marked deterioration in the water
quality of these discharged stormflows despite the enhanced volume
available for dilution.   (Ellis, 1982).

A common low-cost engineering means of limiting the effects of
development on the hydrological response of a catchment and of
maintaining the predevelopment downstream river flow regime, is to
provide temporary storm storage facilities within the urban catchment.
In effect, a detention pond represents an attempt to replace the natural
storage capacity lost through urban development, although since the
latter was distributed throughout the catchment and the former is located
at a single site, the comparison is not totally appropriate.

The design of such detention basins has been conventionally based on
hydrological and hydraulic criteria with their size and capacity
predetermined from selected peak flow rates for a specified recurrence
probability.  The design methods provided by Davis (1963) or Hall and
Hockin (1985) allow the routed peak storm inflows to be delayed and
attentuated through a restricted outlet or overflow capacity.
Unfortunately, the design of what Whipple (1979) has referred to as
dual-purpose detention basins has to resolve the conflicting requirements
of water quantity and quality.  There is a limited amount of post-project
appraisal work available to evaluate the hydraulic design performance of
these control structures (ASCE, 1982) but little consideration has been
given to their trap efficiency in water quality terms.  It is not known
if the designated storm for any particular basin serves to retain the
pollutant inflow loadings through sedimentation or if the release time is
sufficient to regulate and minimise receiving stream quality impacts.
The concept of a "settleable design storm" expressed in terms of specific
pollutant parameters for flood detention basins remains to be formulated
and tested.

POLLUTANT RETENTION IN FLOOD DETENTION BASINS

Whilst short term retardation and attenuation of urban flood flows by

                                   T-I-H-2

-------
storm detention basins may represent  a  best  management practice in
hydraulic terms, it does not necessarily  provide  the most effective means
of retaining or reducing associated pollutant  flows.  Urban detention
basins are normally designed to balance short-duration, high intensity
storm events since these are the  cause  of the  most  frequent downstream
flood activity in the receiving stream.  These basins release the
detained peak inflows over a period of  a  few hours  at most, although
gated and throttled outlets can retain  water for  rather longer periods.
However, it can be argued that it is  the  aggregate  effect of storms of a
return period less than that of the design storm  for flood control that
is of most importance in terms of water quality.   Although the major
flood damage may result from the  larger more infrequent storms, the bulk
of the contaminant loading is carried by  the small  and medium sized storm
events.

In addition, the proportion of pollutants incorporated within the "first
flush" of the design flow that will be  detained by  the flood storage
basin can be extremely variable.   Fig.  1  demonstrates the
runoff-pollutant timing curves for a  flood flow of  10 year return period
within a 350 hectare separately sewered urban  catchment in Hendon, NW
London.
            100
     Total

    Pollution

      load
            80
            60 -
40
            20
                                          • Heavy Metal*

                                        — —Suspended Solid*

                                       	 Total Nitrogen

                                       	Total Phosphate
                     20     40    bO     BO    100

                        % Total Runoff Volume
      Figure 1.  Runoff - Pollutant Timing Curves
Metals  exhibit  the greatest propensity towards a first flush effect  and
this might  be expected given their known affinity to solids.  The        •
greatest  deviations from peak flow are shown for dissolved constitutuents
although  delayed  'last flushes' for solids, cadmium, zinc and
hydrocarbons  are  not uncommon  (Revitt et al, 1982;  Morrison et al,
1984) and the random variability of pollutant parameters in urban  runoff
is  well documented.  (US EPA, 1983.)  Therefore, the initial two hours of
the first flush design inflows to the urban detention basin can contain

                                  T-I-H-3

-------
 between 60 to  80% of  the total  storm pollutant  mass.   (Ellis,  1979).
 However, the partitioning of  pollutants  between solid  and dissolved
 states implies that  those which remain in the  latter phase  will  not be
 amenable to control  through any solids retention and this may
 substantially  reduce  the notional  retention percentages  quoted.   Sizeable
 fractions of nitrogen (60 to  75%)  and phosphorous (30  to 55%)  have been
 shown by Grizzard and Randall (1978)  to  exist  in the dissolved phase of
 urban stormflows.  Runoff studies  by Morrison  et al  (1984)  in  NH London
 and Gothenburg, Sweden have shown  that as much  as 63%, 77%,  60%  and 32%
 of  total zinc, cadmium, lead  and copper  respectively are held  in the
 dissolved or easily exchangeable form.  Therefore, the form  or speciation
 of  the pollutant  in storm flows needs to be ascertained  to  adequately
 evaluate the retention efficiency  and design performance of  the  detention
 basin as a quality control structure.

 A further and  fundamental complication to the  quality  control  performance
 of  urban flood detention basins is in the particle size  association of
 the storm runoff  pollutants.  Ahtough only some 8 to 25% of  the  total
 solids loading to the basin will be of particle size less than 150
 microns, it is this fine size fraction which accounts  for much of the
 pollution potential.   This size range can typically include  30 to 50% of
 algal  nutrients,  30%  of heavy metals, 50 to 60% of hydrocarbons, 25% of
 the oxygen demand and 20 to 40% of total coliforms.  (Sartor et  al, 1974;
 Ellis, 1976).   Given  the low  settling velocities of such fine
 contaminated particles and the  strongly  turbulent  nature of  both inlet
 and outlet flows, it  is highly  probable  that such  material will  not be
 retained within the basin but will be routed in suspension directly
 through  the structural facility.   The high flow rates  maintained during
 flood  flows also  stimulate disaggregation of organically bound floes in
 the stormwater suspension (Ellis et  al,  1982) and  this further
 exacerbates the quality control problem.

 Despite  the anomalies presented by partitioning and size association,
 analysis and appreciation of  the pollutant retention efficiency of flood
 detention  basins  must proceed through considerations of  sedimentation
 rates  and  sediment related quality parameters.

 TRAP  EFFICIENCY OF FLOOD DETENTION BASINS

 The average concentration of  total suspended solids (TSS) in urban storm
 runoff varies  between 100 and 600  mg  I    (Ellis,  1982; US EPA, 1983)
 although  individual low  return  storm  flows  can  average as much as 1700 mg
 I   .   Settleable  solids  greater  than  150 microns,  as determined by
 settling through  a 1.83m  column of 0.3m  diameter for 24  hours, comprise
 between  60  and  90% of  TSS  and possess  specific   gravities ranging between
 0.8 and  2.6.   The settling velocities  of  stormwater runoff solids_as
 determined  under  these  static conditions varied  from 0.001 cm  sec~  for
 fine silts/clays  up to  8.5 cm sec   for  coarse  sand/grit particulate of
 1/2 mm.  The modal stormwater suspended  sediment diameters of  15 to 20
microns  (Ellis  et al,  1982) and the 150 micron  cut-off definipg
 settleable  solids possess  settling velocities of 0.01 cm sec
 respectively.   These  theoretically determined values do  require some
adjustment  to  allow for  the relatively turbulent flow regime conditions
obtained in the detention basin.   It has been suggested that effective
 settling velocities,  for  the  computation of net pollutant removal, may be
as  low as one-ninth of those determined under quiescent  laboratory
conditions.  (Northern Virginia Planning District Commission, 1980.)

                                 T-I-H-4

-------
A number of methods are available to predict  the pollutant  trap
efficiency of detention basins of which  the oldest  and simplest is that
of Camp (1945).  This method, like the more recent  but complex  modelling
approach of Ferrara and Hildick-Smith (1982), utilises the  geometry and
hydraulics of the basin together with computed settling velocities.  The
three parameter sediment removal function  of  Camp is
  r) = function
                 v y
vsy
                                    1/6'
v n  g
and expresses removal as a function of  surface  loading  with  a  turbulence
function equivalent to V^/V  .  The removal  efficiencies can  be read
from the graph (Fig.2) following calculation  and insertion of  the
appropriate values.
             **•'       60
           Efficiency
                       40
                        0
                        0.01
     0.10
1.0
10.0
                     Figure 2. Detention Basin Trap Efficiency

Table 1 shows pollutant  settleability and  trap efficiencies  for a number
of test runs and storm events;  the  latter determined  by the Camp method
using data for a small urban detention pond draining 14.9 hectares of the
larger separately sewered, residential catchment  in NW London mentioned
earlier.

The tabulated results illustrate  the potential utility of the Camp
methodology, which in conjunction with adequate hydraulic criteria can
provide a dual purpose approach to detention basin design.   The design
also emphasise the relatively  low efficiencies of existing structural
designs in providing a quality control, particularly when subject to
turbulent flow regime conditions.  This conclusion is  confirmed by the
field observations and from independent studies (Ferrara and Witkowski,
                                  T-I-H-5

-------
   TfBLjM

   SETTEJBIUTY_ANO.TRJP.EFFlCI|NCIES.OF_OETENTION_B*SIN.£OLLyTANTS
   Pollutant
   Phosphate
   Total
   Coliforms
(1)
Imhoff
Settleability
(Average
X Removal)
(3)
2 Hour
Removal as
X of 24 hour
Removal
(3)
X Trap
Efficiency
"•isq..'

(4)
X Trap
Efficiency
(¥.1M/9)

(5)
Observed
X Trap
Efficiency

TSS
BOD 5
Hydrocarbons
Lead
Cadmium
Total
73
32
67
62
26
46
                85
                         34


                         13


                         18


                         30


                         15


                         20



                         60
                                     62


                                     48


                                     60


                                     59


                                     18


                                     41
                      42


                      28


                      35


                      42


                       2


                      10
6-84
(249) - 86

2V - 48
      abed
    ,cde
20 - 60


10 - 40
(16) - 70

(48) -
      cd
                                                         10 - 78
                                                         20 - 70'
                                    abf
                                                         7-73
   NOTE:  Numbers in parentheses indicate negative trap efficiencies; data from following references:-
*Ferrara and Witkowski, 1983

bGrizzard et al, 1982
C0ally et al, 1983
dtfliippte and Hunter, 1981
                                               *EUis et al, 1985
                                                Ferrara «nd Hildick-Smith, 1982
1983;   Grizzard et al, 1982;   Dally et al, 1983),  although some recently
derived modelUng approaches  would indicate very much higher trap
efficiencies in respect of  TSS.  (Amandes and Bedient, 1980;  McCuen,
1980.)   However, application  of  the Camp method with  settling velocities
for.  particle sizes within the range 20 to 60 microns  produce extremely
low  or  event negative trap  efficiencies and the validity of this is borne
out  from the extreme ranges observed in the field.  Therefore, the
particle size association of  pollutants is of  importance to the removal
efficiency, despite assertions occasionally found  in  the literature that
pollutant removal shows little correlation with sediment removal (Whipple
and  Hunter, 1981; Kuo and Ni, 1984).  Nevertheless, the low computed and
observed trap efficiencies  would imply that factors other than the
immediate storm hydra lies and pollutant settleability are operating
within  the detention basin  to determine the resultant outflow quality.
The  most obvious of these would  be the exchange of pollutants between the
deposited and reducing or anaerobic basal sediments to the overlying
water phase of the detention  basin.  It is well documented that trace
metals  nutrients and hydrocarbons  accumulate in detention and flood
storage basin sediments, (Wignington et al, 1983;  Hvitved-Jacovbsen et
al,  1984;  Ellis et al, 1985)  and  during storm plug inflows these
pollutants can be released  and mobilised from zones of maximum
accumulation such as occur  near  outlets which experience the longest
periods of standing water.  This resuspension and mobilisation phase will
also transfer some of the pollutant load to the dissolved phase,
particularly in the case of pollutants such as cadmium, zinc or nitrate
which have high exchangeable  potential.  It may be that the short
retention time of materials in the basin during storm events would imply
that substantial transformation  of particulate to soluble form is not
likely.   However, the high degree  of turbulence would facilitate exchange
and  in  any case this argument could not be applied to 'wet1 basins where
flushing of soluble material  generated during the preceeding dry period
                                   T-l-H-6

-------
would inevitable take place.

CONCLUSIONS

Considerations of the potential pollutant removal efficiency of urban
storm detention basins require a coupling of their particle size and
settling velocity characteristics.  If this relationship can be
successfully and reliably determined it can be used in conjunction with
hydraulic and geometrical criteria to provide a dual purpose procedure
for detention basin design.  The operational quality performance of any
basin, however, will also be influenced by post-design criteria which
depend on complex biogeochemical exchange mechanisms between pollutants
in the accumulated basal sediments and the overlying water phase.  Proper
and regular maintenance, including sediment removal, is therefore
essential if the projected quality design limits are to be achieved.

Whilst it may be arguable to conclude that urban flood detention basins
may have only a limited pollutant control function, it must be admitted
that with proper maintenance they can only help improve receiving stream
quality.  In addition, they can and do provide substantial intangible
aesthetic, as well as occasional direct recreational community benefits
despite and irrespective of any adverse water quality conditions.
REFERENCES

Amandes, C.B. and Bedient, P.B  (1980).  "Stormwater detention in
developing watersheds".  Journal Environmental Engineering Division4
ASCE, 106, 2, pp. 403-4197
American Society Civil Engineers. (1982). "Proceedjngs^of^Conference^on
Stormwater Detention Facilities".  Urban Hater Resources Research
CouncU~~ASCE7~New"York7
Camp., T.R. ((1945).  "Sedimentation and the design of settling tanks".
IlSDSSSi J2Q5-£2£ll£j?D-§2£l£iy._PJyiL,§n.9iDe.e.£§z PP 895-936 .
Dally, L.K., Lettermaier, D.P., Surges, S.J. and Benjamin, M.M. (1983).
Technica  Report 79, Hater Resources Series, Dept. Civi  Engineering,
University of Washington, Seattle.
Davis, L.H. (1963).  "The hydraulic design of balancing ponds and river
storage ponds".  Chartered_Mun2cJBa_l_Engjneer, 90, pp 1-7.
Ellis, J.B. (1976K  "Sediments and water quality of urban Stormwater".
ysifr_§sryj£fs^ 80' PP- 730-734.
Ellis, J.B. (1979).  "The nature and sources of urban sediment and their
relation to water quality".  In: Hollis, G.E. (Edit): "Man^s^Imgact_on
i!i§-!J^£2i29i£2i_£^£i§_lD_ii!§_yi!$i"  PP- 199-217., Geobooks, Norwich.
Ellis, J.B. 7?9827.  "Benefits and blights of urban Stormwater quality
control".  In: Featherstone, R.E. and James, A. (Edits): "Urban_Drajnage
Systems", pp. 39-51, Pitmans, London.
ElITs7"j.B., Hamilton, R and Roberts, A.H. (1982).  "Sedimentary
characteristics of suspensions in London Stormwater".  Sedimentary
GeoJ-ogy., pp. 147-154.
ITIis, J.B., Revitt, D.M. and Gavens, A. (1985)  "Polyaromatic
hydrocarbons in sediments of an urban catchment".  lDi££D3iJ2D§i..:i2y£D2i
§Qyi!2D2§Qi!iJ5D§i2£i£2i_£t£2!iii£y.' ^, (In Press).
Ferrara, R.A. and Hildick-Smith, A. (1982).  "A modelling approach for
Stormwater quantity and quality control via detention basins".  Water
Resources Bulletin, 18, 6, pp. 957-981.
                                 T-I-H-7

-------
 Ferrara,  R.A.  and Hitkowski, P.  (1983).   "Stormwater quality
 characteristics in detention basins".   Journaj._|nvjronjnentaJ._Engjneer2Qg
 £iyi2i2D^.£§£l' 109' 2' PP-  428-44.
 GrTzzard,~T7j7 and Randall,  C.H.  (1978).   "Occogujn/Four_MUe_Ruj}_Runoff
 P.2iiyi22Q_§ly2^" •  Report/ Dept.  of  Civil Engineering, Virginia
 Polytechnic  Institute, Blacksburg, Virginia.
 Grizzard, T.J., Heand, B.L.  and  Randall,  C.H.  (1982).  "An evaluation of
 Stormwater management ponds  for  the  control of urban runoff pollution".
 In:  Featherstone, R.E. and James, A.  (Edits):  "Urban_Drainage-S^stems"i.
 pp.  135-148, Pitmans, London.
 Hall,  M.J. and Hock in, D.L.  (1980).  "§ujde_to_the-des2gn>_of-storage_Bonds
 i2£-f i222L£2Dl£2i_iQ-B2£iIy..y£2§Qi;J£9^£i!i£il!!!§D£_§£§2§ri  Techni ca I Note
 TOO, Construction Industry Research  &  In?ormation Association, Bristol.
 Hvitved-Jacobsen, T., Yousef,  Y.A.,  Hanielista, M.P. and Pearce, D.B.
 (1984).   "Fate of phosphorous  and nitrogen in  ponds  receiving highway
 runoff".   §cjence_Totaj._Env2ronment, 33,  pp. 259-270.
 Kuo, C.Y.  and~N:i7"'wTY7""(T9847T'"""Pollutant trap efficiency in a detention
 basin".   In: Balmer, P., Malmquist,  P-A and Sjoberg, A. (Edits): "Urban
 §£2£E-££iiD52£" •  PP- 21-28,  Chalmers University of Technology,
 Gothenburg.
 McCuen, R.H. (1980).  "Hater quality trap efficiency of Stormwater
 management basins".  Hater_Resources_BuJUetij2, 1/ pp. 15-21.
 Morrison,  G.M.P., Revitt, D.M., IITis, J.B., Balmer, P., Svensson, G.
 (1984).   "Variations of dissolved and  suspended solid heavy metals
 through an urban  hydrograph".  Environmental Technology Letters^ 7, pp.
 313-318.                                          .........
 Morrison,  G.M.P., Revitt, D.M., Ellis, J.B., Svensson, G., Balmer, P.
 (1984).   "The  physico-chemical speciation of Zn, Cd, Pb and Cu in urban
 Stormwater".   In: Balmer P., Malmquist, P-A. and Sjoberg, A. (Edits):
 "y£25D-§i2£2-2£2iD29S"  PP-  989-1000,  Chalmers University of Technology,
 Gothenburg.
 Northern  Virginia Planning District  Commision. (1980).  "Gujdebook_for
 §££££DiD9-H£2§D-D2D;E2Di_B2ii!dii2D_!D§D§9£!D£Di-§i£§i£9l£S-i  Met ropo I i tan
 Hashington Council of Governments, Hashington, D.C.
 Revitt, D.M.,  Ellis, J.B. and Oldfield, F.  (1982).  "Variation in heavy
 metals of  Stormwater suspended solids  in  a separate  storm system".  In:
 Yen, B.C.  (Ed it):  "U rban_S t or mwa t e £_2uajlit^z>_Management_and_P J. annjng"^
 pp.  49-58, Hater  Resources~PuBncatTons'"Ltd.7  LittIeton,""Colorado.
 Sartor, J.O.,  Boyd, G.B.  and Agardy, F.J.  (1974).   "Hater pollution
 aspects of street surface contaminants".   J2y£D2i_y§i££-£2iiyii2D
 ££^§£J?ii2G/ ^/ PP-  458-467.
 US~ EPA~7T 983 )   Fi^na]. _ Regg r t ^of _t h e^Na t |on wide_U r ban^Runof f __Pr og r am .
 Vol.1. US  Environmental Protection Agency, Hashington, D.C.
 Hhipple, H. (1979).   Dual-purpose detention basins.   Journ^Water
                                                         ""~
Higinton  Pj,RandalcH.andGrzzardTj   (1983.   "Accumulation
of selected trace metals in soils of urban runoff detention basins".
Hater Resources Bulletin, 19, 5, pp. 709-718.
                                T-I-H-8

-------
                  USING THE UNDILUTED EVENT MEAN CONCENTRATION
                            TO DETERMINE RUNOFF LOADS
        Lei and L. Harms and Marsha Smith, Professor and Research Chemist
                        Department of Civil Engineering,
          South Dakota School of Mines and Technology, Rapid City, SD
                                  ABSTRACT

     A three year urban runoff study was conducted on Rapid Creek near Rapid
City, SD.  A new parameter, the undiluted event mean concentration (UEMC), was
introduced in order to directly relate water quality changes to the quality of
runoff water.  Thirty rainfall runoff events were monitored with about 20 data
sets being complete enough to allow the calculation of UEMCs.

     Six water quality sampling stations were installed to monitor water quality
changes in Rapid Creek resulting from urban stormwater runoff in the Rapid City
area.  Five of the stations were in-stream locations on the creek while the
sixth station was located on the drainage channel from a 2,000 acre (809 ha)
watershed.  All sampling sites were gaged for flow.  Precipitation for the
sampled events ranged from 0.08" (0.20 cm) to 2.99" (7.59 cm) and the urban
discharge ranged from near zero to about 470 acre-feet (579 x 10^ m^).
Concentrations of instream constituents varied widely throughout the study, with
the majority of the constituents increasing in concentration at the downstream
stations.

     The UEMC is calculated for a particular event by subtracting out the
upstream load.  To obtain valid UEMCs for various parameters, careful deter-
mination of baseline flow and quality is needed as well as accurate deter-
minations of flow and water quality during the runoff event.  A statistical
analysis of UEMCs will be presented for ammonia, chloride, COD, total lead,
total phosphorus, and suspended solids.  Results show that data are positively
skewed with an error in the UEMC of about 5 percent.
                                  T-II-A-1

-------
THE  STUDY  AREA

General

     The study area  lies  within  the  physical  boundaries of Rapid City, South
Dakota.  Figure 1  indicates  the  locations of  the six sampling stations.  Rapid
City is located at the  foothills  to  the Black Hills in western South Dakota and
is the county  seat for  Pennington  County.  Rapid City is the largest community
in western South Dakota and  is the business center for much of the surrounding
area.

     Between 1940 and 1960,  Rapid  City experienced a dramatic growth rate
resulting  in a 1960  population of  42,399, up  from 13,844 in 1940.  A more
stable, much slower  growth occurred  in the next two decades, giving a 1980 cen-
sus  figure of  46,492.

Sampling Locations

     The six sampling stations are indicated  on Figure 1.  Ideally, a sampling
station should have  been  located just upstream of all urban development, and
just downstream from the  city.  Station 1 essentially satisfies the upstream
station, but some urban influence will manifest itself because of an increase in
density of dwellings and  upstream  highway traffic.  It was not hydraulically
possible to establish a sampling station downstream from the study area before
other non-point sources of runoff would enter Rapid Creek.  Consequently, a
theoretical station, No.  56, was developed which mathematically establishes
downstream values based on mixing the loads present at Sta. 5 (the last station
on Rapid Creek)  and  Sta.  6 (the last major urban drainage into Rapid Creek).

     Drainage  areas  ranged in size from 33,730 acres (13,650 ha) at Sta. No. 1
to 1,610 acres  (650  ha) at Sta. No. 5.  Land  usage varied from 96 percent non-
urban at Sta.  No. 1  to  19 percent non-urban at Sta. No. 5.  Complete land use
information can  be obtained  from a document by Harms et al. (1983).

TECHNICS

Field Methods

     The majority of the  water samples were collected using automated equipment
but  some samples were collected manually, primarily at the upstream stations.
Baseline samples were taken  at normal flow conditions periodically.  Manual
methods were used to obtain  these samples.  Even with the automated equipment
in-pi ace, every  effort was made by the field  crews to be on site during the
runoff events.

     Manually  collected water samples were obtained by wading the streams and
collecting depth integrated  aliquots using between 10 to 15 verticals across the
channel.   In the first year  of the study, manually collected baseline and precip-
itation runoff  samples were  collected directly into new, one-gallon (3.8 L)
plastic milk containers.  During 1981 and 1982, samples were collected by use of
DH-77 depth-integrating samplers.  After collection the samples were transferred
to new, one-gallon (3.8 L) plastic milk containers.  All  sampling equipment and

                                  T-II-A-2

-------
H
I
CO



r




I
1
\ Q-
P ^ i
\ " © r /
^ ^^ 	 ^^_Z C4)
LIME ( ^ 	 ^^
CREEK ^.^ U WATER DQWN ^
Xj^ PLANT TOWN \^
r* r-----xy._.
\ A
\ r \ -' ^^ A
CANYON X^ r--1" r\
CO
*:
o
oc
<
Q.
w IRRIGATION
DITCH
xfe""


RIM ROCK V ^< J \ ,-^~-*~S ,^' \ tx
HKHWAY ^jtf^-S / f— __^ ,' S^
N

— >
\/ w^ ^ -^
r (T^BFISH ' ^^--^^--
^\ /\ HATCHERY ^^



              RAPID CREEK
\
                                                                   LEGEND
     DRAINAGE DIRECTION
    ' SAMPLING STATION
         Figure 1. •• Map of Rapid Creek and Major Drainage Basins within Rapid City Proper; Sampling Site Locations.

-------
containers were rinsed  three  times with native water prior to taking the first
aliquot.  The  sample was  immediately  cooled to 4°C.  Date, time, and stage were
recorded  immediately before and  after  sample collection.

     The  automated  sampling equipment, referred to as an urban hydrology moni-
toring system  by USGS,  incorporates a  microprocessor based system control unit
to receive a record on  site and  to control the automated water sampling device.
The system would switch on to  storm mode when a select stage, corresponding to a
definite  discharge, was reached  in-stream during a storm event.  The system
would make continuous recording  data  (time, date, stage, accumulated rainfall,
and sequential sample number)  at a pre-determined time interval ranging from 30
seconds to 1 hour.  If  a  rapid rise in stage occurred, the timing sequence was
over ridden and additional samples were collected.  Samples were pumped into
new, one-gallon (3.8 L) plastic  containers which were stored in a refrigerator
unit, at  4°C,  and later,  as soon as possible, transported to the South Dakota
School of Mines and Technology (SDSM&T) environmental engineering laboratory.
Under most conditions,  samples were delivered to the lab within three hours of
collection.  A sketch of  this  equipment is shown in Figure 2.

Laboratory Methods

     All  sample preservation and preparation were done in the laboratory. Flow
weighted  composites were  normally made for each station using the appropriate
aliquot from each discrete sample.  Volumes for each aliquot were determined by
calculating the volume  under the hydrograph that each sample represented,
baseflow  included.  A computer program was developed to compute the correct ali-
quots, Harms and Smith  (1983).

     Composites were produced  using a USGS cone splitter, USGS (1980). The
splitter  splits any sample into  10 equal aliquots to the nearest one percent.
The cone  splitter was evaluated  for accuracy prior to use.  After some practice
the error averaged about  3 percent which was the sum of the cone splitter and
analytical error.  After  the composite was properly consituted, the cone
splitter was used to split the composite sample into individual containers for
storage and/or preservation.

     A quality assurance  (QA)  plan was developed and approved by EPA.  Standards
(both known and unknown), duplicates, blanks, and spikes were routinely ana-
lyzed.  Values which did  not satisfy the QA requirements were not entered into
the data  base.
UNDILUTED EVENT MEAN CONCENTRATIONS

     The undiluted event mean concentration, UEMC, was selected to indicate the
quality of runoff which entered the stream.  The UEMC was calculated for each
parameter from laboratory data for a particular event by subtracting out the
portion associated with the baseline flow, as follows:

                         QBL CBL + QRO UEMC = q EMC
                 and,             _
                           UEMC = q EMC - qBL
                                    QRO

                                  T-II-A-4

-------
                                                 RAIN
                                                 GAGE
REMOTE
RAIN
GAGES
  Figure 2. •• Automated sampling station.

-------
                where,

                             = Base  flow  preceeding runoff events, cfs

                             = Assumed base flow concentration for the
                               parameter  under consideration.  Value is
                               assigned from baseline quality obtained
                               from  samples collected when runoff was not
                               affecting  the quality, mg/L.

                             = Mean  runoff flow, equal to in-stream mean
                               flow  recorded during a runoff event - base
                               flow  (Q -  QBL), cfs.

                        UEMC = Undiluted  event mean concentration, a
                               calculated approximation of the mean con-
                               centration of a parameter in the runoff,
                               mg/L.

                           T) = Mean  in-stream flow during a runoff event,
                               flow  weighted; i.e., (Q)(Runoff duration)
                               Volume of  water passing a station during
                               event, cfs.

                         EMC = Event mean concentration.  Obtained by
                               laboratory testing of a flow weighted com-
                               posite sample, mg/L.
                           (Note:  cfs x  2.8317 x 10"2  = m3/s)

A simple example may best illustrate the  UEMC:

     Given:  Station No. 3
             Runoff Event No. 5
             Base flow = QBL = 86 cfs (2.44 m3/s)

             Mean in-stream flow = Q = 110 cfs (3.15 m3/s)
             Instream suspended solids = EMC = 60 mg/L
             Base flow suspended solids = CBL = 8
  To Find:   The concentration of suspended solids in
             the runoff above Sta. No. 3, UEMC.
     UEMC = H EMC - QBL CBL = (110)60-86(8)
            _      110-86
                 QRO

          = 246 mg/L           ans.

     The above procedure is not applicable at Sta. 56 (the theoretical
downstream location) because the event duration was not exactly the same at
Station 5 and Station 6.  The UEMCs for Sta. 56 were calculated using the dura-
tion recorded at each station, as shown on the following page:
                                 T-II-A-6

-------
            UEMC56 =| (UEMC)(Vol R0)| 5 + | (UEMC)(Vol  ROj 6
Vol
                                       + Vol
          where:
                UEMC = as described above, mg/L
              Vol RO = Volume of runoff (base flow subtracted)
                       at each station, acre-feet (1.2335 x 103 m3).

     Because of the inherent error in sampling and measuring procedures.  Some of
the calculated UEMC values for Stations 1 through 5 yield a negative  value.
Obviously this is impossible as a negative 3.5 mg/L of chloride could not be
present in the runoff, for example.  Statistically, however, these negative
values are just as valid as positive values and they were not discarded during
subsequent calculations.  Generally negative UEMC values were viewed  as an indi-
cation that that particular constituent was very low in the runoff of the event
under consideration.

     Frequently, only a minor change in the measured concentration or flow would
have produced a positive value.  For example, the UEMC for Station 1, Storm 1
shows a -21.5 mg/L for a chloride concentration.  If the baseflow used in the
UEMC calculation was in error by only 3% (a very acceptable field measuring
error), the UEMC would calculate to be a positive 14.2 mg/L.  Thus, even though
some of the UEMC values compute as negative quantities; they appear to give
reasonable concentrations.  Approximately 3.7% of the UEMC values for Sta. 1
through 5 yield negative concentrations.
                TABLE 1 — Log Normal mean values for UEMC, mg/L
Sta. No.
1
2
3
4
5
6
56

0
0
0
0
0
0
0
NH3.
N
.26
.12
.09
.12
.11
.18
.15
COD
12
52
91
113
95
190
119
ci-
5.7
4.8
4.3
5.5
5.8
5.5
5.9
TKN
0.56
1.30
1.14
2.42
2.19
3.87
2.56
N03
0.35
0.38
0.35
0.60
0.46
0.52
0.51
Total
P
0.04
0.25
0.29
0.62
0.55
1.48
0.80
Diss.
P
0.01
0.03
0.03
0.06
0.04
0.08
0.05
SS
22
289
350
693
595
2,443
1,038
Tot
Res.
240
533
552
771
621
2,786
1,161
VSS
3
33
35
81
67
228
104
     A summary of the log normal mean values for the UEMCs is presented in Table
1.  Although mean values can be misleading statistics, the data clearly show an
increase above normal runoff in the stream (Sta. No. 1) as urban runoff enters
Rapid Creek.  When values from Sta. No. 5 are compared with the theoretical con-
centrations downstream at Sta. No. 56, the impact of the pollutants from the
Robbinsdale watershed (Sta. No. 6) are very apparent.

     A statistical analysis was done for the UEMC data sets for each parameter.
Summaries of the statistics for the computed UEMC values for ammonia and chemi-
cal oxygen demand (COD) are presented in Tables No. 2 and 3.  Ammonia is repre-

                                 T-II-A-7

-------
 sentative  of a  dissolved  species  and  shows  some negative values while the COD is
 representative  of  a  parameter  associated with solids in suspension and does not
 show  any negative  values.
                 TABLE 2 -- Summary of UEMC for Ammonia-Nitrogen
Parameter
Mean, mg/L
Log mean, mg/L
Median, mg/L
Min., mg/L
Max., mg/L
No. samples
No. negative
values
1
0.60
0.26
0.26
-0.11
3.80
15

2
2
0.36
0.12
0.19
0.02
1.96
18

0
3
0.18
0.09
0.16
0.02
0.78
19

0
Station
4
0.23
0.12
0.18
0.02
1.32
19

0
5
0.22
0.11
0.13
0.02
0.62
21

0
6
0.26
0.18
0.16
-0.02
1.00
20

2
7
0.22
0.15
0.13
0.26
0.61
19

0
                       TABLE 3 « Summary of UEMC for COD
      Parameter
1
Station
  4
   Mean, mg/L         53       69      102     137      112     214      135
   Log mean, mg/L     12       52       91     113       95     190      119
   Median, mg/L        2.5     50       78     106       85     171      100
   Min., mg/L          2.5     11       47      33       21      87       51
   Max., mg/L        359      276      251     377      339     599      367
   No. samples        15       17       19      18       21      21       20
   No. negative
     values            0        0        00        00        0
Runoff Loads

     Rainfall runoff loads for each event were calculated by using the UEMC for
each parameter.  The UEMC was multiplied by the volume of runoff to give the
total load to the stream in pounds (0.454 kg).

     Obviously, there is a dramatic increase in material entering Rapid Creek as
the stream moves through Rapid City.  Table 4 shows these increases as a percen-
tage of the load entering Rapid City (at Sta. No. 1).  The higher percentages
are associated with those constituents which fluctuate with suspended solids,
the most predominate being total lead.  Lead was almost entirely tied to
particulates, and very little dissolved lead was detected.
                                 T-II-A-8

-------
                       TABLE 4 - Increase in Runoff Loads
   Parameter
Percent Increase
Sta. 1 to Sta. 56
Parameter
Percent Increase
Sta. 1 to Sta. 56
Ammonia-Nitrogen
COD
Chloride
Total Kjeldahl N
Total Lead
Nitrate + Nitrite
633
24,400
12,900
2,530
104,000
2,300
Total P
Dissolved P
Sodium
Suspended Solids
Volatile SS
Total Residue
14,000
800
5,980
63,500
61,400
4,730
An important consideration is that these loads are deposited in the stream
within a relatively short time-frame.  Downstream from the Rapid City urban
area is the municipal wastewater treatment facility.  Assuming that the facil-
ity treats its sewage to normal levels, the runoff loads show that it would
take three months for the treatment plant to discharge the same amount of
solids that a moderate storm event would cause to be washed into the creek.  It
would take almost three years for the effluent to contribute solids equal to
those deposited in Rapid Creek in one day by a 2.5 inch (6.35 cm) rain (Event
No. 21).


CONCLUSION

     If used in conjunction with accurate flow measurements and representative
water samples, the undiluted event mean concentration is an effective parameter
for computing runoff loads.


REFERENCES

Harms, L.L., Smith, M. and Goddard, K. (1983).  Urban Runoff Control in Rapid
City, South Dakota.  Sixth District Council of Local Governments, Rapid City,
SD.

U.S. Geological Survey (1980).  Equipment and Supplies - New Sample Splitter
for Water Quality Samples, Technical Memorandum No. 80.17, USGS, Reston, VA.

Harms, L.L. and Smith, H., Jr. (1983).  A compositing Program for Water Quality
Sampling.  Water Engr. and Management. 130, 12, pp. 39-40.
                                 T-II-A-9

-------
                  UET WEATHER IN-STREAM WATER QUALITY EVALUATION

                        John J.  Warwick, Ph.D., P.E.
               Assistant Professor  of Environmental  Sciences
                     The University of Texas at Dallas
                    P.O. Box 830688, Mail  Station BE-22
                        Richardson, Texas 75083-0688

             Contributing Authors:  David Getz and David Kibler


                                 ABSTRACT

The magnitude and duration of non-point  source discharges  are  stochastic  in
nature.   For this reason,  in-stream water  quality  standards  should  be  based
upon an  acceptable frequency of violation  rather  than an absolute  level  which
should  never  be exceeded.   The   effect  of  various  pollution abatement
methodologies  could then be evaluated  based on their ability  to  meet the
prescribed  water quality criteria and associated  acceptable frequency  of
violation.  This  type of approach requires characterization  of receiving water
responses  to a  large number of storm events.

A project  was conducted for the city of Altoona, Pennsylvania, to quantify the
degradation in water  quality of the Little Juniata River  as a result  of
combined  sewer overflows.   Dissolved oxygen was  chosen  as the principal
indicator  of in-stream  water  quality.  Runoff flows  and associated  biochemical
oxygen demand  (BOO)  loadings were  calculated for the urban and surrounding
rural areas by the Army  Corps of Engineers' Storage,  Treatment, Overflow,
Runoff Model (STORM).  A three year  period  from 1975  through 1977, containing
209 individual storms events, was chosen  to  characterize the statistical
behavior of in-stream dissolved oxygen  concentrations.

A one-dimensional,  steady  state,  dissolved oxygen model  was utilized  to
calculate spacial variation of dissolved oxygen concentration.   Model  input
included  STORM output (first flush flow  and  BOD loadings) for each  storm
event.  First flush  values were  chosen to simulate worst  possible conditions.
The dissolved oxygen model  calculated the critical or  minimum dissolved oxygen
value occurring  within the river  study area for each storm.  A  cumulative
frequency  distribution of minimum in-stream  dissolved oxygen  values  was then
calculated  via a  simple  ranking procedure.

Results  indicate  that the seasonal operation of CSO facilities may  be  a  viable
option,  with real  economic savings.  The impact  of supplying  various levels  of
detention storage for the urbanized area of Altoona was also  investigated.
Diminishing  returns (smaller  decreases in the  frequency of  violation)  can  be
expected with increasing  levels of detention storage.


Keywords:   Urban runoff pol 1ution, nonpoint pol1ution, nonpoint pol 1ution
abatement, stormwater quality,  stormwater models,  water quality models
                               T-II-B-1

-------
 INTRODUCTION

     Combined sewer  overflow  (CSO)  discharges pose  a serious threat  to
 receiving  quality  in  many  urban  drainage  basins.  The  combined discharge  of
 untreated  sanitary  "dry  weather" flow and  urban runoff  can introduce
 significant  amounts  of pollutants,  resulting in  severe degradation  of
 receiving water quality.

     In 1981, the Pennsylvania  Department of Environmental Resources (DER)
 directed the City  Authority of  Altoona to analyze  the effects  of  combined
 sewer  overflows in the wastewater management  plan  being developed for the
 city.  Altoona,  a city of approximately 60,000 people, serves as the  headwater
 region for two rivers:  the Little  Juniata River and  the Beaverdam-Frankstown
 Branch of the Juniata River.  The respective  watershed  areas are  343 and 395
 square miles.  Water quality in the two rivers is very sensitive to the impact
 of CSO discharges  since  the  discharges  constitute a significant  portion of the
 total  flow  in upstream reaches of the  rivers during periods  of heavy  rainfal 1
 or snowmelt.

     This  report will  focus on  the study performed in the Little Juniata
 River.  The study was  undertaken to satisfy two  principal objectives:  1)  to
 estimate the frequency of violation of dissolved oxygen  standards as  a result
 of existing combined sewer overflows; and  2)  to  evaluate  the potential
 effectiveness of alternative CSO  abatement  measures.   Mathematical  models were
 used to simulate the generation of CSO  discharges in  Altoona and their effect
 on in-stream dissolved oxygen concentration.   The models  were used to-eval uate
 abatement measures  such  as   expanded sewage treatment plant  capacity and the
 utilization of stormwater detention basins.  The effectiveness of an abatement
 measure was expressed  in two interrelated  ways:   1)   by  its  ability to reduce
 the biochemical  oxygen demand (BOD)  loadings  attributed to combined  sewer
 overflows; and 2)  by its ability to  reduce  the frequency  of  violation  of
 established dissolved oxygen standards during  wet-weather events.
MODEL FORMULATION

     The mathematical modeling effort was divided into two phases:   a land
phase and an in-stream phase.   In  the  land  phase, the Army Corps  of Engineers'
Storage  Treatment, Overflow, Runoff  Model  (STORM)  was used (6).   STORM
performs a continuous  simulation  of the rainfal 1-runoff and snowmelt-runoff
processes on the land surface.  It calculates the quantity of runoff generated
in a drainage  basin and the  quality of the  runoff  in terms of 5-day  BOD
concentrations.   The STORM model was modified in two minor areas:   (1)  to
better  represent  street cleaning operations,  changes in  the  sweeping
efficiency function  were  made;  and  (2)  the closure action of shear-gates at
each diversion  structure  was  incorporated  in the model.   The  STORM model is
wel 1-documented for its runoff and pollutant loading calculations  and no
further discussion is provided.

     The second model used was the Distributed  Oxygen Analysis Program  (DOAP),
a model developed specifically  for this  project.  In this model, the f 1 ow
generated by STORM  is mixed with the background flow  in  the  river, and the
total  flow  is tracked downstream.   The  model  calculates in-stream dissolved
oxygen  concentrations along the  length  of the river using steady-state
equations that perform an  oxygen balance.   The  oxygen balance calculations

                               T-II-B-2

-------
provide a simulation  of the  consumption of in-stream oxygen by the  bacterial
breakdown  of organic material present  in the combined sewer overflow  and  other
waste inputs.

     The oxygen mass bal ance equation used in the stream model is the wel 1 -
known steady-state form:



     0 =   .2.L [expjt .  S.i expgt] + DQ expgt                           (1)
           k2-K1           S2
where:
          U2
      j  =  -  [1 - Sj]                                                  (2)
          2E
          U2
      g =  --  [i - s2]                                                  (3)
          2E
              4 k,E
         -  1  +  --«-                                                    (4)
               U2
              4 k2E
and   So  =  1  + -----                                                    (5)
               U2


     The  variables  present in Eq. (1) to  (5) are defined as:  D = resultant DO
deficit (mg/1); D0 = initial DO deficit  (mg/1); E = longitudinal dispersion
coefficient  (mi 1es2/hour);  kj =  deoxygenation coefficient  (I/hour); ko =
reaeration coefficient (I/hour); L0 = ultimate BOD (mg/1); t = elapsed flow
time (hours);  and  U = flow velocity  (miles/hour).

     The  effects  of photosynthesis,   benthal  oxygen  demand, and  plant
respiration  are  not included in  Eq.  (1).   It  is  expected  that  little
photosynthetic activity  occurs  during runoff events due to cloudy weather and
turbid  water  conditions.   Similarly, plant respiration  may  have an
insignificant effect because of the increased volume of flow in the  river.
The potential  effect of benthal oxygen demand on DO concentrations  can be
quite significant, however. During wet weather, high  flow velocities can act
to scour  sediements that can include 1arge quantities of organic material.
This  phenomenon  and its  corresponding  sediment  oxygen  demand  are  not
represented in the present study.  There is no evidence that benthal oxygen
demand plays a significant role in depleting the DO in  the Little Juniata
River.

                              T-II-B-3

-------
     Controlling parameters in the DOAP model  are:  (1)  ki,  the in-stream
 deoxygenation coefficient;  and,  (2)   k2,  the  reaeration coefficient.  Because
 all  DO depletion was  assumed  to  take place through biological  oxidation, the
 deoxygenation coefficient,  k]_, was a sensitive parameter.  It was  estimated as
 a  function of flow rate using a  formulation by Wright and McDonnell (7):


          kl - \ Q Y2                                                 (6)


 where  kj = in-stream  deoxygenation coefficient at 20° C (I/day); Q = steady
 state flow rate (cfs);  Y]_ and Y2  are  empirical  coefficients to be evaluated by
 calibration against  observed DO data  from in-stream  measurements during a
 storm.  These were primary parameters in fitting the model  to  the Juniata
 River  system.

     The reaeration coefficient,  k2,  was estimated  by use of  an  equation
 developed by Owens et  al . (5) for shallow streams:


               21.7  U°'67
                                                                        (7)
where ko = reaeration coefficient at 20°C (I/day); U = mean stream velocity
(ft/sec); and H = mean stream depth (ft).  Both  k}  and k2 were  adjusted for
temperature using standard correction formulas with 9  values of 1.068 and
1.024,  respectively.

     Since many  inflows occur along the  rivers, the in-stream  flow rate,
velocity and  depth  will  vary greatly  with distance.   In  the case  of  a
headwater receiving stream,  this  variation can be several times greater than
the base discharge during  dry weather periods.  Consequently, it is important
to represent the  spatial distribution of flows and pollutant  discharges to the
headwater receiving stream.  In the present  study, this has been accomplished
by developing depth discharge and velocity-discharge  relations for  each  reach.
With changes in  flow  rate at each inflow point,  the estimates  of mean flow
depth and velocity are updated.  These revised  stream values in turn are used
to re-calculate kj, k2 and the flow time for each sub- reach  (distance between
inflows).   A  sub-area definition  sketch  is shown  in Figure 1 for the 343
square mile  Little Juniata River portion of the total Juniata River watershed.


FREQUENCY ANALYSIS

     The strategy of analyzing the  frequency of occurrence of low DO levels to
evaluate CSO abatement alternatives  has been introduced  previously by  Medina
(3).   The DO frequency  analysis performed  in DOAP is  possible  due to the
model's  ability to  generate DO profiles for a  large  number of separate  runoff
events.   One DO value  from each event is  used in the frequency analysis.  This
is the  minimum  in-stream DO concentration  present, irrespective of  its
location  in the  river.  Thus, the magnitude  of the DO sag is considered, but
its location is not.  The minimum  DO  value  from each  event is then  stored for
subsequent frequency  analysis.  This set of DO minimums is  then ordered and


                              T-II-B-4

-------
H
I
      Figure 1: Little Juniata sub-area definition sketch,

-------
 assigned a  plotting position using the formula:   P = lOOm/N, where P is the
 percentage of events  exceeding a  specified DO  level,  m  is  the  rank  in
 descending order,  and N is the total  number of events.
 MODEL APPLICATION

     The  period of analysis was 1975 through  1977  --  a  three-year period
 experiencing above-normal  rainfall distribution  with the watershed in a
 condition close to that  projected over the life of any CSO  abatement project.
 The definition  of  an individual  "event" is based  on the  return to zero  storage
 as employed by STORM.   The timing of individual  "events" is thus dependent on
 watershed storage and sewer system capacities,  as  well  as on  the  hourly
 precipitation  record.  To reduce  the dependence  between events, a minimum
 inter-event time  of six hours was  used for separation  purposes.  Using this
 event  definition  criteria,  STORM showed that 209  CSO events occurred in
 Subarea  A (urban)  during  the three-year  simulation period.  The number of
 events  in Subareas B-H  (non-urbon)  varied  from 120  to  183,  depending on
 watershed characteristics.  Obviously, larger amounts  of  precipitation are
 required to produce runoff in the more  highly pervious rural watersheds.
 Thus, many smal 1  storms that cause runoff in Altoona (Subarea  A) produce no
 runoff in other subareas.  The first flush or  first hour flow rates  in  the 209
 events ranged from 11  cfs to  389 cfs.  BOD5 concentrations  in the CSO  ranged
 from 9 mg/1 to 302 mg/1.   The wet-weather runoff from Subareas  B-H  varied
 between  zero and  several thousand cfs during the simulation  period, with
 typical  BODg concentrations of !S  mg/1  to 10 mg/1.

     Model parameters  in  STORM were adjusted  for runoff volume  and pollutant
 (BODc)  load by  calibration  against  nine  individual storms on three urban sub-
 areas where  gaging instrumentation was  installed.   The details  of STORM
 calibration and  extension to  the Little Juniata   system  in  Figure 1 are
 presented elsewhere (1,2).   The in-stream DO model,  DOAP,  was  then structured
 to simulate point  source  loading  from  each subarea.  Base flows were  estimated
 from long-term quarterly averages.  These were  assumed  to be at  the same
 temperature  as the wet-weather inflows and   also  were assumed  to be  DO
 saturated.  Sewage  treatment plant  discharges were obtained  and  introduced in
 DOAP at  their respective  locations.

     In-stream  DO  samples  collected during wet-weather  events were  then used
 to calibrate DOAP.  Samples were  collected at different times during an event
 at several accessible  locations.  During calibration,  efforts were made to fit
 the  calculated DO profile  to  the  observed  DO data.   Calibration  was
 accomplished  by  adjusting the  values  of YI  and Y^ in the deoxyyenation
 coefficient equation (6).   Since an attempt is made to simulate  the worst in-
 stream condition  during  each storm event, the DO profile was fitted to the
minimum  DO concentration  measured at each  sampling  station.

     The wet-weather event that occurred on  July 27, 1982 was  selected for
model  calibration  since the DO data collected  on  that  day clearly illustrated
 the impact of combined sewer overflows.   The  best  possible  fit  of the DO sag
 point  was  obtained  for  values of  YJ  and  Y£ set at  14.5 and  -0.37,
 respectively.  Several  other wet-weather events sampled  in 1982 were  also used
 for model  verification purposes.  Calibration and verification of  the DOAP
model was successful  in terms of  matching the magnitude of the dissolved
oxygen sag.  The DOAP  model was  not  very  successful  in matching the  entire

                              T-TI-B-6

-------
measured  dissolved oxygen  profile.  Discrepancies  of  up  to  2.0 mg/1 were
observed at downstream  locations.

     The discrepancy between the calculated profile and the measured  data  is
caused by  several  factors.  First,  due to  problems  in  tracking CSO discharges
downstream, the samples  col lected during an  event may not  reflect the true
minimum DO concentration  that occurred at  a specific location.  Secondly, the
results from DOAP depend upon  accurate  wet-weather  flow  volumes  and BOD
loadings as calculated  by the STORM model.  During calibration of STORM it was
observed  that the  predictions  from STORM  may  not  be reliable for any
individual event,  although the model matched observed  values reasonably well
over  a  series  of  events.  Variation  in the  velocity-discharge  relations  is
another possible  source of error.  Thus, matching the magnitude  of the DO sag
is believed to be compatible with  the  planning-level  application of the DOAP
model in this study.
RESULTS

     The mathematical models, STORM and DOAP,  were  used in tandem to simulate
the effectiveness of various measures in reducing the water quality impacts  of
combined  sewer overflows on the Little Juniata River.   The results are
presented in this section for:   (1)  existing conditions;  (2)  selected CSO
storage treatment alternatives; and, (3)   seasonal analysis of DO  levels.   It
should be noted that the results described below are for first-hour flow rates
and BOD loadings.   This  initial  hour  of  each event was  adopted  as the most
severe loading period  in an effort to represent first flush impacts on the
receiving stream.


Existing Conditions

     As a means of establishing a reference or  bench-mark for comparison  of
alternative CSO abatement alternations, existing conditions  were  analyzed.
Results show that under existing  conditions (existing  diversion,  no storage,
and a  6.77 M6D wet-weather treatment capacity) combined  sewer overflows
frequently cause  water  quality conditions in violoation of the DER  dissolved
oxygen standard.  During the simulation period, in-stream  conditions during  65
of the 209 events violated the standard established  for the  Little Juniata
River.  The cumulative  frequency curve  for  the 209  events  under  existing
conditions is  shown  as  a  dashed  line in Figure 2.


Storage Treatment Alternatives

     A number of storage/treatment/diversion combinations were selected for
analysis with the DOAP  model  to determine their effectiveness in protecting
water quality in  the Little Juniata  River.  Figure 2  shows the effectiveness
of different storage capacities  with  existing flow diversion and  existing
treatment capacity.   Diminishing  returns (smaller decreases in the frequency
of violation) is  shown  with  increasing levels of detention storage.   Similar
analyses were completed for modified  flow diversion  and  a doubled treatment
capacity.  The results from all  of the analyses are  summarized in Table 1.
Results for each  storage/treatment/diversion combination  are expressed  as  an
annual  frequency of violation of the DER  dissolved oxygen  standard.  The

                              T-II-B-7

-------
     0

     ci

     Q
     LU
     u.
     o
     y
     Q.
     CD
     z

     Q
     y
     Ld
     o
     X
     Ul
     z
     Li

     Ul

     u.
     0
2   3   4   5   6   7   8   9   10   11   12  13  M  15


       MINIMUM DISSOLVED OXYGEN CHG/L)
                             E « Existing Condition


                           .01 » Storage Capacity (Inches)
Figure 2: Minimum dissolved  oxygen cumulative  frequency curve.
                                 T-II-B-8

-------
Table 1:  Frequency of Violation of DO Standard Under Various
          Storage/Treatment  Conditions, Little Juniata River
Treatment Storage Treatment
Type of Capacity
Diversion (MGD)a (in)b (MG)C
Existing 6.77 0.00
0.01
0.05
0.10
0.20
Modified 6.77 0.00
" " 0.05
0.10
Modified 13.54 0.00
0.05
0.10
0.00
0.29
1.47
2.94
5.87
0.00
1.47
2.94
0.00
1.47
2.94
No. of Violations
Per Yeard
21.7
19.7
11.3
7.0
4.3
19.0
11.7
7.7
17.7
9.7
7.3
     = million gallons  per day; 1.0 gal. = 3.785 liter
°1.0 in. = 25.4 mm
CM6 = million gallons;  1.0 gal. =  3.785 liter
     DO value below 5.0  mg/1  from  February 15  to  July 31;  any DO value below
   4.0  mg/1  during the  remainder of the year.
present frequency  of  violation,  21.7 events per year, is  significantly  reduced
by the  larger  CSO storage capacities.   Modified flow diversion and expanded
treatment capacity are effective in combination with  small  storage  facilities,
but do  not  significantly affect  the  frequency of violation  when storage
capacities are large.


Seasonal Effect

     In an effort  to  observe  seasonal trends  in  the  minimum DO concentrations,
each of the three years in the simul at ion period was divided into quarters.
Results  from the  seasonal  analysis  are  presented in  Figure 3.   It is  evident
that wet-weather water quality conditions are worst  during the  third  quarter


                                T-II-B-9

-------
o •««
. IvVr
Q :
* Qft-
Q ai!j~
UJ :
r" flcu
UL QV^
o :
lil 7ft-
111 /g-
Q. :
w ^
0. 5CU
(/) ^*r
2 i«:
y l0:
UCL;

u.
0 (


A
I ^
S
^







) t



•~-~.








2



	

vxv






<

	

kM
I
V
.3
'N





A

• ' !•— •



\?
•«.
]
\
*•%_



s

T




"^
"s,

-.


[
\
\


'N-^

ie



~\
\
\
i.



^v
V



it




*-s
>
>•».



N —



12





\
\


\
\


13





1
N
\
i
—.1 .
i
\
t
i
t
"-^->


H









t
i%
*»
^

IS
                            MINIMUM DISSOLVED OXYGEN CM6/L)


            1 - Quarter 1:  January 1 - March 31
            2 - Quarter 2:  April 1 - June 30
            3 - Quarter 3:  July 1 - September 30
            4 - Quarter 4:  October 1 - December 31
Figure 3: Seasonal  analysis of minimum dissolved oxygen cumulative
          frequency curve.
                               T-1I-B-10

-------
(July 1 to September 30).  In  general, CSO discharges  have  little  effect on DO
levels  during the first and fourth  quarters, although a  few  violations of the
DO standards are  observed  in  the first quarter.  The seasonal  operation of CSO
facilities may provide adequate water quality  protection while significantly
reducing operating expenditures.


CONCLUSIONS

     Several conclusions can be drawn from  the  results  obtained in the  study.
Combined sewer  overflows from Altoona presently introduce  substantial amounts
of biochemical  oxygen demand  in  the Little  Juniata  River.   During  warm
weather, the BOD loading is sufficient to cause frequent violations of the
established dissolved oxygen  standards.  Due to the predominantly  rural nature
of the  study  area downstream  of Altoona,  wet-weather inflows  dilute the
impacts from combined sewer overflows.  Combined sewer overflow abatement
measures  that  include  storage facilities, expanded wastewater treatment
capacity,  and modified CSO  diversion  have been  shown  by the modeling results
to be effective to varying degrees.   In  general,  storage facilities show the
most promise in protecting receiving water quality.

     The results  from  this study suggest that the Pennsylvania DER address the
issue of modifying its water  quality  standards  to include  provisions for wet-
weather conditions.   Appropriate wet-weather provisions  must be based on the
significance of  the  frequency, magnitude, and duration of DO depletion in
regard  to  the health of aquatic life.  Meaningful  cost-benefit abatement
analyses will depend  on a better understanding of the significance of low in-
stream DO levels  during wet-weather periods.  This is of critical  importance
to the selection  of alternative CSO abatement facilities.
ACKNOWLEDGMENT

     The investigation  reported here was conducted under  funding  from the EPA
Municipal  Wastewater Treatment  Works Grant Program.  The  authors wish to
acknowledge Gwin Dobson & Foreman, Consulting Engineers,  Altoona, PA for
permission  to publish the study results.


REFERENCES

1)   Gwin,  Dobson and  Foreman Inc.,  1980.  Combined Sewer Overflow  Analysis,
     Altoona,  Pa., Altoona, Pa.

2)   Gwin,  Dobson and  Foreman Inc.,  1983.  Combined Sewer Overflow  impact on
     Dissolved  Oxygen in the Little Juniata River and the  Frankstown  Branch of
     the Juniata  River, Altoona, Pa.

3)   Medina, Miguel A., Jr., 1980.   "Continuous Receiving Water Quality
     Modeling for Urban Stormwater Management,"  Proceedings  of the  national
     Conference  on Urban Stormwater and Combined Sewer Overflow Impact on
     Receiving  Water Bodies,  EPA,  pp. 466-501.

4)   Medina, Miguel A., Jr., 1979.   Level  III:  Receiving Water Quality
     Modeling for Urban  Stormwater Management, EPA-600/2-79-100.

                              T-II-B-ll

-------
5)   Owens, M., Edwards, R., and Gibbs, J., 1964.  "Some Reaeratlon Studies in
     Streams,"  Int. Journal  of Air and Water Pollution. Vol. 8, pp. 469-486.

6)   U.S.  Corps  of Engineers,  1977.   The Hydrologic Engineering Center,
     Storage,  Treatment, Overflow, Runoff  Model  (STORM) Users Manual, Davis,
     Ca.

7)   Wright,  Raymond M.,  and McDonnell, AJ., 1979.  "In-stream  Deoxygenation
     Rate  Prediction,"  Journal of the Environmental Division, ASCE,  Vol.  105,
     No. EE2,  pp. 323-33TI
                              T-II-B-12

-------
Potential and actual  loading  in a small mixed-use  belgian  basin.
Yves DELEU, attach^ at the  Institute  for Hygiene & Epidemio-
logy, Radioactivity section.
This paper emphasizes  the  result of a  study of  the  load
of the Dyle River  in the center of Belgium. This  little  basin
presents a high population density, and  the ground  is  used  for
farming, industry  and  urban purposes.  The  population  is  main-
ly concentrated along  the  riversides.  The  first step  of  the
study was a potential  load approach, on  the basis of  two-year-
ly population, culture and crop, cattle, sheep, pig and  poult-
ry census and non  point source  loading values  in  the  lite-
rature. The second one was the work on four automatic  sampling
stations, on time-mixed sample mode, for one and  half  year,
in the upper basin of  the Dyle  river.  So it has to be  consta-
ted a sharp contrast between  the relatively constant  potenti-
al loading and the quite variable weekly mixed  samples para-
meters values, emphasizing the  preponderant role  of the  potenti
al loading for the gross average load  value and this  of  the
short time meteorological conditions for the variations.
                              T-II-C-1

-------
 Introduction.
 The study of the relationship between pollution potential and
 actual  loads seems to be a greatly difficult question, becau-
 se  of a lot of either less controlled or even unknown parame-
 ters  implicated in their estimation.The socioeconomic and de-
 mographic changes in the given area complicated the approach.
 A great discrepancy exists in the estimation of the  unit po-
 tential loads as computed by Cmernik ( 1976 ) and Volleweider
 ( 1968  ),  who attempted to find a conrmon denominator  between
 the many and diverse values of particular  potenial  load  of
 grass,  cattle, corn and beet field, poultry, industry, depen-
 ding on the slope and the pedology of the ground, a.s.o., ...
 A review of the literature over more than 10 years, suggested
 me  that, despite the revealed works of Cmernik and Vollenwei-
 der, the particular potential loads census is far from termi-
 nated,  because the descriptive parameters of  the many possi-
 ble situations  ( some of them are numbered above )  existing
 for a sole parameter potential effect, as  nitrate,  are  not
 well systematically defined yet. A second discussion element,
 is  the  applicability of the literature data,  which  are rrost
 from an american source, to the european coastal   continental
 theatre, of which Belgium is a part.  For  this  latter scene,
 Vollenweider showed typical values of potential loads (1968).
 The american ground use seems to be of  relatively  the  same
 intensive effect on the soil as the european continent one is.
 This is despite the particular history of  these   soils,  the
 difference of climatic intensities,  and other extrinsic fac-
 tors differentially happening on the  both sides of the Atlan-
 tic Ocean.

 We  had  the opportunity to study a little drainage  basin  for
 three semesters in the years 1978-1979,  with a small prototy-
 pe  sampling network system,  using standard industrial  refri-
 gerated samplers,  but with a rrodification, for remote working
 control  by phone lines.  So,  from week  to  week,   a  complete
 historical  data serie has been recorded  for approximately one
 and a half  year,  at four locations (  a,  b, c and  d on the map
 figure  1  ).  The data were 2-hour,  day and week means ( Glaze-
makers  and Dele,  1980,  Dele 1984 ).  Independent quantitati-
 ve  and  rain data completed the data set.  A full  work  on the
 pedological,  socio-economical   and agricultural   factors  of
 this basin   has been performed,  on the basis   of   statistical
 data,with  the aim to compare the measured loads with the pic-
 tures resulting from the census.
The Upper Dyle basin.  (  figure  1  )

The Dyle Basin is a  little drainage basin at  the east ofBrus-
sels, with three main  affluents:  the  Senne,  where  Brussels
stands, the Demer with Hasselt  as main  town and the Dyle   it-
self, with Louvain and the quasiconurbation Wavre-Ottignies-
                             T-II-C-2

-------
    Fig. 1:  The  Higher Dyle  Basin.
1 .  GENAF'PE
2.  LASNJXS
3.  WAT1XR1 .OC)
4.  LA HU1..PE
r5 .  SI NT-GENES'I US RODE
6.  11 OP! 11..A ART
7.  OVERUSE
8.  HULDENDERG
9.  SINT-JORIS-WEERT
10.  GRKX DOICI.'JMJ
11.  CHAUMONT-GISTOrX
12.  WAL.HAIN
13.  CllA.S'lKl-S
14.  VJL.I FRS LA Vll.l.F.
1 ~>.  COURT SA1 N'l ET J EN NE
16 .  MONT SA I NT Gil I BURT
17.  OTTJGNIKS l.OUVAJN !.A N'HUVE
18.  RIXENSART
19.  WAVKK
                      T-II-C-3

-------
 Louvain-la-Neuve-Court-Saint-Etienne, and then a lot of  sub-
 sidiaries  and municipalities of diverse importances. Its sur-
 face  is  some 2700km2,  with which this study concerns  599km2,
 or  22 5K, called Upper  Dyle basin,  from the spring to the vil-
 lage  of  Sint-Joris-Weert.  The pedology  of  the  upper  basin
 shows the  predominance of  calcarous sands ( table 1 ),  cove-
 red with  loamy sand,  then  supporting the culture horizon. The
 landscape  is various,  with hills  and  valleys, and not truly
 plain surface. Some part of this country is called " Ardennes
 Braban\onnes ".  An estimated 385K of the above mentionned sur-
 face  is  covered with  forestry.  The basin is inhabited by some
 165000 inhabitants, with activities varying  along the basin,
 from  predominently farming at the South to equally agricultu-
 ral and  residential at the North of the partial basin. So,the
 dwelling density differs  from 100 in the predominently  agri-
 cultural zones to 1500 in  the rrostly residential ones: Rixen-
 sart  and Waterloo ( //// 3 and 18 on the map figure 1 ), for e-
 xample. The potential  population is estimated  as  180000  to
 200000 inhabitant-equivalents ( IE ) ( Dufour 1975, De Schep-
 per 1976  ).  Some industrial activity exists, but this  momen-
 tarily figure of some  50000 IE's  ( De Schepper 1976 ) is now
 different  from the present one. Some industries are typically
 bound to farming and  then  seasonal as sugar industry  at  the
 end of each year; some others have now disapppeared ( as  pa-
 per and metal  )  where  they were traditional and new ones  are
 appearing  in quite different zones. From this point of  view,
 this  area  is in  constant evolution. The presented pictures a-
 re  thus valuable for  the years  1978 and 1979. The  communica-
 tion  network is  relatively dense,  with any sorts of ways from
 the truck  path to the  great runways, as this one from Brussels
 to  Namur (  E41 ) and a dense railway. Thus, in a  word,  this
 area  is  typically a mixed-use basin,  with  high dwelling and
 use density, but on a  few  dozen of kilometers ( 20 to 30 mi.)
 along the  river  axis.  As another  singularity of this area as
 other  ones  in  the neighbourhood of great cities, a great num-
 ber of people  travel daily  for work to Brussels, Namur, Cha-
 rleroi or Gembloux so  that certain municipalities can be seen
 as  "sleeping zones" of these main  attractive towns.The poten-
 tial  load  computations don't take   into  account  these daily
 transhumances.
Material and methods.
The sampling program was  started  at  four  points  in   the  area
( a,b,c and d on  the map  figure 1  ),  and  performed with stan-
dard  industrial Buehler PRP/12T type  samplers, with   standard
refrigeration at  4[C and  a  12  2.5  liter  (2/3  USGal)  high den-
sity PE bottles distributor.

                             T-II-C-4

-------
table 1 Pedogeology on the Upper Dyle Basin underground.
  Age             Nature of the formation              Tickness
Pleistocene         Loamy sand (agr.soil)               20m  (60ft)
Brusselian            Coarse quartz sand                25m  (80ft)
Ypresian                  Clayed sand                   40m( 130ft)
Landenian       Fine glauconiferous sand                70m( 250ft)

Masstrichtian                   tuff                    15m  (50ft)
Senonian                       mamas                   15m  (50ft)
Silurian     schists and quartzophyl lades               30m(100ft)
Loy W ( 1978 ).

-------
For  particular  technical  reasons (Deleu 1981),  it was prefer-
red  to  sample  in function of  time,  in place of  in function of
flow. For  the  same reasons, no conservation reagent  has been
added;  only cold at 4[C is used, though it was  reconmanded by
Standard methods ( 1978 ) and Belgian standards (1981),  among
many others.The analysis  delay has  been of one  week after the
last sampling.  The samples were analyzed according to the me-
thods of the Institute  of Hygiene and Epidemiology (IHE  1976)
for  a total of  16 parameters.  From  this set,  6 are discussed
here, because  they seems  to be the  mostly referred in the li-
terature.  They are N03-N, total N, Total P, O-PO4-P, SS  and
ODD. Flow  data  came from the  l£R telemetred hydro logical net-
work jointly operated by  the  former  Noyau  Administratif  de
1'Eau and  by the Rural  Engineering  Department of the  Faculty
of Agronomy,Catho lie University of  Louvain at Louvain-la-Neu-
ve.  ( Persoons  1979, Bazier,  1980 ).  The Meteorological   data
came from  the Royal Institute of Meteorology (   IRM )  (  Van
Diepenbeek 1980 ).  The  data are treated on a Hewlett -Packard
HP97 computer  ( Hewlett-Packard 1977 ) and on a Tandy -  Radio
Schack  TRS80 model 100  portable computer ( Radio-Shack 1983).
For  the estimation  of  the  potential  loading,  it  was   seeking
to different  sources,  either  already written  studies  over  the
subject ( De  Schepper  1976, Laurent  1980, Oovaert 1983 ),  in-
formations graciously  given by  the Ministry ofEconomicAffairs
( 1979,1980  ) and the  National  Institute of Statistics (1979,
1980  ) or still, the known composition  and  quantitites of  the
main  spreaded fertilizers  and manures   during  the asolement
period . Thus,  there exist numerous  approaches  of the  poten-
tial  pollution  loading estimation and comparison. But because
the  statistical  informations  are  far more complete and giving
6 months pictures,  and so  permitting to see the possible va-
riations in the potential  loads along 2 years,  then  the  re-
sults and discussion about potential  pollution  mainly will
concern these statistical  informations.
Results.
Figure 1 is a contour map with  the administrative   limits   of
the municipalities  in the upper Dyle Basin.The  pictures  refer
to the name listing of  these municipalities. The small  I.S.O.
letters indicate  the four sampling  locations.
Figure 2 is an analysis of  the different 6 month censuses  for
the population and human activities  in  the basin (figure 2 t)
and for the culture  ( figure. 2b-),   and the   information   is
treated according to Berlin ( 1977  ). The activities  are  re-
censed following the river  axis and  following   the  values of
                             T-II-C-6

-------
NACE coded industrial activitie
          o  o  o  o  o  o
   FIGURE 2  : human acMvities in the Upper Dylo Basin

-------
 the pollution load unit conversion coefficient for industrial
 used waters ( Moniteur Beige 1974 ).

 The lisibility of such pictures is somewhat particular.  Each
 subfigure owns its own scale. The full black square expresses
 100% value (in which the true value may be either inserted or
 drawn on the upper side ), the other values are  proportional
 to the width of the band.  Under 5%,  which  is  pictured with
 either a sharp line or a blank, the values are  expressed  in
 the closest round fraction of a disk, as  7/8, 3/4, 4/8, 1/2,
 3/8, 1/4, 1/8,... A fraction of a thenth or less is indicated
 with a little sharp vertical upper line.A fraction of a twen-
 tieth less is drawn with a little sharp vertical lower line.
 When it is zero,  the figure is O. The occasionnally  seasonal
 activity ( which is expressed with a 25 %  variation  between
 two consecutive 6-month actvities ) is shown by  partitioning
 the square figure in two triangles ( upper: surnner, June, lo-
 wer: winter, december) with a diagonal,in wich both triangles
 the above mentionned code is also  valuable.  The  industrial
 and service activities are distributed according to the  N/*CE
 code,into categories numbered from 0 ( farm and natural acti-
 vities ) to 9 ( services ) ( Gewestelijke  Ontwikkelingsmaat-
 schappij voor Vlaamse Brabant 1979 ). The relative quantities
 (in percent) for  a given N/°CE category is written in front of
 the concerned municipality. Contrary to the other  comparison
 items, showing a vertical  comparison line between municipali-
 ties,  along the river axis, the distribution is here horizon-
 tal for the same municipality.  By this way,the repartition of
 the human activities and the residential zones becomes  quite
 well seen, with the dynamics of the  changes  along the river
 axis and the year.

 The industrial activities  are also taken into account as  po-
 tential pollution load on  a twin basis according to the  N/CE
 distribution on one side,  and according  to  the standardized
 industrial equivalent-inhabitants (HE),  which is  deduced on
 the legal  definition of the equivalent-inhabitant.  (  Moniteur
 Beige 1974 ).  The definition of the IE has been discussed el-
 sewhere (  Deleu 1983,  Vanderborght 1981 ).  Unfortunately, the
 lawmaker didn't use the HflCE. industrial activity distribution
 code and considered a little amount of the  classical   indus-
 try,  with  an arbitrary insistance over unattended details.The
 distribution of  the recensed activities is then dependent for
 a  given part to the subjectivity of  the user.But the  numerous
 chosen N/"CE categories  are sufficient enough to  minimize the
 possible estimation errors.  The resulting IE are converted to
 potential  pollution loads  according  to the I E legal   defini-
 tion.  It becomes  then  apparent  that  such detailed approach of
 the potential  polluting EH,  even if  misestimated for  the abo-
 vementioned reasons,  is much higher  than  the preceeding esti-
mations  (  Dufour  1975,  De  Schepper 1976,  Oovaere 1981  ).

 Thus,  owing to the  loss of  data due to non recensed ground u-
                              T-II-C-8

-------
ses  (ways, aso.,...) and  the  uncertainty  of  some  of   the  in-
dustrial activities, the  potential   pollution   estimates must
be considered as partial  ones,  even  if  this  part  represents
the  great majority of  the true  estimation.
Figure 3 shows  the variations of potential  (  a)  computed  from
the  literature minimum values, b) computed  from  the  literatu-
re maximum values, c)conrputed from  the  non  industrial  equiva-
lent  living population,  including the recensed cattle, d)com-
puted from the  industrial equivalent population,  as   recensed
following the NAZE code  ) and actual 6  months  loadings (e)  at
the  four sampling sites  and  for the six analyzed parameters.
The  values are expressed in metric  ton/6 months.  The  lecture
code  is the same as  for  figure 2. Here, the seasonal activity
is visible by itself, because the shown values are   6-monthly
ones.
Figure 3 presents data corrected with   respect  to  the   length
of a semester  in weeks, e.g.26 weeks/haIf-year,  and  converted
from kg/s to metric  tons/6 months, with  a  time  conversion  co-
efficient from 86400 second  a day  to week  of  7  day and  a  se-
mester of 26 weeks. For each partial basin,the  picture  is  the
expression of  a specific  income  load in   the   river   for  the
concerned parameter over  6 months.
 Figure 4 presents a distribution of  the  actual  loads  between
the runoff  ( white area  ),  the hypodermic (delayed)  dischar-
ge ( hatched area) and the  ground discharge  (black area).
The picture is drawn on  the same scheme as  the preceding  one.
The symbols are  the same as above excepted one more  :  a squa-
red one indicates insufficient data.When  a   particular  dis-
charge is estimated equal to zero,  the figure shows  a   O.   If
the discharge value is under  5%, a  as above   fractionned  disk
is drawn.over the zone,with the same  drawing  code of the  dis-
tribution.The square hatched zone of  the  disk is without  mea-
ning.
The values are expressed with respect to  the  percent!le  dis-
tribution;  100 % is the  value of each squared subfigure.
Discussion.
It is first worth saying  that every measured   values  conside-
red here are mean values, not point values.  The   total   grab
sampling number  is within the 36000/year.station.  The results
and discussion are dramatically affected with  such data   col-
lection because  the majority of the past discussions  are   re-

                             T-II-C-9

-------
H
I
                PARAMETER
               POTENTIAL POLLUTION
               LOWER LOAD ESTIMATE

                      T /6 months

                  CULTURES
             B)  POTENTIAL POLLUTION
                HIGHER LOAD ESTIMATE

                      T /6 months

                  CULTURES
C)  POTENTIAL POLLUTION
   LOAD ESTIMATE

       10* T /6 months
NONINDUSTRIAL EQUIVALEN
POPULATION (people-<-cattle
             D)   POTENTIAL POLLUTION
                 LOAD ESTIMATE

                        T/6rponths

             INDUSTRIAL EQUIVALENT
             POPULATION.
             ACTUAL POLLUTION
             LOAD  ESTIMATION
                         T/6 months
NN  N  N
NN  N  N
NN  N  N
NN  N  N
NN  N  N
 NN  N  N
 NN  N  N
 NN  N
                                  FIGUi
-------
H
I
n
      NO-N
     1978
1979
     N
n
      total-N
1_978

1
1979
      N
      N
N
n
            total- P
1978
 1979
                            N
N
EC
n n
                  o-PO-P
         1976
                                        jMm
      1979
N  N
N
N
DD
                          SS
      ma
           samp
      ing sites
   FIGURE4:  percentile distribution  of actual  loads  between ground discharge, hy-
            podermical discharge  and runoff

-------
 ferring  to  grab samples,  which are taken  at   different  time
 frequencies,  or with a more or less complex relation with the
 flow.

On  the other  hand,  by the census recensed surfaces are  lower
 than  the above  cited picture of 599 km2. It remains  about 85
km2 ( 1* %  )  out of the census, which are taken back  as  not
recensed, for example some routes  and forestry.
The pathways  and the forestry are  taken into  account by them-
selves for  the  actual load computation, but in the  potential
 load  estimation not, excepted some  forestry,  as   rural ones
 ( National  Institute for  Statistics 1979, 1980 ).

On  the figure 3,  the effect of the variations from a semester
 to  the other  is remarkable, for the six concerned   parameter,
even  in  the case of the actual pollution loads. The alternan-
ce  in the loads is  quite  apparent, giving  an  impression  of
predominance  of the seasonal activities on the otherones.This
same  variation  character  of both potential and actual loads
suggests that the potential pollution pressure also ownsfrac-
 tal properties, as  the actual ones have (Pearsons  coefficient
values up to more than 150 % ).  By the way, it was shown else
where (  Dele  1984 ), that the quality  variations  follow the
 laws  of  fractal curves, or the Mandelbrot theory ( Mandelbrot
 1975  ) from which the most remarkable property is  the conser-
vation of the standard deviation  variations   at  every scale
these variations  are examined.  The load pressure  was assumed
constant because the normal way of life  of  the  populations
and the  indefatiguable industrial  activity. Actually, the po-
tential  loading is  quite  more dynamic and variable than the a
priori computations can show. Thus, it seems  to be worth wor-
king  the calculations at  smaller scales than  the abovementio-
ned 6 months, for greater informational precision.

The second  remarkable element is,  shown  in the figure 3,  the
great difference  between  potential  (  figure 3 a minimum, 3 b
aximum,  3c  non  industrial EH, and  3d  industrial EH )  and  ac-
tual  ( figure 3e  )  loads  estimations,  in the  order  of 1000 to
1.
The import  from population and livestock appears much more im-
portant  than  the  culture  import. The  actual load appears clo-
ser of the  higher culture load estimations and a few tenths
of percents of  the  population and  livestock E H and its legal
conversion.
Furthermore,  the  potential  estimation  is but  partial,because
not including the   non  recensed  sources,though the actual  es-
timation  concerns everything passing   through the  Dyle.  Logi-
cally, both are estimated with the  same amount of   population
and ground  use, minus what  wasn't  recensed in  the  potential
estimate.
The most  frequent (  more   than 95 % in frequency )   polluting
load import  is  concormuttant  with  occurrence  of rain,superfi-
cially appearing  to  have  the central  role in  load   transport
and then which  could  explain the  alternance   from  a  semes-
ter to the another.   But the few percents  (  5  to 30  % )  of  the

                              T-II-C-12

-------
actually  transported  loads  by  net   rains   in  the  case of  some
parameters,forbit  to  consider  its  role  as   of   the  most   im-
portant. In  other wards,  the role of  the rain  seems less  being
the cause of  transportation than its   throttle.  In  non   rai-
ning conditions, the  ground discharge   withdraws  the greatest
part (from  80 up to  100  %)  of  the  total charge coming  up  to
the river,  despite of  the  lower  flow, with stagnation and se-
dimentation  in  the waste pipes. When raining,the  runoff with-
draws a great part (  from  30 up  to 70 % )  of  the  total  charge
and the resulting  sedimented loads are   pushed down with the
runoff and  then become active  in the   river water, giving the
well known  pollution  peaks.  But, an analysis  of  the base  line
of these events shows  a  great  contribution of the hypodermic
discharge to  the polluting  load  (up  to  80  %),  with a time de-
lay of about one week.The  remaining of  the charge (up to  75%)
is taken  in  charge with  the ground discharge  and   constitutes
the basis of  the loading.  Sometimes, for nitrate-N,the  ground
discharge does  not exist.

The effect  of a seasonal activity  on the water quality  as the
yearly sugar campaign  (  Muller 1966  ),  may be  quite expressi-
ve and dramatically change  the parameter values even for  are-
latively  long time length:  two months,e.g.one  third of  semes-
ter ( Glazemakers  and  Deleu, 1980  ).  But  as  it could be  seen
on the figure 3e,  this  local effect  remains  invisible  in the
whole load mass, despite the time  rate.

The dramatic difference  between  the measured  charges and the
potential load  computation  results will not  lead to conclude
that the earlier work must  be  repeated. It evently  could  be
computed according to  the  unit values of Vollenweider (1968),
but the parametrization  of  these unit values  is not known yet
for the Upper Dyle Basin. This difference  overall  underlines
a great difference in  the approaches   in   measuring  and   in
computing.

On the other side,it  could  also  be the  sign of a  great  reten-
tion potency of the polluting  load in  the  basin.  This  reten-
tion could  be either  physical  (  long time  sedimentation,   at-
mosheric elimination  of  NH3 ( Loehr 1974  ) for example,  ) or
depending on a  long delay  discharge, or degradative ( natural
purification  ), or still the three phenomena  together as   Be-
noit  ( 1972  )  showed  in the case  of natural waters. The  form
of the purification processes  could be  rrore general and comp-
lex than assumed by Benoit,  it is  to say that   they occur not
only in the watercourses. Then,  at the  next step, the measu-
red load  is  only a small fraction  of  complex   origin  of the
initial load computed  as potential load.   Further, it appears
that this purification process is  of quite high   efficiency,
on a time delay on only  a  few  days in  the  river  for the   run-
off ( Bazier  1981  ),  one week  for  the   hypodermic  discharge.
On quantitative basis, the  known low values of runoff coeffi-
cient ( .1  to .32  following De Pelsmaker ( 1980  ), .01  to .15
following our own  less rigourous estimations  ) are assumed to

                             T-II-C-13

-------
 fill  a part of the process conditions. In river water,the in-
 tensity of these degradation for NH4  is  high ( 80  to  95 %
 oxidation )( Curtis, Durrant and Harman 1975 ). Joining these
 two values, 0.01 for the runoff coefficient  and .05 for  the
 degradation, one can estimate the remaining NH4-N load to the
 .05 % of the initial potential load.This is within the actual
 load  of the NH4-N relative parameter N03-N.The final explana-
 tion  lies thus in a combination of degradation  and quantity.
 It  could maybe remain elsewhere,but with greater uncertainty.
Conclusions.
The  above presented results indicate an unexpected great dif-
ference between potential and actual loads.This great  diffe-
rence  between room calculations and field measurements  indi-
cates  a retention potency of the polluting charge  along  the
time and resulting in a quite efficient loss of load. A begin
of complete explanation can be presented in a  combination of
degradation chemistry and hydrology.

The  alternance of the results from 6 to 6 months, indicates a
seasonal  effect on both potential and actual pollution loads
showing the possible short term variation  of  the  potential
load like the actual load variations.

During this study, the use of high density data,  as automatic
permanent samplers can collect, appear  quite  potent  in  the
approach of long term complex water quality phenomena.

Acknowledgements:  The autor is quite indebted to  Mr E. Glaze-
makers,  Technical  Engineer, who was in  charge of  managing the
sampling network,  to SSrs 3.  Sadones and C.Van Eerdenbrugghe,
Technical  Engineers, who helped the work and accomplished pa-
rallel  programs on the Dyle R.
He wants  too to thank Mr. Bouquiaux, Chief,  Environmental De-
partment,  who supported the project, Mr.  Ir.  E. Laurent,  for-
mer  Director of the NAE, for  the command of   the   project and
for  the  quantitative values,  Mr Van  Diepenbeek,   Attach6  at
the  Royal  Institute for Meteor logy,for  the meteorological da-
ta,  Mrs.Bavin and Mr.  Coevelier,of the  National  Institute for
Statistics,  for the numerous  census listings,  Prof.  L.De Bac-
ker, former  Dean,  Prof.  E.  Persoons and Mrs  G.  Bazier, Assis-
tant,  of  the Rural  Engineering  Department,  Faculty of Agrono-
my,  University  of  Louvain la  Neuve, for critical   discussions
and  practical  helps,  Mrs.  Braun and Mr. Bastin of  the IBW for
helpful discussions and data,  Dr.  P. Lejeune,   Head,  Radioac-
tivity  section,  for his encouragement,  Mrs.Martur-De Vre   for
the  revew the manuscript,  and Tandy computer Centre  for   his
kind help  for  printing the  manuscript.   This  project  was  sup-
ported as  an extern program of  the N/°E,Dyle Corrrnission,Minis-


                             T-II-C-14

-------
 try  for Public Health  and Family.
                        Bibliography.
Bazier G  (  1980  )  in Monographic du Bass in de  la  Dyle, E.
Laurent ed., Ministere de  la Sant£ Publique, Noyau
Administratif de  1'Eau, Corrmission Dyle, Brussels,  1980  (
Dyle Basin Monograph, E. Laurent ed., Ministry  for  Public
Health and  Family, Water Administration Cell, Dyle
Conrmission  1980  )( in French ).

Belgian standards  (1980) Echantillonnage des eaux -   ^
Principes ge"ne"raux NBN T 91-051, Belgian Insitute for
Standardization  ( Water sampling - Fundamentals  ),  Brussels
1980 ( in French  ).

Belgian standards  (1980) Echanti1lonnage des eaux -
Reconmandations g£n£rales  pour  la conservation  et la
manipulation d'<§chanti lions NBN T 91-052, Belgian Insitute
for Standardization ( Water sampling  - General
Recommandations  for samples conservation and manipulation  )
Brussels  1980.

Benoit RJ ( 1972  ) Self-purification  in Natural Waters,  in
Water and Water Pollution  Handbook, L. Ciaccio  ed.  M. Dekker
New York  1972.

Bertin 1977 La graphique et le  traitement graphique de
1'information, Flamnarion, Paris 1977. ( in French  ).

Buehler PRP/12T  type sampler (  in German ).

Curtis, Durrant and Harman ( 1975 ) Nitrification in  rivers
of the Trent Basin, Water Research 1975, 9, 268.

Deleu Y ( 1981 ) L'<§crianti llonnage appliqu6
1'e'chanti llonnage automatise" special  licence memo ire,
Louvain-la-Neuve  1981 ( in French ).

Deleu Y ( 1983 ) L'expression du bilan massique du  bassin
moyen de  la Dyle en termes d'Equivalents-Habitants  (
manuscript  ) ( The net load yield expressed in  Inhabitant-
Equivalent  units  )( in French ).

Deleu Y ( 1984 ) Le caractere fractal des courbes de
qualite" de  1'eau Tribune du CEBEDEAU  1984, 37,n[489,  pp90-
101 ( Fractal curves in water quality. ) ( in French  ).

De Pelsmaker ( 1980 ) Ame"nagement hydraulique du  Bassin  de
la Dyle (Dyle Basin Hydraulic Management , a monograph,
Agricultural Hydraulics ), Ministry for Agriculture, Mons
1980 ( in French  ).

                              T-II-C-15

-------
 De Schepper  et al.  (  1976 ),  The chemische en
 fysicochemische verontreiniging van de Dijle, IHE,  Ministry
 of Public Health and  Family,  Brussels, 1976 ( The chemical
 and physicochemical pollution of the Dyle )(  in Dutch ).

 Dufour 3 ( 1975 ) La  collecte des eaux use"es  dans la Valise
 de la Dyle,  in Le Cycle de 1'Eau  Louvain-la-Neuve, 19 mars
 1975, Socie"te" de Ge"nie Rural, Louvain-la-Neuve, 1975, ( Used
 Waters collection in  the Dyle Valley, in The  Water  Cycle at
 Louvain-la-Neuve, March 19th, 1975, Society for Rural
 Engineering  ).

 Gewestelijke Ontwikkelingsmaatschappij voor Vlaams  Brabant:
 Bedrijven in Vlaams Brabant adresboek 1979 Gewestelijke
 Ontwikkelingsmaatschappi]  voor Vlaams Brabant, Vilvoorde
 1980 ( Regional Development Association for Flemish Brabant
 :  Trade and  Industry  in Flemish Brabant addressbook 1979
 RegionaT Development  Association for Flemish  Brabant )( in
 Dutch ).

 Glazemakers  E and Deleu Y ( 1980 ), unpublished results.(
 in Dutch )

 Hewlett-Packard ( 1976 ) Stat Pac I,co Hewlett-Packard
 Singapore 1976.

 Hewlett-Packard ( 1976 ) Standard Pac, co Hewlett-Packard
 Singapore 1976.

 IHE ( 1976 ) Metoden  voor  de  fysicochemische  analyse van
 water, Institute for  Hygiene  and Epidemiology, Ministry for
 Public Health and Family,  December 1976,  Brussels (
 Physicochemical  analysis methods for waters )( in Dutch ).

 Laurent E. (  1980 ),  in Monographic du Bass in de la Dyle,
 E.  Laurent ed.,  Ministere  de  la Sant£ Publique, Noyau
 Administratif de 1'Eau,  Conrmission Dyle,  Brussels,  1980 (
 Dyle Basin Monograph,  E.  Laurent ed.,  Ministry for  Public
 Health and Family,  Water Administration Cell,  Dyle
Comnission 1980  )(  in French  ).

 Loehr  RC ( 1974  ) Characteristics and comparative magnitude
 of  non point  sources  JWPCF 1974,  46,  8,  1849.

 Loy W (  1978 ) Hydrogeological  guide-book - The Dijle
Valley.  Hydrographica 1978, p25.

Mandelbrot B.  (  1975  )  Les  objets fractals: forme,  hasard et
 dimenesion,  Flanrmarion  Paris  1975 (  The fractals  Things:
 Shape, Randomness and Dimension,  Flanmnarion,  Paris  )( in
 French ).

Ministry  of Economic  Affairs  (  1979  ) Occupation  des
 terrains  @ 1'usage  de 1'industrie,  de 1'artisanat ou  des
 services,  situation au  31.12.1978, Ministry of Economic
Affairs,  Administration  of  industry,  economic  expansion and
 foreign  investments,  Brussels  1979 (  Soil  occupation  and
 industry,  manufacturing  and servicing,  situation  at
 12.31st.1978  )(  in  Dutch and  French  ).

                              T-II-C-16

-------
Ministry of Economic Affairs  ( 1981  ) Occupation des
terrains   1'usage de  1'industrie, de  1'artisanat ou des
services,  situation au  1.1.1980, Ministry of Economic
Affairs, Administration of  industry, economic expansion and
foreign investments, Brussels 1981 ( Soil occupation and
industry, manufacturing and servicing, situation at
1.1st.1980  )(  in Dutch  and French ).

Moniteur Beige ( 1974 ) Arrt£ Royal du 23 Janvier 1974
fixant les minima de charge polluante preVus aux articles 9
™ 2, 3 et  20 b de la  loi du 26 mars 1971 sur la protection
des eaux de surface centre  la pollution. ( King Arretee of
January, 23th,  1974 stating,  the minimal polluting loads,
foreseen in articles 9  ] 2, 3 and 20 b of the Law of March
26th,  1971, upon the surface water protection against the
pollution,

Moniteur Beige ( 1974 ) Arrt£  minist<§riel du 15 fe"vrier
1974 determinant en execution de  1'article 4 "2, de la loi
du 26 mars  1971 sur la  protection des eaux de surface centre
la pollution ,  la valeur des coefficients de conversion en
unite's de charge polluante pour  les eaux provenant des
entreprises industrielles ou autres.  ( Minister Arretee
determining, for the accomplishment of the Law of March
26th,  1971, upon the surface water protection against the
pollution,  the conversion coefficients values in pollution
loading units  for used  either industry or others waters ) (
in Dutch and French ).

Muller ( 1966  ) Treatment of mixed domestic sewage and
industrial wastewaters  in Germany CECD report Pris 1966.

National Institute of Statistics  ( 1979 ) population census
1978 -municipalities of Sint-Genesius-Rode, Hoeilaart,
Overijse, Huldenberg, Genappe, Villers-la-Vilie, Chastre,
Court-Saint-Etienne, Walhain, Ottignies-Louvain-la-Neuve,
Mont-Saint-Guibert, Grez-Doiceau, Chaumont-Gistoux, Sint-
3oris-Weert, Lasnes, Waterloo, La Hulpe, Rixensart ( in
Dutch and French ).

National Institute of Statistics  ( 1980 ) population census
1979 -municipalities of Sint Genes ius-Rode, Hoeilaart,
Overijse, Huldenberg, Genappe, Villers-la-Vilie, Chastre,
Court-Saint-Etienne, Walhain, Ottignies-Louvain-la-Neuve,
Mont-Saint-Guibert, Grez-Doiceau, Chaumont-Gistoux, Sint-
Joris-Weert, Lasnes, Waterloo, La Hulpe, Rixensart ( in
Dutch and French ).

National Institute of Statistics  ( 1979 ) agricultural
surface uses census 1978 (  listing )  - municipalities of
Sint-Genesius-Rode, Hoeilaart, Overijse, Huldenberg,
Genappe, Villers-la-Vilie, Chastre, Court-Saint-Etienne,
Walhain, Ottignies-Louvain-la-Neuve, Mont-Saint-Guibert,
Grez-Doiceau,  Chaumont-Gistoux, Sint-Joris-Weert, Lasnes,
Waterloo, La Hulpe, Rixensart (  in Dutch and French ).

National Institute of Statistics  ( 1980 ) agricultural
surface uses census 1979 (  listing )  - municipalities of
Sint-Genesius-Rode, Hoeilaart, Overijse, Huldenberg,.

                             T-II-C-17

-------
  Genappe, Villers-la-Ville, Chastre, Court-Saint-Etienne,
 Walhain, Ottignies-Louvain-la-Neuve, Mont-Saint-Guibert,
 Grez-Doiceau, Chaumont-Gistoux,  Sint-3oris-Weert,  Lasnes,
 Waterloo, La Hulpe, Rixensart (  in Dutch and French )

 National Institute of Statistics ( 1979 ) cattle,  porcine,
 horses,  and farm yard census 1978 ( listing ) -
 municipalities of S in t -Genes ius-Rode, Hoeilaart, Overijse,
 Huldenberg, Genappe, Villers-la-Ville, Chastre, Court-Saint
 Etienne, Walhain, Ottignies-Louvain-la-Neuve, Mont-Saint-
 Guibert, Grez-Doiceau, Chaumont-Gistoux, Sint-Joris-Weert ,
 Lasnes,  Waterloo, La Hulpe, Rixensart ( in Dutch and French
 National  Institute of Statistics  (  1980 )  cattle,  porcine,
 horses, and farm yard census 1979 (  listing )  -
 municipalities  of Sint -Genes ius-Rode,  Hoeilaart, Overijse,
 Huldenberg, Genappe,  Villers-la-Ville, Chastre, Court-Saint-
 Etienne,  Walhain, Ottignies-Louvain-la-Neuve,  Mont-Saint-
 Guibert,  Grez-Doiceau, Chaumont-Gistoux, Sint-Joris-Weert,
 Lasnes, Waterloo, La  Hulpe,  Rixensart  ( in Dutch and  French
Cmernik 3M (1976)  The Influence  of  Land Use  on  Stream
Nutrient Levels, EPA-600/3-76-014,  US Environmental
Protection Agency, Washington  DC.

Oovaert A.  et  al.  (  1983 ) Optimal isatie van de
water zuiver ing door  mi dd el van kaarten met de biologische
kwaliteit van  de water lopen  -  Bekken  van de  Dijle  tot
Wrechter, Groep voor Toegepaste  Ekologie vzw, IHE  Brussels
1983  (  Optimization  of water purification with  mean of
watercourses biological  quality  maps  - Drainage basin of the
Dijle River at Wrechter  ( Belgium  - Working  Group  for
Applied Ecology )( in Dutch  ).

Persoons E. (  1979 ) Application du systee LGR  en  Belgique,
in Problemat ique et  gestion  des  eaux  int^rieures, Calembert
ed. Derouaux,  Liege  1979 ( LGR systems applications in
Belgium in  Inland  Water  Problematics  and Management )(  in
French  ).

Radio-Shack (  1983 ) TRS-80 Model  100 Portable  computer
User Book,  Tandy Corporation,  Fort  Worth 1983.

Standard methods ( 1978  ) for  analysis of waters and
wastewaters APHA-AWW 1978.
Van Diepenbeek ( 1980 ) meteorolgical  data for  Beauvechain,
Heverlee and Ottignies for years 1978 and 1979  ( listing )

Vollenweider (  1968  ) Recherches sur  rame"nagement de 1'eau
- Les bases scientif iques de 1 'eutrophisation des  lacs et
des eaux courantes sous  1 'aspect partiel  du  P et de N comme
facteurs  de 1'eutrophisation, OCDE  Paris  1968 - Scientific
fundamentals of the  eutrophication  of  lakes  and  flowing
waters, with particular  reference to  nitrogen and phophorus
as factors  in  eutrophication. CECD  Paris  1968.

                             T-II-C-18

-------
              MODELING  SEDIMENT  DELIVERY  FROM  FIELDS TO CHANNELS;
          A CASE FOR FIELD SCALE DATA AND HIERARCHAL DATA STRUCTURES

                                 Kenneth Baun
                           Environmental  Specialist
                  Wisconsin Department of Natural  Resources

                                   ABSTRACT
There is a broadscale awareness of and concern regarding the water quality
problems caused by nonpoint source pollution.   There is also a growing
committment to abating this type of pollution.  In order to direct nonpoint
pollution control activities in the most cost-effective manner,  it is
necessary to understand the erosion, sedimentation and sediment transport
processes in a watershed.  This understanding is best developed through the
use of a mathematical model.

There are several watershed models available today that can be used to
evaluate erosion, sedimentation and sediment transport.  They vary
considerably in their input requirements and output abilities.  These models
generally require that a watershed be subdivided into many small, square,
cells.  The unit of data is a grid cell.  Data for the model is then
collected, entered into the computer, and analyzed on a cell by cell basis.
From the modeling experiences associated with the Wisconsin Nonpoint Source
Program, it appears that this grid cell technique is not the best way to
divide, collect and analyze watershed data.

An alternative modeling approach was subsequently defined and is presented.
This approach incorporates field scale data and hierarchal data structures.
Field scale data means that the unit of data is the farm field.   Data is
collected on a field by field basis.  The use of field scale data allows for
more homogeneous cell attributes and more applicable field analysis and
management plans.

The  use of field scale data also necessitates the use of hierarchal data
structures.  Hierarchal data structure refers to the manner with which data is
accessed and processed in a computer program.   Hierarchal data structure
indicates a data storage technique by which all cells are accessed only by
means of their relationship to other cells.  The use of hierarchal  data
structures parallels the downslope movement and aggregation of runoff, and is
the most appropriate data structure for the process being modeled.   Most
significantly, the use of field scale data and hierarchal data structures
enables a model to be used to estimate the delivery of eroded sediment from a
field to a channel.

Key Words: sediment delivery, sediment yield,  delivery ratios, sediment
           trapping, nonpoint pollution.
                                  T-II-D-l

-------
 INTRODUCTION

 The  Wisconsin  Fund  Nonpoint Source  Program was initiated in 1978 to identify
 and  rectify  nonpoint pollution  problems  in the state.  This program, with a
 current annual  budget of  $5,600,000,  provides state cost-share dollars for
 nonpoint controls in targeted areas of designated Priority Watersheds. The
 Nonpoint Source Section of  the  Department of Natural Resources ("the
 Department") has the responsibility for  developing a watershed plan for each
 Priority Watershed.   Each plan  includes  an identification of the nonpoint
 pollution problems  in the watershed,  the water quality objectives, the
 critical  sources (eligible  areas) of  nonpoint pollution, and the applicable
 best management practices.   Each year there are four to six new watersheds
 added  to the program,  each  of which is usually 250 to 750 square kilometers
 (100 to 300  square miles) in area.  The  analytical tools used in the
 identification  of critical  areas, the selection of remedial measures, and the
 estimation of  resultant water qualtiy impacts, must be appicable to large
 areas  in a timely and cost-effective manner.

 Extensive land  use iventories are conducted on each watershed.  Currently, all
 fields within a  quarter-mile of a channel are inventoried for location,
 landowner, soils, topography, land  cover, rotation and other management
 practices.  These inventories are conducted on a field by field basis by local
 county agencies  under contract  with the  Department.  An average annual rate of
 erosion is derived for each  field using  the Universal Soil  Loss Equation
 (USLE).  In the most  recent  round of watershed plans, one of the watersheds had
 over 8500 fields for which  the  USLE was  calculated.  Following these soil  loss
 calculations, the watershed  planner,  in  conjunction with the local agencies,
 selects a target soil  loss  rate and a prioritized list of best management
 practices applicable to the  watershed.   The computer then "applies" these
 practices one at a time to each field with a soil  loss rate above the target
 level,  recalulating  the soil loss rate after application of each practice,
 until  the rate is either below  the target level  or, infrequently, all
 practices have been  applied.  The fields are then aggregated by landowner.
 Landowners are in turn prioritized for eligibility in the program based upon
 the  anticipated  total  reduction in soil   loss.

 This method has  several advantages: The  data base is similar to that used by
 several county and state resource management agencies.  This sometimes allows
 for  shared data  and  reduced  data gathering time.   It enables broadscale data
 collection to be carried out over many large areas simultaneously and requires
 no sophisticated data gathering tools or precise geolocator techniques.   It
 also gives the local agencies a significant role in the planning process,
which  gives them a large measure of "ownership"  in the plan.   This method has
 two  strong disadvantages however.  It does not distinguish  between differences
 in sediment delivery between fields, nor is it capable of estimating watershed
 sediment yield.  Rather it treats all  inventoried fields identically,
 regardless of the ultimate delivery of eroded  sediment from each field to
 receiving waters.
                                  T-II-D-2

-------
Consequently, it does not distribute limited money for nonpoint  source
controls in the most cost-effective manner.  In lieu of a  better  tool, it
relegates this phase of the program to a function of soil  erosion control
rather than water quality control.

Recognizing this, in 1983 the Department procured a Federal  grant to
incorporate the ANSWERS model into  it's planning and evaluation  activities.
ANSWERS is a distributed parameter  hydrologic and sediment detachment and
routing model, intended for primarily agricultural  watersheds  (Beasley and
Muggins, 1982).  A distributed parameter model,  unlike a  lumped  parameter
model, divides a watershed into many discrete cells.   ANSWERS, like several
other distributed models (MODANSW [Park and  Mitchell,  1983]  and  AGNPS LBOSCH
et al,1983]) is a grid based model.  A square grid is  placed over the
watershed, and a row and column locator is assigned to each  cell.  For the
ANSWERS model, the location (row and column), slope,  direction of slope, soil
type, land cover and channel class  (if any)  of each cell   must be designated.
Several variables are required for  each soil  type, including total porosity,
field capacity, infiltration descriptors, antecedent soil  moisture, and USLE
"K" value.  Several  more variables  are required for each  land cover, including
rainfall interception capacity, surface cover, surface shape descriptors,
Manning's N, and USLE "C" and "P" values. Several  of these  variables
fluctuate seasonally, requiring several  data bases for different crop stages.

ANSWERS, like other distributed models,  is event specific, i.e., it models
runoff and sediment concentrations  for single events.   It also operates in a
stepped time mode, i.e., under event simulation, the conditions  of each cell
are calculated and updated every few minutes for the  duration of the event.
Because of its distibuted nature and stepped time mode, ANSWERS  requires large
amounts of data and is computationally intensive for large watersheds.

The primary advantages of a distributed model like ANSWERS is that it can
estimate net sediment erosion/deposition on  each cell, total watershed
sediment loadings, and the overall  impact of changing land management on
various cells on sediment loadings.  However, there are two  drawbacks to a
distributed model like ANSWERS that, indirectly, are the  subject of the rest
of this paper.  The first is the difficulty  and inapproprlateness of
collecting and processing land based data, especially for large  areas, in a
grid based format. The second is the inability of a computer internal grid
based data structure to estimate the eventual deposition  or  delivery of
sediment from individual fields to  receiving waters (as opposed  to the net
sediment movement on each cell).  In response to these drawbacks and to the
needs of the Wisconsin Nonpoint Source Program,  an outline of a  proposed
sediment delivery model, incorporating field scale data and  hierarchal data
structures, is presented.
                                  T-II-D-3

-------
 DISCUSSION

 Field scale  data

 The  data  format that  has most frequently been used in watershed modeling of
 sediment  erosion  and  delivery has been the grid cell approach.  Using this
 method, a grid  is placed over a map of the area being modeled, which is then
 divided into numerous identically sized cells.  The cells are thereafter
 identified by row and column numbers and they become the unit by which data is
 collected, entered into the computer, and processed. The utility of the grid
 cell  approach can be  examined in light of its attributes relative to those of
 a field scale data approach.

 To its credit,  the grid cell method enables an easy locator to be associated
 with  each cell, i.e.,  any given row/column designation can be readily located
 on a  map.  Therefore,  the input or output of the model can be easily graphed
 and visually displayed.  Further, in the case of ANSWERS and MODANSW, it
 allows for a more accurate determination of runoff direction from each cell,
 i.e., the direction of flow is highly variable and can be proportioned between
 two adjoining cells.   In the case of AGNPS, the direction of flow must be
 designated to one of  four directly adjacent cells or four diagonally adjacent
 cells.  Lastly, because of the rigid and simple relational locators, it makes
 programming  and modifying grid based models a relatively simple task.

 There are of course a  few major drawbacks in using the grid cell approach.
 One relates  to the size and homogeneity of the cells.  The size of the cells
 should be a  function of the diversity and the size of the area being modeled.
 The smaller  the cells, the more homogeneous their contents, the greater their
 number.

 Tradeoffs will often have to be made, sacrificing the homogeneity of smaller
 cells in  favor of the  practicality of larger ones.  In the 727 hectare
watershed that the Department calibrated and evaluated the ANSWERS model  on, a
 cell   size of one  hectare was employed.   Using this same cell  size on the
 Priority Watersheds would require generally 25,000 to 75,000 cells for each
watershed, obviously a prohibitive number on any account.  Using larger cells,
 perhaps ten  hectares, would alleviate that problem but would result in another
 one.   In much of Wisconsin's diverse topography, ten hectare cells in a grid
 system would yield very high diversity in the many cells that encompased two
or more very diverse land covers, slopes, soils, etc..   The authors of the
AGNPS model   (Bosch et al, 1983)  suggest using a four hectare  (ten acre) cell
size for watersheds less than 800 hectares (three square miles), and 16
hectare (40 acre)  cell size for larger watersheds.  Given the diversity found
in most of Wisconsin's landscape, cells of the latter size would prove far  too
heterogeneous for accurate representation of their contents.

In a  recent related study by the author,  the soil  diversity of a 1024 hectare
area  of Dane County, Wisconsin was examined.   The soils of the area had
previously been digitized,  and this analysis was performed by computer.  A
grid  matrix composed of one hectare cells was superimposed over the area, and
                                  T-II-D-4

-------
the number of different soil classifications in each cell  was counted.   Only
153 cells had just one soil classification within them.   Of the other 871
cells, 447, 313, 79, 28, and 4 cells contained, respectively, 2, 3,  4,  5,  and
6 different soil classifications.

A second problem related to using grid cell data involves  making field
management decisions based on grid cell model  outputs.   The entire purpose of
running a model is to make field management decisions related to water  quality
objectives.  Yet how are management decisions made when  there are cells
straddling more than one field, or multiple cells in a  field indicating
divergent net soil movement?  Field management is not typically based on or
applied to a small portion of a field, but rather it is  based on and applied
to the entire field.  Grid cells are less than ideal for making field scale
management decisions.

A third formidable problem of the grid cell method relates to referencing  the
landscape to the grid cell pattern that was established.  While it is a simple
though tedious task to overlay a grid on a topographic map and determine
elevations, slopes and flow directions of cells, it is  not a simple  task to
accurately locate grid cell boundaries on unrectified air  photos, soil  maps,
or worse yet, actual landscapes.  Though the problem is  not insurmountable, it
presents significant difficulty in application.

There are alternatives to using a grid cell pattern.  One  very attractive
alternative is using actual field boundaries as the basis  for or unit of data
collection and processing.  The unit of data becomes the farm field.  Data
collected at this scale is, not surprisingly,  termed field scale data.   For
these purposes, a field may be defined as an area of homogeneous land cover,
managed as a unit by the same individual(s), without significant topographic
or soil differences within it.  There are several attributes which make field
scale data collection and processing attractive.

The primary advantage of using field scale data is that  it involves  readily
identifiable, relatively homogeneous, management oriented  data units.   There
is little ambiguity about the boundaries of a field. They can easily be
identified on air photos, soil surveys, and "in the field".   Most
significantly, land cover and land management are homogenous by definition.
Further, field boundaries are often laid out following  the contours  and
channels of topography and drainage and, as such, they  tend to have  more
homogeneous slopes and soils then similar size grid cells  placed on  the
landscape irrespective of naturally occuring features.   Where fields do have
significant topographic or soil differences within them, it may be in both the
modeler's as well as the land owner's interest to subdivide them. Field size
also tends to follow the diversity of the landscape. Where slopes are
uniform, fields tend to be larger, and vice-versa. This  relation coincides
with the inherent preferences of the modeler,  i.e., smaller scale data  where
needed and larger scale data where warranted.   And of course, by definition,
field scale data is most appropriate for making field management decisions.
The land operator is already operating at this scale.  Lastly, this  is  a
common unit of data gathering, and several other agencies  and organizations
                                  T-II-D-5

-------
 collect this type of data in  a similar format.  This common data requirement
 and existence of common data  files  provides  for the opportunity of shared use
 of common resources.

 The use of field scale data has many  appealing  attributes but, as we shall
 see,  it necessitates the use  of hierarchal data structures.

 Hierarchal  data  structures

 The term data structures refers to  the  method by which data is organized and
 processed within a computer model.  There are two types of data structures to
 be considered here:  array and hierarchal.

 An array data structure is  (in this application) a two dimensional
 representation of data with X-Y or  row-column coordinates.  The array
 parallels the grid cell  data  format,  (although  a grid cell format, within
 which runoff is  passed to one adjoining cell rather than two, may as well be
 represented in the computer by an hierarchal data structure).  The array is
 processed sequentially,  proceeding  row  by row,  column by column, through each
 iteration.   Runoff and sediment is  routed out of each cell to adjoining cells.

 There are two problems with array processing of runoff and sediment routing
 data.   First and foremost, array processing makes tracking of sediment eroded
 out of  one  cell  impractical to separate and track independently as it moves
 downslope and merges with sediment  from other cells.  Instead, the net
 sediment movement on each cell  is tracked (deposition plus incoming sediment
 minus detachment minus outgoing sediment).  A simple tally of net increase or
 decrease in sediment on  each  cell is  kept, not  the eventual deposition or
 delivery  of sediment from each cell.

 The second  problem is  one of  efficiency or waste of computer resources.  The
 size  of arrays must be declared (allocated) before the computer program can
 run.  Either  the  arrays must  be declared large enough to accomodate the
 greatest anticipated cell numbers (how  big is the watershed and how small are
 the cells?),  or  the program must be modified repeatedly to fit the individual
 watersheds, or the watershed must be  subdivided to fit the program.  Recall
 also that there are several parameters  for each cell (slope,  land cover, soil,
 flow  direction, etc.), and that this  problem is compounded by the number of
 parameters.   Wasted allocation  can rapidly limit the size of  a watershed which
 a particular  computer  can accomodate.   ANSWERS cleverly transforms its data
 into several  one  dimensional arrays, which alleviates the wasted allocation
which results  from watersheds  not being square, i.e., not completely  filling
 the corners of two dimensional  arrays.

An  alternative to the  array data structure is the hierarchal  data structure.
Hierarchal data structure is a method of data storage and access by which each
cell is accessed by its location relative to other cells.   It does not have
such absolute designations such as "row and column"  or defined locations in an
array.  Instead,  each cell is accessed only through  access to a previous
cell.  The location in computer memory of each record is kept in and  a part of
another computer record.
                                  T-II-D-6

-------
The use of field scale data necessitates the use of hierarchal  data
structures, because the fields obviously are not uniformly rectalInear.
Fields have a variety of shapes and sizes.  They do not conform to a grid
pattern either on the landscape or in the computer.  Rather than being
accessed by row and column in the computer, they are accessed by their
position relative to other fields.

There are a couple of drawbacks to using hierarchal data structures however.
Probably the most significant disadvantage is that it requires that all  runoff
from a field flow to a single adjoining downslope field or channel.  Without
some very complicated programing, which would in execution severly hinder
model performance, it is impossible to split the flow of runoff from one field
to two or more adjoining fields or channels.  In designating a flow route,
this requires making a selection based on dominant flow, worst case
conditions, or some other criteria.  A second less critical drawback is  that
field attributes (the input and output of the model) can not readily be  mapped
by the computer.

While the use of hierarchal data structures has some drawbacks, there are
several advantages of this type of data structure that makes it the prefered
choice in watershed modeling.  To understand why this is so requires a more
in-depth look at hierarchal data structures.

Hierarchal data structure is perhaps more commonly refered to as a tree
structure, because the internal representation of the data approximates  the
structure of a tree (Figure 1.).  A cell which has other cells flowing into it
is called a root.  Each root has one or more cells flowing into it called
branches, each of which in turn may have one or more branches flowing into
it.  Each cell is a branch of the root to which it flows, and the  root of the
branches which flow into it.  All of the cells that flow either directly or
indirectly into a root cell are called daughter cells of that root.  In  Figure
1., cells B, G and H are branches of cell A, and cell A is the root of cells
B, G and H.  All cells other than A are daughter cells of cell  A.
Figure 1.  Pictoral  Representation of a Hierarchal  Data Structure
                                  T-II-D-7

-------
 This tree structure closely  mimics  the  natural flow pattern of surface runoff,
 aggregating  from field  to  field  into  small  intermittent channels, and those in
 turn into larger and larger  perennial channels.

 In  a hierarchal  tree structure,  in  order  to proceed down the tree, each cell
 must have one  data  record  associated  with it indicating the cell to which it
 flows. This  record  must be determined beforehand and be included in the basic
 data file.   In addition, each cell  may  have one or more cells flowing into
 it.   To  proceed  up  the  tree  then, each  cell must have a list of which cells,
 if  any,  flow into it.   This  list does not have to be developed beforehand but
 can  by built in  the computer based  on the root record of each of the cell's
 branches.

 In  addition  to paralleling natural  flow patterns, a significant advantage
 using hierarchal  data structures may  be obtained if the model is programed in
 a computer language called Pascal.  This advantage results from a feature of
 Pascal,  known  as  dynamic data allocation, which allows computer space to be
 allocated as needed, eliminates  wasted  allocation and maximizes computer
 storage  potential.   Incidentally, another outstanding feature of Pascal is its
 ability  to do  something called recursion, which allows a subroutine
 (subsection) of a program  to reference  (call) itself.  This feature is a
 tremendous asset  when using  hierarchal  data structures.

 There is  one other  strong  advantage of  using hierarchal data structures, which
 is the ability to track sediment from a field to u channel.  This feature may
 best  be  described in an overall  discussion of modeling sediment and runoff,
 using the combined  strengths of  field scale data and hierarchal  data
 structures.

 Modeling  using field scale data  and hierarchal  data structures

 Natural  flow patterns may be subdivided into and modeled as two distinct but
 integrated subsystems.  There can be  one hierarchal  subsystem representing the
 channel  segments, and another one representing fields flowing to the channel
 segments.  In  the channel  subsystem,  a channel  segment must flow to another
 channel segment,  but it may have either fields, channels or both flowing into
 it.  On the other hand, a field  can flow to another field or channel, but it
 can only have  fields flowing into it, not channels.   Clearly the dynamics of
 sediment routing  operate differently  in both subsystems.  Also the distinction
 serves the purpose of identifying the extent of eroded sediment from an upland
 field that is delivered to a channel  (as opposed to the mouth of the
watershed).  The  need for and practicality of this limit will  be made clear
 shortly.

When modeling a watershed using hierarchal data structures, one begins at the
 root of the entire watershed, i.e., the mouth of the main channel.   For each
 iteration or time segment, the model proceeds up the tree,  stopping at each
cell to calculate sediment detachment, transport and deposition,  routing the
excess back down  to its root, then proceeding up each of its branches, in turn
repeating this process.   The cells of Figure 1.  for example would be processed
                                  T-II-D-8

-------
 1n alphabetical order.  Therefore, for a main channel outlet that has two
 branches coming into it, one branch, and all of the watershed draining into
 it, will be processed during each interation before the analysis proceedes up
 the other branch.  If the branch in question is a channel, then a group of
 paramaters describing that channel, and equations governing channel  transport,
 can be called in.  If it is a field, then the parameters must describe the
 field, and the equations governing overland detachment, deposition and
 transport must be used.  In either case, the excess runoff and sediment is
 routed to the cell's root, be it a field or channel, and the cell's branches
 (if any) are in turn processed, be they fields, channels, or a combination of
 both. By looping through several iterations over time, it is possible to track
 the net sediment dynamics in each of the field and channel segments.

 If the analysis ended at this point, the major differences between this and
 current models would be the units of the input data, the manner of cell
 processing, and the units of the output data.  The output of the model could,
 in a parallel fashion to others, indicate net sediment gain or loss on each
 field and channel (not cell), as well  as total watershed sediment loading.
 However, using the components of this model, it is possible to improve on this
 output significantly. With field scale data and hierarchal data structures it
 is possible to track the amount of sediment eroded off of a field that is
 delivered to the channel network.  This is possible because the length of
 overland flow prior to channelization is generally relatively short.   The SCS
 (1972) indicates that it is seldom more than 1000 feet.  In the small
 watershed that ANSWERS was calibrated on, overland flow was seldom across more
 than two or three fields before channelizing. In that watershed, the greatest
 number of fields of consequtive overland flow was five. These five fields were
 narrow, contour cropped, and on the side of a long gently sloping hill.

 This rather rapid channelization of overland flow means that the tree
 structures representing fields in the computer are typically quite short,
 i.e., there are a lot of them but they do not branch very far.   As a
 consequence, it is practical  to do two things in the computer that would allow
 a program to track sediment downs!ope.  The first is to, for every source
 field, keep a linked list (a one dimensional tree) representing the sediment
 in suspension on each field over which overland flow from the source field
must traverse before reaching a channel.  Fields adjoining a channel  would
 have only one record in their linked list.  Fields with two intervening fields
between themselves and receiving channels would have three records in their
 linked list. Sibling cells (branches with the same root) would have  the same
 number of records in their linked list.   Each of their records would,  in
 parallel, refer to the sediment originating from their respective cells in
 suspension on coinciding downslope cells. The sediment in suspension on each
 field at any one time would be equal  to the sum of the sediment in the
coinciding cell of the linked list of each of its daughter cells.   Again,
using Pascal and dynamic data allocation, the required storage capacity  would
be minimized.

The second component necessary to track sediment downslope is a process,  not
 an item.  Whenever sediment dynamics on a given field are calculated,  a
                                  T-II-D-9

-------
 traversal  through  each  of  the cell's daughter cells is required.  The purpose
 of  this  traversal  is  to modify the record on each of the daughter cell's
 linked list corresponding  to the field in question.  On the record of the
 linked list corresponding  to the field, it would proportionally decrement the
 sediment load  for  deposition (if any) and downslope transport.  Then it would
 increment  the  adjoining downslope record for inflow from the upslope record.
 When  the root  cell flows to a channel (as opposed to another field) the
 downslope  transport represents delivery to the channel.  By this means
 sediment delivery  from  a field to a channel can be tabulated.  The reason that
 this  is  possible but  limited to overland flow is that the computer storage and
 execution  effort required  with each cell to perform these calculations
 increases  linearly with the number of daughter cell it has.  A field seldom
 has more than  a few daughter cells, but a channel can have hundreds to
 thousands.

 The above  description outlines a proposed flow of control for a suggested
 sediment delivery model.   References to specific methods for calculating
 overland and channel  flow  and sediment transport are noticeably lacking, but
 they  can be  selected  from  several in existence.

 The data that  would be  necessary to run the proposed model  includes two
 detailed types of inventory (field and channel), as well as the selection of
 appropriate design events.  The field inventory would require field by field
 descriptions,  including all of the data presently collected in the Piority
 Watersheds to  estimate  erosion using the USLE, plus two additional data
 records.   One  would be  the overland flow distance of the field.  This, in
 conjunction with the field acreage, would be needed to estimate field
 dimensions, as flows are usually calculated in unit volume per unit time per
 unit width.  The second additional  data record would be a designation of the
 field or stream to which each field flows.  Anticipating that at some time in
 the not  too distant future the Department may procure such a model, the
 Nonpoint Source Section is evaluating a method by which these additional
 pieces of  information may  be obtained.

 The second part of the  inventory consists of a channel  inventory.   The channel
 inventory would require a  channel sepent identifier, a physical description
 of the channel segment  (width,  profile,  gradient and vegetation if any),
 length of the  segment,  and a designation of the channel segment that this
 segment  flows  to.  This would not be difficult for perennial streams, but
 could become tricky when applied to small  intermittent streams, grassed
waterways, roadside ditches, etc..   The  present channel inventory  that is
 undertaken in  Priority  Watersheds,  mainly to evaluate habitat and  biotic
 potential, would have to be upgraded significantly to accomodate this data.


CONCLUSIONS

The Wisconsin  Nonpoint  Source Program requires a model  capable of  estimating
the sediment load from  fields to streams,  and sediment transport within
streams, under current  and projected field management conditions.   This model
                                  T-II-D-10

-------
must accept data that can be collected for large areas in a reasonable amount
of time and with a high degree of reliability.  Lastly, the output of the
model must be appropriate for making field scale management decisions based on
water quality objectives.  It is argued that the proposed model  could
demonstrate all of those attributes.  Finally, it is suggested that
appropriate agencies consider funding the implementation of such a model.


REFERENCES

Beasley, D. B. and Muggins, L. F. (1982). ANSWERS Users Model.
EPA-905/9-82-001.

Bosch, D. D., Onstad, C. A., and Young, R. A.  (1983).  A Procedure For
Prioritizing Water Quality Problem Areas.  A presentation at the 1983 Summer
Meeting of the American Society of Agricultural  Engineers, Montana State
University, Bozeman, Montana, June 26-29, 1983.

Park, S. W. and Mitchell, 0. K. (1983). MODANSW (A Modified ANSWERS Model)
Users Guide. Dept of Agricultural Engineering, University of Illinois at
Urbana-Champaign.

Soil Conservation Service (1972).  National  Engineering Handbook, Section 4,
HYDROLOGY.
3622U
                                  T-II-D-11

-------
          EUTROPHICATION PROBLEMS IN NORTH CAROLINA AMD
            MANAGEMENT APPROACH FOR NONPOINT SOURCES

   William A. Kreutzberger, George T. Everett, and Alan Klimek
         North Carolina Department of Natural Resources
                    And Community Development
                     Raleigh, North Carolina


    The State of North Carolina has been experiencing
eutrophication problems in several coastal rivers for the past
decade.  These problems are characterized by extensive surface
blooms of blue-green algae during June through September.  Two
recently filled multi-purpose reservoirs in the lower piedmont
have also been exhibiting signs of eutrophication although only
minor blooms of nuisance algae have as occurred yet. Extensive
monitoring and research efforts have been conducted in the
watersheds of these eutrophic systems to characterize the
magnitude of nonpoint sources of nutrients and sediment from
urban, agricultural and forested watersheds.  The effectiveness
of "best management practices" (BMP's) has also been evaluated.

    In an effort to mitigate the occurrence of nuisance blooms
one coastal river and prevent the occurrence of these problems on
the piedmont reservoirs, the State has classified the watersheds
of these systems as "Nutrient Sensitive Waters"  (NSW).  This
classification provides the regulatory authority to limit the
input of nutrients from point sources.  However, agricultural and
silvicultural activities are explicitly exempted from this
regulatory authority.  Although urban areas are not exempted from
this regulatory authority, there are limited approaches for
controlling nonpoint nutrient inputs from existing urban areas.

    Despite the limited regulatory authority, the State has used
the NSW classification to educate the public, local government
officials, and the General Assembly about nonpoint problems.  The
threat of point source nutrient controls has been used to force
local governments to plan and control runoff from new urban
expansion.  It was also a major factor in motivating the General
Assembly in 1984 to provide $2.5 million for the voluntary
implementation of cropland and animal operation BMP's in NSW
watersheds.  These efforts in NSW watersheds are the first test
case for the effectiveness of nonpoint source planning and
control measures on a large scale in North Carolina.

    Keywords;  Eutrophication, nuisance algal blooms, nutrient
sensitive waters  (NSW), nonpoint source pollution control,
agricultural runoff, urban runoff, future development.
                             T-II-E-1

-------
INTRODUCTION

    Freshwater sections of several coastal rivers in North
Carolina have been experiencing severe eutrophication problems
since the early 1970's.  The Chowan River (Pig. 1) was the first
of these rivers to bring public and regulatory attention to
eutrophication problems with the development of extensive surface
mats of blue-green algae.  These surface blooms have occurred
nearly every year since 1970 and in 1972, 1978 and 1983 surface
mats covered nearly all of the 30 km long lower river during part
of the summer and fall  (N.C. DNRCD, 1982a).
        Fig. 1 Nutrient Sensitive Watersheds in N.C.

    N. C. inland lakes and reservoirs have a wide range of
trophic conditions.  Trophic state classification surveys of
nearly 80 bodies of waters conducted in the mid 1970s and early
1980s indicate that 30 to 40 percent of these lake/reservoirs are
eutrophic or hypereutrophic  (Weiss and Kunzler, 1976; N.C. DNRCD,
1982b).  Extensive surface blooms of blue-green algae have not
occurred in lake/reservoir systems. However, many water treatment
plants using these systems as raw water supplies periodically
have to modify treatment to minimize taste and odor problems.
Two recently impounded reservoirs, B. Everett Jordan Lake in the
Upper Cape Fear River Basin and Falls Lake in the Neuse River
Basin (Fig.l), are already considered the most eutrophic
reservoirs in N. C.  These reservoirs are planned as water
supplies and recreational areas for the rapidly growing
Raleigh/Durham/Chapel Hill area and there is considerable concern
over the impact of severe eutrophication on uses of the lakes
(N.C. DNRCD, 1983a).
    The N. C. Division of Environmental Management  (DEM)
has developed a watershed classification entitled "Nutrient
Sensitive Waters"  (NSW) as a regulatory means to deal with
eutrophication.  The NSW classification provides the
regulatory authority to limit inputs of nutrients to
background levels, if necessary.  The authority is effective in
controlling nutrient inputs from point source discharges, however
the North Carolina General Assembly has passed legislation which
has exempted agricultural activities (other than concentrated
                              T-II-E-2

-------
feed-lots) and silvicultural activities from this of the
authority.  Urban runoff is not exempted from the DEM authority
provided by the NSW classification but there are limited
approaches for effectively reducing nutrient inputs from existing
urban areas.
    The Chowan River watershed has been classified as NSW since
1979 while the watersheds to Falls and Jordan Lakes were
classified NSW in 1983.  Point source nutrient controls (nitrogen
and phosphorus in the Chowan River watershed and phosphorus
controls in Falls/Jordan watersheds) are currently being
implemented.  Nonpoint source control implementation has
initially been slow, however a special funding program provided
by the N. C. General Assembly is making progress on agricultural
Best Management Practices  (BMP) implementation in all three NSW
watersheds.  Also, a unique local/state government cooperative
approach seems to be making progress in minimizing runoff
problems from urban expansion as well as showing some potential
to control some existing urban runoff problems in the
Falls/Jordan watersheds.

CHARACTERIZATION OF EUTROPHICATION PROBLEMS

Chowan River

    The Chowan River is located in Northeastern North Carolina
(Fig. 1).  It is formed by the confluence of the Blackwater and
Nottway rivers near the Virginia/N.C.state line and stretches.
about 80 km south to Albemarle Sound.  The watershed is
approximately 12,700 km  with 73 percent in the Piedmont and
inner Coastal Plain of southeastern Va.  The watershed in North
Carolina is entirely in the coastal plain and the Chowan is
tidally influenced over its entire N. C. length.  Despite the
tidal influence, the Chowan River is essentially a freshwater
system except during periods of extreme drought.
    The Chowan River can be divided into two district regions.
The upper portion of the river, from the Virginia state line to
Holiday Island  (48 km) is a relatively narrow  (90 to 1200 m in
width), sluggish river with depths up to 12 m.  The lower river
from Holiday Island to Albemarle Sound  (32 km) ranges from 1200
to 3000 m. wide and behaves like a shallow  (mean depth 4.0 m),
rapidly flushing lake.  Surface mats of blue-green algae occur
primarily in the lower river due to the large surface area,
shallow depth, and lake-like tendencies (N.C. DNRCD, 1982a).
    Detailed monitoring by DEM has been used to estimate nutrient
inputs to the lower Chowan River from 1977 to 1981  (N.C. DNRCD,
1982a);

    Annual Flow         Total Nitrogen        Total Phosphorus
    Lt^/sec)             (kg x 10^/yr)           (kg x 10^/vr)

    Ave    Range         Ave     Range        Ave       Range

    133.7  34-242       40.3     6.8-75.1     4.3       0.8-7.5
                             T-II-E-3

-------
    These estimates indicate that the N:P ratio is in a range
where either nitrogen or phosphorus could be limiting.  Detailed
research studies on phosphorus and nitrogen uptake, nitrogen
fixation, and sediment nutrient recycling confirmed that nitrogen
and phosphorus limitations varied both seasonally and as a
function of the phytoplankton community.  Phosphorus was
identified as the primary nutrient requiring control since the
major sympton of the eutrophication problem was surface blooms of
nitrogen fixing blue-green algae in the lower river (Kuenzler,
et.al., 1980).  A 30 to 40 percent reduction in phosphorus inputs
was recommended based on modelling analyses by DEM (N.C. DNRCD,
1982a) and confirmed by research results  (Paerl, 1982).   A 15 to
20 percent reduction in total nitrogen inputs was also
recommended since nitrogen was considered a major factor in
occasional blooms of Microcystis spp..a non-nitrogen fixing
blue-green, in the upper portion of the Chowan  (Witherspoon and
Pearce, 1982).
    The magnitude of blue-green blooms seems to be related to
hydrologic conditions (Pig. 2).  The most severe blooms seem to
occur after above average spring tributary flows and moderate
summer flows.  If high river flows are sustained through the
summer, the flushing action minimizes blue-green growth.
Likewise, if stream flow is extremely low, insufficient nutrients
are delivered to the lower river to sustain algal growth.  For
example, during 1980-81, a 15 month drought limited blue-green
growth in 1980 from a lack of nutrients. By the summer of 1981,
saltwater had moved up the Chowan River above (Holiday Island)
completely eliminating the growth of nuisance species of
blue-green algae (N.C. DNRCD, 1982a).
               «•

               Ml
               * "-
               id
               S
               5 31.
               3
               S. II.
               a oj
                           SI  IS 101 12S  ISO IIS 200
                      CltWM RIVER flOW ( i3/SEC)
         Pig. 2. Relationship between flow, residence time,
         nutrient delivery, and algal bloom magnitude in the
         lower Chowan River(N.C. DNRCD, 1982a).

Palls and Jordan Lakes

    Falls and Jordan Lakes are located near the edge of the lower
Piedmont and lower Coastal Plain in N. C.  (Figure 3).  Falls Lake
                             T-II-E-4

-------
is a 5000 ha impoundment of the Neuse River which9was filled in
February 1983 and has a watershed area of 2000 km . The reservoir
is 39 km long with a mean depth of about 4 m. Jordan Lake is an
impoundment of two major tributaries of the Cape Fear River - Haw
and New Hope Rivers and was filled in September 1981.  The
watershed to Jordan Lake is 4464 km2, the lake area is about 5500
ha, and the mean depth is approximately 5 m.
                                     UNE
                 SCUE
         Fig. 3. Falls and Jordan Lake Watersheds.

    There has been considerable controversy over both
impoundments. Both reservoirs are U. S. Army Corps of Engineers
projects for purposes of flood control, recreation, water supply,
wildlife protection and downstream water quality control.
However, due to extensive urban, industrial and agricultural
development in both watersheds the trophic state of the lakes was
a major issue.  Also, the suitability of Jordan Lake as a water
supply is still being examined due to the presence of 170
permitted wastewater dischargers as well as extensive urban and
agricultural runoff in the watershed (N.C. DNRCD, 1985).
     Water quality monitoring conducted during the first few
years after filling the reservoirs has confirmed some of the
predictions of lake quality.  Jordan and Falls Lakes are the most
eutrophic reservoirs in N.C. although the trophic status has not
yet impaired uses of the lakes  (N. C. DNRCD, 1984b).  Chlorophyll
a, levels in both lakes have at times exceeded 200 to 300 mg/m
However, this has been the result of blooms of phytoplankton
which are not nuisance species  (e.g. Chrysoch roroulina spp^.)  (N.C.
DNRCD, 1984a).,  Chlorophyll a. levels have been averaging between
40 to 80 mg/m  during the summers of 1983 and 1984.  There seems
to be a trend towards more dominance of some blue-green species
in Jordan Lake, but more time is needed to confirm this trend.
These reservoirs have relatively short average retention times of
7 and 270 days in the Haw and Mew Hope arms of Jordan Lake,
respectively, and 120 days in Falls Lake. Already there have been
several occasions where extreme flows have flushed substantial
phytoplankton population from the lake  (N.C. DNRCD, 1983a; Weiss
et.al., 1984; Francisco et.al., 1984).
    Table  1   summarizes inflow information and estimated
nutrient inputs to the lakes for mean annual flow conditions.
This information indicates N:P ratios in the range of 5:1 which
                             T-II-E-5

-------
 indicates more nitrogen  limitation than  in the Chowan River.
 The general abundance of nutrients and nitrogen limited
 conditions are major reasons  for  the concern over future problems
 from nuisance nitrogen fixing blue-green algal species.
  Table  1.  Estimated Nutrient  Inputs to Falls and Jordan Lakes
      Prom Instream Monitoring  Results  (N.C. DNRCD, 1983) .
               Mean Annual     Total Nitrogen   Total Phosphorus
Watershed     Flow  (m^/sec      (ka x IQ-Vyr)     (kg x IQ-Vyrl
Falls               22.5             77.9              13.5

Jordan - total      48.6             227.0              55.6
   Haw River Arm    39.3             181.8              45.7
   New Hope River
     Arm             9.3             45.2               9.4
NUTRIENT SOURCES

Chowan River

    The Chowan River watershed is predominantly rural in both
N.C. and Va. It's watershed is 80 percent forested and wetland
areas, 19 percent agricultural and 1 percent urban.  The total
population in the watershed was a little more than 200,000 in
1980.  Municipalities are generally quite small. The largest
dischargers in Va. are Emporia (0.6 MGD), Blackstone (0.5 MGD),
and Franklin (0.8 MGD), while in N.C. Ahoskie  (0.8 MGD) and
Edenton (1.0 MGD) are the largest municipalities.  There is one
large industrial discharger in the Va. watershed.  Union Camp is
a pulp and paper facility located near the state line which
discharges in excess of 380 x 10bl/day  (100 MGD) from holding
ponds only during the months of December through March when the
Chowan River can assimilate the waste.  In N.C., there was a
large fertilizer manufacturer (C.F. Industries) in operation
until 1983 which was initially blamed for the occurence of
nuisance blooms in the early 1970s.  C.F. Industries did
discharge significant quantities of nutrients in the late 1960s
and early 1970s.  This discharge ceased by 1975 and thereafter,
this company contributed nutrients to the river primarily through
air fallout, site runoff and contamination of groundwater.  (N.C.
DNRCD, 1982a).
    Figure 4 depicts nutrient budgets for the Chowan River
watershed in N.C. (N.C. DNRCD, 1982a).  Agricultural runoff
(50%), forest and wetlands (22%), direct precipitation (-5%) and
urban runoff (0.5%)  account for about 78% of the phosphorus
inputs from the N.C. watershed.  Nonpoint source contributions of
nitrogen are similar, 80 percent.


                             T-II-E-6

-------
                    rim
                   NITIICIN
                1,411, Oil RC-VI
                                         FIIESI 1GIIC
                                         IPS  NP $
                                         28.3*5,9.1%
                                         FIIEST  ACIIC
                                         NPS   IPS
                                         22.1%  49.1%
                   rim
                 PIISPIMIS
               11$, Ml 1C-VI
        Fig. 4. Land use/point source nutrient budget for the
        Chowan River Watershed in N.C.(M.C. DNRCD,  1982a).

Falls and Jordan Lakes

    In contrast to the Chowan River watersheds, portions of the
Neuse and Cape Fear River basins draining to Falls  and Jordan
contain some of the most rapidly developing areas in N.C. The
Jordan Lake watershed includes the urban centers of Greensboro,
Burlington, Chapel Hill and the southern part of Durham with a
total population in 1980 of about 475,000 people. There are 10
major municipal dischargers in the watershed releasing nearly 300
x 10 I/day  (80 MGD).  In addition, there are several major
industrial dischargers and a total of 170 permitted dischargers
in the watershed. Urbanization is continuing at a rapid pace in
this watershed with the projected population for the year 2000 to
exceed 600,000 people.
    Development in the Falls Lake watershed is not  currently as
extensive as in the Jordan but the area is also growing rapidly.
Major urban areas include Hillsborough and northern Durham while
the northern portion of the Raleigh metropolitan area is  also
pushing into the watershed.  The 1980 population was about
150,000 people and is estimated to exceed 210,000 people  by the
year 2000.  There are two major municipal dischargers and a total
of approximately 60 permitted dischargers in the watershed.
    Based on analyses of land use and point source  data nonpoint
sources account for approximately 45 percent of the nitrogen and
40 percent of the phosphorus into Jordan Lake and 70 percent of
the nitrogen and 57 percent of the phosphorus into  Falls  Lake
(Figure's 5 and 6).  Nutrient inputs seem to be readily
controllable through point source actions.  However, the  lakes
currently have such high level of nutrients that the systems may
remain eutrophic even if all point source inputs of phosphorus
were eliminated  (N.C. DNRCD, 1983a; Francisco et.al., 1984).
Future development in the watershed could also dramatically
                             T-II-E-7

-------
increase nutrient contributions from urban runoff which are
already substantial.
                     urn
                   NITIICEN
                IU.MI it -i i
                          PHUT
                          SltttCE
                          30-5%
   FIIEST  ACIIC
   IPS   IPS
   23.7%  29.1%
                     Tin i
                           POUT    URI»I   PRECIP
                           SOURCE   UPS    IPS
                           42.6%  22.6%  1.8%
                              X
   FBIEST  URIC
   IPS   IPS
   7.2%  25.8%
                U2,flflfl«6-ri
         Fig. 5.  Land use/point source nutrient budget  for the
         Palls Lake Watershed  (N.C. DNRCD,  1983a).
                     TITU
                    NITINEI
                MIMIIIC-VI
                           P0IIT    IRMI PRECIP
                           SOURCE   IPS  IPS
                           54.9%  6.2% 1.4%
FIIEST   «Giic
IPS    IPS
13.0% 24.5%
                     T8TU
                  PIISPNOIBS
                 (30,001 KG YR
                          PIIIT   URIAI  PRECIP  FIREST   IfiRIC
                          SIIIICE  IPS   DPS   IPS    IPS
                          59.5% 12.6% 0.6%  4.2%   23.1%
         Fig. 6. Land use/point source nutrient budget  for the
         Jordan Lake Watershed  (N.C. DNRCD,  1983a).

NONPOINT AND POINT SOURCE MANAGEMENT

Chowan  River

    Nonpoint source control  initiatives began slowly.  When the
Chowan  was classified as NSW in 1979, agricultural agencies were
unwilling to admit to any contribution to eutrophication problems
of the  Chowan.  This unwillingness along with legislation
limiting the ability of the  DEM to require  nutrient reductions
                                T-II-E-8

-------
from agricultural and silvicultural operations severely limited
options to reduce nutrients in N.C.
    With these restrictions, the State took two major
initiatives; 1) Actions were started to eliminate or minimize all
point source contributions of nitrogen and phosphorus and 2)
Research/education efforts were begun to verify agricultural
contributions to the nutrient enrichment problems, identify the
effectiveness of various best management practices,  and to offer
technical assistance for BMP implementation.
    Point source efforts were relatively straight-forward in  that
all the municipal facilities were small and required upgrading.
Facilities were promoted on the priority list for construction
grants and all facilities began to conduct non-discharging to
land application, wastewater disposal.  These systems will all be
operational by 1987.  Small industrial discharges were required
to meet total nitrogen and phosphorus limits of 3 and 1 mg/1,
respectively, and several enforcement steps were taken to limit
nutrients in runoff and seepage from the abandoned C.F.
Industries fertilizer manufacturing site.  These actions have
eliminated point source nutrient contributions as an issue in the
North Carolina watershed, and have focused attention on
agricultural contributions  (N.C. DNRCD, 1982a).
    A three-year research study from five watersheds from 200 to
2800 ha in the Chowan basin in N.C. was conducted from 1979 to
1982 (to.examine agricultural runoff)  (Humenik et.al., 1983),
The study was not able to quantify before and after  effects from
BMP implementation due to the short research time.  However,
valuable survey information on existing management practices  and
producer attitudes towards nonpoint source controls  were
obtained.  The information was utilized in developing supportive
publicity, educational and technical assistance programs for  the
entire Chowan watershed. Estimated annual yields from the
forested  (control) watershed and the four agricultural watersheds
(averaging 50 percent cropland) are shown on Table 2.
    Despite the failure of the research to demonstrate BMP
effectiveness, these studies in combination with other State
action helped to involve agricultural agencies in the development
of a nonpoint control program.  Also, the study indicated that
estimates of nutrient contributions from agriculture shown in
Fig. 4 may have been somewhat conservative.  In addition, survey
results helped to target cropland requiring control  measures  and
animal operations requiring waste management systems.  BMP
effectiveness was demonstrated as part of research in other parts
of N.C. and demonstration farmswere set up to demonstrate
structural BMPs , provide education on fertilizer application,
serve as advocates for soil testing, and demonstrate animal waste
management systems.
                             T-II-E-9

-------
Table 2.  Sediment and Nutrient Yields from Five Watersheds in
the Chowan River Watershed  (Humenik et.al., 1983).


                     Annual Yield  (kg/km-^-
Parameter          Forested Watershed    Agricultural Watersheds
                                             Aye.      Range
Suspended Sediment
Total Nitrogen
Ammonium-Nitrogen
Nitrate Nitrogen
Organic Nitrogen
Total Phosphorus
13,700
135
8.8
17
101
13
71,940
627
28
275
303
114
4972-400,400
110-1331
1.1-73
6.6-672
84-1071
19-351
    At the end of 1982 and early 1983, this put the agricultural
control program in a position to begin implementation.  Based on
the overall nutrient reduction goals of 30 to 40 percent for
phosphorus and 15 to 20 percent for nitrogen from watershed areas
in N.C. and Va. as well as the reductions being obtained from
point sources, goals of 30 percent effectiveness for phosphorus
and 20 percent effectiveness were set for the agricultural
control program (N.C.  DNRCD 1982a).  However, there was little
or no special funding to encourage implementation of BMPs so the
program was entirely voluntary.  The implementation program had
to rely on existing staff from the N.C. Agricultural Research
Service and N.C. Soil and Water Conservation Districts to provide
technical assistance.  The capability of the N.C. Department of
Agriculture to greatly expand soil testing in the Chowan
watershed was also insufficient.  Cost sharing funds from the
ASCS also had to be relied upon for structural BMP
implementation.

 Falls and Jordan Lakes

    When Falls and Jordan Lakes were classified as NSW in October
1983, this clearly indicated the intent of DEM to regulate point
sources of nutrients beginning with phosphorus. Initial policies
for the lake were to limit all new facilities to a total
phosphorus limit of 1 mg/1 and encourage land application
wastewater treatment systems whenever possible. This applied
primarily to rural subdivision developments on both watersheds
but also is applicable to any expansion of existing wastewater
facilities.  Existing facilities in the New Hope arm of Jordan
Lake were issued compliance schedules to meet a 1 mg/1 phosphorus
limit in 3 years.   All other facilities were issued letters
indicating that the watershed was NSW and nutrient limits might
be established in the future.
    Clearly substantial phosphorus reductions to both lakes could
be accomplished based on budgets in Figure 5 and 6 by requiring
phosphorus removal at point sources.  However, contractors
studying the lake for the Corps of Engineers have expressed
doubts over the estimated nutrient contributions from point


                             T-II-E-10

-------
sources in the Jordan watershed, the effectiveness of phosphorus
reductions for controlling eutrophication,  and the need to
control eutrophication since no uses are currently being
impaired. (Francisco et. al., 1984).  This  opinion was considered
in developing the initial control strategy  for point sources.
There has been little disagreement that actions are necessary  to
encourage proper planning to minimize future nonpoint source
contributions of sediment, nutrients and toxicants.  All three
pollutant categories are a major concern relative to water supply
usage of the lakes.
    Public concern over the future of these lakes put state
agencies in a favorable position to get cooperation from local
governments to protect the lakes.  The threat of point source
controls on phosphorus has also assisted in getting cooperation
from government groups in the upper part of the watershed which
do not get recreational and water supply benefits from these
lakes.  Committees were formed with state and local officials
and, with assistance from DEM staff and local/regional planning
groups, guidance was developed for local actions to protect the
lakes from future nonpoint source inputs. Essential elements of
these action plans are as follows;
       Local sedimentation and erosion control plans,
       County soil and water plans for high erosion areas
        (cropland),
    -  Requirement of storage of first 1/2" runoff from new
       development,
       50 foot stream buffers,
       Designation of water quality critical areas near the lakes
                 6% impervious limit
                 Restricted commercial/industrial development
                 No sewer extentions,
       12% impervious limit for unsewered,  non-critical areas,
    -  30% impervious limit for sewered, non-critical areas,
       Designation of limited industry areas.

    In addition to these efforts to minimize nonpoint source
impacts from new development, DEM is continuing to investigate
means to reduce runoff pollutants from existing urban areas.
Using loading rate information from the Nationwide Urban Runoff
Program study in Winston-Salem, N. C. and nationwide, portions of
the urban areas have been targeted for controls(N.C. DNRCD,
1983b)  .  Stormwater detention basins for the 30 percent of urban
areas with highest pollutant yields could reduce metals inputs
from urban runoff by as much as 50 percent and nutrient inputs by
20 to 30 percent  (N. C. DNRCD, 1985).  This information is being
provided to local planning agencies and funding is being
considered for urban detention basin demonstratJon projects.

NSW LEGISLATION

    In June 1984, the N. C. General Assembly approved special
funding for the implementation of nonpoint source controls in  NSW
watersheds.  Approximately $2.5 million/year was provided to the
county soil and water conservation districts through the N. C.
Division of Soil and Water.  The bulk of this funding will be


                              T-II-E-11

-------
used for 75 percent cost-sharing of animal waste management
systems and cropland BMP's.  A portion of the funding was also
used to hire an additional staff member for each of the counties
in NSW watersheds to provide technical assistance.
    During the first six months of its existance this program
concentrated on publicity and getting agricultural producers
interested in the program.  A good portion of the cost-sharing
funds were used for animal waste management.  Specific targeting
of BMPs for high erosion rate cropland has not occurred during
the first year since it takes some time to get producers involved
in this type of voluntary/cost incentive program and it is
important for continued funding to utilize all funds provided.
Targeting of highly erodable cropland for BMPs is planned for the
second year of implementation.
    In addition to agricultural cost-sharing funds, additional
funds were provided for enforcement of state sediment and erosion
control laws.  Also, special research funding is available to
local governments to investigate means to control nutrients from
municipalities.  Research will likely center on biological means
to reduce phosphorus at existing wastewater facilities.  However,
these funds may be used in the future to investigate the
effectiveness of detention ponds for urban nonpoint source
pollution control.

FUTURE DIRECTIONS

    The NSW watersheds are the first test case for the
effectiveness of nonpoint source planning and control measures on
a large scale in North Carolina.  At present, initial nutrient
control needs have been established and a program has been
developed for agricultural land use.  A program is being further
developed to control urban runoff.  For the Chowan River Basin,
the State of Virginia has just developed a plan for point and
nonpoint source controls for protection of their watershed and is
beginning to be implemented.

    The future success of the nonpoint control programs in NSW
watersheds depends on continued cooperation between state and
local agencies, continuing legislative funding, and
research/study efforts.

LITERATURE CITED

Francisco, D.E., C.M. Weiss and P. H. Campbell (1984).  Jordan
    Lake Water Quality in 1984.  Presentation at the N. C. Water
    Control Federation and American Water Works Association
    Conference.  Raleigh, N.C. November 13, 1984.
Humenik, F.J., B.A. Young, and F.A. Koehler (1983). Agricultural
    Nonpoint Source Case Studies in N.C. Ill: Chowan River
    Priority Watershed. Dept. of Biol. and Agric. Engineering,
    North Carolina State Univ., Raleigh, N.C.
Kuenzler, E.F., S. Mozley, H. Paerl, and A.M. Witherspoon (1980).
    Chowan River Nuisance Algal Bloonms and Probable Causes - A
    Position Paper Prepared for the N.C. Dept. of Natural
    Resources and Community Development.  UNC WRRI, Raleigh, N.C.


                             T-II-E-12

-------
N. C. DNRCD (1982a).  Chowan River Water Quality Management Plan.
    Division of Environmental Management, Raleigh, N.C.
N. C. DNRCD (1982b).  North Carolina Clean Lakes Classification
    Survey.  Division of Environmental Management, Raleigh, N.C.
N. C. DNRCD (1983a).  Water Quality Discussions of Falls of the
    Neuse and B. Everett Jordan Lakes. Division of Environmental
    Management, Raleigh, N.C.
N. C. DNRCD (1983b). Nationwide Urban Runoff Study in
    Winston-Salem, N.C. Division of Environmental Management,
    Raleigh, N.C.
N. C. DNRCD (1984a).  1983 Annual Summary of Phytoplankton in B.
    Everett Jordan Reservoir.  Division of Environmental
    Management, Raleigh, N.C.
N. C. DNRCD (1984b).  Ambient Lakes Monitoring Report - 1983.
    Division of Environmental Management, Raleigh, N.C.
N. C* DNRCD (1985).  Toxic Substances in Surface Waters of the B.
    Everett Jordan Lake Watershed.  Division of Environmental
    Management, Raleigh, N.C.
Paerl, H.  (1982).  Environmental Factors Promoting and Regulating
    N9 Fixing Blue-Green Algal Blooms in the Chowan River, N.C.
    UNC WRRI, Report No. 176, Raleigh, N.C.
Witherspoon, A.M. and R. Pearce  (1982).  Nutrient and
    Multi-species Criteria Standard for the Chowan River, N.C.
    UNC WRRI, Report No. 187, Raleigh, N.C.
Weiss, C.M. and E. J. Kuenzler (1976).  The Trophic State of
    North Carolina Lakes.  UNC WRRI, Report No. 119, Raleigh,
    N.C.
                             T-II-E-13

-------
                  REDUCING BACTERIAL NON-POINT POLLUTION
                    IN TILLAMQOK BAY.  TILLAMOOK.  OREGON

                              James A. Moore
                            Associate  Professor
                                    and
                                J.  R.  Miner
                                 Professor
                    Agricultural Engineering Department
                          Oregon State University
                          Corvallis, Oregon, USA

                                 ABSTRACT

High coliform counts in the waters of Tillamook Bay closed oyster
harvest.  A review of possible sources of pollution revealed that sewage
treatment plants, failing on-site domestic septic systems and the
county's 18,000 dairy cows were the primary candidates.

Extensive water sampling confirmed all three contributed.  Major
improvements were made in the sewage treatment plants and the on-site
systems but the nonpoint contribution from the 100 plus  dairy operations
proved to be challenging.

A computer program was developed to evaluate management  practices and
predict the movement of organisms from the cow to the Bay.  This
assisted dairy operators in selecting the specific component to improve
or add to their unique operation to make the greatest positive impact on
water quality.  State and federal agency personnel also  used the model
to identify those operations with the greatest need and  distribute
federal Rural Clean Water Program cost share monies to assist in
upgrading facilities.

Several local citizens' groups, clean water committees,  state and
federal agencies worked with the oyster growers and the  dairymen toward
the common goal of improved water quality.  The program is still
underway but recent sampling suggests an improvement in  Bay water
quality.
                               T-II-F-1

-------
 INTRODUCTION

 There  are many  documented  water  quality  problems  in  the United  States.
 Over the past 15  years  nutrients,  primarily  nitrogen, have received the
 greatest attention  as pollutants.  Currently several cases of bacterial
 pollution from  non-point sources have  drawn  attention to these  problems
 and to the general  lack of research  effort in this area.

 Organisms have  been responsible  for  closures of oyster harvesting areas
 in a number  of  bays and estuaries.   One  such closure occurred in
 Tillamook Bay in  1977 (1).  Tillamook  Bay is on the  Pacific Coast in the
 northwest corner  of Oregon (Figure 1).

 This closure brought attention to  the  problem and focused the efforts of
 several state and federal   agencies  and  local groups.  Water quality
 sampling conducted  by the  federal  Food and Drug Administration  (FDA) (2)
 and the state Health Division (HD) and Department of Environmental
 Quality (DBQ) during the seventies showed that bay water quality did not
 meet standards  of the National Shellfish Sanitation  Program (NSSP) (6)
 (3).   In general  these  studies showed  that a potential hazard did exist
 as high levels  of total and fecal  coliform bacteria  were present in
 oyster beds  regardless  of  weather  and  tidal  conditions.

 From reviewing  the  studies  and input from involved agency personnel, •
 three  major  sources  of  bacterial pollution have been identified.  These
 are the five sewage treatment plants in  the  watershed, failing  on site
 domestic sewage systems and livestock wastes.  A  two step plan  was
 proposed to  address  the pollution  from these sources.  First to reduce
 the bacterial input by  implementing  technical and institutional options
 such as improved  sewage treatment, land  zoning, and  installing  best
management practices on the dairy  farms  (5).   The second part was to
 develop a bay closure scheme when  failing equipment  or heavy rainfall
would  provide conditions that would  excess the sanitation standards set
 by the NSSP.

The remainder of  this paper will address the efforts to reduce  those
 organisms originating from the dairy operations.  There are some 18,000
dairy  cows on 118 farms in  the watershed.  These  dairies support the
Tillamook cheese  plant  which converts  25% of all  the milk produced in
Oregon into  cheese.  The coastal marine  climate brings 150 inches of
rainfall to  the upper part  of the  watershed  and 90 inches of
precipitation in  Tillamook.  Most  of the rain falls  during the  winter
while  the summers are dry.  Within the project area  there are 23,500
 acres  of agricultural lands.  The  dairies which occupy almost all of the
agricultural land are located on the tidal flood  plains and river
terraces.   These  flat uplands are  dedicated  to pastures that are utilized
by grazing during the dry 6 months of the year.   Cows' access to the
streams for watering purposes provide opportunity for fecal organisms
contamination during the summer  pasturing season.  Manure storage and
spreading are the waste management practices  of concern during  the winter
season, when cows are confined in  and around  the  buildings.
                                T-H-F-2

-------
H
I
u>
                              111
                              o
                              o
                                    SEWAGE TREATMENT PLANT LOCATIONS
                                                                   0123
                                                                           -*.  I	
                    Figure 1.   Sewage treatment plant  locations on the five rivers  in

                               Tillamook Bay, Tillamook,  Oregon.

-------
Personnel front the Soil Conservation Service (SCS) and Oregon State
University  (OSU) provided leadership in a group which developed a list of
best management practices.  Many of the practices on the list have been
proven effective in reducing nutrient pollution but were only speculated
to be effective in reducing bacterial pollution.

A research  project was funded to assist in evaluating manure management
practices and help operators select effective practices to reduce
organism movement and escape.  This research developed a computer model
to evaluate manure management practices.  The evaluator was developed
using coefficients in the literature to estimate die-off, losses and
movement in runoff.    This model follows organisms on a daily basis,
from the cows through manure handling to land spreading and possible
escape in runoff.

A flow chart of the model is shown in Figure 2.  The dairyman inputs the
cow numbers, manure collection techniques and other farm specific
information.  Organims die-off begins in storage and varies slightly
between dry and liquid storage systems.  If the operator elects to spread
manure on his pasture land the application method and rate influences the
number of organisms available to runoff.  Seasonal environmental
conditions  influence the die-off rate of bacteria both in the storage
system and  on the land surface.

Twenty-nine years of weather records were evaluated to generate a wet,
dry and average year in order to evaluate different management
practices.  Utilizing the daily rainfall and irrigation as hydraulic
input and interial drainage as outflow, a moisture balance is kept for
lands receiving manure.  Once a precipitation event occurs the antecedent
moisture conditions and field soil characteristics are utilized to
calculate if and when soil saturation will occur and runoff will begin.
A percentage of available bacteria are moved into the soil with
infiltration and are considered removed from the system in terms of
impact on surface water quality.  When saturated soil conditions occur a
first order kinetic relationship is utilized to calculate the bacteria
that are moved overland in the runoff water.  If the operator is using a
buffer strip a reduction of the transported bacteria occurs.  The reader
is directed to a paper entitled "Modeling Dairy Waste Management Systems'
Influence on Col i form Concentration in Runoff" for details of the model
Using the model on a large number of farms generated the following
guidelines for reducing the bacterial pollution potential of dairy
operations and land applied wastes:

    (1)  Storage of wastes will increase management options and reduce
         pollution of surface waters.

    (2)  Installing tile will improve soil drainage, reducing runoff
         events and the movement of bacteria.

    (3)  Save well drained soils for waste receiver sites during high
         rainfall periods.

                               T-II-F-4

-------
              Steps
           animal wastes
           defecated &
           collected	
           storage bac-
           teria die-off
           wastes with-
           drawn & spread
           land surface
           bacteria die-
           off
           infiltration of
           bacteria
           quantity  of bacteria
           in runoff unchanged
       Input
Condition/Management
     input herd
     size & ragmt.
     practice
                                               application rateI
                                               spreading method |
     function of
     seasonal en-
     vi ronmental
     factors
     characteristics
                                               antecedent.
                                               soil moi K t u r e
                                               slope and
                                               crop ruiidit ion-
                                               quantity oi  bacteria
                                               in runoff reduced
Figure 2.  Flowchart of daily bacterial transport model.

                             T-II-F-5

-------
     (4)  Utilize  buffer  strips  between  land  spreading  areas  and  streams.

     (5)  Use  low  spreading  application  rates  over  large  areas when
         possible.

     (6)  Reduce barnyard runoff by minimizing lot  area,  guttering
         buildings  and diking.   This  source  is the greatest  pollution
         potential.  This means keep  the  clean water clean by rerouting
         and  collect and hold dirty water.

Obviously management is  a critical ingredient of any successful  water
pollution reduction program.  Utilizing windows or dry weather periods
when they occur to  spread manure is perhaps the most important management
choice made by dairymen.

The  many positive actions of those involved in the project;  agency
personnel, local  clean water committees,  oystermen and dairymen  were
instrumental  in presenting  a successful grant application to the federal
Rural Clean Water Program.  This brought  over 2 million  dollars  of cost
share funds into  the program to assist  dairymen in constructing  and
installing best management  practices.

Some 69 operations  have  signed  contracts  to implement  best management
practices as  spelled out in their farm  plan.   By the end of  November .1984
some 60 dairy operations  had spent over $3,690,000 in  construction costs
to install new pollution control facilities.   Approximately  two-thirds of
these funds were  cost share funds provided by the  federal government.

The  model and a broad based evaluation  system was  used to evaluate every
dairy which signed  up for assistance  and  cost share funds in order to
establish where the funds should be spent to  make  the  greatest positive
impact.  The  model  was also utilized  to select which practice(s) or
component(s)  should be installed to have  the  greatest  contribution in
reducing bacterial movement to  surface  waters.  Table  1  shows how each
component can be  ranked when evaluated  using  the computer model.  As
suspected, the precipitation level influences  the  effectiveness  of
management components.   That is,  drain  tile will be more effective during
a high rainfall year, etc.

On site farm  visits and  educational programs  were  conducted  by members of
the  DEQ, SCS  and  Extension  Service personnel.   Publications  have been
written specifically for  dairymen in  this project  area.  Once the project
was  underway  tours of newly installed practices and facilities were
conducted.  Local,  state, and national  press  were  utilized to tell the
story of the  efforts and  results  of the project.

The  program is a  success  in terms of moving the dairymen to  improve their
facilities.   The  program  began  in 1980  and while much  construction has
been accomplished, the entire program is  not  expected  to be  complete
until 1990.
                               T-II-F-6

-------
          Table  1.  Model comparison of different management procedures (each examined using a 100 acre field of pasture during the winter
                    with the specified characteristics).
H
I
M
M
I
"ij
Example
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16 •
17
18
19
20
Storage
(days)
0
0
0
20
20
20
40
0
0
0
0
0
0
0
0
0
0
20
20
0
Management
System
D=dry;
S=semi ;
L=liquid
D
S
L
D
S
L
D
0
D
S
L
D
0
S
L
D
L
D
L
D
Application
Rate
(tons/ac)
10
10
10
10
10
10
10
40
10
10
10
10
10
10
10
10
10
10
10
400
Buffer
Strip
(yes.no)
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
N
Y
Y
Y
Y
Y
Y
Y
N
Drainage
Tile
(yes.no)
N
N
N
N
N
N
N
N
N
N
N
N
Y
Y
Y
V
Y
Y
Y
N
Soil Type
(loam, clay)
L
L
L
L
L
L
L
L
C
C
C
L
L
L
L
C
C
C
C
C
Avg. ppt.
Net Runoff
F£
(xlO9 org)
1,848.5
1.388.2
1,829.9
440.0
330.3
453.2
154.3
7,393.8
2.128.8
1,598.6
2.157.8
4,261.2
838.3
616.8
696.3
1.697.9
1,624.9
440.5
439.7
212.873.0
Dry ppt.
Net Runoff
FC
Rank (xlO9 org) Rank
15 1,186.8 3
10 891.1 1
14 1,188.1 2
4
2
6
1
19
16
11
17
18
9
7
8
13
12
5
3
20
Heavy ppt.
Net Runoff
FC
(xlO9 org) Rank
2.685.6 3
2,016.8 1
2.647.0 2


















-------
The ultimate success of this program will be measured by sampling the
water quality in the bay.  A comprehensive sampling program has been
planned to begin in the fall of 1985.  Routine samples to date indicate
an improved quality in the streams and bay.
                               T-II-F-8

-------
                                REFERENCES
Anonymous  (1979).  Tillamook Bay bacterial study; work plan.  Department
of Environmental Quality, Water Quality Division.  75 pp.

Anonymous.  (1981).  Tillamook Bay Drainage - Oregon Basin Fecal Wastes
Management Plan. Oregon Department of Environmental Quality, Waste
Quality Division.

Carr, V. E. and staff.  (1976).  Tillamook Bay, Oregon:  Pollution source
evaluation with classification and management considerations - May 1976.
Prepared by: Northeast Technical Services Unit, Davisville, R.I., for The
Department of Health, Education and Welfare, Public Health Service and
the Food and Drug Administration - Shellfish Sanitation Branch.  66 pp.

Carr, V. E., Furfari, S. A., and Miescier, J. J.  (1978).  Sanitary
survey of shellfish waters - Tillamook Bay. Oregon. November-December
1977.  Prepared by: Northeast Technical Service Unit, Davisville, R.I.,
for The Department of Health, Education and Welfare, Public Health
Service and the Food and Drug Administration - Shellfish Sanitation
Branch.  143 pp.

Moore, J. A., Grismer, M. E., Crane, S. R. and Miner, J. R.  (1983). .
Modeling Dairy Waste Management Systems' Influence on Coliform
Concentration in Runoff.  Transactions of the ASAE. 26. (4), pp. 1194-1200.

Oster, Dennis (1975).  Land and water use guidelines for development of
the Tillamook Bay estuary.  Tillamook Bay Task Force and Tillamook County
Board of Commissioners.  252 pp.
                                T-II-F-9

-------
               Abatement of Nonpoint Pollution of Semi-arid
          Streams Using Livestock Grazing Management,  Vegetation,
                  Instream Flow Structures, and Beaver

Quentin D. Skinner, Jerrold L. Dodd, J. Daniel Rodgers, Michael A.  Smith

Associate Professors, Department of Range Management,  University of
Wyoming, Laramie, Wyoming, USA

                                 ABSTRACT
An important but often ignored contribution to nonpoint pollution in the
semi-arid western United States is sediment.  Accelerated erosion and
sediment in stream flow is often thought caused by livestock grazing on
rangeland in less than excellent condition.  Erosion of watersheds is
often recognized by stream channel condition.  Poor condition streams
are often downcut and show active bank rilling, bank sloughing, and
piping.  Lateral movements of gullies away from downcut streams through
upland ecosystems are, in addition, contributing sources for increased
sediment pollution.  Equally important, when downcut streams occur,
associated riparian vegetation is often reduced in areal distribution,
disappears, or is replaced by pseudoriparian plant species.  Loss of
riparian vegetation and the hydrologic support of these limited but
valuable ecosystems may further reduce stream channel stability and
sediment trapping efficiency of nonpoint source pollution transported to
the stream by overland flow and ground water.

This paper addresses theory and methodology being applied in two
research programs on perennial and ephemeral streams designed to:   1)
promote water storage and vegetation yield, 2) control nonpoint source
pollution, 3) reverse desertification of riparian zones and, 4) advance
the state of knowledge for management of stream side ecosystems
subjected to user pressure.

Livestock grazing treatments, instream flow structures, willow
management, and beaver activities are being used to manipulate stream
flow velocity.  Reduced velocity causes sediment deposition on channel
banks where it is then stabilized by vegetation.  Encroachment of stream
banks should cause aggradation of stream channels, promote water
spreading over or under dry flood plains, and increase area and health
of riparian zones.

Key words:  Nonpoint Pollution, Sediment, Riparian, Desert, Streams, Best
Management Practices
                            T-II-G-1

-------
 INTRODUCTION

 Riparian Zones

      Riparian zones are wetlands supported by a high water table because of
 proximity to surface or subsurface water.  They are characterized by
 distinct soils and plant communities with high productivity as well as
 species diversity.  They normally occur as an ecotone between more xeric
 lowland communities and aquatic ecosystems (Brown 1978).   Strictly
 speaking, riparian zones are dominated by plant species that depend on
 sub-irrigation throughout their growing season.  When sub-irrigation
 does not occur riparian communities are replaced by more xerophytic
 communties.

 Importance

      The riparian zone has a high level of use by a wide variety of
 interests as noted by Busby (1978), Johnson (1978), Tubbs (1980), Thomas
 et al. (1978), Haugen et al (1980), and numerous others.   Apparent
 causes for concentration of multiple uses in riparian zones are the
 vegetation species diversity, productivity, and proximity to open water.
 High species diversity in the riparian zone is reported by Campbell and
 Green (1968), Brown et al. (1978), Ewel (1978), Haugen (1980), Speck
 (1981), and Kauffman et al. (1983a).  Diversity is associated with zones
 of variation in soil moisture (Miller 1979).   Higher plant biomass
 production due to increased soil moisture is  described by Thomas (1978),
 from research by Minore (1970), Minore and Smith (1971),  and Miller
 (1979).  Flowing water wetlands have higher rates of gross primary
 productivity and net biomass production than still water  wetlands (Brown
 et al. 1978).

      Riparian areas provide a buffering ecotone between the aquatic and
 dryer lowland and upland ecosystems.  The beneficial effect of this
 ectone to stream condition, water quality, and fish populations has been
 reported by Odum (1978), Jahn (1978), and Haugen et al.  (1980).   Water
 quantity and quality as well as physical features of streams also depend
 on characteristics of the watersheds drained  by each stream (Jahn 1978).
 Characterization of watersheds and riparian zones is therefore of
 importance in evaluating user impacts.

_User_

      A complex interrelationship exists between soils, water,  and
 riparian vegetation (Miller 1979).   Impacts compacting  the soils,  such
 as livestock grazing and recreation, may reduce infiltration,  lower
 water tables,  and increase erosion (Peterson  1950,  Schmidly and  Ditton
 1978,  Meehan and Platts 1978,  and Thomas 1978).   Methods  to increase
 water harvest on channelized streams will also affect the soil moisture
 regime and vegetation (Peterson 1950, Jahn 1978,  Campbell 1970,  McCall
 and Knox 1978)  through dewatering the riparian zone.  Such impacts  as
 loss  of plant vigor and changes in species composition  in plant
 communities  result from livestock grazing and trampling,  roads,  mining,
 and recreation  (Peterson 1950,  Gunderson 1968, Martin 1978, Thomas  1978,
 Schmidly  and Ditton 1978,  and  Speck 1981).  The  loss of capacity to bind

                            T-II-G-2

-------
 soil  from a  decrease  in vegetation  increases  the potential for erosion,
 loss  of  stream channel stability, and  results in sedimentation within
 channels.

      Peterson  (1950), Busby  (1978), Meehan and Platts  (1978), Thomas
 (1978),  Roath  (1982), Kauffman et al.  (1983b), and Kauffman and Krueger
 (1984) indicate an  increase  in erosion has resulted from livestock
 grazing,  water development,  and  recreational  use of riparian zones.
 Channel  instability indicated by sloughed stream banks results in wider
 channels,  more shallow water depths and altered flow characteristics
 (Platts  1981).  Gunderson  (1968), Duff (1978), Platts  (1981) and
 Kauffman et  al (1984b) reported  channel widening due to impacts of
 livestock grazing.  Haugen et al. (1980) included mining, water
 development, and  timber harvest  as user impacts that altered stream
 channels.  Sediment from erosion results in a loss of habitat for
 aquatic  organisms such as  invertebrates, reduces fish production, and
 alters water quality  (Thomas 1978, Duff 1978,  Gunderson 1968, and Platts
 1981).   Thus impacts  to streams  caused by users of riparian zones have
 been  well documented.

      This paper addresses  two research programs designed to improve
 riparian zones of an  ephemeral and  a perennial desert steppe stream by
 1) promoting water  storage and vegetation yield season long, 2)
 controlling  nonpoint  source  pollution  and, 3)  advancing the state of
 knowledge for  management of  stream  side ecosystems subjected to user
 pressure.

 Theory for Reclamation of  Cold Desert  Steppe  Streams

      Reclamation  of degraded streams should be initiated on hydraulic
 units where  treatments will  most likely reverse impacted channel
 conditions.  Criteria  for selecting hydraulic  units should allow
 experimental replication on  similar stream reaches or between streams.
 Our study sites are selected based  on  seven geomorphologic and hydraulic
 criteria.  The criteria used are as follows:   1) the stream reach
 gradient must  be  low, 2) the stream reach should have a mature meander
 pattern,  3)  the channel morphology of  the stream should have a developed
 or developing  flood plain, 4) stream flow should be losing to
 surrounding  alluvium  during  high flow, 5) channel damming within the
 selected stream reach should cause maximum response upstream, 6) a high
 potential exists  for  water spreading to adjoining flood plains as
 reclamation  progresses and,  7) the  stream reach selected should be
 typical  of those  found and similarly managed  for grazing in the
 immediate area.

_S tream_ Grad ient

      The success  or failure  of reclaimation of degraded stream channels
 and adjoining  riparian zones depends on deposition of sediment on stream
 banks or bottoms  if dams are constructed.  Channels are usually wide
 compared to  mature  natural conditions.  Streams with high sediment load
 usually  cross  areas with steep gradients but  lose sediment in areas of
 low gradient because  of a  drop in flow velocity.  If one wishes to
 achieve  maximum deposition of sediment for reclamation of stream


                             T-II-c-3

-------
 channelsi  reclamation should start on reaches with  the  lowest gradient
 along the  stream course.

 Stream Maturity

      Mature stream reaches are those which meander  from valley wall to
 valley wall through a developed flood plain  of  deposited alluvium.
 Maturity is associated with low stream gradient.  Alluvium is present
 because of a decrease in  flow velocity and meandering is a natural
 physical flow response across low gradient areas.   Because of
 meandering, valley flood  plain width is maximum.  Meandering also
 increases  the length of stream per lenght  of valley floor.  This
 increases  the contact of  water with banks  and the length of time of
 exposure of water to a valley reach.   Reclamation of degraded streams on
 reaches where meandering  occurs should subject  more land area to
 flooding and promote riparian habitat.

 Flood Plain
      Flood plains  along  a  stream  course  reflect response of flow regime
 to  channel width.  Degraded desert  streams that have been downcut tend
 to  widen because of wet-dry cycles  that  cause bank slumping.  Piping and
 bank erosion also  add  sediment  to the channel bottom.  Slumped banks and
 eroded  sediment deposited  on  the  channel bottom are flushed downstream
 during  periods of  high flow.  This  flushed sediment is often more than
 is  contributed to  stream flow by  scouring of channel banks (Leopold et
 al.  1964)  As downcut channels widen, streams during low flow begin to
 meander within the confinement  of high banks and deposition of sediment
 occurs  on  low banks.

      Stream dynamics and channel  morphology related to sediment
 transport  and delivery are generally reviewed in Leopold et al. (1964),
 Morisowa (1968), Graf  (1971), ASCE Task  Committee (1975) and Simons and
 Senturk (1977).  Hydraulics of  sediment  transport in stream flow is
 further reviewed by Parker (1978a,b).  Andrews (1983) studied bedload
 transport  and deposition within a stream in western Wyoming and
 clarified  the relationship between flow  regime and bank deposition of
 sediment.   Shumm (1963) was able  to show trapping of fine sediments
 by vegetation on banks in eastern Colorado.  Andrews (1983) work related
 to Shumm's by showing  that build  up of bank sediment occurs before
 stream  bottoms rise.   During periods of  low flow sediment is first
 deposited  on banks.  Built up banks cause the channel to become
 narrower.   Bed load is then moved dowstream by increased flow velocity
 because of the new narrow channel.  If aboveground and belowground
 biomass does  not solidify the new bank sediment before the next high
 flow event, bank accumulation will be transported downstream as
 suspended  sediment or  bed load.  This response to high flow maintains
 the channel at a given width and depth corresponding to a yearly average
 flow regime.  The width and depth of a stable channel however
 corresponds to the mean average bank full flow (Andrews 1983).  Stable
banks are  not eroded because high flows overtopping stable banks and
 dissipated  over a well vegetated  flood plain reduces flow velocity.
 This hydraulic process is explained in part by Parker (1978a,b).
Andrews  (1983) points  out that although a stream maintains an average

                           T-II-G-4

-------
 width, it will move laterally from year to year thus  fitting  Leopold  and
 Lanbein's (1966)  description of meandering streams.   The presence  of
 undercut banks along stable stream systems are evidence of  lateral
 movement of meanders.

      Bank first sediment deposition,  vegetation encroachment  into  the
 new bank sediments, and reduced flow  velocity by dewatering or ponding
 are important points to consider when reclaiming degraded streams.
 Vegetation which  is allowed to become well established  within newly
 deposited sediment on banks creates channel bank stability  and a
 roughness factor  to decrease flow velocity during flood stages.  Reduced
 flow velocity will cause flood waters laden with sediment to  deposit
 some of its load.   Sediment dropped on vegetated banks  will be further
 stabilized by regrowth of plants.  This process will  continue to narrow
 a channel until it adjusts to a mean  width sufficient to meet the  annual
 flow regime and is often observed in  desert streams within  the
 confinement of high banks.  When this happens the channel is  mature with
 a developed and stable flood plain.

      Reclamation  of downcut streams in areas where flood plains are
 developed and well vegetated can most effectively be  accomplished  in  low
 gradient stream reaches.  Meander patterns can further  be broken down
 into straight and meander reaches.  Manipulation of stream  flow by
 darning to decrease flow velocity should be implemented  on straight
 stream reaches.  This reduces the chance of losing dams because of
 channel lateral movement along the meander's concave  curve.   In
 addition, belowground and aboveground biomass offer resistance to
 erosion and flow  velocity when instream flow structures are utilized  to
 create ponding and entrapment of bed  load.

 Lpsing_Strearns

      Stream channel transmission loss of water downstream during high
 periods of flow to surrounding alluvium should cause  a  decrease in flow
 velocity and consequently a loss in sediment.  Sediment deposition should
 be maximized in mature stream reaches in areas of low gradient where
 meandering increases time in travel of flow downstream.  Transmission
 loss of stream flow in desert streams has been shown  by Lane  (1970).
 Loss of stream flow should be maximum near the mouth  of a drainage basin
 or in larger basins of losing desert  streams because  of increased  travel
 time of water during any one runoff event (Glymph and Holtan  1969).
 Location of manipulative practices to reclaim degraded  streams based  on
 loss of flow and  sediment deposition  should therefore be placed to
 maximize water travel time after considering the previously discussed
 criteria.

JDamming

      Damming by instream flow structures, like check  dams or  trash
 collectors, and biological damming by beaver or constrictive  channel
 dams caused by encroaching banks and  riparian zones may cause:  1)
 reduced flow velocity, 2) stable bed  load and 3) storage of water  in
 banks proximal to the dam.  Heede (1978 and 1982) discusses reclamation
 of gullies by raising a local base level of an ephemeral stream reach to
 decrease gradient slopes upstream using a variety of  check  dams.   The

                            T-II-G-5

-------
 lower gradient reduces sediment transport.   Deposition occurs upstream
 in a wedge shape until reaching an elevation change upstream equal to
 the elevation of the dam.  Following Heede's 1978 and 1982 research,
 dams should be placed downstream just above a tributary junction.   To
 achieve restoration of riparian habitat however,  the dam should also be
 located on a stream reach having a low gradient where meandering occurs
 and a stable floodplain exists.  The dam will then cause bank deposition
 of sediment and maximum channel filling within the upstream drainage
 network.
      Water spreading across flood plains formerly left dry because  of
 stream downcutting and occupied by pseudoriparian vegetation decreases
 flow velocity.   Shrubby type vegetation may cause sediment to be
 deposited in rows downstream from woody plants parallel to the main
 stream channel.   These rows of deposited sediment may form many smaller
 channels (braiding) which may carry water during high flow events.
 Braiding may also occur in wide channels when vegetation establishes on
 the channel bottom and sediment deposits downstream.   Once braided
 channels become  stable because of vegetation, multiple check dams can be
 established to maximize water spreading over larger areas.

      Water spreading increases riparian zone area by  promoting return
 flow as runoff and ground water to the main channel.   Where degraded
 streams have downcut, widened, and formed a mature flood plain within
 high banks, water spreading is limited.   However,  constrictive damming
 by encroached riparian zones into a main channel,  instream flow
 structures, and  beaver dams often cause peak flows to flood over high
 banks upstream thereby creating desired riparian habitat.   Wider
 alluvial valley  floors are desirable to maximize water spreading.

jGrazing JJ.espj)nse_

      Knowing how to maintain riparian zones at desired conditions when
 used by livestock and wildlife necessitates accumulating verifiable
 information on:   1) how behavior of animals using  uplands and riparian
 zones under any  one grazing management strategy vary,  2)  which season of
 use best maintains riparian habitat,  and 3)  how different stocking  rates
 affect vegetation and stable banks.   Information gathered about these
 three questions  provide input for developing sound strategies to manage
 livestock in cold desert regions where the area of riparian zones is
 small in comparison to surrounding uplands.

      Grazing by  livestock on degraded downcut streams  that  have a mature
 flood plain confined within high banks may be beneficial.   High banks
 channelize stream high flow causing an increase in velocity and
 prevent water spreading.   Hoof action by livestock may  round or flatten
 high banks by trampling.   Hoof imprinting may cause depression storage
 and increase infiltration.

 Approach Taken for Reclaiming Cold Desert Steppe Streams

      Two field research facilities were  selected to study reclamation of
 degraded desert  streams and associated riparian zones.  One  is located

                            T-n-G-6

-------
on an ephemeral stream (15 Mile Creek) in North Central Wyoming near the
town of Worland.  The second, a perennial stream (Muddy Creek), is
located in south central Wyoming between the towns of Rawlins and Baggs.
Both study areas were selected by following the basic criteria
previously discussed.  High concentrations of sediment are present in
flow of both streams.

Ephemeral^ Channels-Facilities

     The  15 Mile Creek facility is divided into 5 pastures.  Pasture 1,
downstream, is going to be used for the study of watershed improvement
practices designed to deposit sediment and improve riparian habitat.
Pastures  2, 3, and 4, located upstream are being used to study the
effect of spring, summer, and fall grazing on stream channel morphology
and vegetation.

     Pasture 5 serves as a control and is located upstream from the
other four pastures.  Livestock behavior studies are being conducted
inside grazed pastures as well as outside to document seasonally
selective use of upland and riparian habitats.  Livestock use studies in
a mature  riparian zone and utilization of associated riparian plants is
being conducted in two separate pastures placed on a second stream near
to the 15 Mile Creek facility.  Utilization levels on plant species in
relation  to season of use and stocking rate are being determined.

Epjietaeral_ Channel Mea_su_remen£s^

     Change in stream channel morphology is being monitored using cross
section techniques.  Permanent cross sections placed on meander and
straight  stream reaches are located downstream, within, and above the
pasture facility.  Changes in channel depth and in cross-sectional area
are being determined.

     Comparative vegetation analysis are being conducted upstream and
downstream from the pasture system.  Within the pastures, permanent
vegetation transects located at channel cross section locations are
established to monitor change in production and shrub density.
Vegetation transects extend from uplands to within 50' of the low
channel bank, from 50" to the low bank, from the top of the low bank to
the interim channel, and along the edge of the interim channel.  Degree
of encroachment of vegetation (above and belowground) across the interim
channel is also being measured.  Core samples on low banks are also
being taken to determine belowground biomass.

     Monthly monitoring of soil moisture using neutron scattering
techniques follows vegetation transects from upland to low bank edge in
all pastures.  In addition, soil moisture data is being collected across
meanders  and straight sections.  Water table depth is recorded and
precipitation monitored continuously during spring, summer, and fall.
Substrates on cross-sections have been core-sampled and analyzed for
particle  size distribution and textural classification.  We intend to
use these data to characterize sediment deposition on banks and with
belowground biomass examine bank resistivity to erosion.


                            T-II-G-7

-------
     Willow propogation and planting  trials have been initiated within
 the interim channel of  the downstream pasture  (1).  Survival relative to
 planting depth and  irrigated-nonirrigated treatments are being
 evaluated.   The purpose of the planting trial  is to study the
 feasibility of forming  a channel plug of willows to filter sediment from
 flow and raise bed  load level.

 Perennial Stream-Facilities
     The Muddy  Creek  facility is divided into six hydraulic stream reach
units  including areas above and below designated treatment areas.  The
first  hydraulic unit  upstream demonstrates stream headcut problems
encroaching  into riparian zones formed by beaver dams and good riparian
zone habitat management.  Because of waterspreading and constrictive
damming, this riparian zone has caused sediment deposition and
maturation of the stream side zone upstream.  Existing differences in
livestock grazing management within this unit illustrate negative
impacts of inappropriate cattle grazing on willow communities and beaver
damming activity as well as the subsequent deterioration in bank
stability.

     Downstream, (Unit 2) below active headcutting, channelization has
occurred but meandering flood plains are confined within high banks.
Willows are  present but limited and need rest from grazing to stabilize
flood  plains.   In addition, recent sediment deposited should be planted
with willows to encourage beaver damming as well as provide potential
sites  for installing  instream flow structures to raise water tables,
cause  deposition of sediment on flood plains, and stabilize bed load.
This channelized stream reach, hydraulic unit number 2, is planned as a
research area for documenting the hydrologic response affiliated with
reclamation  of  cold desert perennial streams.

     Hydrologic data will be collected for three years to develop a
water balance of this degraded stream and riparian zone.   Three more
years are also planned to stabilize flood plains using willows.  Four
additional years will be devoted to using instream flow structures and
beaver to raise the water table and bed load profile over a distance of
approximately three miles and six feet change in elevation.  Pertinent
questions to be answered are:

     1) How much water is stored underground by natural damming and
     riparian zone improvement practices?
     2) What change in downstream flow regime occurs in the form of
     prolonged release of water season long?
     3) What water is lost to downstream users during periods of need
     and right?
     4) What water balance and bank stability response occurs because of
     watershed improvement practices like brush control (burning,
     spraying) and range fertilization?
     5) What is the significance of good riparian zone management to
     abatement of nonpoint and point source pollution?
     6) What water balance,  bank stability, and vegetation response
     occurs because of livestock and wildlife grazing on  improved
     riparian zones because  of watershed improvement practices?

                           T-H-G-8

-------
     7) What are the up and downstream effects of riparian zone and
     watershed improvement practices?
     8) What are the economic costs and benefits of improving degraded
     riparian zones?

     Research methodology initiated after three years of hydrologic
monitoring of this three mile stream reach will be based on pilot
research being conducted downstream.  Water leaving the designated
research facility (Unit 2) flows into the third hydraulic unit.  The
downstream portion of hydraulic unit three, approximately three miles of
a 6 mile stream reach, is characterized by filling with sediment caused
by waterspreading and beaver damming downstream in the fourth hydraulic
unit.

     Hydraulic unit number three is in various stages of plant
succession and dissected by braided stream channels caused by flooded
shrub vegetation and deposited sediment.  Mud flats are evident.
Willows are virtually absent, perhaps caused by historic early spring
livestock grazing.  This reach has been fenced to manipulate spring
grazing and promote willow establishment.  This unit will serve as an
area for initiating riparian watershed improvement practices on a
response area caused by good riparian zone management downstream.  Grass
and willow planting trials will be initiated on exposed sediments.

     Hydraulic unit number four is a stream reach approximately ten
miles long and has been managed to maintain riparian habitat.  The main
stream has one diversion to spread water and return flow enters the
braided main channel.  Channels are dammed by beaver.  Grass has been
established to stablize sediment and provide livestock forage.  Willows
are prominent and grazing by livestock occurs during late fall and
winter.

     Next downstream, hydraulic unit number five, approximately three
miles long, is downcut and has an immature flood plain constrained
between high banks.  This unit may reflect response to flow that has
been stripped of sediment by the improved riparian zone upstream.  Flow
in this unit, because of less sediment, should have more power to erode
the existing channel.  No work is currently planned in this area.

     Hydraulic unit number six changes from a stream channel with an
immature flood plain to a channel with mature floodplains abruptly below
the confluence with an ephemeral stream.  Perhaps this change in flood
plain development is caused by an influx of sediment to Muddy Creek
during flow events from the ephemeral channel.  Willow growth on the
floodplains of hydraulic unit number six grades from light just below
the confluence of the ephemeral stream channel to heavy downstream.
This entire unit (three miles in length) has been fenced to demonstrate
the use trash collectors as a tool for reclaiming riparian zones.
Fencing will allow control of livestock grazing trials on streamside
vegetation during different grazing seasons and a wildlife exclosure
will provide information on effects of livestock and wildlife vegetation
on riparian zone vegetation and channel morphology.  Grazing trials have
begun coincident with installation of trash catchers and fences.

                            T-II-G-9

-------
     Trash collectors have been Installed on straight stream reaches
within  three pastures to be used for livestock grazing during spring,
summer, and fall  followed by winter wildlife use.  They have also been
installed in the  wildlife exclosure and in a pasture downstream
adjoining the wildlife exclosure where wildlife grazing but no livestock
grazing will occur.  A control reach is downstream from all fenced areas
and has no trash  collectors.  This entire three mile stream reach is now
in a ponded condition during low flow due to trash collectors.  Beaver
have now used three trash collectors for constructing dams and winter
food storage.

     Hydraulic unit number six was selected based on the criteria
discussed previously for testing trash collectors as a tool to reclaim
degraded streams  and riparian zones.  It is hypothesized these
structures will:  1) promote bank storage of water, 2) cause deposition
of sediment on channel flood plains thereby causing channel narrowing,
3) stablize and raise bed load levels and 4) cause upstream flooding
because of constrictive damming of high flow events and thereby sediment
deposition on upstream banks.  Willow establishment caused by increased
height of watertables and water spreading should promote beaver damming.
This would eliminate the need for further installation of trash
collectors in hydraulic unit number 6.  Planting of willows on new
floodplains in hydraulic unit number five would promote channel
stabilization and sustain trash collector improvement practices upstream
if the hypothesized effects in unit six do occur.

Perennial Stream  Measurements
     Measurements recorded follow those discussed for the 15 Mile Creek
study.  Changes in stream channel morphology will be evaluated using the
cross section technique.  Species composition, production, and
utilization of the channel, flood plains and riparian zone vegetaion are
being monitored.  Bank-stream interflow and flood plain soil moisture in
hydraulic unit number 6 is ready for monitoring to evaluate change
caused by trash collectors.  Willow propogation in planting trials will
be evaluated in hydraulic units numbers 1, 3, and 6.  Trash collector
stability and new installations will be evaluated for collecting
bedload, causing bank flooding upstream, and any rise in water tables.
Installation of equipment to construct a water balance is planned for
hydraulic units number 2 and 2 through 6 inclusive.
                            T-II-G-10

-------
                            LITERATURE CITED

ASCE Task Committee.  1975.  Sedimentation Engineering.  ASCE, Manuals
     and Reports on Engineering Practices, No. 54, American Society
     Civil Engineers, New York, New York.

Brown, Sandra, Mark M. Brinson, and Ariel E. Lugo.  1978.  Structure and
     functions of riparian wetlands.  Proc. of Symposium - Strategies
     for Protection and Management of Floodplain Wetlands and Other
     Riparian Ecosystems.  U.S.D.A. For. Ser. GTR-WO-12:  p. 17-31.

Busby, Frank E.  1978.  Riparian and stream ecosystems, livestock
     grazing, and multiple-use management.  Proc. of Forum-Grazing and
     Riparian Streams Ecosystems.  Denver, Colo.  p. 21-30.

Campbell, C.J. and W. Green.  1968.  Perpetual succession of stream-
     channel vegetation in semiarid region.  J. Ariz.  Acad. of Sci. p.
     86-98.

Campbell, C.J.  1970.  Ecological implications of riparian vegetation
     management.  J. Soil and Water Cons.  25.  p. 49-52.

Duff, Donald.  1978.  Riparian habitat recovery on Big Creek, Rich
     County, Utah - A summary of 8 years of study.  Proc. of Forum -
     Grazing and Riparian Stream Ecosystems.  Denver,  Colo. p. 91-92.

Ewel, Katherine Carter.  1978.  Riparian ecosystems:  conservation
     of their unique characteristics.  Proc. of Symposium - Strategies
     for Protection and Management of Floodplain Wetlands and Other
     Riparian Ecosytems.  U.S.D.A. For. Ser. GTR-W012:  p. 56-62.

Glymph, L.M. and H.N. Holton.  1969.  Land treatment in agriculture
     watershed hydrology research.  In:  Effects of watershed changes on
     streamflow, Water Resources Symp. No. 2, Univ. Texas Press.
     Austin,  p. 44-68.

Graf, W.H.  1971.  Hydraulics of sediment transport.  McGraw-Hill Book
     Co., Inc. New York, NY.

Gunderson, Donald R.  1968.  Floodplain use related to stream morphology
     and fish populations.  J. of Wildl. Mgt.  32(3):   508-514.

Haugen, Gordon, et al.  1980.  Management and protection of western
     riparian stream ecosystems.  Amer. Fisheries Soc. Western Div.
     publ. 24 p.
                           T-II-G-11

-------
 Heede,  B.H.   1978.   Designing gully control systems for eroding
      watersheds.   Environ.  Manage.  2(6):509-522.

 Heede,  B.H.   1982.   Gully control:   Determining treatment  priorities  for
      gullies a network.   Environ.   Manage.   6(5):441-451.

 Jahn, Laurence R.   1978.   Values of riparian habitat to natural
      ecosystems.   Proceedings of Symposium  - Strategies for  Protection
      and Mgt.  of Floodplain Wetlands and Other Riparian Ecosystems.
      U.S.D.A.  GTR-WO-12:   p. 157-160.

 Johnson, R.  Roy.   1978.   The lower  Colorado River:   A western  system.
      Proc. of  Symposium - Strategies for Protection and Management of
      Floodplain Wetlands  and Other  Riparian Ecosystems.  U.S.D.A. For.
      Ser. GTR-WO-12:   p.  41-55.

 Kauffman, J. B., W.C.  Krueger, and  M. Vavra.   1983a.  Effects  of cattle
      grazing on riparian  plant communities.   J. Range Manage.
      36:685-691.

 Kauffman, J.B., W.C. Krueger, and M. Vavra.   1983b.   Impacts of cattle
      grazing streambanks  in northeastern Oregon.  J.  Range Manage.
      36:683-685.

 Kauffman, J.B.  and W.C. Krueger.  1984.   Livestock  impacts on  riparian
      ecosystems and  steamside management implications.   Arevien.  J.
      Range Manage.   37:430-438.

 Lane, L.J.,  M.H. Diskin,  and K.G. Renard.   1970.  Input-output
      relationshps for  an  ephemeral  stream channel system.  J. Hydrology
      13:22-40.

 Leopold, L.B., M.G. Wolman,  and  J.P. Miller.   1964.   Fluvial Processes
      in  Geomorphology.    W.H. Freeman and Company,  San Francisco.

 Leopold, L.B. and W.B. Langbein.  1966,   River meanders.  The Physics of
      Everyday Phenomena,  Sci. Am. 79-9287:28-38.

Martin,  S. Clark.  1978.  Evaluating the  impacts of cattle grazing on
      riparian habitat.  Proc. of Forum -  Grazing and  Riparian Stream
      Ecosystems.  Denver, Colo.  p. 35-38.

McCall,  James D. and Robin F. Knox.  1978.  Riparian habitat in
      channelization projects.  Proc. of Symposium - Strategies for
     Protection and Management of Floodplain Wetlands and Other Riparian
     Ecosystems.  U.S.D.A. GTR-WO-12:  p. 125-128.

Meehan, William R.  and William S, Platts.  1978.  Livestock grazing and
     the aquatic environment.  J. Soils Water Conservation:  33(5):
     274-278.
                            T-II-G-12

-------
Miller, Robert C.  1978.  Relationship between the vegetation and
     environmental characteristics associated with an alluvial valley
     floor.  M.S. Thesis.  University of Wyoming, Laramie, WY.  135 p.

Minore, D.  1970.  Seedling growth of eight northwestern tree species
     over three water tables.  U.S. For. Serv. Res. Note PNW-115, Pac.
     Northwest For. and Range Exp. Stn., Portland, Oregon,  p. 8.

Minore, D. and C.E. Smith.  1971.  Occurrences and growth of four
     northwestern tree species over shallow water tables.  U.S. For.
     Serv. Res. Note PNW-160.  Pac. Northwest For, and Range Exp. Stn.m
     Portland, Oregon,  p. 9.

Morisowa, M.  1968.  Streams their dynamics and morphology.  Earth and
     Planetary Science Series, McGraw-Hill Book Company, New York.

Odum, Eugen P.  1978.  Ecological importance of the riparian zone.
     Proc. of Sym. - Strategies for Protection and Management of
     Floodplain Wetlands and Other Riparian Ecosystems.  U.S.D.A. For.
     Serv. GTR-WO-12:  p. 2-4.

Parker, G.  1978a.  Self-formed straight rivers with equilibrium banks
     and mobile bed.  Part 1.  Th/e sand-silt river.  J. Fluid Mecb.  89:
     109-125.

Parker, G.  1978b.  Self-formed straight rivers with equilibrium banks
     and mobile bed.  Part 2.  The sand-silt river.  J. Fluid Mech.  89:
     109-125.

Peterson, H.V.  1950.  The problem of gullying in western valleys.  In;
     Applied Sedimentation.  John Wiley & Sons, Inc., New York.  p. 407-
     434.

Platts, William S.  1981.  Effects of sheep grazing on a riparian -
     stream envrionment.  U.S.D.A. For. Serv. Res. Note Int. 0-307:  6
Roath, L.R. and W.C. Krueger.  1982.  Cattle grazing influence on a
     mountain riparian zone.  J. Range Manage.  35(1):  100-103.

Schmidly, David J. and Robert B. Ditton.  1978.  Relating human
     activities and biological resources in riparian habitats of western
     Texas.  Proc. Symp. - Strategies for protection and management of
     floodplain wetlands and other riparian ecosystems.  U.S.D.A. For.
     Serv. GTR-WO-12.  p. 107-116.

Schumra, S.A.  1963.  Sinuosity of alluvial rivers on the great plains.
     Geological Soc. of Amer.  Bui. 74:1089-1100.

Simons, D.B. and F. Senturk.  1977.  Sediment transport technology.
     Water Resources Publications, Fort Collins, Colo.


                            T-II-G-13

-------
Speck. J.E.  1981.  A comparative study of the effects of continuous
     season-long and deferred rotation grazing on mountain riparian
     vegetation and water quality.  M.S. Thesis.  University of Wyoming,
     Laramie.  149 p.

Thomas, Jack Ward, Chris Maser, and John E. Rodiek.  19878.  Riparain
     zones in managed rangelands- their importance to wildlife.  Proc.
     of Forum - Grazing and Riparian Stream Ecosystems.  Denver, CO.  p.
     21-30.

Tubbs, Allen A.  1980.  Riparian bird communities of the Great Plains.
     Proc. of Workshop - Management of Western Forests and GRasslands
     for Non-game birds.  U.S.D.A. For. Ser. Tech. Rep. Int-86:  p.
     419-433.
                           T-II-G-14

-------
        MANAGEMENT ALTERNATIVES FOR URBAN STORMWATER

             by Robert Pitt, and Roger Bannerman
              Department o-f Natural Resources
                     Madison, Wisconsin
A. Introduction

     This paper is a brie-f example o-f how suitable
stormwater management alternatives can be designed using
available monitoring information. Milwaukee area urban
runoff has been extensively studied by the Wisconsin
Department of Natural Resources, the U.S. Geological Survey
and the Southeastern Wisconsin Regional Planning Commission
(SEWRPC), among others, for about the past ten years. Two
major research projects were conducted for the International
Joint Commission (IJC) on Great Lakes Research (Bannerman,
et al, 1979) and for the Nationwide Urban Runoff Program
(NURP) for the U.S. EPA (Bannerman, et al, 1983). The IJC
project involved detailed urban runoff monitoring of many
single land use and mixed land use sites, while the NURP
project examined urban runoff at eight single? land use sites
for two years. This information, in conjunction with other
regional and nationwide urban runoff characterization and
control data, was recently used by the Nonpoint Source and
Land Management Section of the DNR to prepare a Construction
Site and Stormwater Management Plan and Model Ordinance
(Pitt 1985). This paper uses this summarized information in
several related examples to show how stormwater management
controls can be designed to meet both flood control and
urban runoff quality objectives.
B. Required Inventory Information

     Eiefore an evaluation of suitable control measures can
be made, certain information about the study area must be
obtained. This inventory information pertains to the
importance of the various source areas in contributing
runoff flow or pollutants to the outfall. Land cover
information is needed to describe the percentage of the
study area covered by various surfaces, such as streets,
roofs, landscaping and parking areas. This information is
usually obtained by direct measurement from aerial
photographs. Additional information is needed to describe
the aspects of these areas that modify their pollutant or
flow generating capabilities. Roof drain connections,,
pavement texture and condition and the type of drainage
system are examples. This information requires field
investigations of representative areas for each land use
subcategory. Table i is an example of land cover information

                       T-II-H-1

-------
for typical medium density residential and industrial areas
recently studied  (Pitt and McLean  196S5) . Landscaping
comprises the most common surface  cover in the single family
residential area, while rooftops and paved parking and
storage areas are the most common  surfaces in the industrial
area. These major differences can  result in significantly
different runoff characteristics.
     Figures 1 through 4 are based on extensive runoff
monitoring of these two study areas and show how the
different areas contributed to the outfall total solids and
lead yields for different rain volumes (Pitt and McLean
1985). Rain has a much greater effect on the relative
contributions for the residential  area. The larger amount of
impervious surfaces in the industrial area reduces the
variations observed. Table 2 summarises this information for
a selection of pollutants for large rains  (fairly stable
source contributions occur with large rains because of the
decreased importance of pervious area rain infiltration).
Roof and paved parking/storage area runoff contribute most
of the runoff volume and most of many of the pollutants in
industrial areas. Flow and pollutant contributions are much
more diverse in residential areas. This information is used
in the following example to estimate the effectiveness of
the different control measures in  reducing runoff flows and
pollutant contributions.
C. Infiltration and Detention Basin Combined Effects in
Areas Having Poor Soil Percolation

1. Hypothetical Site Characteristics,,
     A series of calculations were made? to investigate
possible alternative flow and pollutant reduction options
available for a hypothetical study area. Options
investigated included dry and wet detention basins, roof and
parking/storage area infiltration, and grass waterway
drainages. The benefits examined included flow volume and
rate reductions along with pollutant concentration and yield
reductions.
     These calculations are intended to generally describe
the benefits of alternative urban runoff controls available.
The calculation procedures, along with suitable design
guidelines, are described in the recently prepared
Stormwater Management Plan  (Manual of Practice) (Pitt 1985).
As stated above, these calculations are based on many field
monitoring projects, including several conducted during the
past ten years in the Milwaukee and Menomonee River
watersheds (Bannerman, et al,  1979 and 1983).

2. Flow Estimates.
     Table 3 shows the land use areas in the example
watershed, divided by hypothetical soil percolation
characteristics. Table 4 shows the calculations used to

                       T-II-H-2

-------
estimate the average urban runo-f-f flow rates for a variety
of large storm conditions. These flow rate estimates are
based on the monitoring results of several hundred Milwaukee
area urban runoff events studied by Banner-man,, et al <1979
and 1983). Statistical analyses of these runoff data also
indicated that peak runoff flow rates for large urban runoff
events are about 6.5 times the average flow rates. The peak
flow rates generally only occur for several minutes during
the event.

3. Dry Detention Basin Design.
     Dry detention basin storage requirements were estimated
for" the different land uses. The basins should drain within
a few days to enable control of storms likely to fallow in a
short time. For the critical 100 year storm., basin outflow
rates of between 100 and 200 cfs would result in reasonable
drainage times. Table 5 shows required basin sizes for
various outflow rates. Total basin areas of about three
percent of the watershed area may be required to control the
critical storm. Alternative basin criteria can also be used
to determine the required basin sizes.
     Alternative detention basin design objectives and
associated required basin sizes are summarized on Table 6.
The minimum basin size would reduce the expected peak flow
rate to the assumed channel carrying capacity and would
require about 0.3 percent of the watershed area. The maximum
basin size is designed to allow basin drainage in three days
to obtain a maximum flow rate reduction and would require
about three percent of the watershed area.. A reasonable
basin objective is intermediate to these two sizes and would
reduce the peak flow rates to the average flow rates. A
basin size of about 0.5 percent of the watershed area would
be required. Even though these dry detention basins would
substantially reduce the peak flow rates, they do not change
the total volume of water or the amounts of pollutants
entering the creek and may not benefit flood control or
water quality in downstream receiving waters.

4. Wet Detention Basin Design.
     Many monitoring projects have demonstrated substantial
water quality benefits associated with well designed wet
detention basins (EPA 1983). Wet basins can be used in
conjunction with dry basins  (and/or other urban runoff
control devices) to help meet both flood control and water
quality objectives. Wet basins also do not reduce the total
volume of water discharged to the receiving water, but can
remove much of the settleable solids.
     High control levels of pollutants in wet basins are
only reasonably possible for the more frequent rains. Water
quality is not usually of much concern during very large
infrequent events, such as the 100-year storm. The most
cost-effective benefits can be relealizecl during common,
smaller events. A wet basin about 0.3 percent of the
watershed area, having a wet depth of about four feet and a

                       T-II-H-3

-------
•freeboard height of about  12  -feet would produce about 75
percent,  solids control. This  -freeboard requirement is needed
to reduce the wet pond outflow velocity to desired values
and may  be  larger than desired.  It would be possible to
reduce this freeboard requirement with the use of a surge
dry basin or upland controls.

5. Upland Infiltration Design.
     Upland infiltration can  be  used to reduce the flow
rates to downstream detention basins, but more importantly,
they reduce the total volume  of  water and amounts of
pollutants  discharged. Upland infiltration devices recharge
the local groundwater and  increase critical low flows during
summer dry  months, further benefiting the receiving water.
Care must be taken to protect significant groundwaters, of
course.  Upland infiltration can  also be used only in areas
with appropriate soil percolation conditions. The soils in
the hypothetical watershed are of two percolation classes:
poorly drained and ,/noderatly  well to well drained. The
following calculations only consider the application of
upland infiltration in the better drained soils.
     Table  7 shows the maximum effects that roof and
parking/storage area infiltration may have on flow rate and
volume reductions for several large rains in residential and
industrial  areas. These reductions can be substantial,
especially  for industrial areas  where about SO percent of
the water originates from roofs  and parking/storage areas.
These improvements must be reduced because of the poor soils
in the assumed study area. Roof  runoff infiltration
performance in the residential area would be reduced to
about ten percent, roof runoff infiltration performance in
the industrial area would be reduced to about 15 percent,
parking/storage area infiltration performance in the
industrial  area would be reduced to about five percent and
roof plus parking/storage area runoff infiltration
performance in the industrial area would be reduced to about
20 percent.

6. Grass Waterway Design.
     Grass  waterways may also significantly reduce flow
volumes through infiltration. Table 8 shows the effects of
grass drainages, instead of underground sewerage.  Flow rates
and volumes during the large  100 year storm may be reduced
by as much  as 20 percent by these devices.  Table 9 shows the
total  expected watershed flow rates using these controls.

7. Revised  Dry Detention Basin Design Considering Upland
Infi1trati on.
     Table  10 shows the revised dry basin storage
requirements with these upland controls.  The area
requirements have been slightly reduced,  to about 0.4
percent for  the 100 year storm and for peak shaving,  but the
outflow rates have been reduced substantially for many
conditions.  The revised wet basin area and  freeboard

                       T-II-H-4

-------
requirements are slightly reduced  (down to about  10  -feet
over about 0.3 percent o-f the watershed area).

D. Summary.
     Table 10 presents a summary o-f the expected  water
quality and quantity improvements  -for the various
alternatives investigated. No water quality or water volume
improvements will occur -for the dry basins alone, but
substantial reductions in -flow rates,, water volume,  and
pollutant yields and concentrations will likely occur -for  a
combination o-f control devices.
     In conclusion,, a reasonable amount, of the watershed
land area can be used to effectively control  urban runoff
quantity and quality. Dry basins comprising about 0.4
percent of the? watershed area  (about six feet deep)  can be
used with wet basins requiring about 0.3 percent  of  the area
(having a wet pond de?pth of about  four feet and a freeboard
elevation of about ten feet) and upland infiltration devices
(in areas having suitable soils) to achieve significant flow
and pollutant reductions. Similar  control devices have been
monitored and proven effective during many urban  runoff
research projects. Special restricting conditions (such as
poor percolating soils) can be considered in  the  design of
these control devices.

E. References

     Bannerman,, R. , J.G. Konrad, D. Becker., G.v1.  Simsiman,
G. Chesters, J. Goodrich-Mahoney,  and B. Abrams.  "The IJC
Menomonee River Watershed Study",,  U.S. Environmental
Protection Agency, EPA-905/4-79-029-C, Chicago,  111., 1979.

     Bannerman, R., K. Baun, M. Bohn. P.E. Hughes, and D.A.
Grac2 y k. "Evaluation  of Urban Nonpoint Source Pollution
Management in Milwaukee County,, Wisconsin", U.S.
Environmental Protection Agency, PB 84-114164., Chicago,
111., 1983.

     Pitt, R., and J. McLean.  "Toronto Area Watershed
Management Strategy Study: Number  River Pilot Watershed
Project", Ontario Ministry of the  Environment, Totonto,
Ontario,, 1985  (Draft).

     Pitt, R. "Construction Site Erosion and  Stormwater
Management Plan and Model Ordinance", Wisconsin Department
of Natural Resources, Madison, Wisconsin, 1985  (Draft).
                       T-II-H-5

-------
H
I
33
     u
         THISTLEDOWNS  TOTRL SOLIDS SOURCES
                          Rein (MI)
                     Vve.Ua-> lift

-------
                      F\
H

I
a
i
               EMERY TOTRL SOLIDS SOURCES

-------
H
I
a
00
                THISTLEDOWNS LERD SOURCES
                            Rein (MI)

-------
                   EMERY  LERD  SOURCES
H
a
                     K

-------
  Table 1. Industrial and Medium Density Residential Land Covers (%)
Land Cover                       Industrial             Residential

Streets                             5.6%                     11.4%
Swales                              0                        3.1
Grass strips                        0                        9.2
Concrete walks                      0.4                      3.7
Front landscaping                  10.2                      16.3
Paved driyeways                     2.2                      6.4
Paved parking/storage              19.9                      0>1
Unpaved parking/storage            15.4                      0
Sidewalks                           0.1                      0.7
Back landscaping                    0                        24.5
Undeveloped                        12.4                      0
Railroad                            2.6                      0
Connected roofs                    31.1                      4.9
Roofs draining to driveways         0                        8.7
Roofs draining to lawns             0                        10.9

                       Total       99.9                      99.7
Total Impervious                    47.4                      36.4
Total Pervious                      30.4                      47.6
Total Disconnected Impervious       22.2                      16.0
        Source: Pitt and McLean  1985
                            T-II-H-10

-------
                        Table  2. Sources of Pollutants  During  Large Events
                 Industrial Areas:

                 Constituent
                Front  Drive-
Streets  Walks  Lands,  ways
Paved  Unpaved  Open   Connected
Park.  Park.    Space  Roofs
H
I
I
SB
I
Total solids
Suspended solids
Phosphorus
Kjeldahl N
Phenolics
COD
Fecal colif.
Copper
Lead
Zinc
Flow
Area

Mixed Residential

Total solids
Suspended solids
Phosphorus
Kjeldahl N
Phenolics
COD
Fecal Colif.
Copper
Lead
Zinc
Flow
Area
2%
3
1
1
2
1
2
3
2
1
9
6

Area:
Streets
6
8
3
8
3
8
8
14
12
3
23
10
1%
1
1
1
1
1
3
1
1
1-
1-
1-


Walks
6
4
6
24
17
27
59
4
12
5
3
4
27,
10
10
6
2
2
6
6
3
2
4
10

Front
Lands.
20
25
24
3
1-
4
7
4
2
1-
8
23
3%
4
2
2
5
3
2
4
3
4
2
2

Drive-
ways
8
8
3
2
2
2
1
14
10
2
6
5
26&
42
20
30
58
37
21
50
35
51
30
21

Paved
Park.
14
18
8
6
4
3
2
26
22
5
8
5
20%
20
8
3
8
8
25
12
12
8
3
15
Open/
Rear
Lands .
26
32
29
6
1-
7
14
6
3
1-
15
31
10%
14
10
5
1
3
5
4
2
2
3
15

Connect .
Roofs
14
3
20
36
14
34
6
21
26
59
19
8
36%
6
48
52
23
45
36
20
42
32
49
31
Roofs
to
Drive.
6
2
6
13
56
13
3
10
12
24
9
7












Roofs
to
Lawns
1-
1-
1
2
4
2
1-
1
1
2
2
7
                         Source:  Pitt and McLean  1985

-------
     Table 3. Example Land Use by Soil Type  (acres)
Residential
Industrial
Commercial and Institutional
Open Space

      Total
Poorly       Well to         Total
Drained      Moderately
Soil         Well Drained
1125           1445           2570
 435            235            670
 110            110            220
 125             85            210

1795           1875           3670
       Table 4.  Average Flow by Land Use (cubic feet per second)
Return
Freq .
(yrs)
1
1
1
1
1
1
1
10
100
Rain
Dur.
(hrs)
0.5
1
2
3
6
12
24
24
24
Total
Rain
(in.)
0.9
1.1
1.3
1.5
1.7
2.0
2.3
3.9
5.5
Rain
Int.
(in/hr)
1.80
1.10
0.65
0.50
0.28
0.17
0.10
0.16
0.23
                                 Residential
                                    0.83
                                    0.46
                                    0.27
                                    0.23
                                    0.14
                                    0.081
                                    0.045
                                    0.083
                                    0.13
          Institutional,
          Commercial,  or
          Industrial

                1.1
                0.59
                0.36
                0.30
                0.18
                0.10
                0.056
                0.10
                0.15
Open Space
    0.06
    0.04
    0.03
    0.03
    0.01
    0.01
    0.01
    0.02
    0.05
                           T-li-H-12

-------
       Table 5. Alternative Dry Basin Designs for Hypothetical Example
   1) Minimum design to reduce peak runoff to assumed channel  capacity
      (1500 cfs):
 Storm  Peak
       Reqd.
(24 hr) (cfs)
         (1.)
  1 yr   1100
 10      2000
100      3100
       (acre-ft)

         0
        21
        66
          Reqd.
 Freq.  Inflow  Storage    Surface
          (acres)
            (2)
            0
            3.5
           11
           Reqd. %     Pond
           of 3700     Drain
           acre basin  (days)
             0%
             0.1
             0.3
            0
            1.3
            2.1
   2.) Maximum design to allow pond drainage in  three days:
Storm  Ave.
Freq.  Inflow
(24hr) (cfs)
       Reqd.
       Outflow
       (cfs)
Reqd.      Reqd.
24 hr      Surface
Storage    (acres)
(acre-ft)    (2)
   lyr  170      57
  10    300     100
 100    480     160
                  220
                  400
                  630
                     37
                     67
                    105
                               Reqd. 2
                               of 3700
                               acre basin
                        1.0%
                        1.8
                        2.8
   3.) Reasonable design to reduce  short-term  peak  to ave.
      (peak shaving):
 Storm  Peak
 Freq.  Inflow
 (24hr) (cfs)
         (1)
                                                 flow
Ave.
Inflow
(cfs)
Reqd.
30-min
Storage
(acre-ft)
Reqd.
Surface
(acres)
(2)
Reqd . %
of 3700
acre basin
   1 yr
  10
 100
1100
2000
3100
170
300
480
   38
   70
  108
 6.3
12
18
  Footnotes:
0.2%
0.3
0.5
    (1) Peak flow duration is assumed  to be 30 minutes
    (2) Dry pond depth is six feet
                                T-II-H-13

-------
     Table 6. Effects of Upland  Infiltration on Runoff Rates  (cfs per acre)
Residential Area
Storm
Freq.
(yrs,)
1
1
1
1
1
1
1
10
100
Storm
Dur.
(hrs)
0.5
1
2
3
6
12
24
24
24
Ave . Flow
With No
Infilt.
0.83
0.46
0.27
0.23
0.14
0.081
0.045
0.083
0.13
Ave . Flow
With Roof
Disconnect •
0.67
0.38
0.23
0.19
0.12
0.065
0.037
0.071
0.11
Percentage
Ave . Flow
Reduction
19%
17
15
17
14
20
18
15
15












Industrial Area:

Storm
Freq.
(yrs)
1
1
1
1
1
1
1
10
100

Storm
Dur.
(hrs)
0.5
1
2
3
6
12
24
24
24

Ave . Flow
With No
Infilt.
1.1
0.59
0.36
0.30
0.18
0.10
0.056
0.10
0.15
Roof Dis. :
Hew %
Ave . Reduc
Flow
0.54 51%
0.30 49
0.19 47
0.15 50
0.094 48
0.053 47
0.031 45
0.060 40
0.098 35
Parking
New
Dis.
%
Ave . Reduc .
Flow
0.90
0.50
0.31
0.25
0.15
0.088
0.048
0.090
0.13

18%
15
14
17
17
12
14
10
13
                                                              Roof  and  Parking
                                                              New       %
                                                              Ave.    Reduc•
                                                              Flow
                                                              0.37
                                                              0.21
                                                              0.13
                                                              0.11
                                                              0.064
                                                              0.038
                                                              0.022
                                                              0.048
                                                              0.080
66
64
64
63
64
62
61
52
47
                                T-II-H-14

-------
Table 7. Effects of Swale Drainages on Runoff  Rates  (Percent)
Storm
Freq.
(yrs)
1
1
1
1
1
1
1
10
100
Storm
Dur.
(hrs.)
0.5
1
2
3
6
12
24
24
24
                   Percentage Flow Reducrion  for:
                     Residential      Industrial
                    (3 ft. swales)    (7 ft.  swales)
                        15-
                        15-
                        15-
                        15-
                        19
                        27
                        50
                        33
                        23
14-
14-
14-
14-
14
23
35
18
14-
  Table 8. New Runoff Volumes with Upland Infiltration  (cfs)
Storm
Freq.
(yrs)
1
1
1
1
1
1
1
10
100
Storm
Dur.
(hrs.)
0.5
1
2
3
6
12
24
24
24
Ave. Flow
With No
Infilt.
3100
1700
1000
860
520
300
170
300
480
Ave . Flow
With
Infilt.
2760
1520
900
760
425
230
120
240
380
Peak Flow
With No
Infilt.
NA
NA
NA
NA
NA
NA
1100
2000
3100
Peak Flow
With
Infilt.
NA
NA
NA
NA
NA
NA
750
1600
2500
Percentage
Flow
Reduction
11%
11
11
12
22
22
32
21
21
                 T-II-H-15

-------
       Table 9. Alternative Dry  Basin  Designs  for Hypothetical  Example
                With Upland Infiltration
   1) Minimum design  to reduce  peak  runoff  to  assumed  channel  capacity
      (1500 cfs):
 Storm  Peak
 Freq.  Inflow
(24 hr) (cfs)
         (1)
  1 yr    750
 10      1600
100      2500
Reqd.      Reqd.     Reqd. %     Pond     Pond
                                 Drain    Acres
                              n  (days)   Saved

                                   0        0
                                   1.1      2.8
                                   1.7      4.2
Storage
(acre-ft)

0
4
41
Surface
(acres)
(2)
0
0.7
6.8
of 370i
acre b;

0%
0.02
0.2
   2.) Maximum design  to  allow  pond  drainage  in  three  days:
Storm  Ave.
Freq.  Inflow
(24hr) (cfs)
   lyr  115
  10    240
 100    380
Reqd.
Outflow
(cfs)

Reqd.
24 hr
Storage
(acre-ft)
Reqd.
Surface
(acres)
(2)
Reqd. %
of 3700
acre basin

Pond
Acres
Saved

 38
 80
130
150
320
500
25
53
83
0.7%
1.4
2.2
12
14
22
   3.) Reasonable design  to  reduce  short-term  peak  to  ave.
      (peak shaving.):
                                          flow
Storm Peak
Freq. Inflow
(24hr) (cfs)
(1)
1 yr 750
10 1600
100 2500
Ave.
Inflow
(cfs)

115
240
380
  Footnotes:
                          Reqd.       Reqd.
                          30-min      Surface
                          Storage     (acres)
                          (acre-ft)    (2)
                             26
                             56
                             88
                        4.4
                        9.4
                       15
                               Reqd. %
                               of 3700
                               acre basin
                       0.1%
                       0.25
                       0.4
    (1) Peak flow duration  is assumed  to be  30 minutes
    (2) Dry pond depth is six feet
                                  Pond
                                  Acres
                                  Saved
                        1.9
                        2.6
                        3.0
                                T-II-H-16

-------
   Table 10.  Pollutant Control Alternatives (Approximate Percentage Control  at  Outfall)
Control Total Sus.
Option: solids solids
H
M
M
a
i-1
•vj
1
1
1
2
2
2
3
3
3
4
5
6
6
6
7
7
7
8
8
8
9
9
9
10
lyr
lOyr
lOOyr
lyr
lOyr
lOOyr
lyr
lOyr
lOOyr
lyr
lyr
lyr
lOyr
lOOyr
lyr
lOyr
lOOyr
lyr
lOyr
lOOyr
lyr
lOyr
lOOyr
lyr
0
0
0
0
0
0
0
0
0
90
75
34
25
20
34
25
20
34
25
20
34
25
20
84
0
0
0
0
0
0
0
0
0
60
50
28
19
14
28
19
14
28
19
14
28
19
14
64
Phos
0
0
0
0
0
0
0
0
0
55
45
34
25
20
34
25
20
~ 34
25
20
34
25
20
64
. TKN
0
0
0
0
0
0
0
0
0
35
30
44
35
30
44
35
30
44
35
30
44
35
30
61
Phen.
0
0
0
0
0
0
0
0
0
50
40
36
27
22
36
27
22
36
27
22
36
27
22
62
COD
0
0
0
0
0
0
0
0
0
55
~45
44
35
30
44
35
30
44
35
30
44
35
30
69
Fecal
colif .
0
0
0
0
0
0
0
0
0
70
. -6Q .....
30
21
16
30
21
16
30
21
16
30
21
16
72
Cu
0
0
0
0
0
0
0
0
0
70
60
37
28
23
37
28
23
37"
28
23
37
28
23
75
Pb
0
0
0
0
0
0
0
0
0
70
60
40
31
26
40
31
26
40
31
26
40
31
26
76
Zn
0
0
0
0
0
0
0
0
0
55
45
54
45
40
54
45
40
54
45
40
54
45
40
75
Flow
volume
0
0
0
0
0
0
0
0
0
0
0
32
20
19
	 "32
21
21
32
21
21
32
21
21
32
Ave.
flow
rate
0
0
0
67
67
67
0
0
0
12
0
32
20
19
32
21
21
32
21
21
32
21
21
	 32
Peak
flow
rate
0
25
52
95
95
95
85
85
85
86
74
32
20
19
0
6
40
97
96
96
90
88
88
90
Pond
area
(%)
0
0.1
0.3
1.0
1.8
2.8
0.2
0.3
0.5
0.7
0.3
0
0
0
0
0.02
0.2
0.7
1.4
2.2
0.1
0.25
0.4
6.7 ~
   Codes:
1. Minimum dry detention to meet channel carrying capacity
2. Maximum dry detention to drain in three days
3. Peak shaving dry basin
4. Wet basin for 90% solids control for 1 yr, 24 hr  storm only
5. Wet basin for 75% solids control for 1 yr, 24 hr  storm only
6. Upland infiltration and swales
7. Minimum dry basin plus upland infiltration and swales
8. Maximum dry basin plus upland infiltration and swales
9. Peak shaving dry basin plus upland infiltration and  swales
10. Peak shaving dry basin plus small (75%) wet basin plus  upland  infiltration and swales

-------
         NONPOINT SOURCE POLLUTION OF THE  VENICE  LAGOON:
               PERSPECTIVES OF LONG-TERM ABATEMENT
F. Zingales
Cattedra  di  Chimica,  Facial ta r-  dii.   Ingegneria,   Universita*   di
Pad ova, via Mar sola 9, 35131 Pad ua  (1ta1y).
A. Marani
Dipartimenta  di Spettroscapia,  Elettrochimica  e   Chimica  Fisica,
Universita" di Venezia, D.D. 2137,  30123 Venice?  (Italy).
G. Bendoricchio
Cattedra  di  Chimica,  Facolta"  di   Ingegneria,   Universita"   di
Padova, via Marzolo 9, 35131 Padua  (Italy).
A. Rinaldo
Istituto di  Idraulica "Giovanni  Poleni",  Universitan   di  Padova,
via Loredan 20, 35131 Padua  (Italy).
ABSTRACT

The  paper adresses  the  major  research  topics   and  achievements
related  to  nonpoint source  pollution   (NPSP)   abatement  of   the
Lagoon  of Venice   (Italy) aiming  at. source   evaluation  -for   the
2000  km2 wide  mainland  and  water  quality   forecasting  of   the
receiving Lagoon.

The relevance of the work lies  in the forthcoming need of  setting
control  policies .  for  the   planned  man-made    flood  barrages
regulating the exchange  between the sea and the   Lagoon. In fact,
control strategies should  account for general  health  of the  city
of  Venice, of  the  natural  environment  and   of   the macro--   and
micro-economics.  Hydrological,  chemical, hydraulic   and physical
aspects are worked out in a tentatively synoptic  framework.

Detailed aspects  are considered   for the  control of  urban and
agricultural  NPS,  for  nutrient transports   through  the  Lagoon
system  and  for  possibility  of  controlling  algal  blooms   via
Catastrophe Theory.
KEYWORDS

Non  Point Source  Pollution, Nan  Paint Source   Abatement,
Quality  Modeling,  Tailoring  Runoff Models  to  Recei vi rig
Models, Catastrophe Theory, Algal Blooms, Lagoon  of  Venice.

                            T-II-I-1
Water
Water

-------
 INTRODUCTION
 The  paper   is  centered   on  a   discussion of   the  research
 progress  on  nutrient   Nan  Point.   Source  Pollution  (NPSP)
 abatement   of the  Lagoon of Venice  (Italy).

 The first part   of  the paper summarizes  the physical setting
 af  the system   and  the main  results   of the   hydrological,
 hydraulic and chemical research,  aimed  at selection ofs
 - source distribution and magnitudes
 - coding of overall  mass transports  and balances:!
 ••••• modeling  aspects.

 The second  part  puts forth some  new ideas  on  forecasting of
 the eutrophication   of the  Lagoon,  with special   emphasis on
 modeling of   algal  blooms  via application of   combined  mass
 transport models and  Catastrophe  Theory.
The third part  briefly
long  term  abatement,
texture of the  problem*
                         outlines  some  likely possibilities of
                         given  the  social   and   environmental
THE LAGOON SYSTEM

The Venice Lagoon  (Figure  1)   is   a   complex   tidal  embayment
which  consists  of   interconnected  sloughs,   marshes,   often
meandering  natural   and   man-made   channels,,   Unlike   other
similar  tidal  water  bodies,    t h e L a g o o n   s y stern c a n   b e
represented  by  one  basic  tidal  type,  that of   wel 1--mixed
shallow  waters. The  salinity,   in  the  whole reach,  ranges
•from the salinity  value of sea water near  the  three mouths,
to that  of  nearly fresh water only along a  confined belt  at.
the tributaries outlet.  Nevertheless salinity concentrations
are  marttained at  values  close   to that of  sea  water for-
al most  the  entire   length  span.   At  the   boundary,   minor-
salinity stratification  may occur in  winter  as a   result  of
runoff  during  periods  of  heavy   rainfalls  otherwise the
Lagoon water is isohaline due  to  low fresh water inflows and
strong wind-- and tide-induced  mixing (Bhetti ,  1979),,

The Lagoon receives drainage waters -from a large basin at its
north-western reach,  whereas the   ratio between  drainage area
                                         •four  (2000 Km2  / 50O
                                          its  input  of  fresh
                                          c ample;-; n e t w a r k  o f
                                         agri culture, whereby
and the Lagoon surface adds up  to about
km2). The  system receives  about 40X of
water  from  pumped outlets  draining  a
lowland areas, mainly devoted  to modern
large  and  steadily  growing  urban runoff  pours   into  the
system (Rinaldo,  1982; Cossu et  al.,  1984; Bendoricchio et
al., 1984; Bendoricchio et al.', 1985).

No major tributary flows into the Lagoons in fact this unique
                            T-II-I-2

-------
natural environment has been  preserved through the centuries
by diverting  the estuaries  of two  rivers,  Bile  and Brenta
(the  works  lasted from   1630  and  1850 ca.)  once  flowing
within  the  reach  of  the system.  The  actual   tributaries
(Zero, Dese, Marzenego, Naviglio  Brenta) peak at  about 20-f6O
m3/s without major solid   transport occurences.Input  of fresh
water,   besides  the   mentioned   urban  and   agricultural
components, is  completed  by  other small tributaries  (up  to
27 overall  delivery locations), sewage treatment  plants and
industrial    discharges   surroundings   the  Lagoon   proper
(Figure 1)  (Rinaldo ancJRinaldo, 1983; Cossuet al. , 1984).
                                     Lagoon  is  ruled by  the
                                     s o m e  instances  10-5-15  m
                                      shoals alongside.  The
                                     the forcing   tide  at the
                                      altered with respect to
                                     made  navigation channel
Tidal   propagation throughout  the
hydraulics   of deep   channels   (in
deep)  connected  to   broad shallow
phase delay of  tidal elevations  to
three marine openings (substantially
natural conditions  because of a man
lined  through  the mid Lagoon)  ranges from about  1  hour at
the  city  of Venice  to  2,5  hours  at the  northern  reach
(Goldmann et al., 1975), The tides  in the Lagoon do not show
properties of  standing waves, underlining the  importance of
both  inertial   and  dissipative  effects   (Di   Silvio  and
D'Alpaos, 1972).

Water  circulation   in  the   Venice  Lagoon  system   is  a
combination of tide-driven  and wind-driven currents, whereas
taaroclinic    density-driven    currents   seem     negligible
(Ghetti,1979). Total  water volumes stored within   the system
range from 160 to 510  millions of cubic meters (Mcum), where
the  dominating tidal exchange  (for a gauged  M2 tide)  over-
half period may   amount  to 330 Mcum (through  the Lido  (130
Mcum), Malamocco  (136 Mcum)  and Chioqgia  (64  Mcum) mouths)
(Ministero dei Lavori Pubblici, 1979).

The further lack  of  coastal currents in  the outer Adriatic
sea induces further uncertainty  to environmental studies: it
has been   inferred,  in fact, from  some experimental evidence
that some  15-5-30% of  the water  volumes ebb-released  by the
system are not dispersed and  reenter the mass balance of the
Lagoon for a typical M2 tide  (Cossu et al.,  1984).
NPSP SOURCES
The  task of  predicting  overall  non-point
nutrient for  the actual purposes) into  the
is  demanding   due  to  the  geomorphologic
complexity   of  the   drainage    watershed
agricultural and urban components.,
                                             sources   (mainly
                                             receiving Lagoon
                                               and  hydraul :l c
                                               both  in   its
Urban runoff  tied to about. 1,000,000  inhabitants within the
whole  mainland,  is  noteworthy   and  mainly  non-point   in
nature.   Only  250,000  units are  collected  through  sewage
                            T-II-I-3

-------
 systems  and  sewage  treatment  plants.   The  remnants  load
 directly   the  drainage network  and   -flow  into the  Lagoon»
 Furthermore, the  inhabitants of  the   city  of  Venice  and of
 the  islands   plus seasonal  presences  due   to intense  tourism
 amount  to  100,000 -further  equivalent  inhabitants,  untreated
 at present  stage.  Hence, total  urban IMPS Nutrient   load has
 been estimated in  about 4,000  t/year of   total Nitrogen and
 1,500 t/year of  Phosphorus  (Bendoricchio et al.,1985).

 As far  as other  NPSP  components  are  concerned,  the  nature of
 the  watersheds  and the  range of  the  land uses  have  called
 •for  a  thorough  model   selection  phase, in  view also  of the
 difficulties  of   calibration.  The   possibility  of  use  of
 sera-order        approximation   .     (like        empirical
 runoff-concentration curves)   has  been discarded,  in  view of
 their conceptual  inadequacy (Zingales  et  al.,1984).  In fact,
 significant   differences between  rising  and   falling   stage
 discharge/concentration relations  can'be  fully   explained by
 quite  simple   mass  transfer  relationships    referring  to
 interactions  with channel    bottom  materials and    soil
 adsorption   mechanisms.     In   this   pattern,    pollutant
 load/discharged    total  water   volume relations   seem fit  to
 linear  correlation  scheems,  as  proved  also by  the results of
 smoothing  of  experimental    data   gathered   in a  testing
 watershed  in environment of  the Lagoon, for  which two  annual
 daily  data  collections  had   been carried  out  (Zingales  et
 al.,  1982).

 An   important  result  of such  conceptual   mass  transfer   and
 balance   relations   consists   of   the  validation   of   the
 first-order    approaches   introducing  reference   weighted
 concentrations   in   the  runoff,   or   loading   functions   for
 long-term, or  event-based,  time scale (Haith   and Dougherty,
 1976; Haith,  1982$  D i c k er of f  De 1 w i. c h e an d  Ha i. t h ,  1983).

 In view of   these facts,  screening model have   been  deemed to
 the   point for  the preliminary phase? of   the case   study at
 hand,   the  flow-averaged concentrations   being  realistically
 linear  with  respect  to  total  water quantities   at. large  time
 scales.    Incidentally,  this   fact  holds  true also  for other
 indices,   like BOD   pollution  loadings (Whipple and  Hunter,
 1977) and, even more  notably,  for  solid   transport   (Haith,
 1982).

A  second-order class   of   modeling   approaches,  suitable  to
simulation of runoff and dissolved nutrient   losses,  consists
of  the development  of  unit-mass  response  functions   (UPIK'F)
for the source area. Such   approach is either of  experimental
nature  (based, for  istance,   on parameter  indentification  via
constrained  linear   search)   (Jolankai , 1.983) or   of conceptual
nature  (Bendoricchi o  and   Rinaldo,   1981;   Zingales  et.   al . ,
 1984),   or    of    statistical    derivation     (Marani    and
Bendoricchio,  1984). Such approaches,  which portray  in depth
most  features of   time  development   of pollution   loadings,

                            T-II-I-4

-------
rely  on hydrologies! and  chemical  parameters which  require
calibration  and   hence  prove   suitable  to    large  scale
prediction
predictive
excessi ve
placed  in
frequently
given the
the  rapid
            upon -field-size  calibrations. In  fact, although
            errors  of  screening  or UMRF  models  may  seem
           if compared  to  observed values,  they should  be
            perspectives:  errors in  model  predictions  are
            comparable   to  sampling  and  analysis  errors,
           difficulty to  accurately measure NPSP  and, e.g.,
             transformations   (changes  in   crop  practices;
urbanizations) of the drainage areas.

At this  stage, there  seem no point  in trying  to  implement
distributed parameter modeling  approaches  (among the several
outstanding contributions  in the  field) to the  actual case
study.
                                                   either  of
                                                  Phosphorus,
                                                  area due to
                                                  over a t i me;
                                                  nature  and
                                               (Haith,  1982).
The model  adopted, as preliminary screening  phase of source
areas, is a modification  of loading function approach suited
to  the   available  data.  The  waste   loading,
solid-phase or dissolved  (or  total) Nitrogen and
is defined as a nutrient export  from unit source
runoff event on  i~th day, eventually integrated
span.  Combined  effects  of   crop  type,  soil
hydrology  are accounted  far  in the  scheme?
Computational sequences for dissolved and suspended chemicals
are separated at small  time scales, reflecting the different
mechanisms  of  leaching   from  croplands  (runoff-erosion);
nevertheless  both   phases  can   be  treated   jointly  for
long-term  total  nutrient  losses evaluations.  Examples  of
typical average  concentration ranges in the  runoff, as from
the U.S.  and European literature  and of the wide variability
of  these parameters  have been  discussed in  the literature
(Jolankai,1983)„
Table 1 - Yearly urban  (URB) arid  agricultural  (AGR) loads of
          N and P split into Lagoon subsystems  (after  Bendo~
          ricchio et al.,1985).

NUTRIENT
N

P


SOURCE
URB
AGR
URB
AGR
LOADS ON LAGOON SUB-BASINS
(tc
Northern
2 , 200
1 , 900
850
350
Dins/year )
Central
1 , 300
1 , 750
500
300
Southern
750
1 , 500
300
300
LAGOON LOADS
(tons/year*)
4 , 250
5 , 200
1 , 650
950
The  overall  results  of  source  magnitude  evaluation  are?
                            T-II-I-5

-------
 reported   in   Table  i    -for   comparison   among  the   single
 contributions.   A detailed description   of   the  results   is
 shown  in   Figure 2»  The  discussion  of  modeling  and physical
 implications   are reported elsewhere  (Zingales  et  'al.,1984;
 Bendoricchio  et  al„,  1985).

 As  a sole   remark,  the  discretization of   input load delivery
 locations  is  relevant to the   receiving water model.  In fact,
 the limited  number   of  NFS  yields  a  lowered priority  for
 spatial transport modeling, at least.  in  the  scale needed  for
 evaluation of  residence time  within  the  Lagoon.  On  the other
 hand,   integration   of  runoff  and  receiving  water   models
 needs  a   time  discretisation  which forces   the user into
 considering tidal  effects on'dispersion   of   the  water body
 (Rinaldo    and  Marani    1984).   Therefore,    once  long-term
 nutrient   discharge   sites are   pointed   out  via  screening
 models, the further  step is   related to the need of  shrinking
 the time   scale of   event-based  responses   (which,  for  the
 mainland of the  Lagoon,  range from  one day to dozen of days)
 to, at  least, inertidal  scales.  Such a time   scale  is  deemed
 necessary  for  water  quality  modeling of   the receiving water
 body (Rinaldo  and Marani,  1984). Experimental   results have
 allowed calibration   of  conceptual   UMRF  approach  which fits
 the   level of   detail   required for  prediction  of   lagoonal
 conditions (on   the  order of   the  hour)   (Zingales et al„,
 1984).

 In   this   pattern,   the concept   of   reference  weighted
 concentration  of  pollutants in the  runoff has been  linked  to
 the determination of parameters   of  conceptual  UMRF  (Zingales
 et  al., 1984)   portraying  mass  transfer kinetics  and flow
 rates  generation.  Even   though any  assumptions  concerning
 extrapolations of  calibrated  parameters  seem,   in principle,
 arbitrary,  it  is  certainly passible  to tailor  UMRFs to total
 quantities  predicted via  screening models.
NFS POLLUTANT TRANSPORT

One of the foremost issues  of long term NPSP forecasting and
control  is related  to feasible?  integrations of  runoff and
receiving  water  models. The  need  of  tailoring the  tools
suitable  to reliable  mass balances  for large  contributing
areas  and complex   receiving water  bodies is  manifest: in
fact,  the conceptual  chain  of. events  to  be simulated  is
const!tued  by  the  array  of  sources,  the  mechanisms  of
transport, plus  the net  exchange with  a larger  water body
(in  the actual  case, the  Adriatic Sea)  which is,  somehow
questionably, considered  as a  vessel of  infinite capacity,.
Hence,   NPSF'   model   selection    is   affected   by   the
characteristics and   the quality  standard of  the receiving
water  body.  In  fact, the  output of NPSF1 model  simulations
serves as input  for the water quality model,  whose level of
sophistication  needs be  tailored  to foreseable  predicting

                            T-II-I-6

-------
ability and seal ing-timing of NPSP models.  Quality  and  extent
of  the data base, either  for  the   source  areas  or-  -for   the
hydrodynamic  regimes of  the water  body, also  play  a  major
role in model selection.

Remarkable  efforts have  been  put,   in  the  recent past,   in
establishing the  theoretical grounds  for  simulation  of  the*
hydrodynamic and environmental  regimes   of  the  Venice Lagoon.
Hydraulic approaces (summarized   in  thorough  state-of-the-art
contributions: Ghetti,  1979; Di  Silvia and  Fiorillo, 1980)
range from  zero order lumped mathematical  models  to complex
formulations  of  the  water    quality   model.  For  instance
(Ghetti,  1979;  Ghetti et  al.,  1971;  Rinaldo  and  Marani,
1984), the  simpler models  of  tidal  behavior  of   the  Lagoon
work  satisfactorily  with  respect,  to   prediction  of  water-
levels and  global water balances, while detailed  prediction
of  local  features   (velocities  and  concentrations)  needs
further  specializations   (Ghetti,   1979;  Di    Silvio   and
Fiorillo, 1980;  Leendert.se, 1967; Cheng and  Walters,  1.982).,
Detailed   modeling  of  movements   and  transformations    of
intensive parameters  (Dejak  et al.,  1981; Jorgensen,  1983;
Leendertse and Gritton, 1971)   like  temperature,  salinity or
nutrient  concentrations, induced  by  the  decay  or  reaction
properties of given tracers, coasts   further  uncertainty   to
the prediction ability of the theoretical tools,,

All of  the dispersion  models  considered   are, in  some way,
based  on mass,  momentum and   energy balances,   ranging from
microscopic transport to  macroscopic  intertidal   phenomena.
Albeit perhaps, the dispersion  type  of description  is locally
good for the Lagoon, it need not  be  true at the  scale  of  the
entire Lagoon. In other words,  the traditional  equations   of
turbulent  convection-diffusion   mass  balance?  might not   be
useful  for  describing  fluid  motions  that are on  a  scale
comparable with the  system itself (Himmelblau  and  Bishoff,
1968). In view also of these fact, the   representation  of  net-
long term mass transport in a model  is inseparably  related to
the spatial structure of the  physical   system   (Shanahan  and
Harleman, 1984) and needs be resolved via spatial lumping   of
appropriate level (Rinaldo and  Marani, 1984),,

Furthermore  the  net  transport   of  mass  by  time-varying
currents like  those tide-  or  wind-driven  within   the  Lagoon
of Venice, must consider  properly velocity fields  other than
the  mean  velocity  in   fixed   vertical   sections.  Stokes'
drifts (dependent  on the  time and   the distance  scales of
the  fluctuating  currents) must  also   be  accounted for   in
tidal  period-averaged  net   transports,   according  to   the
Eulerian  framework (Longuet-Higgins,  1969). Quite  recently
direct calculations of  Lagrangian residual circulations have
also been shown  to be feasible -for  large   tidal  water  bodies
similar to  the Lagoon of  Venice (Cheng and  Casul1i,  1982).
Lagrangian  random  walk-type   simulations  have   also been
proposed for far-field dispersion (Chin  and Roberts,  1985).

                            T-II-I-7

-------
 Furthermore:
 a)  -while the mechanism of  leakage of urban and  industrial
 point   sources    is  not.   affected  by   the  hydrodynamic
 conditions  and is  characterized  by  relatively high  local
 concentrations,    NFS    tracers    are   characteristically
 discharge  dependent  as  far  as  either   concentration  or
 total quantities  are concerned.   The  higher   dissolved  or
 solid-phase  loads  are  carried,  in fact,  by the   floods,
 which  at  the same time,   yield  important,  dilution  of  the
 loading.   Low  concentrations,    difficult.   simulations   of
 movements   and  transformations    of  its  components,   time
 variability  (dependent   upon     hydrologic   and  hydraulic
 parameters)  and  large total    quantities  involved   in mass
 balances  are  therefore the main .features of NPSPj
 b)  -  to simulate long-term concentration variations, a lagoon
 water quality model  must  consider the influence  of net mass
 transport.  Long-term   lagoon  water quality  modeling should
 therefore  implement  fast   and   efficient  algorithms,  and
 subtidal   time discretization.  Certain  discontinuities   at.
 the  sea-lagoon  interface  need   be  accounted for,  because
 gener rally  the   outflowing  water   holds   higher  nutrient.
 concentrations  than  the   inflowing   ones,,   It  seems   that
 dispersion models, as  the  one    already  implemented    for
 environmental   studies  on  the    Lagoon  (Di    Silvio    and
 Fiorillo,   1980$    De j ak   et  al.,   1981),   are  tailored  to
 simulations of   current and  tracer   patterns in  a specified
 tidal  or  wind  condition only for  relatively  short simulation
 pattern.

 At  the present  stage,  it   is  therefore clear  the difficult in
 meeting     all     requirements     of     prediction/simulation
 characterized  by   short, time  step and long  time span.

 In  view   of the  fulfilling of    such twofold  objective,  the
 rules for  the  modeler  seem to follow some speculations:
 1)  -  detailed   studies   of    simplified   schemes    of    two--
 dimensional    tidal   propagations   have   been  carried  out
 (Rinaldo   and   Marani,  1984). The most  competitive tool  for
 lumping   hydraulic  fluxes   into any   level  of   aggregation is
 related   to  filtering   of    velocity  arid  elevation  fields
 calculated  by   advanced  finite  element techniques  (F'utti  et.
 al.,  1985);
 2)  ~  multiple-vessel    spectral   models  serve   at  best  the
 overall purposes  of  long-term  forecasting of   dispersion  at.
 spatial scales  like  that,  of  the  Lagoon and   time  scales like
 those  forced  by  NPSP  inputs  and  by   eutrophication  processes
 (Rinaldo.and Marani,  1984);
3)  - the   spatial    discretisation   of   the system  of   cells
 (vessels)   simulating  the  water   quality  variations  over  the
horizontal  landscape can   be   linked  to  the finite  element,
grid of hydrodynamic  simulation  (Putti  et. al „ ,  1985);
4)  ~ theory  of   signal  processing   and  finds  of   numerical
spectral  analysis  induce   important   simplifications in   the


                            T-II-I-8

-------
mathematics and  in  the  computational  loads.

Due to    limited field  data  available  -for  the   Venice  Lagoon
system, to date  hybrid  modeling  techniques (mainly  referring
to  model calibration   via  more  sophisticated  models   runs)
have been implemented.
EUTROPHICATION OF THE LAGOON

Eutrophication  phenomena   imply  the  most,   ostensibile   and
undesirable  facets  of   phytoplanktonic  growth   and  often
represent  an  ultimate  concern  for  the   environmentalist.
Albeit the phenomenon  has  long been known   within  the  Lagoon
(as some toponyms referring  to algae might  suggest)  alarming
frequencies and  spatial spreading of algal   blooms have been
occurring  only   in  recent  years.  Among    the   noteworthy
spinoffs   of   the   modified   trophic   conditions,   fish
death-blows  (either   in   fisheries  or  within    the  tidal
system),   odors  and   often   limitation   to  recreational
activities are perhe^ps to be mentioned.

While the  "cause" in plankton-based  systems  (certainly  due
in most   part to the  increase in nutrient  release  from  the
drainage area)  seems amenable to numerical   simulation, model
selection  and  application  for  eutrophication  effects   is
demanding,  due**the  lack   of continuous  data on  a  complex
system which  is further characterized by  tidal oscillations
of  the parameters  and marked  hydraulic: advection  (Rinaldo
and Marani, 1984).

In  the  actual case  of  Venice,  for  which the   scheme   of
homogeneous   well-mixed    interconnected    vessels    proved
satisfactory, a  feasible approach for simulation   has  turned
out to refer to Catastrophe  Theory  (Zeeman,  1978;  Poston  and
Stewart, 1978).  The foremost advantages of   such an  approach
are:
— step by step refinements  of the qualitative description;
- effective  representation  of  the  processes  and  overall
  reliable forecasting of blooming phenomena.

Catastrophe, within environmental systems,   means a jump from
an  initial equilibrium  state  to another   one,, usually   far
from  the former,  induced  by slow  changes   of the  control
variables.  Analytical  manipulations  of  the discontinuity
allow application of the  theory  if certain  requirements  are
met.   Such   constraints    seemingly   fit   eutrophication
phenomena.  The -recent  1i t er a ture?   h as,   i n  fact,   proved
noteworthy successful  in application  to the field  of algal
population growth (Dubois,  1979;  Duckstein  et al.,  1979;  Van
Nguyen and Wood, 1979; Kempf et al ., , 1984),.

Catastrophe  models may  be  applied to  bimodal  (Voinov   and
Svirezhev,  1984),  piecewise   continuous,   hysteretical   or

                            T-II-I-9

-------
 divergent processes  provided that  control  variables  can be
 clearly  pointed outs   algal  levels  in coatstal   or estuarine
 waters  are amenable to  control   via  Chlorophi11-a  (a-Chl)
 concentration while, if the bloom  only is to be  modeled,  the
 control   parameters   may  be  cast  into  a  growth  factor,
 specialized  by   an  index  of  nutrient  content   (Liebig's
 limiting factor concentration in  most applications   to date),
 plus   an  index  combining . the effects  of   temperature  and
 insulation.  During the bloom,   in fact,  bacterial  reactions,
 net .zooplankton  assumptions  and   other biological   events do
 not affect  substantially the  overall  processes,,

 As far as phytoplanktonic mass growth is concerned, a typical
 bimodal    trend  is   shown  to   exist.   This   behavior   is
 characterized by a state of  survival at  low  concentrations
 (10-5-30 ppb  of a-Chl)  and a. further one at. high concentrations
 < 100-5-300 ppb of a-Chl)  which  is proper to algal blooming.  The
 back-jump from high to low conditions is  patterned on  paths
 other  than  that labelling  the blooms.  Furthermore, evolution
 patterns of  the phenomena show typical  hysteretical trends of
 catastrophes.

 The  only catastrophic   surface portrayed  by two   control  is
 Whitney's  cusp,  i.e.  a  cubic equation  dependent   upon  the
 catastrophic   variable.  In  this  context,  the   catastrophic
 dependent  variable  is  the   phytpp 1 ankton  mass  measured a.s
 concentration    of   a-Chl;   the  splitting   variable     is
 concentration of  limiting  nutrient;   and the normal  variable
 is  a    lumped  parameter  of  light  intensity    and  water
 temperature.

 Albeit   Whitney's  model   does not  seem  very sophisticated
 indeed  if   compared to   other schemes (Swartzman and  Bentley,
 1979),   in  simulating   all  phenomena   connected to   the  algal
 blooms,   nevertheless  its structure allows a  certain  degree
 of accuracy   in  portraying most,  critical  events and  a simple
 graphical layout.

 As  experimental  evidence  supporting  these  ideas,  a  clear
 bimodal   trend  and   an  overall  catastrophic oecurenee,  with
 change of trend  on the   24-th  day,  is  shown by the  results
 of Figures 3 and  4.  A  field  experiments  campaign of  chemical
 and  biological   nature  (Zingales et    al.,1983)    has,   in
 •fact,   been     carried    out   aiming   at   pointing  out   the
 parameters controlling  the  eutraphi cation of  the   Lagoon   of
 Venice.  In   particular,  attention   has  been focusing  on  the
 water    bodies  adjacent    the   industrial   sites  of  larger
 size and   marked   by   limited  tidal   circulation.  Existing
 temperature  gauging  have  also  been   checked. The  campaign,
 performed during  spring  time/due  to its    potential embedding
of  algal blooms,   lasted  up   to   25  days  and considered   as
potential    indices:    pH;    turbidity;    dissolved   oxigen;
salinity;  water  temperature;  and   sylicon,  ortophosphate,
nitrogen  (in various forms) and Chlorophyl1-a  concentrations.


                            T-II-I-10

-------
Algal blooms  have  been   detected  only   at  the  station quite
apart  -from the   industrial   sites,  due   to  the presence  of
Ammonium-ion  levels  toxic  to algal  life.    The -foremost data
(temperature, total  Nitrogen  and   Ohlorophyl 1 --a)  gaged at the
two  stations involved in   blooming  phenomena are reported in
Figure 3. Figure  4  further  illustrates,  -for  the station near
Tessera,  the plot   of sequences   of  equilibrium  paths on  a
cusp surface.

The normal  control  variable  (water  temperature)   undergoes a
five-degree  excursion connected   with the   deeper insulation
and  wind-calm,   Nonethel ess,   the   high   de?gree  of   mixing
experienced  by   lagoonal  canals,   river   outlets   or  the
openings to the sea,  where  tidal convection is strong, yields
a generally  critical determination  of the   parameters unless
•for shallow low-current sites.,

The second control  variable turned  out  to  be total  Nitrogen.,
Concentration changes have, in  fact,  experienced a noteworthy
increase on  the  10-th day  (right  at a  30  mm  rainfall  event.
in  the  area)  released    most  likely   by   croplands  just
fertilised.  A  decrease  in   salinity   and  an  increase  in
turbidity have also  been observed,.

Algae populations,   the -sawdepedent •-'ar3 ab 1 e of  the mode 1 are
formed  by  BOX   diatoms   (the  most  common  species)   plus
chlorophiceae and   cyariophyceae  in  much  smaller proportions.
Certain  correlation  exists   between  maxima  of Chi orophyl 1 --a
and   limiting  n u t r i e n t    c o n <::: e n t r a t i o n s „   I n   f a c t,  a ••• • C h 1
concentrations    of   250   ppb   against    maximum   Nitrogen
concentrations  of   3 ppm   are  shown  At   blooming   stations
(Tessera) whereas otherwhise  (S. Siuliana)   the  maxima reach
120 ppb  (a~-Chl)  and  2 ppm   (Nitrogen),, The  amplitudes  of
Chiorophyl1-a  concentration   oscillations   may  be   further
related   to  tidal   convection   and,    partially,   to   the
analytical method,,
PERSPECTIVES OF LONG-TERM ABATEMENT

Nutrient NPSP pollution abatement of the  Lagoon   of  Venice  i.B
feasible (may be  at low cost, practices)   provided  that  major
sources  be  pointed out  and  controlled..   The  logical   step
forward  is   hence  related  to  a  detailed  knowledge of:
nutrient   release   mechanisms;    nutrient  movement   and
trasformations  within the  drainage network  and within the
L a g o o n  proper;  a u t a d e p u r a t i ve   p rop er t i es  of  the  La g oa n
system. The  tool is primarily a  continuous data acquisition
system   focused  on   the  environmental    aspects  of   the
phenomenon which  is, to date, still  discussed.  Top-priority
is therefore  to be  granted to  creation  and  maintenance  of
such  data  base  and  on the  related  verification  of the
theoretical progresses, namely generation, propagation,  near--

                            T-II-I-11

-------
 and far-field mixing and dispersion model'
                       abatement  has  also   been
                             ;  upon completion  of
                               f i r a t  a n d   s e c o n d
                              total  domestic  NPS
                              Nitrogen  and  1,500
                              overall abatement
Domestic  pollution
 (Bendoricchio et  al.,  1985)
sewage   systems    (and  via
wastewater  treatment), the
out to be 3,400  t/year  of
Phosphorus. In practice, the
about, 20% lower levels. Sea outfall for  industrial
point sources are being planned, with the  further
   consi dered
  the planned
   stages  of
   load turns
   t/year  of
will range at
    and urban
    aid  of a
 realised  by-law  reduction  in  the content  of  polyphosphates
 within  commercial  domestic  detergents from 87. to 5%,,

 Urban runoff   pollution  abatement, albeit  not  yet  considered
 in  the    array of  studies  in progress,  may benefit  from the
 parallel  r e s e a r c. h  p r • o g r e s s  in the field,,
      *

 Agricultural   releases   may  be   controlled  via  a   set  of
 techniques of  increasing diffusion,   mainly referring to'slow
 release  -fertilisers   and   controlled drainage  (Bkaggs  and
 Billiam,  1984).  In   fact,  50%  abatements in  Nitrate losses
 have  been  measured  without losses   of   production   through
 Nitrate  reduction   via s u ta m er g ed   dr a ins,  an  increasing1y
 papular practice  in the  zone,,   These  techniques added to  the
 results  of  the   oncoming   studies of  the  Best  Management
 Practices commissioned   by  the Regions del  Veneto  (the local
 Government) should guarantee consistent  improvements,,

 Aids  to  reduction   of  .eutrophication  phenomena  may  also
 yielded  by   modifications   of  tidal    circulation  within
 suffering shoals via  revitalization of dead water zones,,

 The basic idea underlined   by the actual   work is  to single
 out  the  relative importance of each  elementary  mechanism
 affecting  the  eutrophi cat. i an   of   the    Lagoon    and    to
 establish the theoretical   grounds   for   an  .assessment.  of
 control   policies for   the   foremost   causes. Hence,   the
 mentioned abatement approaches,   efficient in  a cost/benefit
 perspective, have  been outlined.

 The study is  focusing on   the fate of nutrients in   surface
 waters  as a   major contributor   to the  actual  decay,   since
 toxic   substances  of    industrial ,,   a g r i c u 11 u r a 1   a n d  u r b a n
 source    may   just   requires   a .thorough  control   even  in
 commercial routes™

 As a  final  comment,  a noteworthy   environmental   aspect,  as
 far as  .either nutrients  or  toxic substances  are  concerned,
 is  r e1 at ed    to   lagoonal   sed i men ts«  They  ha ve  n ot   been
 mentioned   insofar    because  they   might   require   ad-hoc
 interventions  on low-current    zones   and  mechanical  remotion
 in some instances.
Through the complex set. of planned interventions, among which

                            T-II-I-12

-------
only   a  part already  operational,  the  complexity of  the
problems  can  be  inferred  and  hence  the  fascination  to
researchers, professionals and, may be, politicians.
ACKNOWLEDGEMENTS

The researches are supported by funds granted by the National
Research Council (CNR), the Ministry  of  Public  Instruction
(MPI), and the Regione Veneto.
                            T-II-I-13

-------
 REFERENCES

 Bendoricchio G. ,  and A. Rinaldo, Un  rnodello   matematico   del
 dilavamento di  sostanze chimiche da terreni  agricoli,,  Atti  &
 Memorie del 1 'Accademia Patavina di SS.LL.AA.,  XCIII,  137-5-154,'
 1982.

 Bendoricchio B. ,  C.  Camis,  and  S«  Alessandrini,  Evoluzione
 temporale del 1 *inquinamento da sorgenti  diffuse  di   origine
 agricola.  Inquinamertto, 25 (12) , 61*64., 1984,,

 Bendoricchio S.,  L.  Agostinettq, and  S. Alessandrini, • Nutri-
 ent!  di  origine  diffusa provenienti- dal  bacirto  scolante   in
 Laguna  di  Venezia.  Acqua Aria, 2, 177-5-184, .1.985.

 Cheng R.T., and  V»  Casulli, On Lagrangian  Residual   Currents
 With  Applications   to  South  San Francisco Bay, California.
 Water Resour.  Res.,  1.8, 1652*1662, 198.2.

 Cheng R«  T., and  R.  A.  Walters, Modeling of Estuarine  Hydro	
 dynamics and Field  Data Requirements.  In:  Finite  Elements  in
 Fluids,  Vol.  4,  Edited  by  R. H.  "Gallagher,  J. T.   Helen,  C.
 Taylor,  and  D.  C,,   Zienkiewicz, 'pp.' 891108, .J. Wiley ?< Sons,
 New York,  N.Y.,  1982.

 Chin DuA.,  and J.W,  Roberts,  Model of  Dispersion  in   Coastal
 Waters.  J.  Hydr.  Eng.,  Ill, '12128, lc?85.

 Cdssu R.,  E. de   Fraja  Frangipane, R.   Donazzolo,  and  A.  A.
 Orio, Elementi per  il  bilancio di nutrient.! nella   Laguna   di
 Venezia.  Ingegneria  Ambient-ale, 13,. 80191, 1984.

 Dejak C. ,  1. Mazzei  Lalatta,  and M.  Mol in, Dispersions di in--
 quinanti  nella  Laguna  di   Venezia  (modello  bidimensionale
 avvetti vo~-di f f usi vo) .  Proceedings of  the Seminar "Process!  di
 mescol amenta in arnbienti  lagu'nari e lor a model  1 &z i one" , Roma,
 (1), 25155,  1981.

 Dickerhoff  Delwiche  L,  L.,and D.  A,,  Haith, Loading  Functions
 for Predicting Nutrient Losses From Complex Watersheds.. Water
 Resour.  Res.,  19, 9511959,  1983.

 Di Silvio,  G. , arid L.  D'Alpaos, II comportamento del la Laguna
 di Venezia  esaminato con  il metodo propagatorio  unidimensio-
 nale. Istituto Veneto  di  Scienze,  Lettere  ed  Art! -  Rapporti
 e Studi,  
-------
Regulation  o-f  Rivers.  State  of  th
ing, 7, Copenhagen,  7511-758,  1979.
the Art in  Ecological Model
Duckstein L. , J.  Cast! ,  and  J,  Kempf ,  Modeling  Phytoplankton
Dynamics Using Catastrophe Theory.   Water   Resour.   Res.,  15,
1189*1194,  1979.

Ghetti A.,  L. D'Alpaos,  and   R.  Dazzi,  La   regolazione  delle
bocche della Laguna  di Venesia  per  1 'attenuazione delle acque
alte indagata con metodo statico.  Atti  e Memorie del 1 * Accade-
mia Patavina di SS.LL.AA., XVII,  11-5-41,  1971.

Ghetti A.,  Etudes concernant  les  problemes hydrodynamiques de
la Lagune de Venise. Proceedings   o-f   the   XVII   Congress   o-f
IAHR, Cagliari  (Italy),  (6),  56-5-94,  1979.

Goldmann A., R. Rabagliati,  and  P.  Sguaszero,  Characteristics
of the Tidal Wave in the Lagoon  of   Venice.   IBM  Tech.  Rep.
CRV009/5 13-3539,  47, 52  pp. ,  1975.

Haith D. A., Development and   Testing   of  Watershed  Loading
Functions   for  Nonpoint Sources.  Proceedings of  Thirteenth
Annual Pittsburgh  Conference   on   Modeling   and  Simulation.
Part 4, 1463-5-1468, 1982.

Haith D. A., and  J.  V. Dougherty,   Non-Point Source Pollution
from Agriculture  Runoff.  Journal  of  the  Environmental  Engi-
neering Division, ASCE,  102,  1055-5-1069,  1976.

Himmelblau  D. M. , and K.  B.  Biscohoff,  Process  Analysis  and
Simulation: Deterministic Systems.  J.  Wiley  & Sons, New York,
N.Y. , 348 pp. , 1968.

Jolankai G. , Modelling of Nonpoint  Source  Pollution.  In: S.E.
Jorgensen (Editor),  Application  of   Ecological   Modelling   in
Environmental Management, Part  A.,  Elsevier   Scientific  Pub-
lishing Co., Amsterdam  (N) ,  283-5-385,  1983.

Jorgensen S. E. ,  Application  of  Ecological  Modelling  in Envi-
ronmental Management, Part A. Elsevier  Scientific  Publishing
Co., Amsterdam (M) ,  735  pp.,  1983.

Kempf J., L. Duckstein,  and  J.  Casti ,  Relaxation Oscillations
and Other   Non-Michael i an Behavior  in   Slow-Fast Pyt op lank ton
Grow Model.  Ecological Modeling,  23,  67*90,  1984,.
Leendertse J. J., Aspects of a Computational  Model  for  Long-
Period  Water-Wawe  Propagation. Rep.  RM-5294-PR,  Rand Corp.,
Santa Monica, California, 1967.

Leendertse J. J., and E. C. Gritton,  A Water-Quality  Simula-
tion  Model  for  Well-Mixed  Estuaries  and  Coastal  Sea,  2,
Computation  Procedures,  Rep. R-708-NYC,  Rand   Corp.,  Santa
Monica, California, 1971.

                             T-II-I-15

-------
 Longuet-Higgins  M. S. , On the Transport  of  Mass  by  Time-
 Varying Ocean Currents, Deep-Sea Res. ,  16, 431-5-447, 1969.

 Marani  A. ,  and G.  Bendoricchio, Models of Statistical Distri-
 butions for NPSP Concentrations. Proceedings of International
 Conference  on  "Agriculture and  Environment  1984". Venezia,
 11-5-15 June, 1984.

 Ministero dei Lavori  Pubblici, Le correnti di marea nella La~-
 guna  di  Venezia.  Istituto  di   Idraulica  del 1'University  di.
 Padova,  Padova,  1979,

 Poston,  T.  and I.  Stewart, Catastrophe Theory  and Its Appli-
 cations, Pitman,  London,  408 pp.,. 1978.

 Putti M. , A.  Rinaldo,  and A,  Marani, Long Term Water  Quality
 of  Tidal Water Bodies   Via  Mixed Spettral-FE Methods. Forth	
 coming,  1985.

 Rinaldo  A., Manpoint  Source Pollution from Agricultural  Run-
 off:  a  Large  Scale  Application  of  Mathematical  Modeling.
 Proceedings  of   Thirteenth Annual   Pittsburgh  Conference on
 Modeling and Simulation.  Part  4, 1570*1.574,  1.982.

 Rinaldo  A., and  A.  Rinaldo, Determinations  dei  singoli reca--
 piti  delle  quantita di fosforo .e asoto  che provengono annual---
 mente in Laguna  di  Venezia dal proprio  bacino scolante  rela-
 tivamente ai  territori agricoli.  Atti del Convegno  di  Studi
 "Lagunai, fiumi,  lidi;  cinque  secol i di  gestione delle  acque
 nelle Venezie",  Ministero dei  Lavori Pubblici, Magistrate al-
 le  acque, Venesia,  10-5-12  giugno 1983.

 Rinaldo  A.,  and  A.  Marani,  Runoff .and  Receiving Water Models
 for NPS  Discharges  into the Venice   Lagoon.   Proceedings  of
 International  Conference  on   "Agriculture   and  Environment
 1984". Venezia,  llflS  June, 1984.

 Skaggs R. W. ,  arid J, W. Gilliam,  Modeling Subsurface Drainage
 and  Water   Management Systems  to  Alleviate Potential Water
 Quality Problems. Proceedings  of  International  Conference  on
 "Agriculture   and   Environment  1984".  Venezia, 11-5-15  June,
 1984.

 Shanahan P.,  and  R. F.   Harleman, Transport   in  Lake  Water
 Quality Modeling. J. Environ.  Eng.,  ASCE, 110,  42-f57,  1984.

Swartzman G.L., and R.  Bentley,   A   Review and  Comparison of
Phytoplankton  Simulation   Models.   ISEM   Journal,  1,  30*82,
 1979.

Van Nguyen  V., and  E.  Wood, On the Morphology of Summer Algae
Dynamics  in  Non-stratified   Lakes.  Ecological  Modeling,  65
 117-5-131, 1979.

                            T-II-I-16

-------
Voinov  A.A.,   and   Yu.fl.   Svirezhev,   A   Minimal  Model  of
Eutrophication  in Freshwater  Ecosystems.,  Ecological Modeling,
23, 2777-292,  1984.
   i»
Zeeman E. C.,   Catastrophe  Theory:  Selected Paper  1972-1977,
Addison-Wesley  Publishing,  Reading,   Massachusetts,  U.S.A.,
700 pp.,  1978.

Zingales  F.,  S. Alessandrini,  G.   Bendoricchio, C.  Comis, A.
Marani, F.  Pianetti,  A.  Rinaldo,  and  C.  Sartori-Baratto, A
Model of  Nanpoint Source Pollution  in  the  Agricultural  Run-
off. Proceedings  of   Thirteenth  Annual  Pittsburgh Conference
on Modeling and Simulation,, Part  4,  1575-5-1580, 1982.

Zingales  F. et  Al.,  Indagini  Idratermodinamiche e  Biologiche
per la valutazione dei  riflessi  ambientali del funzionamento
a piena potensa  della Centrale.  Technical  Report--E. N.E. L.  ,
Venesia,  1983.

Zingales  F., A. Marani, G. Bendoricchio,   and  A. Rinaldo, A
Conceptual Model of  Unit-Mass  Response Function -for  Nanpoint
Source  Pollutant  Runoff.  Ecological   Modeling,  26, 285v311,
1984.
                            T-II-I-17

-------
I
»-*
1
»-*
CD
                       Figure 2 - Mainland  draining into the Lagoon  o-f  Venice: spatial
                                  distribution o-f nutritent loads.

-------
      ucow «r ima 
UONH V HMO 
-------
rsj
o
      Figure 1
                         THE VENICE LAGOON SYSTEM
            In block,  largo navigation  channels,  tha  Chioggia,
            Malamocco, Lido mouthsj  cantor  tha  city of  Venicoi
            towards tho mainland boundary olougho and marches.

-------