Chesapeake Bay Program
   Watershed Model Application to
   Calculate Bay Nutrient Loadings

   Final Findings and Recommendations
    Prepared by the Modeling Subcommittee
       of the Chesapeake Bay Program
               July 1994
              CBP/TRS 157/96
              EPA 903-R-94-042

-------
                                                      EPA
                                                July 1994
       CHESAPEAKE BAY PROGRAM

   WATERSHED MODEL APPLICATION TO
   CALCULATE BAY NUTRIENT LOADINGS

 Final Findings and Recommendations
      Anthony S. Donigian, Jr.
          Brian R.  Bicknell
        Avinash S.  Patwardhan
       AQUA TERRA Consultants
Mountain View, California 94043-1004
           Lewis C.  Linker
    Chesapeake Bay Program Office
      Annapolis, Maryland 21403
           Chao-Hsi Chang
           Robert'Reynolds
    Computer Sciences  Corporation
      Annapolis,  Maryland 21403 >
           Project Officer

          Mr.  Robert Carsel
          Assessment Branch
  Environmental Research Laboratory
        Athens, Georgia 30613
    CHESAPEAKE BAY PROGRAM OFFICE
U.S. ENVIRONMENTAL PROTECTION AGENCY
             REGION III
        ANNAPOLIS,, MD  21403

-------
                                   FOREWORD
The  EPA  Chesapeake Bay Program  (CBP)  has sponsored  a  series of  projects to
develop and implement a 'Watershed Model'  of the Chesapeake  Bay drainage area
that could be used effectively  to estimate nutrient loadings to the Bay and to
evaluate the impacts of agricultural Best Management Practices (BMPs).  A phased
approach was planned to develop the Watershed Model and implement its calibration
and refinement. The U.S. EPA Hydrological Simulation Program -  FORTRAN (HSPF) was
selected as the framework for the CBP Watershed Model.

Phase  I  of  the   effort   was   designed   to  improve  the  nonpolnt  loading
representation, refine the  data  input  to  the Model,  and perform a preliminary
calibration to available water quality data for the 1984-85 period..  This period
was  used as the basis for the 40% nutrient loading reduction goal defined by the
1987 Chesapeake Bay Agreement.

Phase  II  was  designed to focus  on a better representation of the  effects of
agricultural BMPs,  including application  of  the detailed Agrichemical Modules
(AGCHEM) of HSPF to allow a more deterministic, process-oriented approach to BMP
analysis and evaluation.  AGCHEM provides a nutrient balance approach to modeling
cropland  areas so  that  sensitivity to  nutrient inputs can be represented.  In
addition, HSPF code enhancements were performed to allow specific consideration
of sediment-nutrient and bed interactions  within the stream channel.  The Phase
II Watershed Model was then applied for the 1984-87 period to the Bay drainage
area to include sediment erosion, -sediment transport,  and associated nutrients,
in addition to the current modeled water quality constituents, to provide input
to the Chesapeake  Bay 3-D Water  Quality Model.

This report presents the results of the combined Phase I/Phase II effort.  Model
calibration results are presented for hydrology, nonpoint loadings, -and instream
water quality for numerous sites  throughout the Bay drainage.  In addition, model
estimates  of Fall  Line loads  for various  nutrient  forms are  compared  with
regression  estimates used  in the calibration of the Bay Model:   The results
indicate that the Watershed Model is a valid representation of nutrient loadings
to  the Bay, and provides  a framework  for assessing the impacts  of  land use
changes and alternative nutrient and agricultural management practices.

Since  the completion of  Phase  II,  the Watershed  Model has  been  used to: 1)
determine  the  distribution of  point  and  nonpoint  source  loads  and  the
controllable and uncontrollable portions of the loads; 2)determine the quantity
of loads  reduced under  different management  actions;  3) determine the nutrient
loads  to  the  Bay under different Clean Air Act scenarios;  and 4) quantify the
loads  under future  (year  2000 conditions).    These  loads were used  as input
conditions  for the Chesapeake Bay Water Quality Model. Following this Foreword
is a brief technical summary of the  status and use of the Watershed Model for
management  and planning in  the  Chesapeake  Bay  Region.  Work  is continuing  in a
current Phase III effort to  extend the simulation period,  investigate and improve
the  plant nutrient and reservoir simulations,  and refine ..the calibration for
recent water  quality conditions.

-------
     THE  CHESAPEAKE BAY WATERSHED MODEL

                L.C. Linker*, G.E. Stigall*. C.H. Chang**, and A.S. Donigian***

                 * U.S. Environmental Protection Agency, Chesapeake Bay Program
        ,                       410 Severn Ave. Annapolis, MD 21403
                                " Computer Sciences Corporation
                               410 Severn Ave. Annapolis, MD 21403
                                   •" Aqua Terra Consultants
                           2672 Bayshore Pkwy. Mountain View, CA 94043
ABSTRACT
   The Chesapeake Bay Watershed Model is designed to
simulate nutrient loads delivered to the estuary under different
management scenarios. The nutrient loads are differentiated
into anthropogenic loads amenable to management, and
nonanthropogenic loads  which are considered to be
uncontrollable. The model divides the Bay basin into sixty-
three model segments with an average area of 260300
hectares. Hydrology, sediment and nonpoint source loads
  Deluding atmospheric deposition) are simulated on nine
 .jid uses. In the river reaches, sediment, nonpoint source
loads, point source inputs, and water supply diversions are
simulated on an hourly time-step. Paniculate and dissolved
nutrients are transported sequentially through each segment,
and ultimately to the tidal Chesapeake Bay.. The input data
used  for development of the Watershed Model water is
reviewed. Model scenarios of existing loads, Chesapeake
Bay Agreement loads, year 2000 loads, and loads under limit
of technology controls are described. The model will be used
to assist the Chesapeake Bay Program in the restoration of
Chesapeake Bay water quality through  nutrient load
reductions.

INTRODUCTION

   "Decision makers who assess which environmental
control strategies to implement are particularly wary of two
possibilities:
   1. Reducing  waste inputs  to a  water body and
   observing little or no improvement in water quality
   (the environmental engineering equivalent ofbuilding
   half a bridge) or,
   2. Mandating control actions that are subsequently
   shown to be [excessive]...(the environmental
   engineering equivalent of building a bridge  to
   nowhere)."
                            Robert V. Thomann
   The Chesapeake watershed drains the waters (and nutrient
loads) of seven mid-Atlantic States  from New York to
Virginia. Six major basins of the watershed are the
Susquehanna, Potomac, Rappahannock/York, James, West
Shore Chesapeake, and East Shore Chesapeake (Figure 1).
Taken together, the three largest basins of the Susquehanna,
Potomac, and James, comprise 80% of the total basin area.
Land use in the watershed is predominately forest (60%),
cropland (20%), pasture (9%), and urban land (1.0%) (Figure
2). By the year 2000, urban land in the watershed will
increase to 13% due to the conversion of forest (2% decrease)
and cropland (1% decrease).
     Western Shore

     Eastern Short

     Rappahannock/York Rivers

     Susquehanna River

     Potomac River

     James River
Figure 1. Chesapeake Bay watershed: major basins.
                                              ii

-------
THE CHESAPEAKE BAY WATERSHED MODEL.
 ,_ o.

2
 o

ฃ 0.

•c  .
 o

 &0.



I
   0
          a
         6-




         4-




         2

               (n-8)
                    (n-6)
(n-5)
                                    3        4

                                    Segment
                                  PO4    --A--ORGP
   Figure 5. Watershed average phosphorus transport
          
-------
                                           _THE CHESAPEAKE BAY WATERSHED MODEL
                                      Susquehanna Basin
       segment 6a->,


       segment 6b->
>segment 5a->segment 4a->
                                           —>segment &->segment 2—>segment 1—>discharge to tidal Bay
     segment 6c->segment 5b->segment 4b->

                                       /
                  segment 5c->segment 4c->

Segments 6a, 6b, 5a, and 4a represent the East Branch Susquehanna. Segments 6c, Sb, and 4b represent the West Branch Susquehanna.
Segments 5c and 4c represent the Juniata Branch and Segments 3,2, and 1 represent the Susquehanna main stem.
   Not surprisingly, phosphate P is rapidly attenuated
through incorporation into riverine  sediment by
phytoplankton uptake or adsorption to particulates.  This
removal mechanism is dramatically reversed during periods
 * high flow and scour of sediments.  During a high flow
 -riod in November, 1985 (peak discharge 33 times the
average discharge of the 1991 record) in the Upper Potomac
segments, the phosphorus discharged was 1.8 times that
amount input by annual edge-of-streara loads.

   As can be seen from Figure 5, attenuation of nutrient
loads is dependent on the nutrient form and the time or length
of transport Otherfactors of attenuation arenutrient limitation
and the presence of reservoirs in the system which in some
cases greatly increase the basin residence time. Examination
of the nutrient concentrations reveals these systems to be
consistently under phosphorus limitation. The slope of the
phosphate P attenuation is -0.125. The slope of the organic
P attenuation is -0.030.

   Average transport losses of ammonia, nitrate, and organic
nitrogen are diagramed in Figure 6. Organic nitrogen, as in
the case of organic phosphorus, includes biomass, labile and
refractory organic and paniculate inorganic nitrogen.
Ammonia is removed quickly (slope = -0.105) in riverine
transport by incorporation into organic  nitrogen through
biomass uptake and adsorption to particulates. Ammonia is
 "so removed through nitrification. Nitrate (slope = -.014)
 -moval mechanisms include uptake by biomass and
denitrification. The transport time in the basins is correlated
with increasing organic nitrogen primarily through conversion
of inorganic nitrogen.
                                  MANAGEMENT SCENARIOS

                                     Several key scenarios werecompleted in order to develop
                                  die basic inventory of loads under specific management
                                  conditions.

                                  BASE CASE SCENARIO
                                     This scenario is the base case yean 1985 loads to the
                                  Chesapeake Bay.

                                  BAY AGREEMENT SCENARIO
                                     This scenario represents the nutrient loads to be reduced
                                  by the year 2000 under the Bay Agreement.  The reduction
                                  is a 40% reduction of controllable nutrient loads. Controllable
                                  loads were defined as the base case loads minus the loads
                                  delivered to the Bay under the conditions of an all forested
                                  watershed. In other words, controllable loads are defined as
                                  everything over and above  the total  phosphorus or total
                                  nitrogen loads that would have come from an entirely forested
                                  watershed. In this definition, point source loads are considered
                                  entirely controllable.

                                  LIMIT OF TECHNOLOGY SCENARIO
                                     The limit of technology scenario is defined as all cropland
                                  in conservation tillage; the Conservation Reserve Program
                                  fully implemented; nutrient management,  animal waste
                                  controls, and pasture stabilization systems  implemented
                                  where needed; a 20% reduction in urban  loads; and point
                                  source effluent controlled to a level  of 0.075 mg/1 total
                                  phosphorus and 3.0 mg/1 total nitrogen. An important aspect
                                  of the limit of technology scenario is that it determined the
                                  feasibility of the 40% nutrient reduction.
                                                viii

-------
THE CHESAPEAKE BAY WATERSHED MODEL.
YEAR 2000 SCENARIO
    This scenario represents the growth in population and
the projected changes in land use by the year 2000. The
controls in place in the base case scenario are applied to the
year 2000 point source flows and land use representing the
loading conditions without the nutrient reductions of the Bay
Agreement

1991 PROGRESS SCENARIO
    This  scenario applies the actual reductions made in
nitrogen and phosphorus by the year 1991.
              Figures 7 and 8 show the base case total phosphorus am
          total nitrogen loads relative to the Bay Agreement, the"'
          of technology, the year 2000, and the 199 1 Progress
     Chesapeake TP loads by scenario.
       4
              In all basins, the limit of technology phosphorus loads
          are less than the Bay Agreement loads. For nitrogen loads
          only the point source dominated basins of the Potomac
          James, and the West Shore have limit of technology loads
          appreciably less than Bay Agreement loads. For nonpoint
          source dominated basins of the Susquehanna, Rappahannock
                    York, and the East Shore, the limit of technology
                    loads are essentially  equivalent  to the Bay
                    Agreement loads.

          Smquduuma  PCXOOMC   June*   lUplTork Weft Shore EUL Shore

      • Base S1991 Progress BBay Agreement BLinit of Tech. QYeซr2000
                        The Year 2000 scenario has increased loads
                     of phosphorus and nitrogen for all .basins. Th.
                     point source dominated basins of the Potomac,
                     James, and West Shore, which experience the
                     greatest urbanization in the Year 2000 scenario
                     have the greatest increase in nutrient loads.

                        The 1991 Progress Scenario indicates the
                     phosphorus reductions are on track to reach th
                     year 2000 goal The nitrogen reductions posted
                     by the 1991 progress scenario are m'
                     Overall progress in nitrogen reduction^
                     need to increase by a factor of 3.7 in order to
                     reach the year 2000 goal
Figure 7. Chesapeake total phosphorus
loads by scenario.
       Figure 8. Chesapeake total
       nitrogen loads by scenario.
Chesapeake TN loads by scenario.
  70
      Swquetanm Potomac    Jtmei   Rip/York Wett Shore E** Shore

 • Base ฃ31991 Progress ฎBซy Agreement 0 Limit of Tech. OYeir 2000
                                              ix
                                                                                               CSC.MOU:

-------
                                             .THE CHESAPEAKE BAY WATERSHED MODEL
    Another aspect of the Progress Scenario is that the Bay
Agreement reductions become nutrient load caps after the
year 2000.  The Watershed Model is envisioned as the tool
which will track the changes in watershed loads to ensure the
caps are not exceeded in any basin.           .
CONCLUSION

    The Watershed Model is a successful water quality
simulation which quantifies the nonpoint source and point
source nutrient loads from all basin sources. The model was
essential in establishing a consistent method of accounting
for the nutrient loads, among all sources, and among the
basin jurisdictions of the Bay Program. The model is used to
examine the level of controls achievable from different
management practices. The model has examined management
practices, which when combined into strategies of nutrient
control, will meet the Chesapeake Bay Program water quality
' objective of a 40% reduction in controllable nutrients in the
Chesapeake watershed and do so in a cost-effective and
equitable manner.
ACKNOWLEDGMENTS

    The authors wish to acknowledge the state, regional, and
federal members of the Chesapeake Bay Program Modeling
Subcommittee for their essential guidance and direction
throughout  the development of the Chesapeake Bay
Watershed Model, and in particular Dr. Robert Thomann for
his expertise and experience gained through 23  years of
water quality modeling work on the Chesapeake and its
tributaries.                                .
REFERENCES
Donigian. Jr.. A.S., and H:H. Davis (1978). Users Manual for
   Agricultural Runoff Management (ARM) Model. U.S. EPA
   Environmental Research Laboratory. Athens. GA.

Donigian, Jr.. A.S., B.R. Bicknell, L.C. Linker, J. Hannawald,
   C.H. Chang, and R. Reynolds (1990). Watershed Model
   Application to Calculate Bay Nutrient Loads: Phase I
   Findings and Recommendations. U.S. EPA Chesapeake
   Bay Program; Annapolis, Mb.

Donigian, Jr., A.S., B.R. Bicknell, A.S. Patwardhan, L.C.
   Linker, C.H. Chang, and R. Reynolds (1991).  Watershed
   Model Application to Calculate Bay, Nutrient Loads:
   Phase II Findings and and Recommendations, U.S. EPA
   Chesapeake Bay Program, DRAFT. Annapolis, MD.

Conservation Technology Information Center (1985) 198S Crop
   Tillage Data Base. West Lafayette, IN.

Hartigan. J.P. (1983).  Chesapeake Bay Basin Model Final
   Report prepared by the Northern Virginia Planning District
   Commission for the U.S. EPA Chesapeake Bay Program.
   Annapolis, MD.

Bicknell, B.R., J.C. Imhoff, J.L. Kittle. Jr.. A.S. Donigian
   Jr., and R.C. Johanson. 1993. Hydrological Simulation
   Program - Fortran: User's Manual for Release 10. U.S.
   EPA, ERL, Athens. GA. EPA/60O-R-93/174.

National Atmospheric Deposition Program (1982); (1983);
   (1984); (1985); (1986); (1997). (IR7)/National Trends
   Network. NADP/NTN Coordination Office, Natural
   Resource Ecology Laboratory, Colorado State University.
   Fort Collins. CO.

U.S. Bureau of the Census (1984). 1982 Census of Agriculture.
   Volume 1, Geographic Area Series. U.S. Department of
   Commerce. Washington, DC.

U.S. Department of Agriculture (1984). Soil Interpretations
   Records (SCS-SOI-5 data file) and the National Resources
   Inventory (NRI), Soil Conservation Service, Washington,
   DC.

U.S. Geological Service (1974)  Geographic Information
   Retrieval and Analysis System (GIRAS) 1:125,000 land
   use/landcover (Anderson Level II classification.
   Washington, D.C.
 CSC.MOU.7/93

-------
                                   ABSTRACT
Since  1985  the EPA  Chesapeake Bay  Program  (CBP) has  sponsored  a  series  of
projects to  convert  the proprietary HSP-NPS model  to the public  domain HSPF
model, make it operational on the CBP VAX computer system, and implement a number
of model  refinements so that  the resulting 'Watershed Model'  could be used
effectively  to estimate nutrient loadings to the Bay and to evaluate the impacts
of agricultural Best" Management  Practices (BMPs).   A comprehensive workplan
developed in September  1987  proposed a two-phased approach to address Watershed
Model deficiencies and  produce  an improved and re-calibrated model.  Phase I was
designed to improve the nonpoint loading representation, refine/re-evaluate the
data input to the Model, and perform a preliminary re-calibration  to available
water quality data for the 1984-85 period.  This period was   used  as the basis
for the 401 nutrient loading reduction goal defined by the 1987  Chesapeake Bay
Agreement.

Phase  II was designed  to focus  on  a better representation of the  effects  of
agricultural BMPs, including application  of the  detailed  Agrichemical Modules
(AGCHEM) of HSPF to allow a more deterministic, process-oriented approach to BMP
analysis and evaluation.  AGCHEM provides a nutrient balance approach to modeling
cropland areas so  that sensitivity  to nutrient  inputs can be  represented.   In
addition, HSPF code enhancements were performed to allow specific consideration
of sediment-nutrient and bed interactions  within  the stream channel.  The Phase
II Watershed Model was  then applied to the Bay drainage area to include sediment
erosion, sediment transport, and associated nutrients, in addition to the current
modeled water quality constituents,  to provide  input to the  Chesapeake Bay 3-D
Water Quality Model.

This report presents  the results  of the combined Phase  I/Phase II effort.  Model
calibration results are presented for hydrology, nonpoint loadings,  and instream
water quality for numerous sites throughout the Bay drainage.  In addition, model
estimates  of Fall Line loads  for  various nutrient  forms  are  compared with
regression  estimates used  in  the calibration of the  Bay  Model.   The  results
indicate that the Watershed Model is a valid representation of nutrient loadings
to  the Bay, and provides a framework for assessing  the,  impacts  of  land  use
changes  and  alternative  nutrient  and  agricultural  management  practices.
Refinements and recommendations are provided to improve model simulations and its
use as a tool to  assist in the  1991  Re-evaluation of the 402 nutrient reduction
goal of the Bay Agreement.

This report is submitted in fulfillment of Work Assignment No. 7(B) by AQUA TERRA
Consultants under EPA Contract No. 68-CO-0019 with the EPA Athens Environmental
Research Laboratory.   Earlier  work  for Phase I of the CBP Watershed Model re-
calibration was performed under  EPA Contact No.  68-03-3513.
                                      xi

-------
                                   CONTENTS
                                                                          Page

Foreword	       i
The Chesapeake Bay Watershed Model	. ..	      il
Abstract.	<	     xi
Tables	     xv
Figures.	,	  xviii
Acknowledgments	>.....	   xxii
List of Acronyms	,	;	   xxiy

1.0  INTRODUCTION	 . . .		.......... %	...       1
     1.1  Background	—' — —	 . .      1
     1.2  Overview of the Phase  I/Phase II Efforts ....................       2
     1.3  Phase I Preliminary Findings and Recommendations 	.....       9
          1.3.1  Hydrology Conclusions and Recommendations ......:....... '      9
          1.3.2  Phase I Nonpoint Loading Conclusions and
                 Recommendations	      10
          1.3.3  Phase I Water Quality Simulation Conclusions, and
                 Recommendations	.....	      11
     1.4  Phase II Findings and  Recommendations	      12
          1.4.1  Phase II Hydrology Conclusions and Recommendations....      12
          1.4.2  Phase II Nonpoint  Loading Conclusions and
                 Recommendations.		      13
          1.4.3  Phase II Water  Quality Simulation Conclusions  and
                 Recommendations.	      16
     1. 5  General Watershed Model Recommendations	      20
     1.6  Format of this Report	      21

2.0  KETEOROLOGIC AND PRECIPITATION DATA  BASE DEVELOPMENT PROCEDURES....      22
     2.1  Introduction	      22
     2.2  Regional Meteorologic  Data	      22
          2.2.1  Cloud Cover and Dewpoint Temperature	      25
          2.2.2  Air Temperature and  Wind Speed	      25
          2.2.3  Solar Radiation ..'...•:.,.		...      25
          2.2.4  Potential Evaporation ^			      27
     2.3  Hourly Precipitation  Data '.'.-	'ป	      28
          2.3.1  Development of Observed  Rainfall Data Base  .......	      28
          2.3.2  Basin Segmentation and Development of
                 Weighting Factors	      28
          2.3.3  Generation of  Weighted Hourly Rainfall Records .......      29
                                     •Xii

-------
3.0  LAND USE DATA AND WASTE LOAD INPUTS	 .      31
     3.1  Land Use Data	      31
          3.1.1  Land Use for 1978 			      32
          3.1.2  Land Use for 1985	      33
          3.1.3  Cropland Tillage 	.	.	      33
          3.1.4  Water Acres	33
          3.1.5  Manure Acres	      34
          3.1.6  Urban Land Subcategories  	      35
          3.1.7  Land Use by County ....		      37
          3.1.8  Land Use by Model Segment 		      37
          3.1.9  Land Surface Cover and Erodlbillty Parameters -
                 COVER, SLSUR,  KRER	.....	      37
          3.1.10 Crop Distributions for Conventional (CNT) and
               :  Conservation (CST) Tillage.	      38
          3.1.11 Comparison of Forest Land.	      39
     3.2  Point Sources	;	'._..	      39
    - 3.3  Surface Water Diversions				      44
     3.4  Atmospheric Sources.		...	      45

4.0  Nonpoint Loading Simulation Modules and Simulation Results........      47
     4.1  Overview of Cropland and Non-Cropland Simulation....	      47
          4.1.1  Hydrologic, Sediment Loading, DO, and Water   '
                 Temperature Simulation	      54
     4.2  Simulation of Non-Cropland Areas	      60
          4.2.1  Forest, Pasture, Urban Area Procedures..	   .   60
          4.2.2  Animal Waste Procedures	      63
     4.3  Simulation., of Cropland Areas	.		      64
          4.3.1  AGCHEM Overview	      64
          4.3.2  Soil Nutrient Processes Modeled by AGCHEM..	      68
          4.3.3  Composite Crop Representation............ป............      72
          4;3.4  Development of Model Segment Nutrient Application
                 Rates...;	,...	      79
          4.3.5  AGCHEM Application Procedures for Model Segments	     103
          4.3.6  AGCHEM Simulation Results....		     121
          4.3.7  AGCHEM Conclusions and Recommendations	     135
     4.4  Comparison of Expected and Simulated Nonpoint Loading Rates..     136

5.0  RIVER TRANSPORT AND TRANSFORMATION SUBMODEL.	     151
     5.1  Overview of the Instrearn Model		.'	     151
     5.2  Model Structure and Processes	,	     154
          5.2.1  Hydraulics			     154
          5.2.2  Water Temperature	     155
          5.2.3  Sediment Transport..	     157
          5.2.4  Water Quality Processes.	     161
     5.3  Idealized Channel Network	     174

6.0  HYDROLOGY CALIBRATION RESULTS....			     178
     6.1  Overview and Steps in the Hydrologic Calibration Process.....     178
     6.2  Snow Simulation				     179
          6.2.1  Identification of Areas Needing Snow Simulation	     179
          6.2.2  Snow Calibration and Simulation Results	     182
     6.3  Adjustment of Hydrologlc Parameters for Land Use	     187

                                     xiii

-------
     6.4  Phase I Hydrologlc Re-Calibration for 1974-78	    189
     6.5  Phase .II Hydrology Calibration for 1984-87	    193

7.0  WATER QUALITY CALIBRATION RESULTS	•'.    206
     7.1  Overview and Steps in the Water Quality Calibration Process..    206
     7.2  Sediment Loading and Instream Transport Calibration... v	    207
          7.2.1  Potency Factor Adjustments	    215
     7.3  Water Quality Calibration Results.......;			    215
          7.3.1  Instream Water Quality Concentrations	    216
          7.3.2  Fall Line Load Simulation and Regression Comparison...    220
     7.4  Analysis of Segment and Basin Loads	    221

8.0  REFERENCES			    274
APPENDICES .(Available  from CBPO, Annapolis, MD)

A.   Hydrology Calibration Results
B.   Water Quality Calibration Results
C.   Nonpoint, Loading  Simulation Results
       and Nutrient Application Rates
     Addendum to Appendix C
D.   Meteorologic and  Precipitation Data Development
E.   Land Use and Selected Parameter Values
F.   Point Sources, Diversions, and Atmospheric Deposition

-------
                                    TABLES
                                                                          Page

1.1   Subbasin,  Segment Name, and Segment Numbers of the Chesapeake
      Bay Watershed Model. .	       6

2.1   Correspondence "Between 1974-78 Regional Data and
      Published Data			      23
2.2   Stations Used to Develop 1984-87 Regional Meteorologic Data .	      23

3.1   Estimated Quantities  of Voided Manure from Livestock and
      Poultry.  Normalized  to 1,000 Pounds Body Weight..	      34
3.2   Crop Distributions for Conventional and Conservation Tillage
      for Each Watershed Model Segment	      40
3.3   Municipal Nitrogen Default Percentages of Ammonia, Nitrate
      and Organic Nitrogen, and the Default Concentrations of Total
      Nitrogen.   Defaults Based on Levels of Secondary Treatment	      42
3.4   Municipal Phosphorus  Default Percentages of Phosphate and
      Organic Phosphorus, and the Default Concentrations of Total
      Phosphorus.  Defaults Based on Levels of Secondary Treatment.....      42
3.5   Basin-Wide Average Annual Atmospheric Deposition Loads	      45

4.1   Land Use Categories in the Watershed Model.		      48
4.2   Water Quality Constituents Simulated in the Watershed Model	      49
4.3   HSPF Modules used to  Simulate Nonpoint Loadings from Each
      Land Use in Phase II			      53
4.4   Crop Cover Values for all .Regions and Practices	      75
4.5   Fertilizer Application Rates Supplied by the NRTF		...      82
4.6   Manure Application Rates Supplied by the NRTF.	      84
4.7   Percentage of Cropland Receiving Fertilizer and/or Manure
      as Supplied by the NRTF.	      86
4.8   Estimated Percentages of Multi-State Segments in Various States..      90
4.9   Annual Nitrogen Atmospheric Deposition Loads	      93
4.10  Model Procedures for  Timing and Placement of Fertilizer/
      Manure Applications and Atmospheric Deposition.	     100
4.11  Example Calculations  of Final Nutrient Application Rates	     101
4.12  Example Calculations  of Timing and Placement of Final
      Nutrient Application  Rates			     102
4.13  Final Calculated Model Segment Nutrient Application Rates for  • •   .
      Conventional Tillage	     104
4.14  Final Calculated Model Segment Nutrient Application Rates for
      Conventional Tillage	     106
4.15  Final Calculated Model Segment Nutrient Application Rates
      for Hay		     108

                                     .xv

-------
 4.16  Typical Nitrogen Balance  for Major Crops When Nitrogen is
       Applied at Agronomic  Rates.	     112
 4.17  Typical Phosphorus  Balance  for Major Crops When  Phosphorus
       is Applied at Agronomic Rates	     113
 4.18  Crop Nutrient Uptake  Targets	     114
 4.19  Nutrient Transformation and Sorption Parameters  used in the
      .AGCHEM Simulations	     119
 4.20  Plant Nutrient Uptake Rates used in AGCHEM Simulations	     120
 4.21  AGCHEM Results for  Juniata  Segment 100  (HT Composite Crop)	     122
 4.22  AGCHEM Results for  Juniata  Segment 100  (LT Composite Crop)	     123
 4.23  AGCHEM Results for  Juniata  Segment 100  (Hay)	     124
 4.24  AGCHEM Simulated Nutrient Runoff Losses for  Selected Model
       Segments	     128
 4.25  Summary of Observed Concentrations from Selected Field
       Monitoring Stuflies		....     131
 4.26  Summary of Rappahannock  (Segment 230) Simulated  Concentrations
       by AGCHEM	..		     132
 4.27  Summary of Patuxent (Segment 340) Simulated  Concentrations
       by AGCHEM.		     133
 .4.28  Summary of Monocacy (Segment 210) Simulated  Concentrations
       by AGCHEM...	...:...	'.		     134
 4.29  Expected Ranges of  Land Use Loadings Developed in Phase I..-7	     138
 4.30  Mean and Range of Loading Rates for Each Land Use from
       Various Sources for the Chesapeake Bay  Region.....	     140
 4.31  Nutrient Expert Coefficients from Beaulae et al.  (1980)......	     141
 4.32  Watershed Model Segment 60  Annual Loading Rates	     142
 4.33  Watershed Model Segment 180 Annual Loading Rates.		     143
 4.34  Watershed Model Segment 280 Annual Loading Rates		     144
 4.35  Summary of Average  Annual Loading Rates for  Each Land Use in
       Selected Model Segments;				....     145
 4.36  Overall Simulated Mean and  Range of Loading  Rates from
       Selected Segments Listed  in Table 4.35		     149
                    t                   .                   • •                •
 5.1   Water Quality Constituents  and Processes Simulated in
       the Watershed Model				     162
 5.2   Instream Water Quality Parameters	     175
 5.3.  Channel Reach Information.	   -  177

 6.1   Elevation, Temperature and  Snow Depths  for Watershed
       Model Land Segments	     180
 6.2   Land Use Cover Values Used  in the Watershed  Model for
       Non-Cropland Areas	;	     188
 6.3   Watershed Model Hydrology and Water Quality  Calibration
       Stations	,....•	     190
 6.4   Summary of Chesapeake Bay Watershed Hydrologic Calibration
^      vs. Simulated Flows	,     191
 6.5   Summary of Watershed Model  Hydrologic Calibration Seasonal
       R2 Values, 1984-1987..			     205

 7.1   Total Annual Suspended Sediment Loads at Selected Locations	     209
 7.2   Susquehanna River Fall Line Loads:  Comparison of Regression
       and Model Estimates	     222

-------
7.3   Potomac River Fall Line Loads:  Comparison of Regression
      and Fall Line Estimates	     223
7.4   James River Fall Line Loads:  Comparison of Regression
      and Model Estimates	 .	     224
7.5   Patuxent River Fall Line Loads:  Comparison of Regression
      and Model Estimates	     225
                                     xvii

-------
                                      FIGURES
                                                                            Page

  1.1   Chesapeake Bay watershed and major above fall line
        subbasins	      3
  1.2   Above fall line watershed model segments.	      4
  1.3   Below fall line watershed model segments	      5

  2.1   Meteorologic regions and principal stations in the
        Chesapeake Bay Watershed Model	     24
  2.2   Diurnal variation of air temperature	     26
  2.3   Diurnal variation of wind speed 		     26
  2.4   AFL Daily and Hourly Precipitation Stations	...-...•	   28,1
  2.5   BFL Daily and Hourly Precipitation Stations.-..		   28.2

  4.1   PERLND structure chart.		•			.....     50
  4.2   I^PLND structure chart • impervious land-segment application           .
        module	;..........	     51
  4.3   Hydrologic processes simulated by the PWATER module of HSPF	     55
  4.4   Flow diagram for SEDMNT section of PERLND application module.....     57
  4.5   Flow diagram for PQUAL section of PERLND application module......     61
  4.6   Nutrient storages and transport modeled by the AGCHEM Module.....     65
  4.7   Soil profile representation by the AGCHEM Model..		     67
  4.8   Nutrient transformations simulated by. the AGCHEM Module	     69
  4.9   Crop COVER region boundaries	     74
  4.10  Composite crop COVER values for Segment  100 - conventional
        tillage......			i..	     77
  4.11  Composite crop COVER values for Segment  100 - conservation
        tillage		.		 T......     78
  4.12  Flow chart for development of model input nutrient '        '     '
        application rates			     80
  4.13  Composition of nutrient inputs to AGCHEM Model		.......     95
'  4.14  Plant nutrient inputs - Juniata Segment  100	    125
  4.15  Comparison of nutrient losses • Juniata  100.	    126
  4.16  Nutrient runoff losses - Juniata 100	.	....................    127

  5.1   RCHRES structure chart	    152
  5.2   Structure chart of  the RQUAL section of  the RCHRES module.	    153
  5.3   Flow diagram for HTRCH section of the RCHRES module	    156
  5.4   Flow diagram of inorganic sediment fractions in  the SEDTRN
        section of the RCHRES module.	    158
  5.5   Simulated bed shear stress in Reach 290.	    160
  5.6   Instream nitrogen dynamics in the Watershed Model.	    164
  5.7   Instream phosphorus dynamics in the Watershed Model......	    165

            ;                           xviii

-------
5.8   Flow diagram of dissolved oxygen in the OXRX section of the
      RCHRES module		,				     166
5.9   Flow diagram of BOD in the OXRX section of the RCHRES module	     167
5.10  AFL model segments and river channel reaches	     176

6.1   Watershed model segments for snow simulation	     181
6.2   Simulated and observed snow depth for  Model Segment 20, 1984-87..     184
6.3   Simulated and observed snow depth fpr  Model Segment 50. 1984-87..     185
6.4   Simulated and observed snow depth for  Model Segment 160, 1984-87.     186
6.5   Simulated and observed flow frequency  for Harrisburg, PA
      Segment 80, 1984-87.	     195
6.6   Simulated and observed daily flow for  Harrisburg, PA,
      Segment 80, 1984-87	     196
6.7   Simulated and observed flow frequency  for Conowingo
      Reservoir, MD,  ^Segment 140, 1984-87....	     197
6.8   Simulated and observed daily flow for  Conowingo Reservoir
      MD, Segment 140, 1984-87			     198
6.9   Simulated and observed flow frequency  for Potomac River,
      Washington, DC, Segment 220, 1984-87			     199
6.10  Simulated and observed daily flow for  Potomac River,
      Washington, DC, Segment 220, 1984-87.		...     200
6.11 . Simulated and- observed flow frequency  for James River,
      Cartersville; VA, Segment 280,  1984-87.....	     201
6.12  Simulated and observed daily flow for  James River,
      Cartersville, VA, Segment 280,  1984-87.	     202

7.1   Suspended sediment concentration • Susquehanna River at
      Conowingo	*	......ป.........;..............;.     210
7.2   Suspended sediment concentration - Potomac River at Chain
      Bridge..		.	.			..     211
7.3   Suspended sediment concentration - Juniata River at Newport	    212
7.4   Suspended sediment concentration - Patuxent River at Bowie	     213
7.5   Suspended sediment load - Juniata River at Newport....;	     214
7.6   Simulated and observed water temperature (ฐC),  Susquehanna   •
      River at Harrisburg,  PA, 1984-87.	     226
7.7   Simulated and observed dissolved oxygen (mg/1), Susquehanna
      River at Harrisburg,  PA, 1984-87.			..r.....    .227
7.8   Simulated and observed suspended sediment (mg/1), Susquehanna
      River at Harrisburg,  PA, 1984-87.	     228
7.9   Simulated and observed nitrate (mg/1), Susquehanna River
      at Harrisburg,  PA, 1984-87....				......     229
7.10  Simulated and observed ammonia (mg/1),, Susquehanna River
      at Harrisburg,  PA, 1984-87.	     230
7.11  Simulated and observed particulate nitrogen (mg/1), Susquehanna
      River at Harrisburg,  PA, 1984-87		     231
7.12  Simulated and observed total nitrogen (mg/1),  Susquehanna  •     ,
      River at Harrisburg,  PA, 1984-87.	,.	     232
7.13  Simulated and observed phosphate (mg/1), Susquehanna River
      at Harrisburg,  PA, 1984-87		.,		     233
7.14  Simulated and observed particulate phosphorus (mg/1).
      Susquehanna River at Harrisburg, PA, 1984-87	     234
                                     xix

-------
7.15  Simulated and observed total phosphorus (mg/1),  Susquehanna
      River at Harrisburg, PA, 1984-87	     235
7.16  Simulated and observed total organic carbon (mg/1),  Susquehanna
      River at Harrisburg, PA, 1984-87	     236
7.17  Simulated and observed chlorophyll A (ug/1) ,  Susquehanna
      River at Harrisburg, PA, 1984-87		.	     237
7.18  Simulated and observed water temperature (ฐC), Susquehanna
      River at Conowingo, MD, 1984-87	     238
7.19  Simulated and observed dissolved oxygen (mg/1),  Susquehanna
      River at Conowingo, MD, 1984-87.................		     239
7.20  Simulated and observed nitrate (mg/1), Susquehanna River
      at Conowingo, MD, 1984-87		     240
7.21  Simulated and observed ammonia (mg/1), Susquehanna River
      at Conowingo, MD, 1984-87	     241
7.22  Simulated and flbserved particulate nitrogen (mg/1),
      Susquehanna River at Conowingo, MD, 1984-87	     242
7.23  Simulated and observed total nitrogen (mg/1),  Susquehanna
      River at Conowingo, MD, 1984-87	     243
7.24  Simulated and observed phosphate (mg/1), Susquehanna River
      at Conowingo, MD, 1984-87	...;				.	     244
7.25  Simulated and observed particulate phosphorus  (mg/1),
      Susquehanna River at Conowingo, MD, 1984-87	     245
7.26  Simulated and observed total phosphorus (mg/1),  Susquehanna
      River at Conowingo, MD, 1984-87...	.7		     246
7.27  Simulated and observed total organic carbon (mg/1),              ."',''..
      Susquehanna River at Conowingo, MD, 1984-87....	• •••	...     247
7.28  Simulated and observed chlorophyll A (ug/1),  Susquehanna
      River at Conowingo, MD, 1984-87..'•..'		-..     248
7.29  Simulated and observed water temperature (ฐC), Potomac River
      at Washington, DC,  1984-87	;	     249
7.30  Simulated and observed dissolved oxygen (mg/1),  Potomac  River
      at Washington, DC,  1984-87	4.	'-,.     250
7.31  Simulated and observed nitrate (mg/1), Potomac River at
      Washington, DC, 1984-87..			,	     251
7.32  Simulated and observed ammonia (mg/1), Potomac River at
      Washington, DC,. 1984-87..........		     252
7.33  Simulated and observed particulate nitrogen (mg/1),  Potomac
      River at Washington, DC, 1984-87.	 . ....-.-		     253
7.34  Simulated and observed total nitrogen (mg/1),  Potomac River
      at Washington, DC,  1984-87.			.	     254
7.35  Simulated and observed phosphate (mg/1), Potomac River at
      Washington, DC, 1984-87		, 1.'	     255
7.36  Simulated and observed particulate phosphorus  (mg/1),
      Potomac River at Washington, DC, 1984-87	     256
7.37  Simulated and observed total phosphorus (mg/1),  Potomac
      River at Washington, DC, 1984-87	     257
7.38  Simulated and observed total organic carbon (mg/1),  Potomac
      River at Washington, DC, 1984-87.	     258
7.39  Simulated and observed chlorophyll A (ug/1),  Potomac River
      at Washington, DC,  1984-87...	     259
7.40  Simulated and observed water temperature (ฐC), James River
      at Cartersville, VA, 1984-87......	     260
                                      OCX

-------
7.41  Simulated and observed dissolved oxygen (mg/1).  James River
      ,at Cartersviile,  VA,  1984-87.....	     261
7.42  Simulated and observed suspended sediment (mg/1),  James River
      at Cartersviile,  VA,  1984-87	     262
7.43  Simulated and observed nitrate (mg/1),  James River at
      Cartersviile, VA, 1984-87	'.	     263
7.44  Simulated and observed ammonia (mg/1),  James River at
      Cartersviile, VA, 1984-87	     264
7.45  Simulated and observed particulate nitrogen (mg/1), James
      River at Cartersviile, VA,  1984-87		....	     265
7.46  Simulated and observed total nitrogen (mg/1), James River
      at Cartersviile,  VA,  1984-87	     266
7.47  Simulated and observed phosphate (mg/1),  James River at
      Cartersviile, VA, 1984-87	     267
7.48  Simulated and observed particulate phosphorus (mg/1), James
      River at Cartersviile, VA,  1984-87	     268
7.49  Simulated and observed total phosphorus (mg/1),  James River
      at Cartersviile,  VA,  1984-87..		     269
7.50  Simulated and observed total organic carbon (mg/1), James
      River at Cartersviile, VA,  1984*87..	     270
7.51  Simulated and observed chlorophyll A (ug/1), James River
      at Cartersviile,  VA,  1984-87......		.		     271
7.52  Simulated and observed-total nitrogen (mg/1), Patuxent River
      at Bowie. MD, 1984-87		.....	     272
7.53  Simulated and observed total phosphorus (mg/1),  Patuxent River
      at Bowie, MD, 1984-87,	:....,.....			     273
                                      xxi

-------
                                ACKNOWLEDGMENTS
A large number  of individuals and organizations contributed to the successful
completion.of this Watershed Model re-calibration and re-application effort.  The
financial support was provided by  the Chesapeake Bay Program with administrative
and  contractual  assistance   through  the  EPA  Athens Environmental  Research
Laboratory  (ERL)  in Athens, GA.  Mr. Robert Carsel was the Project Officer for
the Athens  ERL, and Mr. Charles App was the Watershed Modeling Coordinator for
the first phase of the study;  Mr. Lewis Linker subsequently replaced Mr. App and
assumed both technical and administrative responsibilities on the project.  All
these   individuals   contributed  to  the  project  by   their  assistance  in
administrative   and   contractual  matters;   this  assistance  is  gratefully
acknowledged.                                           .

The project was guided by a number of technical committees that met  on a regular
basis to provide  thorough technical review of all aspects of the study.  These
committees included the Modeling Subcommittee, the Nonpoint Source Committee, the
Nutrient Reduction Task Force (NRTF), and the  Science  and Technical Advisory
Committee.   It would be impossible  to  accurately  acknowledge  the  specific
contributions   of all  the   individuals  on  these  committees,   but  their
participation,  review, and input was a major factor in the conduct of this work.
Mr.  James  Collier,   as   Chair of  the  Modeling  Subcommittee, played a  key
.coordination and  oversight role among the various agencies and committees.

The  NRTF  deserves  special  mention because  its various  state members  were
instrumental in providing the  information needed to develop appropriate input for
the AGCHEM modeling  of the cropland areas, including the nutrient application
rates, crop cover values,  tillage practices,  etc.   Mr. Jin Cox, as  Chair of the
NRTF,  was  a  key force  in   this  effort,  in addition  to  the various  state
representatives including Dr. Robert Summers  and Mr. John McCoy of MDE, Mr. Mike
Flagg of VA DSWC,  Dr.  Elizabeth Gasman of ICPRB,  Ms. Veronica Kasi of PA BUSC and
Mr.  Fred Samadani of  MDA.   the AGCHEM model application  would not have been
possible without  the valuable assistance from  these individuals. The assistance
of Joel  Myers,  SCS is acknowledged for.his  contributions to the  Pennsylvania
agricultural data.   In addition,  Mr. James Hannawald, SCS. played  a major role
in  the  land  use  evaluation,  soil  parameter assessment,   and  crop  cover
representation  and provided invaluable knowledge of agricultural conditions and
practices  within the  Bay drainage.

The CBP Watershed Model Evaluation Group was convened at various times throughout
the  study to provide  expert  technical review  of  the modeling procedures and
results.  Mr. Thomas Bamwell of Athens ERL,  Dr. Alan Lumb of the USGS, and Dr.
Robert  Thomann of Manhattan  College  provided helpful and  insightful review
comments as members of this group.  Their participation is sincerely  appreciated.
                                      xxii
                                      41

-------
In addition to the authors, various individuals at AQUA TERRA Consultants and the
CBPO provided key technical  and staff support roles  throughout the study.  At
AQUA TERRA,  Mr.  John Imhoff was  a key participant in the  development  of the
sediment-nutrient interaction  capabilities  and  assisted in  the  water quality
parameter  review and calibration; Mr.  Jack Kittle assisted  in  the Conowingo
Reservoir operation program, the rainfall generation program, and the ANNIE/WDM
application;  and  Ms.   Dorothy  Inahara and   Ms.   Mary  Shandonay  provided
administrative, word processing, and graphics support.

At CBPO, in addition to  the authors, several individuals provided guidance and
assistance.  Mr.  Ed Stigall, Technical Programs  Director, and Mr. Lynn Shuyler,
Nonpoint Source Coordinator, assisted in bridging the  gap between the technical
aspects  of the  model  and  its  application  to  water  quality problems  in the
Chesapeake Bay.  Their assistance is gratefully acknowledged.  Many  of the staff
of the Computer Scierices  Corporation were essential for the successful execution
of a number of  tasks  throughout  this project, with special appreciation to Scott
McDiarmid and Lacy Williams for satisfying the variety of requisites, large and
small, made of them by HSPF or  it's modelers.  Ms.  Diane Allegre, the Modeling
Intern, is thanked for her many contributions to the appendices and the report.
                                     xxiii

-------
                      TABLE OF ACRONYMS AND ABBREVIATIONS
AFFIX



AFL

AGCHEH


ARC-INC

ARM

ATEMP



BFL

BMP

BOD

BtU

BWSC

C

CBP

CBPO

COVER



CNT

CSC
An HSPF parameter .representing the incorporation of detached
sediment into the soil matrix where it would be unavailable for
transport.                  . .
  X                                                   ..
Above Fall Line.

Agricultural/Chemical section of HSPF.  An HSPF module which
simulates agricultural nutrient loads in detail in PERLND.

A type of CIS software.

Agricultural Runoff Model.               .    .  '  .

An HSPF module used to adjust air temperature for adiabatic
differences between the measuring station and the average. •
altitude of the model segment;                 .

Below Fall Line.

Best Management Practice.                       •    •.  .

Biochemical Oxygen Demand.

British Thermal Units heat metric.

Bureau of Water and Soil Conservation (PA).

Centigrade temperature.

Chesapeake Bay Program.

Chesapeake Bay Program Office.            ,

An HSPF parameter used to specify the degree to which the soils
are protected for rain splash soil detachment by vegetation,
organic debris, or vegetative litter.

Conventional Tillage land use.

Computer Science Corporation.
                                     xxiv

-------
CTIC

CST

3-D Water
Quality
Model


DETS


DSUC

DO

EPA

ERL

F

FHA

FTABLE


CIS

HSPF


HSP-NPS


HTRCH


HYDDAY

HYDHR

HYDR

HYDSDY

IMPLND   .


ICPRB

IQUAL
Conservation Technology Information Center.

Conservation Tillage land use.

The Chesapeake Bay Three-Dimensional Time-Variable Model,  a
coupled hydrodynamic and water .quality model of the Chesapeake
Bay.  The Watershed Model simulates the nutrient loads used as
input in the 3-D Water Quality Model.

An HSPF parameter used to represent detached sediment storage
in SEDMNT.

Department of Soil and Water Conservation (VA).

Dissolved Oxygen concentration.

Environmental Protection Agency.

Environmental Research Laboratory (US EPA).

Fahrenheit temperature.

Federal Highway Administration.

An HSPF input file table which relates water volume to depth,
surface area, and flow rate for a reach or reservoir.

Geographic Information System.     '

Hydrologic Simulation Program - Fortran.  Model code used as,
a basis for the Watershed Model.

Hydrologic Simulation Program - Nonpoint Source - A proprietary
model code.                                        .

An HSPF module used to simulate heat exchange and water
temperature in RCHRES.

App D.

App D.

An HSPF module used to simulate hydraulic behavior in RCHRES.

App D.

Impervious Land simulation section of HSPF.  An HSPF module
used to simulate the pollutant loads from impervious land uses.

Interstate Commission of the Potomac River Basin.

An HSPF module which simulates water quality in IMPLND.
                                      XXV

-------
IWTGAS


K

KPL


KRER

KSER


LU/LC

MAGI

MDA

MDE

MGD

MS1AY

N

NADP

NASA

NHAPP

NITR

NOAA

NPDES

NPS


NRI

NRTF

NUTRX


NVPDC
An HSPF module which simulates water temperature and dissolved
gas concentration in IMPLND.

A USLE parameter used to represent soil credibility.

An HSPF parameter representing plant uptake  rates of nutrients
in AGCHEM.                                          .

An HSPF parameter representing the erodibility of soil.

An HSPF parameter used to adjust the  amount of soil fines
detached by raindrop impact.

Land Use/Land Cover.
   ^
Maryland Geographic Information System.

Maryland Department of Agriculture.

Maryland Department of Environment

Millions of Gallons per Day.

An HSPF module which simulates solute transport in AGCHEM.

Nitrogen nutrient.                   ,

National Atmospheric Deposition Program.

National Aeronautics and Space Administration.

National High-Altitude Photography Program.

An HSPF module which simulates nitrogen in AGCHEM.

National Oceanographic and Atmospheric  Administration.

National Pollution Discharge Elimination System.

Nonpoint Source, a term describing pollutants from diffuse
sources.                    .  .    •

National Resource Inventory;                   .

Nutrient Reduction Task Force.

An HSPF module which simulates primary inorganic nitrogen and
phosphorus reactions in EQUAL.

Northern Virginia Planning District Commission.

-------
NVSI


ORGN

ORGP

OXRX
An HSPF parameter used to represent net external additions or
removals of sediment caused by atmospheric deposition or wind.

Organic Nitrogen.

Organic Phosphorus.

An HSPF module used to simulate primary DO and BOD reactions
in RQUAL.

Phosphorus nutrient.
PANEVAP


PERLND


PLANK

PRECIP


PQUAL


PHOS


PSTEMP


PWATER


PWTGAS


RADCLC


RCHRES



RQUAL


SCS
A FORTRAN subroutine used to develop pan evaporation rate time
series for the Watershed Model.

Pervious Land simulation section of HSPF.  An HSPF module used
to simulate the pollutant loads  from pervious land uses.

An HSPF module which simulates plankton in RQUAL.

A FORTRAN subroutine used to develop an hourly time series of
rainfall data.

Vater quality section of HSPF. An HSPF  module which simulates
water quality in PERLND.

Detailed phosphorus simulation section of HSPF. An HSPF module
which simulates phosphorus in AGCHEM.

Soil temperature section of HSPF.  An HSPF module Which
simulates soil temperature in PERLND.

Water budget section of HSPF.  An HSPF module used to simulate
the water budget in PERLND.

Water temperature section of HSPF.   An  HSPF module which
simulates water temperature and gas concentrations in PERLND.

A FORTRAN subroutine, used to develop solar radiation time
series for the Watershed Model.                   •

Reach and Reservoir section of HSPF.  An HSPF module used to
simulate transport and  transformation in the river reaches and
reservoirs.

An HSPF.module used to  simulate water quality constituents in
RCHRES.                                  .

Soil Conservation Service.              •
                                     xxvii

-------
SEDMNT
An-HSPF module which simulates sediment to the edge of stream
in PERLND.
SEDTRN

SLSUR


SLSED


SNOU


SOLIDS


TAUCD


TAUCS


TO

TOC

TP

UAN

USDA

USGS

USLE

VIRGIS

WDM
An HSPF module which simulates sediment transport in RCHRES.

An HSPF parameter that represents the average land slope of a
model segment.

HSPF parameter that represents external lateral input of
sediment from an upland segment.

Snow simulation section of HSPF.   An HSPF module used to
simulate snow accumulation and melt.
  ^     '
An HSPF module which simulates the accumulation and removal of
solids in IMPLND.

An HSPF parameter which specifies the critical shear stress for
sediment deposition in SEDTRN.

An HSPF parameter which specifies the critical shear stress for
sediment scour in SEDTRN.                                .

Total Nitrogen.                 -o  '

Total Organic Carbon.

Total Phosphorus.

Urea, Ammonium, Nitrate liquid fertilizer.  •

United States Department of Agriculture.

United States Geological Survey.

Universal Soil Loss Equation model.

Virginia Geographic Information System.        .

Watershed Data Management  system.   System data files by HSPF.
                                    xxviii

-------
                                  SECTION 1.0

                                 INTRODUCTION
1.1  BACKGROUND

Since  1985  Che  EPA  Chesapeake Bay  Program (CBP) has  sponsored a  series  of
projects to convert the proprietary HSP-NPS model to the public domain HSPF model
(Jchanson et al, 1984),  make it operational on the CBP VAX computer system, and
implement a number of model refinements so that the resulting  'Watershed Model'
could  be  used effectively   to estimate  nutrient loadings  to the Bay  and  to
evaluate  the  impacts of  agricultural Best Management   Practices (BMPs).   A
comprehensive workplan developed in September 1987 proposed a two-phased approach
to address Watershed Model  deficiencies.   Phase I was designed to improve the
nonpoint loading representation, refine/re-evaluate the data input to the Model,
and perform a preliminary re-calibration to available water quality data for the
1984-85 period.  This period was used as the basis for the 40Z nutrient loading
reduction goal defined by the  Chesapeake  Bay Agreement.

Phase  II  was  designed to  focus on a better representation of  the  effects  of
agricultural  BMPs through  application of  the  detailed Agrichemical  Modules
(AGCHEM) of HSPF which allow a more deterministic, process-oriented approach to.
BMP analysis and evaluation.  In addition, HSPF'code enhancements were performed
to allow specif ic consideration of sediment-nutrient and bed interactions within
the  stream  channel  so  that  runoff,  and subsequent:'delivery to. the Bay,  of
dissolved and sorted nutrient  forms can be more  accurately modeled.  The Phase
II Watershed Model was then applied to the Bay drainage area to include sediment
erosion, sediment transport, and associated nutrients, in addition to the current
modeled water quality constituents, to  provide  input to the Bay model.

This report presents' the  results of the combined Phase  I/Phase II effort.  The
Phase II effort, tn addition to the agricultural and sediment-nutrient  emphasis
noted   above,  also  addressed  the  preliminary conclusions  and recommendations
resulting from the Phase I work as discussed in the Phase I  report (Donigian et
al,  1990).   In  summary,  the .Phase I effort Identified the need for a longer
simulation period (i.e.  more observed data),  more in-depth analysis of animal
waste  and point  source  contributions,  .and simulation  problems  with selected
nutrient forms and phytoplankton as issues that needed to be addressed in Phase
II.  Thus, the results presented herein  represent the integrated  efforts of both
phases of the Watershed Model refinement and re-calibration work.  However, the
detailed results of Phase I are not repeated in this  report since they have been
superseded by  Phase  II and  were presented in  the Phase  I report.

The Watershed Model represents the entire  drainage area to the  Chesapeake Bay as
a  series of  land segments  each  with  relatively uniform  climatic  and soils

-------
conditions.  Within each model segment a variety of different land use categories
are  each  modeled with  its own parameter values,  and each  land  use provides
surface and subsurface nonpoint  loadings  to the  stream draining  that  model
segment.  Each model segment also corresponds  to a single channel reach which are
linked sequentially with other channel reaches in other segments to represent the
major and minor river systems that comprise the Bay drainage.  Figure 1.1 shows
the primary subbasins for the areas above the Fall Line,  which include the three
major subbasins  -  Susquehanna,  Potomac, James -  and  a number of smaller river
systems.  Figures  1.2  and 1.3  show the model segments for the Above Fall Line
(AFL) and Below Fall Line  (BFL) areas, respectively;  Figure 1.2 also shows the
primary model calibration stations.   A complete  list of the model segments is
included in Table  1.1.  This brief overview  is  provided to establish a common
basis  for  the  following   sections on the Phase  I/II efforts and associated
conclusions and recommendations. The model segmentation is discussed further in
Sections 4.0 and 5.0V

It should be noted that the  entire Phase  I/II project has been a joint effort
between AQUA  TERRA Consultants and the Chesapeake Bay  Program Office (CBPO),
including the support of Computer Sciences Corporation,  in Annapolis, Maryland.
AQUA  TERRA was responsible  for overall guidance  and review of  the modeling
effort, with the  lead on specific  tasks  such as  meteorologic  data  update
procedures, snowmelt/hydrology calibration, nonpoint parameter evaluation, AGCHEH
application, water quality Calibration guidance and review, and filial model input
sequence  development.    CBPO   has taken  the  lead   on  database   update  for
meteorologic data,  diversions,  point sources, and observed water quality data;
land use re-evaluation; and urban nonpoint source loading assessment. The final
water quality  calibration was  a joint effort with AQUA TERRA preparing final
input sequences  for all basins, developing  a preliminary calibration for the
Susquehanna basin, and providing essentially final calibrations for the remaining
basins  and below  Fall' Line  areas.   CBPO  staff continued  calibration -of the
Susquehanna, performed selected refinements  on  the other basins,  executed the
final runs, and  prepared'the statistical  analyses and appendices of the final
results.  Through this joint  effort CBPO staff have become thoroughly trained in
the application and many  of  the intricacies  of the Watershed Model.

1.2  OVERVIEW OF THE PHASE I/PHASE II EFFORTS

The primary  tasks  in the  Phase I project involved selected model refinements,
updating model input, and preliminary hydrology and water quality re-calibration;
these primary tasks and major sub tasks are listed below:

          WATERSHED MODEL REFINEMENTS

              •  Incorporate snowmelt  simulation            /      '•
              •  Allow variable hydrology for each land use category
              •  Include  animal waste  contributions

-------
   Chesapeake  Bay  Watershed
   and. Ma j or  A'F L  S ub -b-a s i n s
Figure 1.1  Chesapeake Bay watershed and major above fall line

-------
 A F L  Mo'de-1' Segments   and
  C a 1 i b r a ti on  Static ns
Figure 1.2 Above fall line watershed model segments.

                   4     •

-------
           Baltimore
           Harbor
         Severn
                                                              Bohemia
   Anacostla
Chickahominy
                  Nansemohd
                                                   Elizabeth
       Figure 1.3  Below fall  line  watershed model  segments.

-------
TABLE 1.1  SUBBASIN, SEGMENT NAME, AND SEGMENT NUMBERS OF THE CHESAPEAKE BAY
           WATERSHED MODEL
SUBBASIN

ABOVE FALL LINE

SUSQUEHANNA
POTOMAC



RAPPAHANNOCK

YORK


JAMES

APPOMATTOX

PATUXENT
SEGMENT NAME / GROUP
East Branch Susquehanna (EBSUS)
Vest Branch Susquehanna (VBSUS)
Juniata (JUNIATA)
Lower Susquehanna (LOSUS)
Conowingo (CONOW)

Upper Potomac (UPPOT)
Shenandoah (SHENA)
Lower Potomac (LOPOT)

Rappahannock (RAPPA)

Mattaponi (MATTA)
Pamunkey (PAMUN)

James (JAMES)

Appomattox (APPOM)

Patuxent (PATUX)
SEGMENT NUMBERS
10, 20, 30, 40
50, 60, 70
90, 100
80, 110
120, 140

160, 170, 175
190, 200
180, 210, 220

230

235. 240
250, 260

265, 270, 280, 290

300, 310

330, 340
UPPER EASTERN SHORE
BFL_1A
1
BFL_1B

LOVER EASTERN SHORE
BFL_2



VEST CHESAPEAKE
BFL_3A

BFL_3B

^
Coast_l
Bohemia
Chester
Wye
Chop tank

Nanticoke
Wicomico
Pocomoke •
Coast_4

Coast_ll
Coast_6
Gunpowder
Baltimore
Patapsco

360
370
380
390
400

410
420
430
440

450
460
470
480
490

-------
TABLE.1.1  (Continued)
SUBBASIN

BELOW FALL LINE
SEGMENT NAME / GROUP
 SEGMENT  NUMBERS
PATUXENT/MID CHESAPEAKE
BFL_4                   Patuxent
                        Severn
                        Coast 5
POTOMAC
BFL 5
RAPPAHANNOCK
BFL 6     ,
JAMES
BFL_7A

BFL 7B
Potomac
Anacostia
Occoquan
Rappahannock
Coast_8
Great Uicimico
York
James
Chickahomlny
Nansemond
Elizabeth
Coast 9
                                    500
                                    510
                                    520
530
540
550
560
570
'580
590
600
610
620
630
640

-------
          UPDATE MODEL INPUT

              •  Update meteorology and precipitation to 1984-85
              •  Update diversions, point sources, and observed
                  streamflow/water quality data to 1984-85
              t  Re-evaluate land use for 1974-78 and update to 1984-85

          WATERSHED MODEL RE-CALIBRATION

              •  Re-evaluate nonpoint parameters and loadings
              •  Re-calibrate Watershed Model to 1984-85 with model
                  refinements and updated input

The Watershed Model refinements in Phase I focused on areas that had been noted
as specific deficiencies in the earlier applications.  Snow accumulation and melt
had not been included in the first Watershed Model formulation,  and thus was not
included in the HSPF version resulting from the HSP-NPS model conversion study
(Donigian et al, 1986).  Since snow accumulation is significant in much of the
northern portions of the watershed,  simulating snow processes was considered an
important element in year-round simulations produced by the Watershed Model.

Also, the converted Watershed Model used the same hydrologic parameters for all
land uses within a specific hydrologic segment; this was a limitation of the NFS
model which  was eliminated when  the  Watershed   Model was converted  to HSPF.
Thus, allowing  a  different set of hydrologic  parameters for  each land use was
considered  important  to  representing  the  impacts  of  land  use  changes.
Furthermore, animal waste had not been specifically considered in earlier efforts
and contributions from this source needed to be evaluated.

A major portion'of the Phase I effort involved evaluating, refining, and updating
the  required input data  to  the  Watershed  Model.   This effort was  severely
hampered by the lack of detailed documentation and loss of back-up information
for the original Watershed Model  (NVPDC, 1983),  especially for the diversions,
point sources,  and meteorologic  data.   Also,  the accuracy and validity of the
land use information used in the original model was questioned, leading to the
need to re-evaluate the  land use for both  the  1974-78  period and  the 1984-85
period.    Although not  originally  envisioned   as  part  of  Phase  I,  direct
atmospheric deposition of nitrogen was  an added input  to the model (for water
surfaces only) to allow for this source.   To  avoid future problems, such as those
experienced in  this study in updating the model, special efforts were expended
to develop detailed documentation and back-up for all  procedures and analyses
performed in this model development effort.

The  Watershed Model re-calibration  effort  involved both hydrology and water
quality.  The  hydrology re-calibration  was  needed because of the inclusion of
snow simulation and the parameter adjustments and refinement for better land use
representation; addition of an animal waste  land segment and changes to the land
use  also  required a re-assessment of  the  hydrology.   The water  quality re-
calibration  involved  re-evaluation of  the  nonpoint parameters  (i.e.  potency
factors, subsurface concentrations, accumulation/washoff rates) and the expected
loadings  from each land  use represented in the  Model,  and then adjustment of
these parameters and selected instream water quality parameters was performed as

                           •   '         8          .

-------
part  of the  calibration  process.    Although  the  model  had  been previously
calibrated  to  the  1974-78 period,  differences  in the  point loads  and  the
additional  effects  of animal  waste  and  atmospheric  contributions  required
considerable  re-calibration  for  the  1984-85  period  in Phase  I.   Problems
identified during the re-calibration were discussed in the Phase I report, and,
as noted above, helped  to  focus the  Phase II effort.

The emphasis  in the Phase  II  work  was directed  to more  detailed modeling of
agricultural BMPs, addition of sediment erosion and instream  transport modeling,
and incorporation of instream sediment-nutrient interactions in order to better
represent nutrient  delivery to the  Bay.   This effort  involved  replacing the
empirical .potency factor/subsurface concentration  approach with  the process-
oriented Agrichemical modules  in HSPF  for the cropland areas of the watershed.
In  addition,  the  instream module  (RCHRES)  of HSPF  was enhanced  to  include
sediment-nutrient  interactions,  and the resulting enhanced version  was  re-
applied/re- calibrated to the drainage area.  The original  workplan for Phase II
identified the  following major tasks and subtasks:

       DETAILED EVALUATION OF AGRICULTURAL PRACTICES •

          •   Identify/inventory current practices/BHPs
          •   Identify/evaluate field site data
          •   Apply detailed agricultural models                    •      .
          •   Adapt Watershed Model to  reflect field site  results

       SEDIMENT/NUTRIENT INTERACTIONS  IN HSPF RQUAL

          •   Enhance HSPF  RQUAL to include sediment/nutrient interactions
          •   Test model enhancements on selected  subwatersheds
                                                            • •
       APPLICAtlON/RECALIBRATION OF  ENHANCED WATERSHED MODEL

          •   Calibrate  erosion and instream sediment transport
          .•   Apply/calibrate  enhanced  RQUAL
          •   Integrate  Task A results  into Watershed Model                    .
          •   Apply/Calibrate  Enhanced  Watershed Model to  1984-87 data

1.3   PHASE I  PRELIMINARY FINDINGS AND  RECOMMENDATIONS

In this section we briefly  summarize  the primary conclusions  and recommendations
resulting from  the Phase I work, in  terms of hydrology, nonpoint loadings, and
water quality simulation,  in order to set the  stage for  describing the Phase II
efforts and final results.   These findings and recommendations were discussed in
greater detail  in the  Phase I  report.

1.3.1 Hydrology  Conclusions  and Recommendations

       a.  Overall,, the hydrology calibration results for 1974-78 were very good
           to excellent.   The agreement in mean  annual flow volumes and flow
           frequency results was excellent.  Although  some differences remained
           in the daily timeseries flow comparison, the basic  conclusion was

-------
           that the model  was  a good representation of the  observed data for
           that time period.

       b.  The hydrology  results  for 1984-85  were not  nearly  as -close  to
           observed data values as in the earlier 1974-78 period. Both the daily
           and monthly R2 values were considerably lower  at most stations, and
           the frequency curves generally showed greater  deviations with lower
           peak flows and higher base flows than observed.

       c.  The agreement could have  been  considerably improved for 1984r85 if a
           complete re-calibration had been performed,  instead of limiting our
          . adjustments to  the  input evaporation.   The  Phase I recommendations
           for additional efforts in Phase II included the following:

           1.  Extend  the  data base  and simulation .through 1988 or  1989  to
               provide a better basis for evaluating the  model performance for
               this latter period.
                   •                                                       (
           2.  Snow depth  data should be developed for the  1984-89 period to
               check  and  evaluate  the  snow simulation,  and  allow for  any
               additional calibration if needed.

           3.  The Conowingo Reservoir simulation needs to be investigated and
               re-evaluated because the rule curve needed  to be adjusted during
               the 1974-78 calibration.

           4.  The storm of November 1985  in  the Upper  Potomac was  not well
               simulated,  in  terms of  peak flow,  by  the Watershed  Model,
               although the monthly  volume was accurate. Further efforts should
               be made to  investigate  the  problems with'simulating this event
               once the simulation period has been extended.

1.3.2  Phase I Nonpoint Loading Conclusions and Recommendations

       a.  The simulated annual nonpoint loading rates  were within the range of
           expected values, with some deviations  such  as the rates for P04 and
           NH4 from forest, and PO*  rates from pasture.

       b.  Comparing   Conventional   and   Conservation   tillage   segments,
           Conventional produced higher loading rates for  all pollutants except
           N03, .where Conservation is the higher rate.

       c.  The highest rates shown  were for the manure   segment,  followed by
           urban and/or conventional tillage, conservation tillage, pasture, and
           forest.  The  order changed slightly  for individual pollutants.  The
           manure segment loading rates  need to be further evaluated to ensure
           they provide a reasonable means  of representing animal waste inputs.

       d.  Analysis of Total Segment Loads  indicate that  agricultural cropland
           (i.e. both  conventional  and conservation) tended to  be the highest
           loader  of  both Total   N  and  Total  P;  however,   the  percent
           contributions  varied .from  segment   to  segment,  and other  sources
                      •      •                       -
                                      10

-------
           dominated  in selected  segments.    Both  urban  and  animal  waste
           contributions were significant, generally varying between 10X and 30Z
           of TN and/or TP.

       e.   Point sources are still a major source of  TP,  often in the range of
           202 to 302 or more,  and a moderate source of TN.  The assumptions in
           developing the point  source  loads  should  be re-assessed  to  insure
           their representation is accurate.

       f.   The atmospheric  contribution tends  to be only  a few percent or less
           for most  areas,   since only direct, deposition to water surfaces is
         .  considered.  Atmospheric deposition to land surfaces  is assumed to be
           implicitly included in the nonpoint calibration.

       g.   It should^be  noted that these results from  Phase I were based on only
           two years  of simulation.   Loading  rates from nonpoint sources  are
        .  notoriously variable from year to year.  Thus,  the recommendation to
           extend the simulation in  Phase II through 1988 or 1989 (as  discussed
           above for hydrology) is also supported by  the  need to better define
           the simulated range and mean of the nonpoint loading rates.

1.3.3  Phase I Water Quality Simulation Conclusions and Recommendations

       a.   Water  temperature  and DO simulation  was  generally  very good  to
           excellent throughout the Bay drainage.

       b.   The majority of the  nutrient concentrations  were  simulated  quite
           well,  reflecting the general  range of the  observed data  values,
           especially  for  Total N and Total P at the major Fall Line  stations.
   '...".'.       '  .  '                        '           '
       c.   Larger deviations were seen'in the individual nutrient forms, such as
           PO* in the Upper Susquehanna and the James rivers.   The  focus of
           Phase II, on sediment simulation  and sediment-nutrient interactions,
           should help to improve the phosphate simulation and should be further
           investigated.

       d.   The Potomac  simulation  results were significantly better  than  the
           results for the  Susquehanna and James, especially for the phosphorus.
           components and Chi a.  The Potomac data is more extensive than in the
           other  basins,  allowing  a better  representation  of the  expected
           variability  in the observations.               .

       e.   The  river systems  show  considerable variability  within the  Bay
           drainage.   Differences  in P0t,  Chi a, and other  parameter between
           regions and  individual subbasins need to be  further investigated as
           part of Phase II.                             .

       f.   The  storm of November 1985  in  the  Upper Potomac was also  under-
           simulated in terms of particulate associated  nutrients.  In Phase II,
           the  water  quality  simulation  should  be   further'  investigated
           especially in terms of the nonpoint loadings contributions.


                                      11

-------
       g.  Phase II should also  focus  on some of the smaller  subbasins (e.g.
           Patuxent,  Appomattox)  where  further investigation and calibration is
           needed to present a consistent level of simulation throughout the Bay
           drainage.

       h.  For Total  N and  Total P  loads  at  the  Fall Line,  the model  and
           regression annual loads generally agreed within about 20X;  problems
           with Organic P in 1985 (partly due to the November 1985 storm) lead
           to greater differences for that year.

       i.  Nitrate showed the closest agreement between the two load estimates,
           at least for  the three major  Fall Line stations.   Interestingly,
           nitrate dominates the   Total N  and all other nutrient loads in terns
           of total mass,  especially  for the Susquehanna, Rappahannock, and the
           Patuxent,^and to a lesser extent, the Potomac.  In the James, Organic
           N dominates the Total N load and  is  significantly greater  than the
           nitrate load.

       j.  The differences  between  the  simulated and  regression loads  were
           generally much greater for  the minor tributaries  for most  nutrient
           forms, than for the major Above Fall Line basins.

       k.  Monthly load timeseries showing the  simulated and regression loads
           (plus  Standard Error  Bounds)  indicate that  the model  estimates,
           especially for TO and  TP,  generally fell within the envelope defined
           by the two error bounds,

       1.  Simulation of  Total Organic Carbon  (TOO)  should be  considered in
           Phase II,  either to replace BOD or supplement it.

1.4  PHASE II FINDINGS AND RECOMMENDATIONS

The general conclusions of the Phase II -effort are discussed below in terms of
hydrology, nonpoint loadings and AGCHEM simulation, and water quality simulation:.

1.4.1  Phase II Hydrology Conclusions and Recommendations

The Phase II hydrology calibration results indicate the following:

       a.  The  agreement  between  mean  annual flow  volumes  is  excellent;,
           simulated values are within 62  of  observed values.  The year-to-year
           variations shown in the appendix are  greater but they still indicate
           good agreement.                            ,                 •

       b.  The  daily R2  values   for  simulated  and observed  streamflow  are
           generally 0.70 or greater, and usually in the range of 0.75 to 0.84
           for the three major basins.  The monthly R2 values are consistently
           in the range of 0.77 to 0.90 for all subbasins.

       c.  The daily  R2  values for  1984-87  are consistently higher  than the
           corresponding values for 1974-78,  possibly due to the greater number
        .  ' of rain gages used in Phase II.

                                  .12

-------
       d.  Both  the  timeseries  and  flow   frequency   comparisons   for  all
           calibration  sites  show  good   to  very  good  agreement.    Although
           differences  exist,  the  daily  simulated  timeseries and  frequency
           curves track the observed timeseries quite well for most all sites.
           The largest  deviations  are consistently  at the  smaller  subbasins
           where more detailed spatial representation  of  precipitation may be
           needed.    Also,  these differences  are  greater when  reservoirs  are
           present in the smaller subbasins (e.g.  Pamunkey,  Patuxent);' further
           study of the reservoir operating rules and  finer spatial segmentation
           may be needed.
                                     i
       e.  Very little improvement was possible in Phase  II  for simulation of
           the November 1985 storm in the Potomac.  The storm was well simulated
           in the Upper Potomac  and  Shenandoah model  segments, but the recorded
           rainfall amounts,  intensities,  and  spatial distribution in.the Lower
           Potomac could not sustain the  extreme observed storm flows.

       f.  In  Phase  II,  the  Conowingo   Reservoir   hydraulic  simulation  was
           investigated and adjusted to include a minimum summer release rate of
      :     5000 cfs  and minor adjustments to the weekly  outflow  coefficients
           Both the  daily timeseries and  the  frequency  results  show  very good
           agreement between simulated and observed values, with some deviation
           in the low flow simulation since these flows are controlled entirely
           by the operating procedures of the reservoir.

       g.  Seasonal R2 values were calculated by CBPO   based on the  following
           seasons:
                   Season 1   January -  February
                   Season 2   March - May
                   Season 3   June - September
                   Season 4   October -  December
                                                  '•"'••'
           The results show that the winter (Season 1) is consistently simulated
           worst than the other three seasons, and  that the summer and fall
           (Seasons 3 and 4)  are usually the best, at least for the three major
           basins.    For most  of the smaller basins,  the  spring  (Season 2) is
           often the best simulated.

Overall,  the  Phase  II  hydrology calibration  results are  a very  good
representation of the Chesapeake Bay basin hydrology and provide a sound basis
for nonpoint load generation and delivery of pollutant loads to the Bay.  Ongoing
updating of the meteorologic, diversion., and land use  information is recommended
to allow for complete verification of the hydrology using the 1988-91 period.

1.4.2  Phase II Nonpoint Loading Conclusions and Recommendations

       a.  Generally, the simulated  annual loading rates are within the range of
           expected  values, with  some deviations.    Annual rates  for P04 from
           forest  and pasture,  and NH4  from forest occasionally  tend  to be
           toward the lower end  of the defined range.  Annual  P04 rates from the
           cropland areas are somewhat higher than the defined range.


               •                       13

-------
       b.   For non-cropland categories, the Total N and Total P simulated values
           compare  favorably  with both the expected means and ranges.

       c.   For the cropland  categories of Conventional  Tillage,  Conservation
           Tillage,  and Hay,  the  Total  N and  Total P  simulated values  are
           generally close to  the mean lumped cropland data; the Hay values are
           generally  less  than  the  mean, while the  tillage categories  are
           usually  greater than the mean but  well within the observed range.

       d.   Comparing    Conventional   and  Conservation   tillage   segments,
           Conventional produces higher loading  rates for  most  model segments
           for. all  pollutants except  N03, where  Conservation  is  sometimes the
           higher rate.

       e.   The highest rates for Total N and Total P are for the manure segment,
           followed by conventional tillage,  conservation  tillage,  urban,  hay
           land,  pasture,  and  forest.    The     order  changes  slightly  for
           individual  pollutants.

       f.   The manure  segment loading rates are  the most uncertain since there
           is very  little  information on which  to assess their validity.

       g.   For NH3 and P04 from cropland,  the simulated ranges are generally 0.5
           to 4.0 Ib/ac and 0.2 to 2.0 Ib/ac,  respectively;  these  ranges  are
           generally higher than the limited observed data for these forms, but
           they are not unrealistic based on  the general literature.

       h.   Urban pervious - and  impervious areas provide  loadings  which  are
           comparable  to the  Hay and Pasture  categories,  and for some nutrient
           species  (etg.,  NH3)   the   loadings   are  similar  to  .the  tillage
           categories.   Thus, urban land can be  a significant; source of total
           nonpoint loadings  in urbanized model  segments.

1.4.2.1  AGCHEM Conclusions  and Recommendations
    1 .       • -                           -    --
The AGCHEM application to the  cropland areas of the Chesapeake Bay drainage has
satisfied the objectives outlined at  the beginning of Phase  II,  namely to (1)
represent nutrient balances  for  the  major cropland categories in each model
segment,  (2)  allow investigation of  sensitivity  to  climate variations  and
agricultural  practices,  and  (3)  provide a  mechanism to  project impacts  of
nutrient management alternatives  .for  agricultural cropland.  The results  are
consistent  with  expected  nutrient   balances,   observed   ranges  of  runoff
concentrations,  and expected  ranges of nutrient loadings from cropland areas.
The AGCHEM approach to modeling the nutrient  balances on agricultural cropland
is exactly the type of  approach recommended by the Chesapeake Bay Nonpoint Source
Evaluation Panel in their report to the U.S. EPA  and the CBP Executive Council.
In  March  1990,  the   EPA Administrator  convened the  Panel to  "assess  the
effectiveness of current efforts to reduce nonpoint source loadings of nutrients
entering the  Bay system".  They reviewed and evaluated  the effectiveness of a
wide range of nonpoint  source programs,  including program design, implementation,
budgets, modeling and  assessment methods,  and research efforts.  One of the key


                                       14

-------
recommendations of the Panel related to new approaches to nutrient control was
as follows:

       "The Panel recommends that the Bay jurisdictions and the federal
       agencies develop a mass  balance accounting system, where nutrient
       loadings are balanced by the nutrients  removed from the system
       plus  those which  are  introduced and  stored"  (Chesapeake  Bay
       Nonpoint Source Evaluation Panel, 1990).

The AGCHEM application provides an initial assessment of this 'mass balance' for
cropland  areas,- a tool  for assisting  in the  1991  Re-evaluation of  the 40X
nutrient reduction goal of the  Bay Agreement, and a framework for future efforts
on nutrient management based on the mass balance approach.  Recommendations to
improve the use and application of AGCHEM by CBP include the following:
                    *
       a.  A detailed review and refinement of the current fertilizer and manure
           application  rates,  and changes expected  under nutrient management
           alternatives should be initiated since these rates are the starting
           point  for nutrient  reduction-from cropland.

       b.  Detailed site  applications  of AGCHEM,  such  as  in the  Patuxent,
           Nomini, and Owl run watersheds, should  be diligently pursued by
           either CBLO and/or  state agencies to further refine and improve the
           simulated nutrient balances and regional differences.

       c.  Variations in soil nutrient storages, throughout the Bay drainage and
           for  different  land  uses,   should  be included  in  future  model
           /improvements.  .Use of CIS capabilities and soils databases should be
           used to improve spatial definition .and regional variations in model
           parameters.                                                     ,

       d.  In  line with the CBP NPS Evaluation Panel recommendations, future
           Watershed  Model  refinements  should include  application  of  mass
           balance and soil process modeling procedures  for  all  land uses to
           represent nutrient  balances in all areas.

       e.  Recommended improvements to the AGCHEM module to facilitate Watershed
           Model  application and operation,  and refinement of selected nutrient
           processes  include:

               1.  Capability  to simulate  NH3  volatilization from  soils and
                   manure applications.
                                                       •             .
               2.  Modifications   to   simplify  specification  of   nutrient
                   application rates and agricultural operations; more than one-
                   half  of the current  model  input sequences  are  devoted to
                   defining this  input for cropland areas.

               3.  Capability  to specify wet  and dry atmospheric deposition
                   rates  for all nutrient forms.
                                       15

-------
               U.  Refinement of soil phosphorus cycle and processes to allow a
                   better representation of sorption, fixation, and plant uptake
                   mechanisms.   Additional research and algorithm development
                   may be needed to accurately model the soil phosphorus cycle.

1.4.3  Phase II Water Quality Simulation Conclusions and Recommendations

Review of the water quality calibration results indicates the following general
conclusions and recommendations, some of which were  also presented in the Phase
I report:

       a.  Water  temperature and  DO  simulation is generally  very good  to
           excellent throughout the Bay drainage.  In Phase I, water temperature
           was  calibrated at  three  instream sites within  the  basin,  where
           continuous  water  temperature  data was   available,   and  then
           extrapolated  to  the other model  segments.   The  simulation  was
           extended and refined  in Phase II.

           The DO simulation is  generally  very good  for most calibration sites
           but  with  some  deviations  at  selected   sites  for  low  summer  DO
           observations.   At most sites, the simulation is quite good except for
           one or two extreme points whose observed  values may be questionable.
           Some of the low summer values may be due to benthal demands which are
           not currently calculated  in the model.-                ,

       b.  The majority of the nutrient constituents are simulated quite well at
           most sites, reflecting the general  range of the observed data values,
           especially  for Total N and Total P at the major Fall Line stations,
           with  the .notable exception  of the  Susquehanna  Basin  (discussed
           below).  Peak  simulated concentrations are often somewhat higher than
           observed values,  as  would be expected when  comparing a continuous
           simulated timeseries  with discrete, individual observations.

       c.  The Phase II results  are  a  significant improvement over the Phase I
           results due to the water quality parameter re-evaluation, the AGCHEM
           nutrient loading simulations, and the additional calibration effort.
           Although some problems remain,  the results show improved simulation
           of seasonal variations in nitrate and Chi  a, closer representation of
           individual nutrient species, and better agreement with most observed
           concentrations and loads.

       d.  Large deviations  are  still  seen in the individual nutrient forms at
           some sites, such as P0t  in portions of the Susquehanna and inorganic
           nutrients in the minor tributaries.  Chi  a observations were limited
           at many calibration sites, thus restricting the calibration process.

       e.  The Potomac simulation results  are the best of all the major basins
           and  are generally a  good to very good  representation of nutrient
           concentrations throughout the Bay basin.  This could be  partially due
           to the  fact that the Potomac data is more extensive than in any of
           the other  basins,  allowing a better representation of the expected
           variability in the observations and a better  calibration.  The total

                                       16

-------
    nutrient  concentrations,   and  the   individual  species,  are  well
    simulated throughout  the  entire  period.   Concentration peaks  do
    exceed individual observations, especially for NH3 and P04, but they
    are generally within the range of the observed data.

f.   The  Susquehanna  simulation  also  shows  good  agreement  between
    simulated and observed concentrations, but this was accomplished only
    after  intensive  calibration  effort  of  the  entire  basin,  with
    particular focus on the reservoir simulation.  Generally, Total N and
    component concentrations at Conowingo are somewhat under-simulated.
    P04 and Total P concentrations are somewhat over-simulated throughout
    the Susquehanna, possibly due to interactions with sediment and metal
    oxides,  and increased sorption under low pH conditions.

    USGS data* for pH show extremely low values for  two main stem stations
    in the East and West Branch subbasins,  with mean  monthly values in
    the 3.0  to  4.0 range.   Detenbeck and  Brezonik  (1991)  show greatly
    increased P binding by lake sediments for pH in the range of 4.5 to
    6.0,  with  greatly decreased diffusive  P   fluxes  under  these
    conditions.   Consequently,  increased  sorption and  retention,  and
    possibly precipitation  of  P04  from the  overlying  water  may be  a
    reason  for most of  the  observed  concentrations  being  close  to
    detection limits.

    Clearly,  further investigation is  warranted and  should focus  on
    instream processes  in the  Upper and Lower basins  for both nitrogen
    and phosphorus  delivered to the Conowingo Reservoir, and processes
    within the reservoir.  This should be performed in conjunction with
  :  a  thorough  loadings  analysis  to  confirm the  relative  loadings
    contributions from all sources.  Further calibration of the reservoir
    water  quality  is  heeded  to  improve  the  sediment  and  Chi  a
    simulations,  which may  also  assist in  improving  the  phosphorus
    simulation.                               '

g.  The, simulations for the Janes River are  a significant improvement
    over .the Phase I results.   Although the concentration agreement is
    not  as  good  as  in  the  Potomac,  the  Total H  and Total P  are
    consistently within the range of the observations, and the component
    nutrient species  are reasonably well simulated.   The data base for
    the James was much more limited compared to  the  other  sites,  thus
    limiting the  calibration effort.

h.  The  river  systems  show  considerable  variability  within  the  Bay
    drainage, as  noted above concerning the P04 simulation problems; in
    the Susquehanna.  In general, the smaller tributaries and those with
    reservoirs are  not  simulated as well as the larger basins with free-
    flowing  reaches.   When  both these  conditions  occur, such as in the
    Pamunkey and Appomattox  basins,  the differences between simulated and
    observed values usually increase.  In these cases, finer segmentation
    and more investigation into reservoir operations is  needed.  For the
    short-term, simulation differences in the smaller basins should have
                                17

-------
           minimal  impact  on the Bay simulation since they are a small  fraction
           of the delivered nutrient load.

       i.   Although the Patuxe'nt River  falls in the category of a  small  basin
           with reservoirs,  the  simulation results  are exceptionally good due, to
           the wealth of  observed  data  for  calibration  and the point  source
           monitoring that  accurately defined the point  source  loads inx the
           Watershed Model.   The Total P simulation reflects the  impact  of the
           phosphorus ban  in December  1985,  with the 1986-87  concentrations
           being significantly lower than in 1984-85.   The few instances  where
           simulated values are high are due primarily to under-simulation of
           extreme  low flows during late summer.

       j.   The storm of November 1985 in the Upper Potomac was  under-simulated
           in  terms ,of peak flow,  and  many  of  the  particulate  associated
           nutrients were  also under-simulated for this event.  In Phase II, the
           HSPF code was modified to allow user-specified particulate NH4 and P04
           concentrations  for scouring  of the bed sediments; however, much of
           the scoured material  is organic.  Consequently,  the particulate N and
           P  concentrations  are under-simulated  for  this  event  in both  the
           Potomac  and James basins, and since these components dominate  during
           a major  scour  event,  Total N and P concentrations are also under-
           simulated. . Future efforts should consider adding the  capability to
           allow user-specified bed concentrations for organic  components.

       k.   Simulation of  Total  Organic  Carbon (TOC)  was  included in  Phase II
           because  TOC is a  required input  for the Bay model and observed TOC
           values were more  readily available at  more sites than BOD.  BOD is
           still simulated  by  the  Watershed Model because  it  is a  required
           parameter. .BOD  is used in the current version  as an indicator of
           TOC, i.e.  BOD nonpoint and point source  Ipadings are converted to TOC
           by a conversion  factor  which was used as  a calibration factor to
           adjust   the  TOC  simulation.    The  simulation  results  for  TOC  are
           generally quite good  for most sites, and are similar in appearance to
           the organic N and organic P simulations.

       1.   The sediment simulation  results are generally adequate but  could be
           improved with further calibration.  The model tends to over-simulate
           large storms,  and under-simulate  small storms  and  concentrations
          .during   low  flow periods.   Additional   sediment  calibration  should
           concentrate on evaluation of the  simulated sediment loadings and the
           target values  derived from the HRI data.   Also,  reservoir  sediment
           simulations  should   be   further   investigated  with   additional
           calibration.   Finer  model segmentation .and spatial detail of both
           land segments  and stream  reaches  is  probably needed for  improved
           sediment simulation.

The following general conclusions are derived from comparison of the simulated
Fall Line loads with regression estimates:

       a.   For  Total N and Total  P,  the model  and  regression annual  .loads
           generally agree  within  about 251 to 302.   Problems with Organic P

                                      18

-------
           and, to a lesser extent Organic N,  in th Potomac in 1985 (partly due
           to  the  November 1985 storm)  lead  to  greater differences  for that
           year.

       b.  In the Patuxent and the Potomac, the closeness of the concentration
           comparison is also reflected in the Fall Line load ratios.  For both
           these  rivers,   extensive  data  was available for  developing  the
           regression and performing the model calibration.

       c.  Nitrate  generally  shows  the  closest  agreement between  the  two
           estimates,  at  least  for  the three  major  Fall  Line  stations.
           Interestingly, nitrate dominates the Total N and all other nutrient
           loads  in terms  of  total  mass,  especially  for the  Susquehanna.
           Rappahannock, and the Patuxent.
                    • ^
       d.  The differences between the  two estimates are generally much greater
           for the minor tributaries  for most nutrient forms  (with the exception
           of the Patuxent),  than for the major Above Fall Line basins.  For the
           minor  tributaries  and  the  James,  significantly  less  data  was
           available   for  developing   the   regression.  Consequently,   the
           uncertainty of the regression is greater for these basins.

1.5  GENERAL WATERSHED MODEL RECOMMENDATIONS

The Watershed Model provides a framework for quantifying and evaluating nutrient
loadings  to  the Bay, and a mechanism for assessing the impacts of  land use
changes and alternate nutrient and agricultural management practices.  However,
the watershed' modeling component of the Chesapeake Bay Program should  be an
ongoing, evolving process with continuing data collection, model development and
testing,  and calibration/validation efforts.   The current Watershed  Model,
although providing 'a sound basis for current decision-making, needs continuing
updating  and refinement  in  a number  of areas.   As all planning  tools,  the
Watershed Model should not be static.  Continuing efforts should be devoted to
the following areas:

       a.  Further calibration and validation with new data (e.g. 1988-90) with
           specific  focus on problems  areas,  such as animal -waste,  reservoir
           modeling, and P04 in the Susquehanna.

       b.  Further investigation and refinement of reservoir hydrology and water
           quality simulation for model segments with major reservoirs.

       c.  Refinement  and improvement  in urban -and  animal waste  simulation
           procedures and management alternatives.

       d.  'Refinement   of AGCHEM nutrient   balances  based  on  field  site
           applications, regional conditions, and seasonal variations.

       e.  Finer  spatial segmentation and representation  of  land areas  and
           stream  reaches to  provide  more detailed input  to  regional  water
           quality  issues and problems.
                                     H
                                      19

-------
f.  Direct  linkage  between  the  CBLO  capabilities   with  geographic
    information systems  (CIS)  and  model parameter/input  development.
    This is a  major  focus  of current development efforts  in watershed
    management and nonpoint source modeling.                            .

g.  Extend nutrient balance concepts of AGCHEM to simulation of all land
    uses to provide a consistent mass  balance approach for all portions
    of the Watershed Model.

h.  Develop  improved capabilities  in  the  Watershed Model  .for  user
    Interface,  output   analysis,   and   specification   of   nutrient
    applications and agricultural management practices.
                              .20

-------
1.6  FORMAT OF THIS REPORT

Section  2  describes  the  procedures  used  to  update  the  meteorologic  and
precipitation data for the 1984-87 period, while Section 3 briefly describes the
land use and waste load development techniques.  Section 4 describes the nonpoint
loading  simulation procedures used  for both  cropland  and non-cropland  land
categories, including an overview of the AGCHEM  module  and the  application
procedures  used for  cropland  areas; nonpoint  simulation results  are  also
presented in Section 4 for all land uses.  Section 5 provides an overview of the
RCHRES module sections used in the Phase 11 effort,  along with discussion of the
sediment-nutrient capabilities incorporated into the HSPF code.  Sections 6 and
7, respectively discuss the hydrology and water quality results, with the primary
focus on the Phase 11 results.  As noted above, due to the vast  amount of data
and information, developed for and generated by the Phase 11 watershed modeling
work, the majority of the results  and input related information are provided in
the Appendices.
                                      21

-------
                                  SECTION  2.0

                     METEOROLOGIC AND  PRECIPITATION DATA
                         BASE  DEVELOPMENT PROCEDURES
2.1  INTRODUCTION

This section  documents  the procedures  that were  used to update  the regional
meteorologic data  base  and hourly precipitation  data base for  1984-1985 and
subsequently,  1986-1987.  The regional meteorologic data base consists of wind
speed,  air temperature, dewpoint temperature, evaporation, cloud cover, and solar
radiation.  The goal  of  this task was to produce a  data base that was both valid
and, insofar as possible,  consistent with the  1974-78 data with  regard to the
observation stations,  computation and disaggregation methods,  and level  of
detail.  Because documentation of the existing  (1974-78) regional data base was
hot available, this data base was exhaustively analyzed and compared to observed
and generated  data in an attempt to determine  data  sources  and  computational
methods.   Based  on   this  analysis,  observation  stations were  selected and
procedures were developed to update the data base  for  the 1984-85.  Development
of the  hourly  precipitation data was accomplished using modified versions of the
Thiessen procedure, aggregation software, and observed data base used to develop
the 1974-78 data base.  The same procedures were also  followed in completing the
data base for  the  1986-1987- period;  however,: small changes in the  network of
observed stations were necessitated by the discontinuance of several stations.

The final data base is stored in Watershed Data Management (WDM)  System files.
WDM files are  unformatted,  direct access files that may contain  several types of
data and many individual data sets of each type.  The data sets are identified
by unique data set numbers and  other  attributes  such as a name,  descriptive
information, and time  interval.   WDM file format  is  the primary.data storage'
format for the HSPF program.  In addition, several intermediate data formats were
utilized in developing the data base.  These formats are ASCII, HSPF-readable,
hourly and daily formats that are identified in this document as HYDDAY, HYDSDY,
and HYDHR, respectively.  The formats are described in Appendix D.I

2.2  REGIONAL METEOROLOGIC DATA

The 1974-78 regional  met data was compared with published data for major stations
in the basin.   As  a  result of these  comparisons,  the cloud cover and dewpoint
temperature data sets were found  to exactly match the  published  data as shown in
Table 2.1                             .
                                      22

-------
          TABLE 2.1  CORRESPONDENCE BETWEEN 1974-78 REGIONAL DATA
                     AND PUBLISHED DATA
REGION
1
2
3
4
5
6
7
CLOUD COVER •
STATION
Binghamton, NY
Williamsport, PA
Harrisburg, PA
Roanoke, VA (1)
National Airport
Roanoke , VA
Richmond, VA
DEWPOINT TEMPERATURE
STATION
Binghamton, NY
Binghamton, NY (2)
Harrisburg, PA
Roanoke, VA
National Airport
Roanoke, VA
Richmond, VA
         (1)      Region 4 cloud cover data are computed from Roanoke
                  data and monthly factors.

         (2)      Region 2 dewpoint temperatures for March-November
                  are the Binghamton data multiplied by 1.1.
The  other regional  met data  sets  (wind speed,  air  temperature,  potential
evaporation,  and solar radiation) did not match these stations,  but were found
to exhibit other relationships. Generally, these relationships involved constant
monthly factors between the regions.  Since the methodology used to develop these
factors is unknown,  it was  decided to base  the regional data entirely  upon
observed data from representative stations in or near each of the regions.   The
stations selected were those that were used in the 1974-78  data base for dewpoint
and cloud cover, with two exceptions.   The weather stations at  Elkins,  WV and
Dulles  Airport  were chosen as  more  representative  of Regions  4  and  5,
respectively, than Roanoke,  VA  and National Airport.  Table 2.2  contains the
stations used to  develop the 1984-87 regional data base,  and Figure 2.1 shows the
seven meteorologic regions and the locations of their corresponding stations.
             TABLE 2.2  STATIONS USED TO DEVELOP 1984-87 REGIONAL
                        METEOROLOGIC DATA
             REGION
STATION
STATION NUMBER
1
2
3
4
5
6
7
Binghamton, NY
Williamsport, PA
Harrisburg, PA
Elkins, WV
Dulles Airport
Roanoke, VA
Richmond , VA
300687
369728
363699
462718
448903
447285
447201
                                      23

-------
    Principal Meteorologlc Station
Figure 2.1  Meteorologic  regions and principal stations in
            the Chesapeake Bay watershed model.
                               24

-------
2.2.1  Cloud Cover and Dewpoint Temperature

The  Watershed  Model  utilizes  daily  values  of  cloud  cover  and  dewpoint
temperature.  The daily records for the stations listed in Table 2.2 were down-
loaded from the NOAA data tape, reformatted,  and transferred into the WDM files
without any modification of  the data.

2.2.2  Air Temperature and Wind Speed

The  Basin model  utilizes  hourly  values of  wind speed (miles/hour)  and air
temperature (deg F).  The daily records for the stations listed in Table 2.2 were
1) downloaded  from  NOAA data tapes, 2)  reformatted,  3)  distributed to hourly
values using customized software,  and 4)  transferred into the WDM files.

The daily max-min air  temperatures were converted to hourly values based on a
smooth variation with the daily minimum at 6 AM and daily maximum at 4 PM.. This
variation is illustrated in  Figure 2.2.   The  program TMPDIS, which is included
in Appendix D.2,  was used  to perform this distribution.   TMPDIS reads  a daily
data file containing maximum and minimum temperatures in HYDSDY format and the
daily observation hour  (1  to 24).  . The output hourly air temperature data are
written to the output file in the  HYDHR  format,

The daily wind movement (miles/day) data were distributed to  hourly Values using
a slight  diurnal  variation having  a minimum value from midnight to 7 AM and a
maximum at  3 PM.   Figure 2.3  illustrates this variation.   The program WNDDIS
(Appendix D.3) was  used to  perform the  distribution.   This program reads the
total daily wind  movement (miles per day)  from an input file, using the format
HYDDAY, and outputs the hourly wind speed (miles per hour)  to the output file in
the HYDHR format.

2.2.3  Solar Radiation

Analysis  of the   solar radiation data  and the  corresponding cloud 'cover time
series  indicates  that the 1974-78 radiation used by  the Watershed Model was
computed  using a model based  on latitude  and cloud cover.  Such radiation models
generally use curves of clear sky radiation as  a function of  latitude and day of
the year, along  with an empirically derived  correction  factor based on cloud
cover or  percent  sunshine.   Since  measured solar radiation data are no longer
readily available, it was decided  to generate  the 1984-87 radiation data using
a model.

Two simple solar radiation models were compared to the 1974-78 clear sky data in
an attempt  to  match the 1974-78  regional data.  The  model that  more  closely
matched the 1974-78 data was the RADIATION generation utility contained in the
HSP QUALITY program (Hydrocomp, 1977). Using approximate area-weighted latitudes
for each of the regions, this program generated curves that were nearly identical
to the  1974-78 clear sky data.   Small differences (1 to 41)  occurred  in the
spring  and  .autumn months depending on the latitude used in the  computation.
While neither the clear sky radiation nor the cloud cover correction factor were
reproduced exactly, the differences in generated  radiation  are negligible.
                                       25

-------
            60
            60
            40
              Temperature (F)
             0       •      ซ      18      24
                      Time of Day (hours)
Figure 2.2   Diurnal  variation of air temperature.
            16
              Wind Speed (miles/hour)
            14
            10
              06       12      W      24
                      Time of Day (houra)
Figure  2.3   Diurnal variation  of wind  speed.
                       '26

-------
A  FORTRAN subroutine  (RADCLC)  was  developed to  implement  this model.   The
algorithm  is  based on empirical curves developed by Hamon et al., (1954), and
documented in the  HSP  II Manual  (Hydrocomp,  1978)).   Also,  since  the daily
radiation  must be distributed to hourly values for input to HSPF,  the -subroutine
RADDST was adapted from the HSP QUALITY program to perform this distribution.
This routine generates hourly radiation values using the daily total, the month,
and the  day.   These two routines  were implemented in a program that reads the
daily cloud cover data and the latitude for a region, and generates the daily and
hourly solar radiation for each  day of the year.  Listings of the subroutines are
included in Appendix  D.4 along  with a  table  of  inputs  and  outputs  of the
-routines.

The area-weighted latitudes for the regions  are based on the following model
segment  groupings:
                     ^                                           •
      Region          Segments                              Latitude

         1             10  20 30                               41.75
         2            40  50 60 70  90  100                     40.9
         3             80  110  120 140                         40.8
         4             160 170 175  190 200                    39.25
         5             180210220230330340                39.75
         6             265 270                               39.1
         7           .235 240 250  260 280  290  300 310        37.8

2.2.4  Potential  Evaporation

Daily and  monthly totals of potential evaporation from the regional data sets
were compared to observed data from the pan evaporation stations located in the
Chesapeake Bay watershed.  No definitive relationships were observed between the
regional and  recorded data in these  comparisons.  However, comparisons between
the individual  regional data sets showed that the .daily values are related to
each other by monthly factors.   Since the evaporation data were not derived from
observed pan data, but are apparently related .to each other, it was hypothesized
that the data for one region were computed using an evaporation model; and then
the other  regions were  obtained  from the computed time  series and long term
relationships between the regions. This hypothesis was reinforced by comparison
of. the Region 5 data to daily pan evaporation data that was estimated using the
Penman equation.  Dewpoint, air temperature, wind speed, and solar radiation data
from Region 5 were used as parameters in the Penman,model.   When these estimated
values were  compared with the Region  5 data  set,  it  was  found that the daily
values were related by. constant monthly factors  for the years 1974-76.

The Penman (1948) method is a simple, commonly-used algorithm  for estimation of
pan evaporation.   The  method, as  described by Kohler et al. (1955), uses daily
inputs of  solar radiation (Langleys), wind speed (miles/day),  dewpoint (F), and
average  air temperature  (F).  The  average  air temperature is defined here as the
mean of  the maximum and minimum  daily values.  The output from the model is daily
pan  evaporation  in  inches.    This  algorithm  was implemented in a  FORTRAN
subroutine entitled PANEVP  that  reads the appropriate input  data  files,  and
outputs  the daily data in the HYDDAY  format.  The PANEVP subroutine is listed in
Appendix D.5.                               .

                                       27

-------
Development  of  Che  regional  potential  evapotranspiration  data  sets  was
accomplished using the following two-step procedure:


      1.  Regional daily pan evaporation totals were computed using the Penman
          method as implemented in the subroutine PANEVP.

      2.  Factors were developed to account for the pan coefficient and apparent
          overprediction by the Penman  method during  winter months.   These
          factors were applied to the Penman data  to obtain the final potential
          evapotranspiration.            .

2.3  HOURLY PRECIPITATION DATA

Significant effort was devoted to development of the hourly rainfall data for the
watershed model because this data is the principal driving force for the model
and because of the quantity of data involved.  This task Included three principal
subtasks.   First,  a  large amount  of  observed hourly ..and daily data  was
reformatted from NOAA tapes  to WDM format using the ANNIE program and associated
utility software. Second, an iterative procedure, using CIS software (ARC-INFO),
coordinate locations of the  observed stations,  and digitized model segments was
used to develop the precipitation segmentation and weighting factors needed to
combine the data.   Third,  the observed data were combined  and  distributed to
hourly values utilizing software adapted from the procedures used to generate the
1974-78 data base.  The resulting data base consists of 32 data sets of generated
hourly rainfall  corresponding  to individual model segments  or,  in some cases,
several aggregated model segments.

2.3.1  Development of'Observed Rainfall Data Base

This task involved reformatting the  1984-85 and 1986-1987 daily and hourly data
for all rainfall stations located within or near the basin to WDM format (Figure
2.4 and Figure 2.5).  This data was obtained from standard NOAA tape files for
each of the states  in the basin, i.e.,  NY, PA, W, DE,  MD,  and VA.   The ANNIE
program and several additional stand-alone utility programs were employed in this
effort, which involved reading and obtaining an inventory of the NOAA tape files,
selecting stations to extract based  on coordinate  locations, breaking the large
data files into smaller files,  and transferring the data to a WDM file.  Details
of the procedures and software are documented  in Appendix D.6.

2.3.2  Basin Segmentation and Development of Weighting  Factors

A Thiessen analysis was performed to determine the weighting factors to use in
the rainfall generation.  The  stations  and Thiessen  weights from the original
1974-78  rainfall generation  were  reviewed,   and  it  was decided to  redo  the
analysis since several of the original stations were no longer available, and the
model  areas and segments  associated with several of  the  generated rainfall
records were  not documented.   Also, it was  likely  that the  availability of
additional observed stations would improve the analysis.

The final  1984-85  Thiessen analysis was based on the  digitized model segment
boundaries and the coordinate  locations (latitude and longitude) of a selected

                                      28                        .

-------
  A.F L   Daily   and  Hourly
Precipitation  Stations
                           HOURLY PRECIPITATION STATIONS
                           DAILY PRECIPITATION STATIONS
  Figure 2.4 AFL daily and hourly precipitation stations..
                28.1

-------
   BFl   Dai ly   and  Hourly
 Precipitation  Stations
                               HOURLY PRECIPITATION STATIONS

                               DAILY PRECIPITATION STATIONS
Figure 2.5 BFL daily and hourly precipitation stations.

                '28.2

-------
subset  of  the observed  stations.    It was  performed  using the  ARC-INFO CIS
software.  The outputs  from the procedure were  the  observed stations and the
corresponding area-based weights for  each model segment.  The procedure required
a sequence  of steps as described in Appendix D.7.  The resulting Thiessen network
consisted of 32 separate rainfall segments and utilized a total of 173 individual
stations.

When  the 1986-87 data were being generated,  it was  discovered  that several
stations included in the 1984-85 Thiessen network had been discontinued or had
significant  missing  data.    Therefore, these  stations were removed  from the
network and their Thiessen weightings were added to the nearest station in the
set of stations for that  segment.  The differences between the 1984-85 and 1986-
87  Thiessen  networks  are   documented in  Appendix  D.7.   Generally,  these
differences are very small.
                    ^                                               -
2.3.3  Generation of Weighted Hourly Rainfall Records

The final step in creating the hourly  precipitation was to run the PRECIP program
to generate  the  data using  the  Thiessen network and the data base of observed
hourly and daily stations.   The  PRECIP program is  a FORTRAN program designed to
generate an hourly time series of rainfall that is a weighted average of several
observed time series.  For  each execution of the program, which generates one
hourly time series,  the following inputs are required:

      •   Between one and ten observed time series of hourly rainfall
      •   Between zero and  ten  observed time series of daily rainfall
      •   Weighting factor for each of the above time series (sum of all factors
          must be 1.0)

The program then proceeds daily  through the following steps to produce an hourly
time  series:                                   .           -      .

      1.  Computes the daily totals  for all input (observed)  time series (daily
          and hourly).   (Data are obtained from the WDM file.)

      2.  Computes the weighted daily total rainfall using the weighting factors
          and the daily totals computed in Step  1.   (Note that  any station
          having missing data during  the day is eliminated from the computations
          on that day, and the weighting factor for that  station is distributed
          among the  remaining stations in proportion to their original weights.)

      3.  Compares the weighted daily total with each of the daily  totals of the
          observed hourly  time  series.   Selects the  observed  hourly station
          having the closest daily total to  the weighted  daily  total.

      4.  Computes the generated hourly time series by distributing the weighted
          daily  total using the hourly distribution of the station selected in
          Step 3.        .

      5.  Stores the resulting  hourly time series in the  WDM file.
                                      .29

-------
      6.   Writes a summary of  the  data  generation  for  each day that data were
          generated.  If the  program experiences difficulty because of too much
          missing data,  it also writes  an  explanatory message to  the  output
          file, and skips the current day.     .'

During generation of several  of the rainfall records, the PRECIP program skipped
a number  of days on which precipitation occurred  because none of  the  hourly
stations had data on that day.  When the total quantity of skipped precipitation
was judged to be non-negligible, the skipped rainfall was distributed manually
over the day, and the hourly  values were inserted into  the WDM file using ANNIE.

Additional documentation of  the PRECIP program, including operational details,
input formats and a program  listing, may be found in Appendix D.8.
                                       30

-------
                                 SECTION 3.0

                     LAND USE DATA AND WASTE LOAD INPUTS
3.1  LAND USE DATA

This section documents the methods used to provide a 1985 base year land use data
set  for  use in  the•ปWatershed Model.   The 1985  base year was  chosen  to  be
consistent with the 1987 Bay Agreement, and because it was  a recent year that had
sufficient land use information coverage from the different  sources used.  The
Phase  II  Watershed Model  land uses  are forest,  conventional till  cropland,
conservation till cropland, cropland in hay,  pasture land, animal waste areas,
and water areas.  The hay cropland was defined as a separate category in Phase
II and extracted from the total cropland acreages used in Phase I.

There are several methodologies, developed at the state level, for the assessment
of land use.   These  methodologies focus on .different planning and .assessment
issues and at differing levels  of detail. As a consequence, their methods, land
use  definitions,  and base  years  vary  from  state to  state.   Some  excellent
examples  of state level  land use  assessments  include Maryland's  Geographic
Information System (MAGI) , Virginia's Geographic Information System (VIRGIS) , and
Pennsylvania's detailed  County Level  National Resources  Inventory  (NRI) data
base.                 •

Another methodology with partial basin  coverage is  the USGS Land Use and Land
Cover System (USGS LU/LC).. The USGS LU/LC has  consistent coverage for the basin
for the 1972-73 period, except  for the basin area within Hew York.  Other basin-
wide  land use  assessments include  the  U.S.  Census Bureau which has surveyed
agricultural land. use since  the  1800's.   The U.S.  Forest  Service has been
conducting state forestry surveys since the 1940's.  More recently, the U.S. Soil
Conservation Service has  conducted the National Resources  Inventory, a data base
of land use and geographic characteristics,  to the extent possible, consistent
with the objective of obtaining a  basin-wide methodology, all of the above land
use methodologies have been utilized in the development of the 1978 and 1985 data
sets.   •'         '                  •                        •

A consistent methodology of determining land  use  for  the  entire  Bay basin was
developed  which  obtained  particularly detailed  information  on  agricultural
cropland.  The principle sources of information,used provided data on land use
at a county level throughout the basin.  Principle sources were the U.S. Census
Bureau, the U.S Forestry Service,  and  the U.S. Department of Agriculture.  Also
used to advantage were the U.S. Geological Survey LU/LC data for areas of water
(rivers,  lakes,  and reservoirs) and urban land.
                                      31

-------
3.1.1  Land Use for 1978

The first step in the  creation of the 1985 data base was  to review the 1978 data
set used by the early  versions of the model.  The refined 1978 land use data set
provided a base for the update to 1985 land use.

Refinement of the preliminary 1978  land use data was achieved through utilizing
several  existing  sources.    These  sources  include the  series,  Census  of
Agriculture for 1978  and Census of Agriculture for 1982. Volume 1,  Geographic
Area Series (U.S.  Census Bureau, 1980, 1984) published for each state the by the
U.  S.  Department  of  Commerce, .Bureau of  the Census.   The  1978  Census  of
Agriculture provided  consistent, reliable  data on  land  areas of cropland, and
pasture.   Combining  the  major cropland  categories from the  1978  Census  of
Agriculture produced^cropland acreage  figures  by county. Major categories of
pasture land were also aggregated  to obtain  the  total pasture  area.   The 1982
Census of Agriculture was used to determine the total land area of each county
in the basin.                   .

Woodland area was derived from USDA Forest Service Timber Surveys (U.S. Forest
Service, 1974, 1978,  1978, 198Q, and 1982)-  The latest available publications
of the Timber Surveys  were used to obtain  woodland area by county.  Total "other
land" found in the Ag Census was aggregated with the woodland area to create a
land use defined as natural woodlots, planted woodlots,  timber  tracts, cutover
land, and land not harvested, grazed, or used for buildings  or  roads.

Neither the Census of Agriculture nor the Forestry Surveys  quantify  urban land
use.  Urban land area for the preliminary 1978 land use data was determined by
subtracting the cropland, pasture,  and woodland acreage from the total acreage
of each county.   Any  negative differences resulted in  the assumption of zero
urban land area.  Negative differences  were then reconciled to the total county
land area by adjustment of the woodland acreage.

The  resulting  preliminary 1978 land use data  set  was then sent to  the State
Offices of the USDA, Soil Conservation Service for revision and/or confirmation.
All states made revisions.  Each state used its own  methods to revise the county
data set.   Among the methods  used were  work  load  analysis  studies,  National
Resources Inventory data, state planning agencies, and local county field offices
of the Soil Conservation Service.   The  1978 data set after the State  revisions,
is presented in Appendix E.I.  This data  set has four categories of land use on
a county basis, cropland, pasture land, woodland and urban land.
                                               j                 -
Assisting with review and corrections  to the  refined 1978 land use  data base
were:
      David Benner, Asst. State Conservationist, Delaware
      Jeffery  Loser,  State Resource Conservationist, Maryland
      Phillip Nelson,  Asst. State Conservationist,  New. York  .
      Robert Heidecker, State Resource Conservationist,  Pennsylvania
      Willis Miller,  Asst. State Conservationist, Virginia '
      Bruce Julian, State Resource Conservationist, Virginia
      Ken Carter, Soil Conservationist, Virginia
      Dixie Shreve, State Resource Conservationist, West Virginia
      Ernest L. Moody, District Conservationist, District of Columbia

                                      32

-------
 3.1.2   Land Use  for  1985

 To obtain the 1985 land use data, the 1978 data were further processed.  The 1978
 data were sent to the State Offices of the Soil Conservation Service for updating
 to 1985.  The state  contacts reviewed and updated the land use to 1985 as  shown
 in Appendix  E.2.   Pennsylvania land use data  were updated by the  Pennsylvania
 Department of Environmental Resources.

 Assisting with review  and corrections  to the  1985 land use data base were:
       David Benner,  Asst. State  Conservationist, Delaware
       Jeffery Loser, State Resource  Conservationist, Maryland
       Phillip Nelson,  Asst. State  Conservationist, New York
       Robert Heidecker, State  Resource  Conservationist,  Pennsylvania
       Timothy Murphy,  Hydraulic  Engineer,. Pennsylvania
       Bruce Julian,  State Resource Conservationist, Virginia
       Dixie Shreve,  State Resource Conservationist, West Virginia
       Ernest L. Moody, District  Conservationist, District of  Columbia

The 1985 land use data were further  refined as described below.

 3.1.3   Cropland Tillage

 The Watershed Model has  three categories  of cropland; conventional  tillage,
 conservation tillage, and hay land.  Conventional tillage represents fall plowed
 and/or spring  plowed conventional  tilled cropland.    Conservation tillage
 represents those tillage practices  that  result  in a residue cover of at least 30Z
 at the time of planting.                        .

 Tillage information  on a county  level for the 1985 data base  was obtained from
 the Conservation Technology Information Center (CTIC),  West Lafayette,  Indiana.
 The  CTIC  is  a  clearinghouse for  information on  soil conservation  and, in
 particular, cropland tillage  practices.   The  CTIC conducts an annual survey by
 county of acres  of crops grown under different tillage systems.  CTIC  data for
 1985 was the basis for the conventional and conservation cropland distribution
 in the 1985 land  use data set  (CTIC, 1986).

 The CTIC  tillage data were processed  to eliminate  area overestimation due to
 double cropping.   Percent of cropland under conservation tillage was  calculated
 for each county  (Appendix E.3).

 Hay acres were compiled from the  1982 Census of Agriculture from the category of
 harvested, "hay,  alfalfa,  and other tame, small grain, wild,  grass  silage, or
 green  chop*.  The hay  acres were transformed  to a percentage of the Census of
 Agriculture total harvested crop acres,  and the area of hay acres was  determined
 by this proportion applied to  total  model cropland area.

 3.1.4   Water Acres

 Water  acres are defined as the area  in  rivers, creeks,  streams, canals, lakes,
 and reservoirs.  Only nontidal waters of the basin are  considered to  be in this
 land use category.  Tidal  waters  are  included  in the simulation of the 3-D Water
 Quality Model.

                                      33

-------
The water  land use area was  obtained  from the CBPO  Geographical  Information
System (CIS), USGS Land Use/Land Cover  (LU/LC) compiled from NASA high-altitude
aerial  photographs and National  High-Altitude  Photography  Program  (NHAPP)
photographs.  The CIS  was used to directly determine the acres of water in each
segment, with the  exception  of  the  area  of New  York state  which was  not
available.    Data  from  the National Resources  Inventory  (NRI)  were used to
determine acres of water for New York.

3.1.5  Manure Acres

Manure acres is a derived land use which represents the production of nutrients
from manure produced  in a segment.    These acres do not represent  acres of
concentrated  animals,  nor  do they represent  manure piles or manure stacking
facilities, rather the manure acres are used to represent the aggregate of all
of these activities.'

Animal units in each model  segment were estimated from the livestock numbers in
the 1982 Census  of Agriculture.   An animal unit is defined as  1000 pounds of
animal weight.  For the  purpose of this  analysis, this  corresponds to 0.71 dairy
cows, 1  beef cow, 5 swine, 250 poultry  layers,  500  poultry broilers,  or 100
turkeys.   Table 3.1,  modified from Animal  Waste  utilization on Cropland and
Pastureland (U.S.  EPA,  1979)  tabulates the voided manure of  different  animal
types.  The tabulated wet  weight of manure  does not  include  the wet weight of
bedding material, spilled food, or soil.

                     • i       ...          '  •         -•
TABLE 3.1   ESTIMATED QUANTITIES OF VOIDED MANURE FROM LIVESTOCK AND POULTRY.
            NORMALIZED TO l.OOQ POUNDS BODY WEIGHT.   FROM, ANIMAL WASTE
            UTILIZATION ON CROPLAND AND PASTURELAND (U.S. EPA.  1979)

Animal
Type
Dairy
Beef
Swine
Poultry
Poultry
Turkeys

Animals/
Animal Unit
0.71
1.0
5.0
layers 250
broilers 500
100
Tons of



. Wet Manure
Voided
. 14.9
6.7 •
•11.7
9.7
13.1
10.2
P (Ibs)
21
18
37
100
110
84
N (Ibs)
123
61
160
235
390
304
COD (Ibs)
3,340
1,510
2,080
4,353
5,915
4,599
Animal units of poultry,  swine, beef, and dairy were adjusted to account for the
predominant manure handling practices.   Poultry are usually raised in houses.
The houses are  cleaned out completely once a  year  and the manure is spread on
fields.  As the spread manure  is already accounted for   in the  land uses of

-------
conventional cropland, conservation cropland, and hayland, manure from poultry
are not counted in the manure land use.

Likewise, swine are housed during a. significant portion of the year in finishing
sheds.   Manure  is washed daily from  the finishing pens  to  storage ponds and
ultimately applied to cropland as  liquid manure.  As for poultry, spread manure
is  already  accounted  for in the  (mostly)  county  level estimates  of manure
applications to conventional tilled, conservation tilled, andhaylands.  Only the
period that swine  are allowed on ground for breeding and gestation (115 days) are
accounted for as manure acres.

Following Gasman  (1990),  a ratio  of  indoor  weight  to outdoor weight can be
estimated.  Assumptions are 10 sows per boar, an adult pig weight is 350  Ibs, two
litters of 10 pigs per sow per year, and that piglets increase weight from 30 to
250 Ibs in 6 months with an estimated annualized  weight of 140  Ibs per  finished
pig.  Therefore:         '

indoor weight  -  (140 IbWIO finished Digs)+(3SO lb)*(250/365 davsl - 13.6
outdoor weight          (350 lbs)*(l.l adults)*(115/365 days)

Therefore, swine  animal units were adjusted down by the factor of 0.0735,

Animal units of cattle were segregated into two categories.   The first category
is  beef  cattle,  defined here as beef cattle, heifers, heifer calves,  steers,
steer calves, bulls and bull calves.   Animal units of beef cattle are decreased
by  a  factor  of  0.7 to account for the large proportion of cattle that are for
most  of  the  year,  unconfined to  small pastures or  enclosures.    The second
category is dairy cows. No adjustments are made in animal  units of dairy cows.

The total adjusted animal  units were divided by a "compromise  animal  density" of
145 animal units  per  acre,  yielding  the number  of  manure acres.   Manure acres
are listed by model segment in Appendix E.4.

Manure acres were  originally taken from the pasture land use acres.  . As manure
is stored and treated through  the implementation of a control  program', a portion
of  these acres  will  revert  back to pasture acres for simulation purposes.  As
control actions of guttering,  diversions, and manure containment are  not totally
effective, some proportion  of each manure acre will revert to pasture for the
full  implementation of controls and some smaller proportion will remain in the
manure land use to account for control efficiencies less  than  100X.

The overall simulation of manure land  use allows the model to  track agricultural
waste  storage practices,  and  show any  increases  due to  increases in animal
populations or  decreases  due to control actions.

3.1.6  Urban-Land Subcategories

The CIS system was used to differentiate the  urban land into five sub-categories!
The urban land  uses are listed below  as described by Anderson  et al. (1976).

Residential  -  ranging in  density  from very high  in urban cores to  low density
with  units on more then one acre.

                                       35

-------
Commercial  -  including  urban central business  districts,  shopping centers,
commercial strip developments, warehouses, etc.

Industrial  -   including  light  arid  heavy  manufacturing  as  well   as  mining
operations, stockpiles, and spoil areas.

Transportation -  roads,  railroads, airports, seaports,  and facilities associated
with the transportation of water, gas, oil, electricity, and communications.

Institutional - urban parks,  cemeteries,  open land, playgrounds, golf courses,
zoos, and undeveloped urban land in an urban setting.  .

Urban land imperviousness was determined for each model segment based on the five
subcategories of urban land.  The model simulates one urban land use based on the
area-weighted parameter of imperviousness of the above five subcategories.  The
following  lists  the  five  classes  of  urban land with  the reported  range  of
imperviousness and the single  impervious value chosen for use in the model.  The
range of imperviousness, is from  the  EPA  report,  National Urban Runoff Program
(U.S. EPA, 1982) except for the transportation subcategory, which was obtained
from  the  Federal Highway  Administration report,   Retention.  Detention.,  and
Overland Flow for Pollutant Removal From  Highway Stormwater Runoff (FHA, 1988).


                 Reported Range     Impervious Value             .
Land Use        of Imperviousness    Used in Model
residential
commercial
industrial
transportation
institutional
5-60X
65-851
75-851
5-201
27-1001
30X
75X
BOX
10X
50X
Using the proportion of the total urban area in the different subcategories and
the  value   of  imperviousness   described  above,  a   single  area-weighted
imperviousness was determined for the single aggregate urban land use modeled.
Appendix E.5 lists  the percentage of urban land in each of the subcategories and
the calculated imperviousness for the area-weighted total.

The area weighted imperviousness was used to determine the expected urban load
in each model segment. The  degree of imperviousness of the urban area determined
the expected annual urban  load using the method described by Schueler (1987):

L - l(P)(PJ)(0.05 + 0.9)(I)/12]  (C)(2.72)

where:                   '                   /
      L  - expected annual load
      P  - rainfall depth  over an annual period                               .
      Pj - fraction of rainfall  events  that produce runoff
      I  - imperviousness
                                      36

-------
      C  - mean concentration of pollutant based on national NURP average
           (C values were obtained for phosphate,  total phosphorus,  ammonia,
           nitrate, total nitrogen and BOD)

Following Schueler, the annual loads were adjusted  to include baseflow loads and
a  curve  of  expected  annual  urban  loads  was  generated  for   the  range  of
imperviousness encountered.   The urban land uses  in  each model  segment were
calibrated to the expected urban load.

3.1.7  Land Use bv County

The county land use data were converted to a model segment basis. Consistent with
the level of spatial detail of the model,  it is assumed that all land uses are
evenly distributed within a county.   Counties within the Bay basin are shown in
Figures 1-6 (Appendix E.6).

3.1.8  Land Use bv Model Segment

Land uses  by  county are proportioned  by  percent  of the county  in  each model
segment  (Appendix E.7).    The  percent of county  area  in  each segment  was
determined by CIS. With few exceptions, model segments are larger than counties
in  area.   Usually several  counties were aggregated up  to compile the model
segment, data.   Figures 7  through 26 in Appendix E.8 show  the relationship of
model segments to county boundaries throughout the Bay basin. Appendices E.9 and
E.10 list the land uses by model segment for 1978  and  1985, respectively.

3.1.9  Land Surface Cover and Erodibilitv Parameters -  COVER. SLSUR. KRER

COVER represents the fraction of the land surface that is covered by canopy, crop
residue, leaf litter,  etc.  and is subsequently protected from raindrop erosion.
Cover is a one of  the primary determinates of the  generation of  sediment fines
that can be transported by  runoff as part of the  erosion  process.   The twelve
monthly values used in  the model represent the land cover on the FIRST day of the
month.  Cover  is interpolated daily in the model  between the user-defined monthly
cover values.  COVER values for  conventional cropland and conservation cropland
are based on the crop types grown in the segment.   Major crop types  aggregated
from harvested acres  in the 1987 Ag Census  were  used to  obtain unique cover
Values  for conventional  and  conservation cropland  in  each  model  segment.
Aggregated crop  categories  obtained from the 1987 Ag Census include  corn for
grain, sorghum for grain, wheat  for grain, barley for grain, buckwheat, oats for
grain,  rye for  grain,  sunflower  seed,  tobacco,  soybeans,  potatoes,  sweet
potatoes, corn for silage,  sorghum  for silage,  and vegetables.   Section 4.3.3
discusses the development of the COVER values used for the cropland  categories
in  each model  segment.  COVER values for  pervious  land  areas are tabulated by
land use in Appendix E.ll. .

Appendix E.ll also includes the  values  of the SLSUR parameter  for each land use.
SLSUR represents  the -slope for overland flow.   This parameter  influences the
simulation of hydrology and sediment erosion. Land slope data were derived from
the  National  Resources  Inventory  (NRI) data  base.  The  county-based  NRI
distribution of slopes  was  combined with the proportion of different county land
uses  to develop an average  slope for cropland,  woodland and pasture  in each

                                       37                         .

-------
segment.  Urban data were  not  available  from this data set.   The slope of the
urban land was set equal to that of cropland.

The parameter KRER,  a coefficient in the model soil fines detachment equation is
also listed  in Appendix E.ll.   KRER, in conjunction with the COVER parameter
controls the amount of fine sediment detached by  raindrop  impact and is then
available to be  transported  by overland  flow.    It is  usually  estimated  by
assuming it  is equal to  the erbdibility  factor,  K, in the Universal Soil Loss
Equation (USLE).

3.1.10  Crop distributions for Conventional  (CNT)  and Conservation (GST)
        Tillage

Due to computational limitations for modeling at  the scale of  the Bay drainage,
the Watershed Model required the concept of a 'composite crop' representation for
the CNT and  CST categories  in order to evaluate land cover, nutrient application
rates, and expected plant uptake rates.  The  'composite  crop representation'  is
introduced and discussed in Section 4.3.3 as  part of the nonpoint load modeling
for the cropland categories.   In order  to  develop a 'composite crop'  for each
cropland model segment, the crop distribution in the CNT and CST categories  is
needed.  These distributions were developed as follows:

      1.  Total Cropland was determined as the sum of the Phase I CNT and CST
          categories after they had been adjusted for  removal  of the hayland
      \    acres (as discussed in Section 3.1.3 above).
       \                          . •    '  '         •
      2.  The  1987  Ag  census  information   was  used  to  develop  the  crop
          distributions 
-------
                         Corn       Soybeans          Grains

          PA            50.9 X      52.6 X     .       29.6 X
          MD            74.0 X      79.8 X            54.1 X
          VA            67.4 X      66.0 X            50.0 X
          WV            54.0 X      73.6 X            29.1 X
          NY            20.6 X       7.2 X  .           6.6 X

      5.  The acres in each crop/tillage category were then adjusted to maintain
          the same split of  conventional and conservation  acres as the original
          Phase I cropland acres.

      6.'  The Ag Census  information was then used to determine the breakdown of
          corn-grain, and corn-silage acres  (and  percentages)  so  that separate
          parameter values,  especially COVER (see Section  4.3.3), could be used
          for each corn type.

The final 'percentages for the crop distributions for the four major categories
listed above for Conventional and Conservation Tillage are shown in Table 3.2 for
each model segment, both above and below the Fall Line.

3.1.11  Comparison of Forest Land                 .

Forest acreage for 1950, 1978. and 1985  were compiled from the 1950 Timber Survey
Data and  from the 1978  and 1985  County Land  Use Data submitted by  SCS  State
Offices.  Appendix E.12  contains a summary of this data by  county  and the totals
for each state.

The CBPO Cultural Data file for 1950 does not have data for all counties in
the basin.  Forest land  estimates include idle land and wetlands.  This was done
because the model has a limited number of land uses available.
               \          ,                   '
These data give an indication of changes in forest land from 1950 to 1985.  Since
the data for 1950 are limited, however, it is difficult to show
total changes in basin acreage between 1950 and 1978.          .

From the  data available,  forest acreage increased overall  in  the basin  by 8X
between 1950 and 1978.  Forest acreage in the Bay basin decreased by 3X between
1978 and 1985.  These estimates are consistent with previous reports (U.S. EPA,
1983) which ascribe an  increase in forest  land between 1950 and 1978  to a
decrease  in cropland acreage due  to  changes  in agricultural practices.   The
decrease in forest land between 1978 and 1985 may be due,  in part, to increases
in urban land acreage.   .                       .

3.2  POINT SOURCES

Point source inputs to the Watershed Model were developed for  the 1984-87 period.
These  data represent all  loadings from  municipal wastewater and - industrial
facilities  which  discharge  to  channel reaches  in  the  basin.   Point  sources
discharging below the Fall Line were considered to be a direct discharge to the
tidal Bay' and were not  included as part of the Watershed Model input data.


                                       39                                      .

-------
TABLE 3.2  CROP DISTRIBUTIONS FOR CONVENTIONAL AND CONSERVATION TILLAGE FOR
           EACH WATERSHED MODEL SEGMENT
SEGMENT
10
20
30
40
50
60
70
80
90
100
110
120
UO
160
170
175
180
190
200
210
220
230
23S
240
250
260
265
270
280
290
300
310
330
340
ANACOSTIA
BAL.T HAK
BOHEMIA
CHESTER
CHICKAHOM
CKOPTAMK
COASTAL1
COAST ALII
COASTAL4
COASTALS
COAST AL6
COASTALS
COASTAL9
ELIZABETH
GREAT UIC
GUNPOWDER
JAMES
NANSEMOND
N ANT 1 COKE
OCCOOUAN
PATAPSCO
PATUXENT
POCOMOKE
POTOMAC
, RAPPAKANN
SEVERN
UICOMICO
WYE
YORK
CORN
GRAIN
33. n
31. 5X
37. U
52.9X
46.2X
51.3X
49. 7X
42.6X
41. 6X
46.3X
46.1X
49.2X
47.9X
36.3X
35. 5X
40. 1X
37.7X
19. 7X
25. 6X
34. 3X
47.9X
34 .9X
24. 2X
23.8X
30.8X
21. 3X
30.9X
30.SX
24. 5X
9.6X
17.2X
34. 8X
47. 4X
54. U
55. 5X
51. 4X
54. OX
49.2X
25. 6X
31. 2X
37.2X
49.8X
25. 2X
59.4X
52. 7X
24.7X
20. OX
0.2X
21 .4X
49.4X
31. 6X
56. 2X
40.2X
35. 2X
43. 8X
52. 4X
33.3X
31. 5X
22.3X
40. 7X
28.4X
41. 6X
24. 2X
CONVENTIONAL TILLAGE
CORN
SILAGE SOYBEANS GRAINS
34. 3X
54. 2X
45. 7X
5.9X
12.7X.
,19. 6X
11. 2X
18. U
33. 2X
25. 3X
15.5X
24\9X
14.7X
31. 5X
56.5X
35. 4X
28.3X
57.2X
43. 8X
21. 5X
17. U
39.3X
13.5X
2.0X
32.3X
10.2X
67.4X
54. IX
20. IX
23. 1X
24. 1X
10.SX
7.2X
8. OX
S.8X
6.0X
7.U
3.3X
5. OX
2.4X
2.1X
16.0X
O.OX
3.3X
9.4X
0.4X
0.7X
O.OX
O.OX
7.3X
3.U
1.0X
2.2X
27.6X
9.7X
2.2X
0.9X
1.7X
0.6X
3.3X
0.6X
2.5X
2.4X
0.6X
O.OX
0.3X
11. IX
1.SX
4. IX
11 .ex
6.9X
1.2X
2.8X
9.8X
7.5X
7.5X
0.6X
0.8X
1.8X
4. IX
0.8X
3.8X
9.4X
17.5X
8.0X
40.8X
48.2X
16.9X
40.2X
O.OX
O.OX
33. 4X
44. 3X
29.3X
22.3X
1S.2X
13.6X
19.2X
11.5X
15. 6X
21. U
34.6X
27.n
30.SX
13.U
45. ZX
24. 2X
11. n
47.0X
60.2X
S5.0X
41.0X
11. 5X
46.8X
27.3X
17.9X
10.2X
12.2X
25.0X
42.3X
34.9X
43. IX
15.7X
49. OX
26. 4X
44. 8X
31. 4X
. U.3X
16.8X
30. IX
39.7X
24. 9X
27.2X
32.4X
24. OX
• 25. 6X
28.6X
18.SX
29.9X
31. 6X
7.2X
22.7X
29.9X
22.4X
26.9X
34. 9X
17.5X
17.7X
21. 5X
26.1X
20. IX
28.3X
1.7X
15.4X
22.0X
23.0X
29.5X
32.5X
30. 2X
24. 2X
19.5X
31. IX
23. 3X
26.3X
34.8X
38.8X
30.2X
21 .OX
29.6X
13.2X.
26.2X
27.9X
19. IX
- 44.8X
37.6X
31. 7X
18.4X
15.5X
39.7X
27.0X
34. 4X
20. 5X
23.SX
31 .9X
34.0X
40.3X
22. OX
29. 5X
28.6X
TOTAL
100. OX
100. OX
100.0X
100.0X
100.0X
100. OX
100.0X
100.0X
100.0X
100.0X
100.0X
100.0X
100.0X
100.0X
100.0X
100. OX
100. OX
100.0X
100.0X
100. OX
100.0X
100.0X
100.0X
100.0X
100. OX
100.0X
100.0X
100.0X
100.0X
100.0X
100. OX
100.0X
100. OX
100.01
100. OX
100. OX
100.0X
100.0X
100.0X
100.0X
100.0X
100.0X
100.0X
100.0X
100.0X
100. OX
100. OX
100.0X
100. OX
100. OX
100. OX
100.0X
100.0X
100.0X
100. OX
100.0X
100. OX
100. OX
100.0X
100.0X
100. OX
100. OX
100. OX
CORK
GRAIN
42. OX
34. 9X
41 .3X
63. 9X
60.4X
60. OX
58. 7X
52.5X
48.4X
54. 4X
55. U
55.0X
56. U
44.6X
36.8X
45. 8X
44.8X
22 .3X
29.8X
41. 5X
52.4X
38.7X
28.0X
28.4X
34.7X
25.7X
31.2X
33. 2X
28.3X
11. 2X.
20. n
42.5X
53.7X
59. 4X
57.8X
59.6X
58.4X
53. U
27.5X
35. 4X
39.5X
54. 9X
30.7X
S8.4X
59.0X
29.8X
23. IX
0.2X
27.4X
57.5X
36.1X
62.2X
48. 1X
41.2X
•51. 7X
53.6X
32.9X
36.3X
27.9X
49.3X
26.8X
44. 7X
29.3X
CONSERVATION TILLAGE
CORN
SILAGE SOYBEANS GRAINS
42.7X
60. U
50.8X
7.1X
16.6X
23.0X
13.3X
22.3X
38.7X
29.8X
18.6X
27.8X
17.2X
38.7X
58.6X
40.4X
33. n
64.7X
51. OX
26.0X
18.7X
43.SX
15.6X
2.4X
36.4X
12.3X ^
68.0X
58.8X
23.2X
27. 1X
29.0X
12.8X
8.2X
8.8X
6.0X
6.9X
7.7X
3.5X
5.3X
2.7X
2.2X
17.7X
O.OX
3.2X
10.6X
0.4X
0.8X
O.OX
O.OX .
8.5X
3.6X
1.1X
2.6X
32.3X
11. 5X
2.3X
0.9X
2.0X
0.7X
4. OX
0.6X
2.6X
2.9X
0.5X
O.OX
0.4X
14.4X
2.0X
5.2X
1S.OX
9.0X
1.5X
3.5X
12.5X
8.9X
12.2X
1.2X
1.9X
3.9X
6.4X
0.8X
4.4X
14.4X
20.0X
8.3X
44.4X
S4.1X
17.9X
45. 5X
O.OX
O.OX
36.2X
48.7X
33.2X
2S.6X
23.9X
20.7X
27.n
18.SX
23.4X
31 .6X
S1.6X
43.6X
45. OX
17.4X
51.8X
33.0X
18.2X
53.4X
65. 4X
70.3X
49.3X
18.6X
S0.2X
28.4X
29.6X
11.2X
19.9X
35. 4X
56.5X
45. 3X
50.8X
26.4X
64.0X
39.5X
51 .OX
14. 7X
5.0X
7.6X
u. n
21.0X
11. 8X
13.0X
16.2X
11. 3X
12.2X
13.8X
8.4X
14.5X
15.6X
2.n
9.9X
15. IX
12.2X
14.9X
18. IX
8.8X
9.5X
12.0X
15.1X
11. OX
16.5X
0.8X
8. IX
12.3X
13.0X
17.2X
19.2X
14.2X
11. OX
8.4X
15.0X
10.5X
11.8X
15.5X
18.3X
13.3X
10.0X
17.5X
5.4X
12.2X
16.3X
10.7X
29.5X
23.3X
15. 4X
10.2X
8.3X
19.7X
15. 3X
16.9X
8.7X
9.7X
16.SX
20.6X
20.3X
8.6X
13.2X
16.8X
TOTAL
100. OX
100. OX
100. OX
100.0X
100. OX
100. OX
100. OX
100.0X
100. OX
100.0X
100. OX
100. OX.
100.0X
100.0X
100.0X
100. OX
100. OX
100.0X
100. OX
100. OX
100. OX
100.0X
100. OX
100. OX
100.0X
100.0X
100.0X
100. OX
100.0X
100. OX
100. OX
100.0X
100. OX
100.0X
100. OX
100.0X
100.0X
100.0X
100.0X
100. OX
100.0X
100.0X
100.0X
100.0X
100.0X
100. OX
100.0X
100.0X
100.0X
100. OX
100.0X
100. OX
100. OX
100. OX
100.0X
100.0X
100.0X
100.0X
100.0X
100. OX
100. OX
100. OX
100. OX
                                     „ 40

-------
The  Point  Source Atlas  was used as  the  universe of  possible point sources.
Municipal dischargers were  selected based on a flow of 0.4 million gallons per
day  (MGD) or greater.   This criteria  captured more  than 96X of the municipal
point source flow.   The  remaining (approximately) 4X of point source flow was
from numerous small  discharges..

Industrial   dischargers  have  highly  variable   concentrations  of  discharged
nutrients, ranging  from  zero to many times the concentrations associated with
municipal discharges.  Accordingly, Industrial dischargers were included in the
point source data set if the load from the industrial source was equivalent to
the  TN,  TP,  or  BOD load of a  0.4 mgd municipal point  source with secondary
treatment.   Industrial dischargers  were  selected if any of the  following were
true: total  phosphorus load greater than  or  equal to 12 Ib per day; or, total
nitrogen  load greate^r than  or  equal  to  40 pounds per day;  biological  oxygen
demand greater than or equal to  100 pounds per day.

Data for all facilities  that discharge to streams within a model segment were
aggregated to obtain a single  set of point source loads for the corresponding
model reach.  The constituents and units are listed below:

                        CONSTITUENT       UNIT
                        flow             ac-ft
                        BOD              Ibs
                        DO                Ibs
                        ammonia  (NH3)     Ibs
                        nitrate  (N03)     ,lbs
                        qrganic-N         Ibs
                        phosphate  (P04)   Ibs
            .  "          organic-P         Ibs
                        heat             BTU

The  data sets consist of monthly total  loads for each  reach for the  1984-87
period.  It  was determined that monthly values represented the  most appropriate
resolution for the  available data.

Data  for these  .parameters  were  derived  in the  following  manner.   If State
(National Pollution Discharge  Elimination System, NPDES) data were available,
they were used preferentially.   When no State NPDES data were available, data
from the 1985 Point Source Atlas were used.  As a last resort, defaults were
calculated for missing data, as described below.   Permit  Compliance System data
•were found to contain many errors and were not used to  compile  the initial data
set.      •  .• .        •    ••   •            '  . /      ••

Defaults for municipal dischargers were based on  default  concentrations  applied
to the municipal flows.   The record of  flows for municipal dischargers was good,
and  only measured, values of flow were used.   The default value for dissolved
oxygen assumed a dissolved oxygen concentration of 6.25 milligrams per liter for
municipal dischargers.   A flow-weighted load was calculated from the default.'
Tables 3.3 and 3.4 summarize nitrogen and phosphorus  default  concentrations and
chemical-species percentages used when nutrient data were missing from municipal
flows.        '
                                       41

-------
TABLE 3.3 MUNICIPAL NITROGEN DEFAULT PERCENTAGES OF AMMONIA, NITRATE AND ORGANIC
           NITROGEN, AND THE DEFAULT CONCENTRATIONS OF TOTAL NITROGEN.  DEFAULTS
           BASED ON LEVELS OF SECONDARY TREATMENT
Maryland Virginia **
Year
1984
1985
1986
1987
1988
Ammonia/
Nitrate/
Org. N,
percent
50/35/15
50/35/15
50/35/15
50/35/15
50/35/15
Default
Total
Nitrogen
mg/1
18.0
18.0
18.6
18.0
18.0
Ammonia/
Nitrate/
Org. N,
percent
73/11/16
73/11/16
73/11/16
73/11/16
75/09/16
Default
Total
Nitrogen
mg/1
18.7
18.7
18.7
18.7
18.7
Pennsylvania
Ammonia/
Nitrate/
Org.N,
percent
50/35/15
50/35/15
50/35/15
50/35/15
50/35/15
Default
Total
Nitrogen
mg/1
18.5
18.5
18.5
18.5
18.5
**
    Virginia default concentrations and percentages were based on       7
 1  Hampton Roads Sanitation District processes.
Note:  Tabulated point source data for 1988 are npt currently included in the
     Watershed Model.
TABLE 3.4  MUNICIPAL PHOSPHORUS DEFAULT PERCENTAGES OF PHOSPHATE AND ORGANIC
           PHOSPHORUS, AND THE DEFAULT CONCENTRATIONS OF TOTAL PHOSPHORUS.
           DEFAULTS BASED ON LEVELS OF SECONDARY TREATMENT **
Maryland
Year
1984
1985
1986
1987
1988
Default
Ortho-P/
Org. P,
percent
85/15
85/15
85/15
85/15
85/15
Total
Phos.,
mg/1
7.0
7.0
3.0
3.0
3.0
Virginia
Default
Ortho-P/
Org. P,
percent
85/15
85/15
85/15
85/15
81/19
Total
Phos . ,
mg/1
6.4
6.4
6.4
6.4
2.5
Pennsylvania
Default
Ortho P/
Org. P,
percent
85/15
85/15
85/15
85/15
85/15
Total
Phos.
mg/1
8.0
8.0
8.0
8.0
8.0
**  Primary or tertiary treatment level ratio of ortho-phosphate to organic
    phosphate is 45/55.

Note:  The tabulated point source data for 1988 data are not currently included
       in the Watershed Model.
                                      42

-------
The. shift in default values of TP for Maryland in 1985 and for Virginia in 1988
reflects the phosphate ban on detergents that went into effect for Maryland in
December 1985 and for Virginia in January, 1988.                   .

Missing municipal biochemical oxygen demand (BOD) values were replaced with the
average of values for the months  for which data was available.  No defaults were
applied if all BOD data was missing.

A different procedure was developed for industrial load defaults.   Industrial
flows are highly variable and,  when nutrients are discharged, are characterized
by low volume and high concentrations.  In addition, some industrial flows are
very large t such as water used for condenser cooling or hydroelectric generation,
and have no nutrient loads whatsoever.  As a consequence, industrial flows are
assumed to  have zero flow  (and  are,  of course,  not counted  as  surface water
diversions) .  Nutrients from industrial sources were input to the channel as a
load based on the dischargers' reported values.

Since industrial flow is  not quantified, dissolved oxygen loads for industrial
dischargers are not required.   Industrial loads are primarily BOD, P04, organic
P, or NH4.   Industrial loads of nitrate and organic N  are  negligible and not
included.  The total industrial nitrogen load is assumed to be entirely ammonia.
The industrial  phosphorus load is assumed to be 851 ortho-phosphorus and 1SZ
organic phosphorus.

All of  the  data sets generated  as  described above were  then reviewed by the
States of Virginia, Maryland, and Pennsylvania for accuracy.  The reviewed data
sets were then  further processed for  use as model input.

State contacts  who provided input  to the collection and review of  the point
source data are:

Maryland:        Narendra  Pandy
                 Maryland  Department of the Environment
                 2500 Broening Highway
                 Baltimore, Md.   21224

Pennsylvania:    Ken Bartel / Cedric Karper / George Fctchko
                 Pennsylvania Bureau  of Water Quality Management
                 Fulton Building, 12th Floor
  ,               Harrisburg, Pa.  17120

Virginia:        Alan Pollock / John Kennedy / Arthur Butt
                 Virginia  State Water  Control Board
                 2111 North Hamilton  Street
                 Richmond, Va.  23230

The point source dischargers were assigned to appropriate model segments by the
NPDES record latitude and longitude  of each discharger.  Point source loads by
model  segment  were  then reformatted and  stored in  the WDM  files  for  the
appropriate  river  basins.  During model  simulations,  the monthly  values were
divided  evenly  over  all  simulation intervals  (hourly)  in  the  month.   The
following  tables are  included  in Appendix F:

                                      43

-------
    Appendix F.I details the total basin-wide point source loads by year.
    Appendix F.2 displays loads by model segment for each year.
    Appendix F.3 displays loads by basin for each year.
    Appendix F.4 lists the total municipal loads for each parameter by State.
    Appendix F.5 lists the total industrial loads for each parameter by State.
•  Appendix F.6 lists the point source discharges for which data was compiled.

3.3  SURFACE WATER DIVERSIONS              .             ,

Ma j or' surf ace  water diversions  from channel  reaches  are represented  in the
Watershed Model for  the  1984-1987  simulation period.   Data were obtained from
U.S. Geological Survey offices  in  Pennsylvania,  Maryland,  and Virginia (USGS,
-1988) and from the U.S. Army Corps of Engineers report Water Needs Assessment of
the Susquehanna River Basin (COE, 1988).
                    ^
The data were reported in monthly or annual diversions  for  each water user.  If
only annual diversions were available, then equal monthly divisions of the annual
total were assumed.                                      •

Categories of water use  included as surface water diversions were:

domestic water use - water for household purposes.

industrial water use  - water used for .industrial fabrication, processing, washing
and cooling.               -                            .             .

irrigation water use -water used.to assist in growing crops or to maintain
vegetative growth.

public  supply - 'water' withdrawn  by public and  private water  suppliers  and
delivered to groups of users.                         ,                      .

Several  categories  of water  uses were  assumed to  involve  a "once  through"
nonconsumptive use  and were not  included in the model  data  of surface water
diversions.  These categories are hydroelectric,  thermoelectric, and industrial
cooling.  Mining is assumed to be a water  use which provides its own source of
wash or process water and does not involve consumptive use.  Ground water sources
and water that is continually recycled in industrial use or in mineral processing
were also not included in the surface water diversion data.

Commercial water, use is a subcategory of the  public  supply  water use and is not
included  to avoid  double-counting.   National  figures  (USGS  Circular  1004,
Estimated Use of Water in the United States in 1985)  show that  93X of commercial
water  is from public supply (82Z) or  ground water sources  (112).

The  surface water diversions are  assigned to model segments  by the reported
latitude and longitude and. summed for each  segment.  The surface water diversion
data is reformatted and stored in WDM files in time steps of one day and in units
of  cfs.   -The WDM files  are the total surface water diversions  in each model
segment.  Appendix F.7 lists the average annual surface water  diversions in cfs
in each model segment.


                                      44

-------
3.4  ATMOSPHERIC SOURCES

The Watershed Model accounts for  the  atmospheric  deposition of  nitrogen and
phosphorus directly onto water  surfaces for the  1984-87  period of. simulation.
Deposition to water surfaces is explicitly modeled.

Deposition of inorganic nitrogen to land surfaces is explicitly included in the
AGCHEM  land  uses  (conventional cropland,  conservation cropland,  and hayland)
through  the  inclusion of  atmospheric loads  with nutrient applications  of
fertilizer and manure  (see Section 4.3.4).   Atmospheric deposition  to PQUAL land
uses (forest, urban, pasture) is implicitly included by calibration to the annual
loads observed in field measurements.

The wetfall ammonia and nitrate loads for  each model segment were determined by
the use of annual  isopleths produced  by  the National Atmospheric Deposition
Program  (NADP,  1982  - 1987).    The wetfall  nitrate and  ammonia loads  vary
spatially by model.segment with the highest deposition generally in  the northwest
areas of the  bay basin,  (Vest  Branch Susquehanna and Juniata),  and the lowest
deposition in the southeast area of  the basin (lower James, lower Eastern Shore).
The year-to-year variation of the wetfall nitrate  assigned to model Segments for
the 1982-87 period is  about  16X.  The year-to-year variation in wetfall ammonia
is about 44Z  for the same period.

Following Tyler (1988), the  dryfall nitrate loads  are assumed  Co be equal to the
wetfall nitrate.   Accordingly,  the  dryfall nitrate was  set  to the  long term
(1982-1987) average of the  wetfall  nitrate.  The dryfall nitrate  is  spatially
distributed by model segment, but has no year-to-year variation.  Dryfall ammonia
is assumed to be negligible.

Orthophosphate,  organic nitrogen  and organic  phosphorus are not  typically
monitored by  NADP.  Annual  loads of these constituents were  derived  from the.
report'Chesapeake Bay  Program Technical Studies:  A  Synthesis  (U.S. EPA, 1982).
The Synthesis Report gives no indication of year-to-year or spatial variation of
these parameters,  so constant loads  are used for all  years.  Table  3.5 lists the
basin-wide average annual atmospheric deposition  loads.


       TABLE 3.5   BASIN-VIDE AVERAGE ANNUAL ATMOSPHERIC DEPOSITION LOADS
                  (LB/AC/YR)
                            ORGANIC  TOTAL    ORTHO      TOTAL
          YEAR   NH3   N03  NITROGEN NITROGEN PHOSPHORUS PHOSPHORUS
1984
1985
1986
1987
1.95
1.50
1.68
1.70
6.94
6.70
6.86
6.41
6.07
6.07
6.07
6.07
15.0
14.3
14.6
14.2
.143
.143
.143
.143
.566
.566
.566
.566
The  average total  nitrogen  load compares favorably  with other  estimates  of
atmospheric deposition in the basin (Jaworski and Linker, 1991):

                                      45

-------
The data was reformatted as a monthly load to the total model segment water area
and added to the WDM files.   The data is input to the model in the same manner
as the point source  inputs, i.e., the monthly totals are divided evenly over each
hour of the month.   Atmospheric loads by year  and model segment are included in
Appendix F.8.
                                      46

-------
                                 SECTION 4.0

                      NONPOINT LOADING SIMULATION MODULES
                            AND SIMULATION RESULTS
This section discusses the HSPF modules used to simulate the nonpoint loadings
from each of the  land uses considered in the Phase II Watershed Model.  Table 4.1
shows the  land use categories in the  model,  while Table 4.2  lists  the water
quality and nonpoint "loading constituents simulated in Phase I.  Note that TOC
(Total  Organic  Carbon)  and  Chlorophyll A  are  simulated  only as  instream
constituents.

The focus in Phase II  was to provide  a more detailed simulation of agricultural
cropland areas so that a cause-effect relationship could be represented between
the  nutrient  applications  (fertilizer and  manure)  on these  areas and  the
subsequent  runoff loadings  from these land uses. Consequently,  the  empirical
approach  to nonpoint  simulation used  in Phase I was replaced with  the more
detailed mechanistic  approach provided by the Agrichemical Module sections of
HSPF.  Also, a significant element of the Phase II work was the discovery that.
hay cropland was  a major portion of the total agricultural cropland as. defined
in Phase I.  Hay cropland represents SOX of  the  total cropland above the Fall
Line, and 10Z of the cropland below  the Fall Line.  Since'the nutrient loadings
from hay areas  are expected  to be  significantly  lower than  from cultivated
croplands,  the acreage of hay cropland was removed from the total' cropland and
simulated as a separate land use category.

The procedures.used in simulating the nonpoint loadings from both cropland and
non-cropland areas are  discussed in  the  following  subsections,  along with the
estimation of selected input parameters needed for model application at the scale
of the Chesapeake Bay  drainage area.   The loading simulation results for all land
uses are discussed in  the final subsection, Section,4.4.  Further details on the
model equations and algorithms are contained  in the HSPF Users Manual (Johanson
et al. , 1984) and the HSPF Application Guide (Donigian et al., 1984).

4.1   OVERVIEW OF CROPLAND AND NON-CROPLAND SIMULATION

Within  the  HSPF  framework,  the PERLND and IMPLND module sections  are  used to
simulate hydrologic, sediment,  and water quality processes that produce nonpoint
loadings to a stream  from both surface and subsurface pathways. 'The structure
charts  for  the PERLND and IMPLND module  sections are  shown in Figures 4.1 and
4.2, respectively.  Each module section is comprised of subroutines,  or groups
of  subroutines,   that perform the  simulation of individual processes.   For
example, the SNOU module in Figures 4.1 and 4.2 performs the snow accumulation
and melt simulation for both pervious and impervious land areas, while the PWATER
module  in Figure  4.1  performs the hydrologic simulation for all pervious land


                                      47

-------
      TABLE 4.1 LAND USE CATEGORIES
         IN THE WATERSHED MODEL
Simulated  Land Uses:

  Pervious  -
     Forest
     High  Tillage  Cropland
     Low Tillage Cropland
     Hay Cropland
     Pasture
     Urban
  Impervious

  • Animal  Waste
  • Urban

Urban  Categories:
    Residential
    Commercial  and Services
    Industrial
    Transportation
    Institutional
                      48

-------
 TABLE 4.2 WATER QUALITY CONSTITUENTS
   SIMULATED IN THE WATERSHED MODEL
Dissolved Oxygen (DO)

Water Temperature
Sediment
Biochemical Oxygen Demand (BOD)
Nitrate-Nitrogen (NO3-N)

Ammonia-Nitrogen (NH3-N)

Organic Nitrogen

Orthophosphorus (PO4-P)

Organic Phosphorus

Total Organic Carbon (TOO

Chlorophyll A (CHL A)
Total N
Total P
                    49

-------
                         PERLND
                          Simulate
                          a pervious
                          land
                          segment
ATEMP 1 SNOW 1
Correct air S
tempera- s
ture ic

PSTEMPI
Estimate
soil •
tempe'a-
ture(s)



imulate
now and
:e

PWATER 1 SEDMNT!




Simulate Simulate
water , sediment
budget
•
PWTGAsI
Estimate
water
tempera-
ture
gas
and
con-
centrations

PQUAL 1
Simulate
general
quality
constit-
uents


MSTLAY
PEST
NITR
PHOS
TRACER
Estimate
solute
transport







Simulate
pesticides








Simulate
nitrogen








Simulate
phosphorus








Simulate
a tracer
(conserva-
tive)
•' •
                  AGRI-CHEMICAL SECTIONS
                 Figure 4.1  PERIOD structure chart

                              50

-------
                             IMPLND





'
Perform
computa-
tions on a
segment of
impervious
land

ATEMP 1 SNOW I WATER I
(see
module

PERLND)







(see
module

PERLND)





-












Simulate
water
budget
for
Impervious
land
segment










• -

SOLIDS 1 IWTGAS | IQUAL











Accumulate
and remove
solids



















Simulate
water
tempera-
tures and
•dissolved
gas cones.



f













Simulate
quality
constit-
uents
using
: simple
relation-
ships with
solids
and/or
water
yield
Figure 4.2  Structure chart - impervious land-segment application module,

-------
areas.  The specific processes  simulated by  each module are identified in the
figures.

Since  HSPF  has  the ability  to use ซither simple or detailed  (i.e.,  AGCHEM)
approaches  for  nonpoint loading simulation,  different  modules were used for
different land use  categories to combine the simplified simulation approach from
Phase I for  non-cropland areas with the detailed AGCHEM approach for the cropland
categories.    Table 4.3 lists  the  specific HSPF modules used  to  simulate the
individual nonpoint loading constituents from each land use category.

Typically, nonpoint loadings are calculated.in nonpoint models by using one or
more of three basic approaches, which include the following:

       a.  Potency factors and subsurface concentrations
       b.  First-order washoff and subsurface concentrations
       c.  Detailed modeling of land surface and soil processes

In the 'potency factor' approach,  nonpoint loadings  from the land surface are
calculated  as a  function of Che sediment loading rate, which in turn may be
calculated  from the Universal Soil Loss Equation  (Uischmeier and Smith, 1978),
or one of its modifications, or by detailed  modeling of storm runoff and soil
erosion.  The potency factor approach has been used in ~a  number of nonpoinC
loading estimation techniques and models, including Che  NFS model (Donigian and
Crawford, 1976b) which was. Che basis for Che original NVPDC and Phase I CBP
watershed modeling efforts.                .

                1              •                          v
The  'first-order washoff  approach includes  a  daily calculation  of pollutant
accumulation/deposition  on  Che land  surface,  .and  a  subsequent washoff  of
pollutants  for storm events as a first-order  function of Che scorn runoff race.
This approach was originally developed for urban areas where impervious surfaces
provide the major fraction of surface nonpoinC loadings; Che veil-known EPA Scorn
Water Management Model  (SWMM) (Huber and Dickinson; 1988), the Modified NFS Model
used  in the NVPDC  Study,  HSPF, and numerous  other models use Che first-order
washoff approach.              -                                       "   .

Both the potency factor and first-order washoff  approaches are used exclusively
for surface loadings; they rely on user-specified subsurface concentrations to
define the  subsurface pollutant contributions.

The 'detailed modeling' approach involves Che  representation of soil chemical and
biochemical processes that, in  conjunction with hydrologic and erosion modeling,
calculate both the surface and  subsurface nonpoinC loading contributions:  For
modeling  nonpoint  nutrient loadings, this often involves  Che calculation of
mineralization,     nitrification/denitrification,     immobilization,
sorption/desorption, plant uptake,  and other soil nutrient processes as impacted
by nutrient applications (fertilizer and manure) , agronomic practices, hydrologic
conditions,' and soil moisture and  temperature.           •

Although  this approach requires more data and  input than  the other empirical
methods, detailed modeling allows a direct linkage between nutrient application
rates  and agricultural practices, and resulting runoff loadings. The HSPF AGCHEM
modules, CREAMS (Knisel eC al.,  1980), CREAMS-NT (Deizman and Mostaghimi, 1991),

'   .       '                   .      ; '•• 52             •

-------
                                 TABLE 4.3  HSPF NODULES USED TO SIMULATE NONPOINT LOADINGS FROM EACH LAND USE IN PHASE II
en
OJ
Hydrology

Dissolved Oxygen

Water Temperature

Sediment Loading

Organic* Loading

NOj-N

NHj-N

Organic N
       P04-P

       Organic P
                                Conventional
                                   Tillage
PWATER/SNOU

PWTGAS

STEMP

SEDNNT

POUAL

A6CHEN

AGCHEN

AGCHEN

AGCHEN

AGCHEN
Conservation
   Tillage


 PUATER/SNOW

 PWTGAS.

 STEMP

 SEDNNT

 POUAL

 AGCHEN

 AGCHEN

 AGCHEN

• AGCHEM

 AGCHEM
    Hay


PUATER/SNOW

PWTGAS

STEMP

SEDNNT

POUAL.

AGCHEN

AGCHEM

AGCHEM

AGCHEN

AGCHF.N
                                                                            Forest
                                                                Pasture
                                              Urban
                                            Pervious
                                             Urban
                                           Impervious
                                                                                                                                   Manure
                                                                                                                                    Acres
PWATER/SNOW   PWATER/SNOU   PWATER/SNOW   IWATER/SNOW   IWATER/SNOW

PWTGAS        PWTGAS        PWTGAS        IWTGAS

STEMP         STEMP         STENP         IWTGAS

SEDNNT        SEDMNT        SEDMNT          •-             -•

POUAL         POUAL         POUAL         IOUAL

POUAL         POUAL         POUAL         IOUAL

POUAL         POUAL         POUAL         IOUAL
                                                                                 POUAL
                                                                                        POUAL
                            POUAL
IOUAL

-------
 and  other agricultural models use this general approach to varying  degrees  of
 process  representation.

 The PQUAL and IQUAL modules of HSPF allow both the potency factor and first-order
 washoff  approaches to nonpoint loading simulation.  As shown in Table 4.3,  these
 modules  were  used  in Phase  11  to  simulate the  nonpoint  loadings  for all
 constituents from forest, pasture, and urban (pervious and impervious) land uses,
 and  for  the  BOD/organics loading from  the cropland  categories.   The  AGCHEM
 modules  were  used  for nutrient loadings  from the  cropland  areas.   The  nutrient
 loadings from the  animal waste 'manure acres'  were  derived from the  runoff and
 defined  waste concentrations as described below in  Section 4.2.   That section
 also  discusses additional details of the  non-cropland loading  simulation.

 4.1.1 Hydrologic.  Sediment Loading.  DO.  and Water  Temperature Simulation
                     *                         •                    ;
 As shown in Table 4.3, the hydrology, sediment, DO  (Dissolved Oxygen),  and runoff
 water temperature  simulations are based on  the same procedures, utilizing the
 same  HSPF  modules,  for all  pervious . land categories.    The  hydrology for
 impervious  land categories  (i.e.,  urban impervious  and manure acres) uses a
 different module, although the snow simulation is identical.  Since these modules
 are  common  to all  the land uses,  a brief overview of the simulation  approaches
 is provided below  with further details available  in the HSPF User  Manual.

 The hydrologic submodel, PWATER, in HSPF was derived  originally from the  Stanford
 Watershed Model  (Crawford and Linsley, 1966), and subsequent refinements through
 the HSP  (Hydrocomp,  Inc.,  1976) , ARM  (Donigian and Crawford,  1976a; Donigian  et
 al.,  1977),  and NPS  models.   Figure  4.3  is  a flowchart of the .hydrologie
 processes simulated  in  PWATER; the  snowmelt processes are  simulated in the
 separate SNOW modules of HSPF.   PWATER calculates a complete water balance for
 each  land use category  within  the  watershed,  or model segment, by  converting
 input rainfall and evaporation data  into the resulting surface runoff,  changes
 in soil  moisture storages for various portions  of the soil profile,  infiltration
 of water, actual evapotranspiration,  and subsequent  discharge of subsurface flow
 (both Interflow and baseflow) to the stream channel.  During storm  events,
 rainfall is   distributed  among  surface  runoff  and  soil  moisture   storage
 compartments based on nominal storage capacities and  adjusted infiltration rates.
 Between   storm  events,  water  storage in  the  soil  profile  is  depleted  by
'evapotranspiration and  subsurface recharge,  thereby freeing up soil  moisture
 capacity for  rainfall inputs from the next storm (NVPDC, 1983)i

 The SNOW module  requires  input of additional meteorologic timeseries,  that is  in
 addition to  precipitation  and evaporation,  including solar  radiation, air
 temperature,  wind,  and dewpoint.  Energy balance calculations,  in terms of heat
 fluxes,  are performed to evaluate melt  components associated with radiation
 (solar  and terrestrial), heat exchange  due  to  condensation  and convection,
 rainfall temperatures,  and  subsurface  heat.    In   conjunction with the  melt
 calculations,  air  temperatures are used  to  determine  whether  precipitation  is
 falling as  rain or  snow,  and snowpack characteristics  are simulated including
 snow density and compaction, areal coverage, albedo,  evaporation, heat loss, and
 liquid water  storage. The end  result is a continuous simulation of snow depth,
 water equivalent,  liquid water storage, and subsequent release of melt  water  to
 the  land surface.  This melt water release is then received by the PWATER modules

                                       54

-------
Actual  '
  ET  /
                      Potential ET
                      Precipitation
                      Temperature
                      Radiation
                     Wind, Dew point
                 -S/-
                              Snowmelt
     Interception
        Storage
                             Interception
  Subroutine)
       i  r^- '
 (   Input  )

 ("Output  j

' I  Sforage~~~\

 0Decision

  ET-Evapo-
  transpiration
           Lower Zone
             Storage
                               Surface
                               Runoff
                               Interflow
                                                          Overland
                                                            Flow
                                         Upper
                                          Zone
                                        Storage
                                                          Interflow
     Deep or Inactive'
      Groundwater
                                      Groundwater
                                        Storage
                                                                /To
                                                                \  Stream
Figure 4.3 Hydrologic processes simulated by the PWATER module of HSPF.

-------
as input for the soil hydrologic processes shown in Figure 4.3.  The equations
used  in  SNOW are based  on work by  the  U.S.  Army Corps  of  Engineers  (1956),
Anderson and Crawford  (1964),  and Anderson (1968); they  are  identical  to the
equations used in the HSP, ARM, and NFS codels.

For impervious land surfaces,  the IWATER oodule of  HSPF performs the hydrologic
simulation  which includes  only the  processes of  detention  or retention  of
incident rainfall,  evaporation from retention storage,  and overland flow routing
of the rainfall excess.   Retention accounts for surface depressions and ponding
on various impervious  surfaces  (e.g.,  roof  tops,  parking  lots,  road  side
depressions) from which the retained water can then evaporate.  Thus, impervious
runoff simulation would  be analogous to including only the 'interception' and
'overland flow'  portions  of the hydrologic processes  shown in Figure  4.3 (for
pervious surfaces).

The  SEDMNT module  of HSPF performs  the  sediment loading simulation for all
pervious land use  categories,  i.e.,  all land uses except urban-impervious and
manure acres.  Figure 4.4  shows the sediment processes and fluxes simulated by
SEDMNT,  including  detachment  of sediment particles  by raindrop  impact,  net
vertical sediment input (or export),  attachment or  aggregation of fine sediment
particles, washoff of detached sediment, and scour of the soil matrix by overland
flow..  All these processes are simulated in the Watershed Model except for scour
of the soil matrix which requires  information at finer special scale  than was
available for the model segments.

The equations used to produce and  remove sediment are based on the ARM and NFS
Models  (Donigian and Crawford  1976  a,b).   The algorithms  representing  land
surface erosion  in these models were derived  from a sediaent model developed by
Mcshe Negev (Negev  1967) and influenced by Meyer and Wischmeier (196?) and Onstad
and  Foster  (1975).   The equation  which., represents  the scouring of the  matrix
soil, which is not included in ARM or NFS, was derived from Negev's method for
simulating  gully erosion.

Two  of the sediment  fluxes shown in Figure 4.4,  SLSED  and HSVI, are . added
directly  to the detached sediment storage variable  DETS.   SLSED  represents
external lateral input from an upslope land segment.   It is a time series which
the user may optionally specify, this option was not implemented in the Watershed
Model.   NVSI  is a parameter  that represents  any net external additions  or
removals of sediment caused by human  activities or Wind.

Impacts of tillage  operations  on sediment availability are affected in SEDMNT by
re-setting  the Value of DETS to represent an .increased -amount of sediment fines
available for erosion, at  the time of the tillage operation.

The washoff process involves two parts:  the detachment/attachment of sediment
from/to  the soil matrix and the transport of this sediment.   Detachment (DET)
occurs by rainfall.  Attachment occurs only on days without rainfall;  the rate
of attachment is specified by parameter AFFIX.  Transport of detached sediment
is by overland  flow.                                             .
                                       56

-------
 SLSED
 lateral
 input of
sediment
 to sur-
  face
            NVSI
         net  ver tical
          sediment
            input
  WSSD
washoff of
 detached
 sediment
 by water
  DETS
detached
sediment-
 storage
     AFT ix
   sediment
  attachment
                  SOSED
                    total
                  removal
                  of .soil &
                  sediment
                  from sur-
                  face by
                   water
                 DET
              dolachmenl
               of soil by
                rainfall
                 soil matrix
                 (assumed to
                    have  '
                  unlimited
                  storage)  .
                                           scour of
                                            matrix
                                            soil by
                                             water
        Figure 4.4  Flow diagram for SEDMNT section of PERLND application module

-------
Kinetic energy from rain falling on the soil detaches particles which are then
available to  be  transported by  overland flow.   The equation  that simulates
detachment is:

      DET    - DELT60*(1.0 - CR)*SMPF*KRER*(RAIN/DELT60)**JRER           (4.1)
                     i                        '       .
where:                                                                       .
      DET    - sediment detached from the soil matrix by rainfall in
               tons/acre per interval
      CR     - fraction of the land covered by snow and other cover
      SMPF   - supporting management practice factor
      KRER   - detachment coefficient dependent on soil properties
      RAIN   - rainfall in in./interval
      JRER   - detachment exponent dependent on soil properties
                    ^
The variable CR is the  sum of the fraction of the area covered by the snowpack
(SNOCOV), if  any,  and  the  fraction that is covered by anything else but snow
(COVER).  SNOCOV is computed by  section SNOW.   COVER is  a parameter which for
pervious areas will typically be  the  fraction1 of the area covered by ve.getation
and mulch.  It can be input on a monthly basis.

When simulating the washoff of  detached sediment, .the transport capacity of the
overland  flow is estimated and  compared to  the  amount of  detached sediment
available.  The transport capacity is calculated by the equation:

      STCAP - DELT60*KSER*((SURS + SURO)/DELT60)**JSER                   (4.2)

where:                                                 •
      STCAP  - capacity for removing detached sediment in       .
               tons/acre per interval
      DELT60 - hr/iriterval                                          .
      KSER   - coefficient for transport of detached sediment
      SURS   - surface  water storage in inches                            .
      SURO   - surface  outflow of water in in./interval
      JSER   -exponent for transport of detached sediment

When STCAP is greater than the amount of detached sediment in storage, washoff
is calculated by:

      WSSD - DETS*SURO/(SURS + SURO)                                   •  (4.3)

If  the storage  is sufficient  to fulfill  the transport  capacity,  then  the
following relationship  is used:

      WSSD - STCAP*SURO/(SURS + SURO)                                    (4.4)

where:
    .  WSSD   - washoff of detached sediment in tons/acre per interval
      DETS   - detached sediment storage in tons/acre

WSSD is then  subtracted from DETS.
                                      58

-------
 SEDMNT also  simulates  the re-attachment  of detached sediment  (DETS) on  the
 surface due to soil compaction, particle aggregation, etc.  Attachment  to  the
 soil matrix is simulated by merely reducing DETS by a daily rate  input  by  the
 user.   Since the soil matrix  is considered to be unlimited,  no addition  to  the
 soil matrix is necessary when this occurs.   DETS is  diminished at  the start of
 each day that follows a day with  no precipitation by multiplying  it  by  (1.0 -
 AFFIX), where AFFIX is the user defined parameter. This represents a first order
 rate of reduction of the detached soil storage.

 Water temperature of runoff and the DO concentration are calculated by the  PSTEMP
 and PWTGAS modules,  respectively,  for the pervious land  categories.   Actually,
 the PSTEMP module calculates soil temperatures  for  each of the  defined soil
 layers • surface, upper zone,  lower zone, and groundwater zone  -  and the flow
 component originating from that zone  is assumed to  be at the calculated soil
 temperature, except that  the  water tempera'ture cannot be less  than  freezing.
 Thus,  surface runoff occurs at the calculated surface soil temperature,  interflow
 occurs at the calculated upper zone temperature, and groundwater occurs  at  the
 calculated  groundwater  zone  temperature.    In  addition,   these calculated
 temperatures are used to perform adjustments to the input soil nutrient reaction
 rates in the AGCHEM module (see Section 4.3 below)

 For the Watershed Model,  the  surface and upper zone soil temperatures  are based
 on a linear regression with air temperature in  each  time  interval.   The lower
 zone and  groundwater  temperatures are  assumed  to  be the same and  are  input
 monthly values defined by the user; these temperatures  correspond  to  the lower
 root zone and shallow groundwater that contributes baseflow  to the stream.

 The surface soil temperature  and surface runoff  temperature is then used  in  the
. PWTGAS module to calculate the DO concentration of the  overland flow, which is
 assumed to  be at saturation.   PWTGAS  uses the  following empirical  nonlinear
 equation to relate dissolved oxygen at saturation to water temperature (Committee
 on Sanitary Engineering Research,  1960):

       SODOX - (14.652 + SOTMP*(-0.41022 +
               SOTMP*(0.007991 - 0.000077774*SOTMP)))*ELEVGC               (4.5)

 where:
       SODOX  - concentration of dissolved oxygen in  surface  outflow, in mg/1
       SOTMP  - surface outflow temperature in degrees C
       ELEVGC - correction factor for elevation above sea level

 The concentration of dissolved oxygen in the interflow and the active groundwater
 flow cannot be assumed  to be at saturation.   Values for these concentrations  are
 provided by the user.   He may specify a  constant value or 12  monthly values  for
 the concentration of each of the gases in interflow and groundwater.  If monthly
 values are provided, daily variation in values will automatically be obtained by
 linear interpolation between the monthly values.

 For impervious surfaces,  the  procedures  for  runoff water temperature and  DO  are
 identical to those used for pervious surfaces; the  IWTGAS module uses a  linear
 regression to calculate impervious overland flow temperature, which is then used
 with Equation 4.5 (above) to calculate the DO concentration.

                                       59

-------
4.2  SIMULATION OF NON-CROPLAND AREAS

Based  on the  hydrology  and  sediment  loading  simulations,  the  PQUAL module
simulates the nonpoint pollutant loadings, from the non-cropland areas of forest,
pasture, and urban-pervious, while the IQUAL module simulates  these loadings for
the urban-impervious land use  category.   Animal waste contributions from manure
acres are derived from IWATER/SNOW modules and user-defined concentrations, as
discussed in Section 4.2.2 (below).

4.2.1  Forest. Pasture.  Urban Area Procedures

Both PQUAL  and IQUAL allow use  of either  the  potency  factor  or first-order
washoff approach for any of the nonpoint constituents that are simulated.  In.the
Watershed Model, the potency factor .approach was selected for all constituents
from pervious land areas, while the  first-order washoff approach was selected for
all constituents from impervious  land areas within the non-cropland categories.
This was the  approach used in Phase I and is consistent with current nonpoint
modeling procedures.    .                                                    •.

Figure 4.5 is the flow chart for the PQUAL module.  The corresponding diagram for
the IQUAL module is identical to the top half of Figure 4.5,  since pollutant
contributions  from scour,  interflow and groundwater  are  not allowed  from
impervious land areas.

The potency .factor  approach  simply assumes that  the contaminant  loading  is a
direct function of the sediment load, as  simulated by the SEDMNT module in each
simulation timestep.  The relationship is shown as follows:

      WASHQS - VSSD*POTFW

where:
      VASHQS - flux of quality constituent associated with
               detached sediment washoff in quantity/acre per interval
      USSD   - washoff of detached sediment in tons/acre per interval
      POTFtf  - washoff potency factor in quantity/ton .

In Figure 4.5,  the  pollutant  contribution calculated by the  above equation is
shown  as  'associated with detached sediment'.   Potency  factors  indicate the
pollutant strength  relative  to  the sediment removed  from the land surface.
Certain constituents that are strongly adsorbed to sediment particles, such as
metals,  phosphates., organics,etc., actually move with the  eroded  sediment;
whereas, for other  less-highly attached contaminants,  (e.g.,  ammonia, nitrate,
BOD) this approach simply means that sediment loadings are a reasonable indicator
of the expected contaminant load in surface runoff.

Potency factors are specified separately  for each land use category, and can be.
input  on a  monthly basis for pollutants  that demonstrate  a  definite seasonal
variability.    In   conjunction  with the potency  factors,  subsurface  (i.e.,
interflow and baseflow)  concentrations  are needed  to define  the  subsurface
contributions; these concentrations can also be defined on a monthly basis for
seasonal variation.
                                      60

-------




. -







/
I







> <


storage of
OUAL on
surface, (or
direct wash*
oil by
overland
Mow

,1 •
t ' ' 1 ซ '
• storage ol
/QUAL '\
affsoclateq
with soil1 ..
( /Dal r 1 x v'
/
5S5ES9BP>^^**(
storage ol j
QUAL I
associated 1
with 1
Interflow I
aL^Ii^^^j^ai
storage ol
QUAL
associated
with active
groundwater
IwSSM^ftMMi^taJ

s" '
accumulation

direct
washofl
by
overland
Lflowl 30QO

washoll •
JM- '" ' ''l| associated
H storage ol • II with
II OUAL M detached
N associated with D ^v sed- jr
g 	 II
T • • " '"" • ^ ' - (

' <^r ^^^. ftCOAA AAA4

. scour
of QUAL
associated
1 with soil .
matrix


1








total
• . Dull IOW
ol OUAL
trom
surface

^

^^. 100
^^' ^^^ I**W
outflow I -„. .!_...
- 1 r>n AI 1 OU{ HOW
Ol OUAL I _ i /-MI AI
•ซociii.d 1 ฐ'w?,UhA
seSTen, ^erMo,
• • • ^^^ ^K


out now
Of QUAL
wi th
active
ground
l^vvater












ii A


:)



S












1

I •













POQUAL
^
Total
i out Mow
of OUAL



^^^






Figure 4.5  Flow diagram'for PgUAL section of PERLND application module.

-------
In the first-order washoff approach, simulation of surface runoff pollutant loads
involves two components: daily accumulation and depletion (non-runoff related)
of contaminants on the land surface, and washoff by overland flow during storm
events.  The constituent can be accumulated and removed by processes which are
independent of storm events such as street cleaning, decay,  and  wind erosion and
deposition, or  it  can be washed  off by overland  flow.  The  accumulation and-
removal rates can have monthly values to account for seasonal  fluctuations.

When there is surface outflow and some quality constituent is in  storage, washoff
is simulated using the commonly used relationship:

      SOQO - SQO*(1.0 - fXP(-SURO*WSFAC))                                (4.6)

where:              ^              '
      SOQO   - wa,shoff of the quality constituent  from the land
               surface in quantity/acre per interval
      SQO    - storage of available quality constituent oh the surface
               in quant ity/acr^e
      SURO   - surface outflow of water in in./interval
      WSFAC  - susceptibility of the quality constituent to washoff
               in units of I/in.  .               .            '
      EXP    - Fortran exponential function            .

The storage is updated once  a day  to account for accumulation and removal which
occurs independent of runoff by the equation:

      SQO - ACQOP + SQOS*(1.0 - REMQOP)                                  (4.7)

where:                        ;                         .
      ACQOP  - accumulation rate of the constituent,
               quantity/acre per day
      SQOS   - SQO at the start of the interval
      REMQOP - unit removal rate of the stored constituent, per day

The module computes REMQOP and USFAC for this subroutine according to:

      REMQOP - ACQOP/SQOLIM                                    ,          (4.8)

where:
      SQOLIM - asymptotic limit for SQO as time approaches infinity
                (quantity/acre), if no washoff occurs

and                                                                       .

      WSFAC - 2.30/WSQOP                      '        .                   (4.9)

where:                      '

      WSQOP  -rate of surface runoff which results in a 90 percent
               washoff in one hour,  iri./hr
                                       62

-------
Since the unit removal rate  of  the stored constituent (REMQOP) is computed from
two other parameters, it does not have to be supplied by the user.

As with the potency factor approach, subsurface concentrations which can be input
on a monthly basis are used to define the subsurface contributions.

4.2.2  Animal Waste Procedures

To represent  animal  waste contributions within the Watershed Model,  the CBPO
developed estimates of 'manure acres' for each model segment (see Section 3.0)
which are used  to calculate  runoff and associated pollutant loads from animal
waste.   These  areas  are  included in the Watershed  Model  as 'impervious' land
surfaces in order to conceptually model a rain-driven source of pollutants with
constant concentrations.   The available  literature was reviewed  to estimate
reasonable hydrologic parameters and concentrations to use in calculating the
waste contributions.  The following hydrologic parameters are currently used in
the model:

          Retention parameter -     0.35 in.
          Roughness factor          0.15
          Slope         ,      -     2.0 X Above Fall Line
                                    1.0 X Below Fall Line
          Slope Length        -     150 ft.

The most critical of these parameters is the retention parameter which specifies
the surface storage to be satisfied prior to runoff.  Since evaporation occurs
from this storage, it also determines  the total fraction of rainfall  that becomes
runoff.   The value  of  0.35 inches  is consistent with the  limited available
literature from Edwards et al.  (1985, 1983) and Phillips and Overcash  (1976) for
paved animal feedlots; this value results  in about 70Z  runoff of precipitation.
The remaining parameters are consistent with small feedlot  size areas, and they
have no  significant impact on runoff volumes.

The calculated runoff volumes are multiplied by assumed constant concentrations
to calculate  the pollutant  loadings to the stream channel from  the 'manure
acres'.  Initial  values  for concentrations were  derived  from a  review of the
available literature  on animal waste  runoff;  these initial values  were then
adjusted, primarily  reduced, during  the calibration process based on instream
concentrations  and  relative  contributions  from  all sources.    The  .final
concentrations used  in both  Phase I  and Phase II  are listed below, along with
annual per acre loadings corresponding to 25 inches of runoff:  .

                                   Concentration            Annual Loading
                                       (og/D                  (Ib/ac/yr)

          Organic N                     .300                      1698
          Ammonia N                      40                       226
          Nitrate N                      10                        57
          Total N                       350                      1981

          Organic P                      60                       340
          Phosphate P                    10                        57

                            -           63

-------
          Total P                        70                       397

          BOD (ultimate)                700                      3962

Actual segment loadings are based on the actual runoff volumes for each segment;
the  above loadings based on 25 inches  of runoff are  shown to .provide  some
perspective on the general magnitude of loadings associated with animal waste in
the Watershed Model.   These  concentrations are considerably lower  than those
expected  from  feedlots,  primarily because the  manure  acres concept  does  not
represent feedlot  areas, manure piles,  or manure stocking facilities.  . It is
simply  a means  of accounting  for  animal waste  nutrient contributions,  and
allowing  for this contribution to be reduced as treatment and storage programs
are implemented (See Section 3.1.5).

4.3  SIMULATION OF CROPLAND AREAS

The primary objectives  in applying  the AGCHEM modules  for  simulating nonpoint
nutrient  loadings from the cropland areas were as follows:

      a.  Represent  nutrient  (nitrogen  and  phosphorus)  balances,  including
          applications   (fertilizer   and    manure),    crop    uptake,    soil
       ;   transformations,  and runoff loadings for the major cropland categories
          in each model segment.

      b.  Provide a mechanism for investigating.and representing sensitivity to
          climate variations and agricultural practices.

      c.  Provide  a means of evaluating the water quality impacts  of current
          nutrient application  rates and practices,  and projecting  impacts of
          changes  in  rates  and practices  as  potential nutrient  management
          alternatives.

This  section  provides an overview  of the processes simulated by  the  AGCHEM
modules  of HSPF,  and discusses the  general  application  procedures used  in
applying  the modules to the model segments within the Watershed Model.  Details
of the  equations  used are  provided in the HSPF User Manual  (Johanson et  al.,
1984).  Following the overview of AGCHEM processes,  we discuss  some of the key
issues involved in applying the AGCHEM module at the scale of the Watershed Model
segments, including the development of a 'composite crop'  representing a weighted
combination of  the major crops  grown within the Basin, and specification of
current fertilizer and manure application rates, chemical forms, and application
procedures  for  defining model  input.   The  final subsection,  Section  4.3.6,
summarizes the AGCHEM simulation results.

4.3.1  AGCHEM Overview

The AGCHEM modules of HSPF attempt to represent the major  nitrogen and phosphorus
processes occurring within the soil profile  that determine and control the fluxes
of soil nutrients within the soil/plant/terrestrial environment.   Figure  4.6
shows  the soil  nutrient storages,  nutrient application modes,  and  subsequent
movement  of soil nutrients modeled by the AGCHEM module.  Runoff, soil moisture,
and  sediment values calculated  by  the corresponding  sections of HSPF are  used


                                      64

-------
                 .Nutrient
                application
Crop uptake and
denitrification
                Application
                  mode
    soil incorporated
             surface
applied
                 Surface
               storage and
              transformation
                infiltration

                    i
               Upper zone
               storage and
              transformation
                percolation

                    1
                Lower zone
               storage and
              transformation
                                Nutrients with sedment
             (adsorbed organic)

          Nutrients in overland flow
                                       (soluble)
            Nutrients in interflow
          losses to groundwater

                    1
               Groundwater
               storage and
              transformation
           Nutrients in
           Groundwater discharge
                                        To
                                       stream
   Figure 4/6 .Nutrient storages and transport modeled by the AGOHEM
            .module.         :
                                 65

-------
by the AGCHEM module  to provide the moisture storage and transport values needed
for the nutrient  simulation.   Fertilizers,  animal wastes,  plant residues, and
other  nutrient  inputs  (e.g.,  atmospheric deposition,  sludge  application) are
applied in their chemical  form  (i.e., NH<,, N03, P04, Organic  N, Organic P) either
as a surface application or incorporated into the soil.

Figure 4.7 shows the AGCHEM conceptualization of  the soil profile. Nutrients are
stored  in  four  depth  layers:  surface  zone,  upper  zone,  lower zone,  and
groundwater  zone.  . The depths  of each  zone  are defined  by  input parameters
estimated by the model user.   The shallow surface layer is a continuous mixing
zone which  may  be defined functionally as  the  zone  of.interaction of surface
applied chemicals, from which surface runoff, sediment erosion, and associated
chemicals are transported to a waterbody.  The surface zone depth is typically
estimated at about one centimeter,  although values ranging from 0.5 to 2.0 cm
have been used in specific applications with AGCHEM and other agricultural runoff
models.

The upper  zone  extends from the bottom of  the  surface zone to a  depth often
ranging from 10  to 20 cm.   This  usually corresponds to the depth of major tillage
operations and/or incorporation of applied nutrients.  It also defines the depth
used  to calculate the  mass of soil and  nutrients  through  which interflow and
percolating  water pass to reach  the lower  zone  regardless of the method of
chemical application.

The lower zone is the primary source of  plant evaporation,  and  it regulates the
amount of soluble, chemicals that can be introduced into groundwater since the
chemicals must pass through this layer. For agricultural applications, the lower
zone depth is typically set to  the maximum depth of the crop root zone, usually
ranging from 50 to 150 cm.

The groundwater layer represents the depth of shallow groundwater that actively
contributes  baseflow -to  the  stream channel.   It can also be  visualized as a
shallow mixing  depth within the  surface aquifer that controls the  chemical
transformations  and  associated contributions to baseflow  concentrations.   In
reality, this depth  is highly  variable  and  difficult to define; for watershed
modeling purposes, depths of 1. to 3 m are typically used.

In Phase II, the  soil depths for these layers were set to  the  following values
for all model segments:

                                                Bottom Depth
          Soil  Layer.       . Thickness         from Soil  Surface

          Surface               1   cm                .1     cm
    /      Upper Zone           14   cm               15     cm
          Lower Zone          105   cm              120     cm
          Groundwater Zone    152   cm              272     cm

Although variations in the depth of  the  groundwater zone  for each model segment
may be appropriate, and should be considered  in any future studies, the use of
a  constant value was  consistent  with  the  both the  objectives and resources
available for the Phase II effort.

                                      66

-------
   Outflow to
   Stream with:
Layer:
(Ti
     Sediment,
     Surface  runoff

     Interflow
     Ground
      water
                                                               Lower
                                     ?^???T5?JC??!?A5l!3S?^5^5iteซ5?!iFT6'???rS?

  Ground
   water
                Figure 4.7 Soil profile representation by the AGCHEM Module.

-------
4.3.2  Soil Nutrient Processes Modeled by AGCHEM

Within the AGCHEM module, the NITR and PHOS sections perform the reactions and
transformations of nitrogen and phosphorus  compounds, respectfully,, i'n the soil
profile as a basis for predicting the nutrient content of agricultural runoff.
The transformations, nutrient storages, and reaction rates considered by NITR and
PHOS are shown in Figure 4.8.   The nutrient simulation primarily assumes first-
order reaction rates and is derived  from work by Mehran and Tanji  (1974), and
Hagih and Amberger (1974).   The AGCHEM module is a translation and  enhancement
of the soil nutrient capabilities of the  Agricultural  Runoff Management (ARM)
Model (Donigian and Crawford,  1976b;  Donigian et  al., 1977; Donigian and Davis,.
1978).    The  processes  simulated  include  immobilization,  mineralization,
nitrification/denitrification,  plant uptake,  and  adsorption/  desprption (for
which an equilibrium isotherm option is available).
                    *                           '
The nutrient module simulate both nutrient transformations and movement in the
watershed by  water or  sediment.   Transformations of nutrients  determine the
nutrient forms in each soil  layer and  their resulting susceptibility to movement.
Nutrient transport processes are simulated only to the extent that  is needed to
predict r'unoff quality and  quantity.   The model does not consider the generally
secondary movement  of  soil chemicals by concentration  and thermal gradients.
However, it does model the  lateral and downward transport of chemicals by water
from the soil  zones.  Lateral transport of nutrients towards the stream can occur
from the surface  and upper zone storages.  Inflow of nutrients to the  groundwater
zone and nutrient transformations in the groundwater are modeled as  a basis for
predicting the outflow of dissolved  constituents with groundwater flow.

The diagrams in Figure 4.8 show the transformation pathways,  storages, and names
of the reaction  rate input  parameters.  Reaction rates  are input on a per day
basis  for each  soil  layer.    Nitrite  (NO2)  transforms  so  quickly  in  most
agricultural  soils  that it is not considered separately.   The adsorbed phase
represents the nutrients in a complex form along with those adsorbed on the soil:
The plant uptake  rates  (KPL) are input monthly as  a function of the stage of crop
growth.   These monthly uptake  parameters  are adjusted to  represent  the crop
uptake of N and  P from the soil storages and to distribute it. throughout the
growing, season.                                                .

Soil temperature is presently the only environmental factor that is modeled as
affecting the  reaction  rates,  with  the  exception that the transformations are
stopped  entirely at very low  moisture  levels as often occurs  in the surface
layer.   Soil  temperatures  for  the  surface and upper zones  are  determined by
regression equations based on air temperature, while the daily soil  temperature
for the  lower zone  and groundwater  is interpolated from average monthly input
values.  The  input  first-order rate  parameters  are defined as the maximum,  or
optimum, rates at 35 degrees C. .Using the  calculated soil temperature for each
soil  layer  in each model time-step,  the  input rate parameter is corrected for
soil  temperatures below 35 degrees C by the generalized equation:
                                      68

-------
                  N2
        PLNT-N
                         KDNI
         KPLN
                      NO3
                            (+N02)
                             KNI
   NH4 - A
             KADAM
              KDSA
NH4 - S
     Key
                   KAM
N   -  Nitrogen
P   -  Phosphorus
A   -  Adsorbed
S   -  Solution
K   -  Reaction  Rate
      Parameters
*   -  Linear Equilibrium
      Sorptlon  (optional)
         KIM AM
ORG -N
                                          KIMNI
       Nitrogen transformations in NITR module

P04 - A
KDSP*

tfARP*
PLNT - P
4
PO4

KPLP
1/IK4D '
- S

ซ n, nnp P
kTRXP
     Phosphorus transformations in PHOS module

  Figure 4.8  Nutrient transformations simulated by the AGCHEM Module.

                          ,69

-------
      KK - K*TH**(TMP-35.0)                                             (A.10)

where:                                              .
      KK     - temperature corrected  first order  transformation rate in
               units-of per simulation  interval
      K      - optimum first order  reaction rate  parameter
      TH     - temperature coefficient  for reaction rate correction
               (typically about  1.06)
      IMP    - soil  layer temperature in  degrees  C

When temperatures are greater than 35  degrees C, the rate is  considered optimum,
that  is KK is set equal to K.  When the temperature of  the  soil layer is below
4 degrees C or the layer is dry,  no biochemical transformations occur.

The corrected reaction rate parameters are determined every biochemical reaction
interval and  multiplied by the  respective storages as  shown in Figure 4.8 to
obtain the reaction fluxes.  Plant uptake can vary  monthly and can be distributed
between nitrate and ammonium by a user-specified input parameter which designates
the fraction of plant uptake from each  species of N.

Other factors deemed as having a constant influence during the simulation period,
such  as  soil  pH,  are represented by  adjustments  to the input reaction rates.
Also, all input parameters are specified  separately for each soil layer; thus,
variations in characteristics of the  soil prof ile. can be accommodated by using
different reaction rates for the different soil layers.

Although HSPF has been applied to hundreds of watersheds across the United States
and abroad, the detailed agricultural runoff simulation provided by'the ACCHEM
modules has not had the same degree  of application experience.  However, initial
testing and application was performed on. field sites in  Georgia and Michigan as
part of the ARM model development work (Donigian and Crawford, 1976; Donigian et
al., 1977), and subsequent applications  including both field and watershed-scale
sites have been conducted in Iowa (Donigian et al.,  1983;  Imhoff et al., 1983),
Nebraska (Gilbert et al., 1982), Florida (Nichols and Timpe. 1985), and Tennessee
(Moore et al., 1988).
  •                                             •(                       '

4.3.2.1  General Limitations of  AGCHEM

As with any model, AGCHEM has a  number  of inherent  limitations in representing
soil  nutrient processes  that need to  be discussed in order  to  fully appreciate
the model results and their use in the Phase II effort.  Some limitations are due
to current  capabilities of the HSPF code, others are due to the  scale of the
Chesapeake Bay application, while others are a reflection of the current state-
of-the-art  of modeling soil  nutrient processes.   Each  limitation and  its
potential impact is  discussed below:      •

      a.  Ammonia volatilization  is not currently  simulated by the AGCHEM model..
          Volatilization can be  a significant  loss  of nitrogen when applied as
          manure  or  as urea-based  fertilizers.   In order to account for this
          loss,   ammonia  losses  from  manure   applications  were  estimated
          separately .and deducted  from the  total  manure  N application (see
          Section  4.3.4.4).    Subsequent  ammonia losses,  i.e.,  following the

                                      70

-------
    applications,   from  the   soil  were   represented  by   increased
    immobilization/fixation.    Although  this  did  not  eliminate  the
    volatilized N  from  the  soil,  it did reduce soil  ammonia  values and
    subsequent runoff,  and  the  resulting increase  in  soil  organic N was
    minimal.

b.  Urea  is not  modeled as  a  separate  form  of  nitrogen  in  AGCHEM.
    Consequently,  the  urea component  of  .fertilizer  was  assumed  to be
    readily hydrolyzed  to ammonia so that all nitrogen fertilizer in the
    urea form was input  to the model as ammonia (see Section 4.3.4).  This
    is expected to have  very little  impact on  the simulation results, and
    would only affect storms occurring within a few days of a fertilizer
    application.

c. ' Plant uptake is represented as a first-order process in AGCHEM, based
    on the soil nutrient-storages  in the various soil zones.  As described
    above, monthly rates are input by the user for  each soil zone to mimic
    the time-variation  in rpot development and associated plant nutrient
    uptake from the various  soil zones.  The  input*rates are corrected for
    soil temperature and uptake is stopped when moisture levels are zero;
    this primarily occurs in the surface layer.  Thus,  uptake  is a direct
    function  of soil  nutrient  storages  arid  application  rates;  plant
    growth,  root  development,  stresses associated with  environmental
    conditions, etc. are not represented.  Consequently, uptake rates are
    primarily  determined through calibration to  expected annual  crop
    uptake  (see  Section 4.3.S).   The rates and resulting  annual uptake
    need  to be re-evaluated whenever  application rates are  changed in
    order to ensure that'crop needs are satisfied.   •

d..  The  soil  phosphorus cycle  and  processes  represented in  AGCHEM are
    considerably  simplified compared   to  the  complex reality  of  how
    phosphorus behaves  in the  soil environment.   Chemical  fixation and
    precipitation reactions that produce iron  and aluminum complexes with
    phosphate  are  not;  represented, and the  distinction between sorbed
    phosphate and organic. P  is difficult to model,  and somewhat arbitrary,
    since these forms'are not usually reported in field studies .due to the
    difficult analytical problems they present.  In effect,  the sorbed P04
    in  the model  represents  an  'exchangeable  P04'  which  controls  the
    amount  in  solution, while  the organic P effectively represents both
    organic P and tightly-bound P0t not available for desorption.  Section
    4.3.5.3 discusses  the estimation of model parameters based  on this
    view of the phosphorus state variables.  Very few agricultural models
    currently available attempt to represent the complex phosphorus cycle
    in soils; further model development and testing work is needed.

e.  Only  sediment-bound organic  N  and P  are  simulated  by  AGCHEM;  no
    dissolved organics  are modeled.  This is generally  considered to be a
    reasonable  approximation  for surface  runoff contributions,  where
    dissolved organics  are usually small components of  the total nutrient
    load.   In  Phase II,  a BOD  component  from  the  cropland areas is
    included  in the model simulations providing   a  source  of dissolved
    organic material for the instream processes.

                               , 71

-------
       f.  Croo  rotations  are  not  explicitly  represented by  AGCHEM  in the
          Watershed Model.  AGCHEM  is designed to represent a constant annual
          land  cover/land use pattern  for each  model  segment  throughout the
          simulation  period.    Consequently,  crop  rotations  and  associated
          practices  at  the  field scale  are  not directly represented.   The
          implicit assumption is  that, at  the  scale of  a model segment ranging
          from  100 sq. mi: to 1000 sq. mi. or  more, rotations, within *. segment
          will  offset  each other  so that  the  total area  in each land use and
          crop  will remain approximately constant throughout the period.  That
          is, although crops  and practices  on an individual  field may change
          from  year  to year,  the net change  in  crops  and practices  within a
          model segment  will be minimal.

Other  limitations of  AGCHEM,  primarily  related  to  the  large scale  of the
Watershed Model application, are  discussed below in Sections 4.3.4 and 4.3.5.
Recommendations for. improvements to AGCHEM, based on the above limitations, and
refinements  of  the Watershed Model application are discussed in Section 4.3.7.

4.3.3   Composite  Crop  Representation              .       .    '

The Phase II Watershed Model includes 63 individual model segments, comprised of
34, segments above the Fall Line (AFL) and 29 segments below the Fall Line (BFL).
The AFL segments  include simulation of nonpoint  loadings from all the land .use
categories plus the instream flow and water quality constituents, while the BFL
segments provide  the  nonpoint loadings  directly to the estuaries and main Bay
'segments modeled by  the  3-D Water  Quality Model.   In  each segment  .of the
Watershed Model, each of the eight land use categories  (pervious and impervious)
are modeled for flow,  sediment, and all nonpoint loadings; over 500 combinations
of model segments and land use  categories (i.e., 63 segments times 8 land use
categories  equals 504 combinations)  are  simulated.    This is  a  significant
computational burden even with the modern increases in computer processing speed.
Moreover, the computations increased in Phase II due to the addition of  the  'hay'
land use category and the use  of the more detailed computations of the AGCHEM
module for all  the cropland categories.        ,               .

Due  to  this computational  burden and  limitations  inherent in  the. spatial
representation  of land  use  within a model -segment at the scale of the Bay
drainage, it was necessary to develop  the concept  of a 'composite crop' in order
to  evaluate segment•wide parameter values  for  the Conventional  Tillage and
Conservation Tillage categories.   The  'composite  crop' concept is simply  a means
of calculating  selected  parameter values for a model segment by weighting .crop-
specific values by the  relevant  percentages in  each  crop  category.   The crop*
percentages, or crop  distributions,  for both  tillage  land use categories were
developed as described in Section  3,1.10; the percentages shown in Table 3.2 were
used to calculate the  segment parameters.

The key model parameters  affected by the cropping patterns  include monthly land
surface cover  (i.e.,  the COVER  parameter),  nutrient  application  rates, and
expected composite crop  nutrient uptake  values.   This section discusses the
calculation  of COVER  values  for the model segments.  Section 4.3.4  (below)
discusses the development of the  nutrient  application rates and  the associated
                                       72

-------
influence of  che  composite crop  approach,  while Section 4.3.5  describes the
expected crop uptake values  and how they were used  in  the  application of the
AGCHEM simulations.

The COVER parameter represents the fraction of the land surface protected from
direct raindrop impacts by either crop canopy or residue  on the surface.   Its
primary influence in the AGCHEM model is on the calculated sediment loadings for
each land use  category within each model segment.  Monthly COVER values are input
to the sediment model representing the land surface  cover on the first day of
each month;  values for  days during the  month are  linearly interpolated from the
neighboring monthly values.  Figure 4.9 shows the division of the Bay drainage
area into three regions for definition and variation of the monthly COVER values;
the COVER values  for each major  crop for both  Conventional  and Conservation
Tillage, and for the hay cropland  are shown in Table 4.4.  Figures 4.10 and 4.11
show the weighting procedure used  to calculate the composite crop COVER values
for model segment number  100 (within  the Juniata  Basin)  for the Conventional
(CNT) and Conservation (CST) Tillage categories, respectively.

The monthly  COVER values  for CNT, CST,  and Hay  for  all model  segments are
provided in Appendix E.ll, and were estimated as follows:

      a.  Monthly COVER values  for each of  the four  major crop categories and
          hay  were  estimated  for Region  2  in  conjunction with  the  Soil
          Conservation Service  (J. Hannawald, 1990, Personal Communication).

      b.  Values  for Region 1  and Region  3  were estimated by  adjusting the
          expected planting dates  two  weeks earlier for Region 3 and two weeks
          later  for Region 1,  based   on data  from U.S.D.A. Handbook  No.  628
          (USDA.1984)  and  from  State representatives on  the  CBP  Nutrient
          Reduction  Task  Force  (NRTF).   The  resulting  monthly values  were
          reviewed by. the NRTF members.

      c.  The composite crop COVER values for CNT and CST were then calculated
          using the corresponding crop distributions  for each tillage practice,
          as  shown  in  Figures 4.10 and 4.11 for  CNT and  CST in model segment
          100.  The hay COVER values are used directly by the model since it is
          represented as a separate land use category.  The final numbers were
          reviewed by the NRTF.

In reviewing  the  COVER values, please  note the following:

      1.  As noted above,  all COVER values are for the first day of each month,
          and values between months are linearly interpolated.

      2.  There are  only slight differences between the CNT and CST values for
          Corn-Silage, primarily in March and April, since  the stover is removed
          in both practices.

      3.  The values for Grains in June-September are  'dashed' because they are
          not included in the weighting, under the assumption that they will be
          followed by one of  the  other crops.
                                       73

-------
Crop  Cover
Region B ounda r i e s
                         REGION  1
                         REGION  2
                         REGION  3
   Figure 4.9 Crop COVER region boundaries.
               74

-------
           TABLE 4.4  CROP COVER VALUES FOR ALL REGIONS AND PRACTICES
COVER VALUES FOR CONVENTIONAL TILLAGE
CORN GRAIN

REGION 1
REGION 2
REGION 3
JAN.
50
50
50
FEB.
45
45
45
MAR.
0
0
0
APR.
0
0
5
MAY
5
10
45
JUN.
45
50
70
JULY
70
75
88
AUG.
88
93
93
SEP.
93
93
93
OCT.
80
85 ,
90
NOV.
55
70
85
DEC.
55
55
55
CORN SILAGE .

REGION 1
REGION 2
REGION 3
SOYBEAN

REGION 1
REGION 2
REGION 3
GRAINS

REGION 1
REGION 2
REGION 3
JAN.
5
5
5

JAN.
10
10
10

JAN.
40
45
50
FEB,.
4*
4
4

FEB.
5
5
5

FEB.
40
50
65
MAR.
0
0
0

MAR.
4
3
6

MAR.
60
65
80
APR.
0
0
5

APR.
5
4
0

APR.
80
75
90
MAY
5
10
45

MAY
0
0
5

MAY
90
90
90
JUN.
45
50
70

JUN.
3
5
60

JUN.
.. .
--

JULY
70
75
88

JULY
60
65
80

JULY
• ...
, -•-• .
-I-
AUG.
88
93
93

AUG.
82
87
87

AUG.
• --
-'. •
• -
SEP.
93
93
93

SEP.
87
87
87

SEP.

--
--
OCT.
10
10
20

OCT.
65
75
80

OCT.
35
40
45
NOV.
8
8
15

NOV.
15
15
15

NOV.
40
45
50
DEC.
6
6
10

DEC.
15
15
12

DEC.
40
45
50
COVER VALUES FOR CONSERVATION TILLAGE
CORN GRAIN  -
           JAN.  FEB.  MAR.  APR.  MAY
REGION 1    65    60    55    50    50
REGION 2    65    60    55    50    55
REGION 3    65    60    55    50    60
JUN.  JULY  AUG.
 60    70    88
 65    75    93
 70    88    93
SEP.  OCT.  NOV.  DEC.
 93    85    7.5    70
 93    90    80    70
 93    93    85    70
CORN SILAGE
           JAN.  FEB.  MAR.  APR.  MAY
REGION 1     53     335
REGION 2     5     44     4    10
REGION 3     5     5     5     5    45
JUN.  JULY  AUG.   SEP.   OCT.  NOV.  DEC.
 45    70    88    93    10     8     6
 50    75    93    93    10     8     6
 70    88    93    93    20    15    10
                                      75

-------
      TABLE 4.4  CROP COVER VALUES FOR ALL REGIONS AND  PRACTICES  (continued)
SOYBEAN

REGION 1
REGION 2
REGION 3
JAN.
10
15
20
FEB.
9
12
15
MAR.
8
10
10
                             APR.  MAY   JUN.
                               8     8    15
                              10    10    20
                              10    15    60
JULY
60
65
87
AUG.
87
87
87
SEP.
87
87
87 .
OCT.
60
75
80
NOV.
20
25
30
DEC
15
20
25
GRAIN

REGION 1
REGION 2
REGION 3
JAN.  FEB.  MAR.  APR.  MAY
 50    55^   60    75    90
 55    60    70    80    90
 60    65.   80    90    90
JUN.  JULY  AUG.
SEP.  OCT.  NOV.  DEC.
 - -    45    50    50
       50    55    55
       55    60    60
COVER VALUES FOR HAY
HAY

REGION 1
REGION 2
REGION 3
JAN.  FEB.  MAR.  APR.  MAY
 85    85    88    65    65
 85    85    88    93    93
 85   .85    88    93    93
JUN.  JULY  AUG.
 93    93    93
 93    65    65
 93    65    65
SEP.  OCT.  NOV.  DEC.
 88    85    85    85
 73    85    85    85
 73    85    85    85
                                      76

-------
       Percent Coverage - Conventional  Tillage
                    Juniata Segment 100
                    Percent Coverage
   Corn-gr.

   Corn-sll.

   Soybeans

   Grains

   Composite
            % of   o
          Cropland
Corn-gr.
Corn-sil.
Soybeans
Grains
Composite
                                   Time (Months)
   Figure 4.10  Composite crop COVER values for Segment 100 - Conventional.Tillage.

-------
-o
oo
              Percent Coverage -  Conservation Tillage
                           Juniata Segment 100
                           Percent Coverage
           Corn-gr.
           Corn-sil.
           Soybeans
           Grains
           Composite
                   % of
                  Cropland
AugSep
MayJun
        Corn-gr.
        Corn-sil.
        Soybeans
        Composite
                                           Time (Months)
           Figure 4.11  Composite crop COVER valu   *or Segment 100 - Conservation Tillage
                                       '/'

-------
      4.  We've reduced the maximum values (e.g., 96 and  100) which are commonly
          used for some crops to account for borders, non-cropped areas, spatial
          variability  (e.g., dry spots), etc.

      5.  Values on the scale of the size of the model segments will likely be
          somewhat different than those measured in a single field since we need
          to  use average values that are adjusted for variations in practices,
          e.g., 20% fall plowing in CNT, or 75Z of CST areas leave residues.

      6.  Differences of a few percentage points (i.e., 3 to 5) will have little
          or no significant impact on the simulation; the COVER values primarily
          affect the simulated sediment loading rates.   Differences of 10 to 20
          points for specific crops  in  specific months could have a significant
          impact  if the  changes also  result  in similar  differences  in  the
          calculated*Composite Crop COVER values.

4.3.4  Development of Model Segment Nutrient Application Rates
                                                           ซ
The 'key' element in the application of the AGCHEM model to the cropland areas
of the model segments was the development of the nutrient  (fertilizer and manure)
application  rates used as  input to  the model.   This information was developed
over a period of 6 months with continuous interaction and discussions among Aqua
Terra staff,  CBPO,  and the NRTF members.  The outcome  of  this  effort was the
specific values  of  fertilizer  and manure nutrients,  including composition and
methods of application, that were used to'model the' expected nutrient balances
for  each cropland  category,   Moreover, this effort  also  demonstrated  the
relatively high degree of uncertainty in the  information  available in each state
to categorize these rates.   This is a critical issue since any efforts to reduce
nutrient loadings from agricultural cropland must consider current application
rates and procedures as a starting point.  In other words, application rates are
the means of implementing nutrient reduction  from nonpoint source loadings from
cropland.  Thus, accurate,information on current  fertilizer and manure nutrient
applications  is essential  to  any  management approach attempting to  reduce
nutrient loadings from/agriculture.                            •

Development of the model input application rates involved aggregating input from
each  of  the major  CBP states (i.e.,  PA,  MD, VA),   developing  assumptions
appropriate to the scale of the Bay drainage, and calculating.rates corresponding
to the  'composite crop' discussed in Section 4.3.3  (above).   Figure 4.12 is a
flowchart  that  demonstrates  the  information  used  and  steps  performed  in
calculating  the model segment nutrient  application  rates.  The items in 'boxes'
are  the input  data  and assumptions  used,  while   the  'ovals'  represent  the
calculations .performed  in developing the input, rates.  The major data needs and
issues  involved  in  these calculations included the following:

      a.  Fertilizer  and manure application  rates,  procedures,  and timing for
          each major crop,  plus  atmospheric deposition estimates.

      b.  Crop distributions for Conventional and Conservation Tillage.

      c.  Composition  (i.e., organic and inorganic  fractions) of fertilizer and
          manure nutrients.

                                      79

-------
                  Development of  Model
               Input Nutrient  Applications
  Application
  and  Volatil-
  ization Loss
  Assumptions
  Crop
  Distribution
  Percent of
  Cropland
  Receiving
  Fertilizer/
  Manure
  Fertilizer
  and  Manure
  Composition
  Nutrient
  Application
  Timing  and
  Methods
                    Application Rates
                     for Each Crop
                              CBLO
                           Atmospheric
                            Deposition
                             Fertilizer
   Adjust  N
for  Volatilization
    Calculate Composite Rates and
           Crop  Components
      Calculate  Rates  Weighted
     by Percent of Cropland and
          Associated Rates
      Calculate  Composition  of
   Final Fertilizer  and Manure Rates
     Calculate  Timing and
   Placement of Applications
 Model
Segment
 Input
 Values
Figure 4.12  Flowchart for development of model input nufrient application rates.

                                80              .

-------
       d.   Application/volatilization losses of manure nitrogen.

       e.   Model representation of application procedures  and timing.

 The  following  sections  discuss  each  of  these  issues,   except  the  crop
 distributions which were discussed in Section 3.1.10, followed by a description
 of the Lotus spreadsheet used to calculate the application  rates and input values
 for each  model segment.   Appendix C, provides copies of the spreadsheets for all
 model segments; the spreadsheets are also available on diskette  from CBPO.

 In reviewing  the following  sections,  the reader should note that many  of the
 problems  and issues of applying AGCHEM (or any other detailed agricultural runoff
 model) at the scale  of the model segments would not exist  if the model was being
 applied at the field-scale.  At the field level, farmers would be able to specify
 (or data  could be obtained to define)  fertilizer and manure compositions,  timing
 and methods of applications, application rates for individual crops,  etc.  For
 model segments with croplands  approaching hundreds of  square miles,  field and
 farm  level  practices  must  be  adapted  to  define  what   is   'typical'  or
. 'representative' of practices throughout the entire model segment.  Much of the
 effort in developing the application rates and procedures involved the definition
 of  representative  conditions  and practices,  and  resolution  of  conflicting
 information  from  individual  agencies  in  order  to prepare  consistent  and
 compatible model input for all portions -of the Bay drainage.

 4.3.4.1 State 'Supplied application rates

 Table 4.5 lists the fertilizer application rates  for each major crop category for
 each model segment which were developed by the state representatives on the NRTF.
 The corresponding manure application rates are shown in Table 4.6.  while Table
 4.7 shows the estimated percentages for each crop category receiving applications
 of fertilizer only, manure only, and both fertilizer and  manure.         -

 As  noted above, this  information was  initially supplied by  the states  and
 subsequently refined in  NRTF meetings and telephone/fax communications in order
 to clarify how the  rates (and percentages) would be used and interpreted within
 the  framework  of   the   AGCHEM  modeling calculations.    In  reviewing  this
 information, please note the following:

       a.   All  nitrogen   (N)  and  phosphorus  (P)  rates  are  in  terms  of  the
           elemental nutrients, i.e., N and P.  P is calculated as 0.44 x
       b.  Application rates are usually the same for corn-grain and corn-silage,
           with a few exceptions.

       c.  The rates for corn,  soybeans,  and grains were used to  calculate the
           weighted rate  for the composite  crop,  while  the hay rate was  used
           unchanged (except as  discussed below for manure  N volatilization) as
           input to the 'hay' cropland category.
                                       81

-------
       TABLE 4.5  FERTILIZER APPLICATION  RATES SUPPLIED BY THE NRTF (Ib/ac)
MODEL
SEGMENT
10
20
30
40
50
60
70
80
90
100
110
120
140
160
170

175

180
190
CORN
N
120
120
120
120
120
120
120
90
95
95 *
120
120
120
140/31a
150/105b

150/105b

142
150/105b
P
24
24
24
24
26
26
26
37
59
59
44
44
44
22/0*
26/22b
35/26d
* ซ_
26/22b
35/26d
30
26/22b
BEANS
N
34
34
34
34
34
34
34
34
34
34
20
20
20
20
20/18b

20/18b

25
20/18b
P
8
8
8
8
8
8
8
8
8
8
18
18
18
26 b
29/22b

29/22b

19
29/22b
GRAINS
N
34
34
34
34
34
34
34
34
34
34
34
34
34
100
100/80b

100/80b

76
100/80b
P
8
8
8
8
8
8
8
8
8
8
18
18
18
0
35/26b

35/26b

2
35/26b
HAY
N
60
60
60
60
60
60
60
15
0
0
20
20
20
0
36/20b

36/20b

7
36/20b
P
29
29
29
29
26
26
26
35
16
16
26
26
26
79
3Q/1&

SO/IS1*

63
30/16**
200

210
220
230
235

240

250

260

265

270

280

290

300
310
330
340
         35/26
15Q/105b 26/22b   20/18b   29/22b   100/80b  35/26b   36/20b   30/161
         35/26^
144
148
150
150/160ฐ
29
23
22
22/35ฐ
23
20
20
20ฐ
35
31
24
24/29ฐ
43
82
100
100ฐ
21
29
21
31/26ฐ
5
24
60
36/30
-------
       TABLE 4.5   FERTILIZER APPLICATION RATES  SUPPLIED BY THE NRTF (Ib/ac)
                 (continued)
MODEL
SEGMENT
Anacostia
Baltimore
Bohemia
Chester
Coast_l
Coast_4
Coast_5
Coast_6
Coast_8

Coast_9

Coast 11
Chicka.
Choptank
Elizabeth

Gr. Wicomico

Gunpowder
James

Nansemond

Nanticoke
Patapsco
Patuxeht
Pocomoke
Potomac
Occoquan

Rappahannock

Severn
Wicomico
Wye
York
CORN
N
143/25a
160
120
130
130
160
160
180
160

160

158
160/150C
130
160

160

160
160/150C

160

145
160/43*
135/21a
145
144
150

160

160
145
130
160/150a
P
25/Oa
18
18
16
16
35
18
42
35
44d
35j
44d
43
35/22c
44/31d
16
35
44d
3*.
44d
18
35/22c
44/31d
3*.
44d
30
18/0a
35/6a
30
34
22H
31d
35
44d
18
30
16
35/22a
BEANS
N
\ 20
20
20
20
20
20
20
0
20

20

7
20C
20
20

20

20
20C

20

17
20
7
17
13
20

20

20
17
20
20a
P
44
13
18
35
35
29
13
26
29

29

23
29/24c
35
29

29

13
29/24ฐ

29

13
13
11
13
18
24

29

13
13
35
29/24*
GRAINS
N
45
50
80
60
60
100
50
30
100

100

26
100C
60
100

100

50
100C

100

74
50
90
74
95
100

100

50
74
60
100a
P
26
13
18
18
18
26
13
26
26

26

23
26/31c
18
26

26

13
26/31c

26

20
13
20
20
25
31
- _ '
26

13
20
18 -
26/31a
HAY
N
0
30
20
30
30
30
30
0
36

30

7
30/26C
30
30

36.

30
30

30

30
30
30,
30
30
36

. 36

30
30
30
30/36a
P
40
13
26
26
26
25
13
31
25

25

27
25/30=
26
25

25

13
25

25

26
13
13
26
19
.30

25

13
26
26
25/30*
                       44/31ฐ

a - Two fertilizer rates used, manure applied .with both fertilizer rates;
b - Two fertilizer rates used, manure applied with second fertilizer rate only;
c - Two fertilizer rates used, in cases where the second rate separated by a
    slash (/) is not shown means it is same as the first rate;
d -.Phosphorus fertilizer application for corn-silage (Nitrogen application is
    same as Corn Grain)

                                     .83

-------
TABLE 4.6  MANURE NUTRIENT APPLICATION RATES SUPPLIED BY THE NRTF (Ib/ac)
MODEL
SEGMENT

10
20
30
40
50
60
70
80
90
100
110
120
140
160
170
175
180
190
200
210
220
230
235
240
250
260
265
270
280
290
300
310
330
340
   CORN
  GRAINS
   HAY
N
N
N
100
100
100
100
180
180
180
180
150 *
150
180
180
180
109
265
265
107
265
265
104
37
0
0
o •
0
0
0
265
0
0
0
0
111
110
18
18
18
18
32
32
32
32
26
26
32
32
32
27
60
60
23
60
60
22,
9
0
0
0
0
0
0
60
0
0
0
0
26
25
140
140
140
140
140
140
140
140
140
140
180
180
180
0
265
265 .
121
265
265
113
• 37
0
0
0
0
0
0
265
0
0
0
0
111
110
25
25
25
25
25
25
25
25
25
25
32
32
32
0
60
60
25
60
60
24
9
0
0
0
0
0
0
60
0
0
0
0
26
25
140
140
140
140
140
140
140
140
140
140
180
180
180
109
265
265
121
265
265 .
113
37
0
0
0
0
0
0
265
0
0
0
0
111
110
25
25
25
25
25
25
25 .
25
25
25
32
32
32
27
60
60
25
60
60
24
9
0
0
0
0
0
0.
60
0
0
0
0
26
25
0
0
0
0
0
0
0
0
150
150
180
180
180
109
50
50
71
50
50
81
37
0
0
0
0
, o
0
50
0
0
0
0
111
•110
0
0
0
0
0
0
0
0
26
26
32
32
32
27
16
16
16
16
16
18
9
0
0
0
0
0
0
16
0
0
0
0
26
25
                                        84

-------
TABLE 4.6  MANURE NUTRIENT APPLICATION RATES SUPPLIED BY THE NRTF (Ib/ac)
  (continued)
MODEL
SEGMENT
Coast_l
Bohemia
Chester
Wye
Chop tank
N ant i coke
Wicomico
Pocomoke
Coast_4
Coast_ll
Coast_6
Gunpowder
Baltimore
Patapsco
Patuxent
Severn
Coast_5
Potomac
Anacostia
Occoquan
Rappahannock
Coast_8
Gr. Wicomico
York
James
Chtcka.
Nansemond
Elizabeth
Coast 9
CORN
N
173
109
155
170
197 ''
245
241
230 *
0
135
121
109
124
117
114
126
110
67 .
118
.0
0
0
0
0
0
0
0
0
0
P
36
23
38
37
40
51
51
49
0
27
26
25
27
28
29
23
27
19
31
0
0
0
0
0
0
0
0
0
0
BEAN'S
N
173
109
155
170
197
236
246
231
0
135
121
109
124
117
114
126
110
67
118
0
0
0
0
0
0
6
0
0
0
P
36
23
38
37
40
50
52
50
0
27
26
25
27
28
29
23
27
19
31
0
0
0
0
0
0
6
0
0
0
GRAINS
N
173
109
155
170
197
244
238
227
0
135
121
109
124.
117
114
110
110
67
118
0
0
0
0
0
0
0
0
0
0
P
36
23
38
37
40
52
51
49
0
27
26
25
27
28
29
25
27
19
31 ,'
0
0
0
0
0
0
0
0
0
0
HAY
• N
30
20
, 155
170
205
252
253
238
0
135
121
109
124
117
114
126
110
67
118
0
0
0
0
0
0
.0
0
. 0
0
P
26
26
38
37
42
53
54
51
0
22
26
25
27
28
29
23
27
19
31
0
0
0
0
0
0
0
0
0
0
                                        85

-------
TABLE 4.7  PERCENTAGE OF CROPLAND RECEIVING FERTILIZER AND/OR MANURE AS
           SUPPLIED BY THE NRTF

                                    CROP PERCENT (%)
MODEL
SEGMENT
10
20
30
40
50
60
70
80
90
100
110
120
140
160
170
175
180
190
200
210
220^
230
235

240

250
260

265
270

280

290

300
310
330

340

CORN
F M
70 10
70 10
70 10
70 10
70 10
70 10
70 10*
70 15
70 15
65 15
45 25
45 25
4.5 25
77
85
85
70
49
51
68
93
100
79
21b
97
3b
100
93
7b
100
91
6b
98
2b
94
6b
100
100
65

80

F+M
20
20
20
20
20
20
20
20
20
20
30
30
30
12
15
15
30
51*
49a
32
7





•



3a







30
5a
16
4a
BEANS
F M
70 10
70 10
70 10
70 10
70 10
70 10
70 10
60.15
65 15
65 15
45 25
45 25
45 25
100
85
85
73
49
51
76
94
100
79^
21b
97
3b
100
93
7b
100
91
6b
98
2b
94
6b
100
100
70

84

F+M
20
20
20
20
20
20
20
25
20
20
30
30
30

15
15
27
51a
49a
24
6




• ••




3a







30

16

GRAINS
F M-
80
80
80
80
80
80
80
80
80
80
80
80
80
64 24
85
85
70
49
51
64
91
100 ""
79
21b
97
3b
100
93
' 7b
100
91
6b
98
2b
94
6b
100
100
60

75

F+M
20
20 v
20
20
20
20
20
20
20
20
20
20
20
12
15
15
30
51a
49a
36
9









3a







30
10a
16
9a
HAY
F
50
50
.50
50
50
50
50
50
50
50
50
50
50
30
89
89
30
58
58
30
30
100
79

100

100
91
9b
100
91
8b
44
56b
97
3b
100
100
30

30

M F+M








30
30
30
30
30
15
lla
lla
22
42a
42a
24
6


21b






la





<

30

16

                                        86

-------
TABLE 4.7  PERCENTAGE OF CROPLAND RECEIVING  FERTILIZER AND/OR MANURE AS
           SUPPLIED BY THE NRTF  (continued)
MODEL
SEGMENT
Anacostia

Baltimore
Bohemia
Chester
Chickahominy

Chop tank
Coast_l
Coast_4
Coast_5
Coast_6
Cpast_8
Coast_9
Coast_ll
Elizabeth
Gr. Wico.
Gunpowder
James

Nanticoke
Nansemond
Occoquan
Patapsco .

Patuxent

Pocomoke
Potomac
Rappahannock

Severn
Wicomico
Wye
York

CORN
F M
77

83
96
91 2
73
b*
ซ,
84
92 1
100
87
97
100
100
26 9
100
100
71 4
73
27b
31
100
100
82

91

19
94
90w
10b
65
26
85 2
85
15b

F+M
20
3a
17
4
7


16
7

13
3


65


25


69


16
2a
8
1*
81
6


35
74
13


                                     CROP  PERCENT  (%)
                                 BEANS
                               F    M  F+M

                               69  11  20
                                    2   17
                                    1    4
                                    5    7
 81
 95
 88

 73h
 27b
 80      20
 92   1   7
100
 87      13
 96   13
100
100
'26   9  65
100
100
 56  19  25
                               22
                              100.
                              100
                               77

                               92

                               21
                               97
                               90
                               10b
                               65
                               31
                               86
                               85
                               15b
         78
      7  16

          8

         79
          3
         35
         69
         13
                   GRAINS
                 F   M   F+M

                 69  11   20
2
1
 81
 95
 89

 27b
 84
 92   1
100
 87
 96   1
100
100
 39
100
100
 68   7

 73h
 27b
 30
100
100
 82  16

 90   2

 15
 94
 90
 10b
9
2
 75
 23
 77  10
 85
 15b
17
 4
                                                        16
                                                         7

                                                        13
                                                         3
    61


    25


    70


     4

     8

    85
     6
16
75
13
     HAY
  F   M   F+M

 30  20

 30  17
 30   4
 30   7

 27b
 30  14
 30   4
100
 30  13
 30   3
J.QO
100
 38  55
100
100
 30  25
100

 30  34
100
100
 30  16

 30  16
             3
             4
 30
 63

 9V
 10b
 30  35
 30  61
 30  14

 85h
 15b
                                         87

-------
TABLE 4.7  PERCENTAGE OF CROPLAND RECEIVING FERTILIZER AND/OR MANURE AS
           SUPPLIED BY THE NRTF (continued)

                           CROP PERCENT (%)

MODEL            CORN           BEANS          GRAINS
SEGMENT       F   M  F+M      F  M  F+M      F   M  F+M

SEGMENTS WHERE PERCENTAGES WERE DIFFERENT FOR CONSERVATION TILLAGE

160           78      12     100             40  18  12

175
180
210
220
330

340

Baltimore
Chester
Chop tank
Gunpowder
Nanticoke
Patapsco

Patuxent
Pocomoke
Severn
Wicimico

83
66
70
93
65

81

83
91
84
72
32
83

92
19
65v
5
10a
17
, 34
30
7
30
5*
16
3a
17
2 7
16
3 ,25
68
16
la
8
81
35
95

83
73
76
94
70

84

80
90
84
63
28
80

92
23
65
32

17
27
24
6
30

16

3 17
3 7
16
12 25
72
4 16

8
77
35
68

83
60
52
89
50

64

82
85
82
61
23
79

88
100
65
2

17
40
48
11
30
20*
16
20a
1 17
15
18
14 25
77
5 16

4 8

20 16
98
Wye           86   1  13-     86  1   13     80   7  13
a   - Receives manure with the  second fertilizer rate
b   - Receives second fertilizer rate
F   - Crop receives fertilizer  only
M   - Crop receives manure only
F+M - Crop receives both fertilizer and manure
                                        88

-------
      d>  The fertilizer application  rates  for  corn  are relatively consistent
          among  the  various   states  and   model  segments,   whereas  manure
          application rates and fertilizer  rates for  the  other crops are much
          more variable, -even between  neighboring model segments  and within
          model segments that cross state boundaries  (see discussion below).

      e.  In Maryland and Virginia,  different fertilizer rates were provided by
          the states when fertilizer was used alone, and when it was applied in
          addition  to  manure.   The second  (usually  smaller) number  shown in
          Table 4.5 is the rate used when manure was  also applied.

The basic information available to  the  state representatives  on which to base
these application rates varied in detail from state to state.   In general, the
rates for fertilizers,  especially  for  corn,  are  probably more reliable than the
manure  rates.  Much of  the  information  was  extracted from county-level SCS or
Extension Service data, supplemented by 'best estimates' when information was not
available.  In many  cases, estimates were used when a particular crop represented
a minor fraction of the  cropland  in a specific  model segment,  e.g., races for
soybeans in the Upper Susquehanna  were often estimated due tp lack of data, but
they often represented less than 5X of  the cropland.

The rates were then adapted  to model segments prior  to  being .transmitted to Aqua
Terra and CBPO;  the only exception was  for  Virginia  which  provided rates for
subregions within the state which were  then  transformed into the model segment
values  shown  in. the tables.    '.

For model segments  that crossed state  boundaries, a weighing procedure was used
to estimate application rates when the  appropriate information was available from
the adjoining states and the areas were  significant (e.g., more  than 10X of the
model segment).  Table  4.8  shows  .the  estimated  percentages  of the multi-state
model segments in the various  states.  The weighting was  performed for each crop
application rate prior to calculating the model segment composite rate for the
composite crop.  The percentages of croplands receiving  fertilizer, manure, and
both were usually consistent and values  for  one state were adopted.

Although Delaware, New York, and West Virginia include portions of the Chesapeake
Bay drainage  area,  they are not active  members  of  the Chesapeake Bay Program.
Consequently, information was not available  for nutrient applications in these
states  and the rates had to be estimated from the rates  in the other CBP member
states.

The specific  approach and relevant  issues in individual model segments were as
follows:

      Segments 10 and 20: PA rates were  used since application rate  information
      for NY  was not available. Although 67X and 91X of the area of these two
      segments respectively reside in  NY, the cropland (CNT,  CST, Hay) comprises
      20X and 18X,  respectively, of the  model segments.
                                       89

-------
           TABLE 4.8  ESTIMATED PERCENTAGES OF MULTI-STATE SEGMENTS
                      IN VARIOUS STATES                      -      .
MODEL
SEGMENTS        NY         PA         DE         MD         VA         WV
10             67.0%      33.0%
20            ' 91.0%       9.0%
160                       19.0%                 36.0%                 45.0%
170                                                         7.0%      93.0%
175           '      *     13.0%                 19.0%       1.0%      67.0%
180                       36.0%                 23.0%      19.0%      22.0%
200                                                        92.0%       8.0%
210                       23.0%                 77.0%
220                                             33.0%      67.0%
Bohemia                              20.0%      80.0%
Chester                              10.0%      90.0%
Choptank                           •  17.0%      83.0%
Coastal 1                             2.0%      98.0%
Coastal 11                37.0%                 63.0%
Gunpowder                  2.0%                 98.0%
Nantlcoke                            63.0%      37.0%
Pocomoke                              6.0%    /  89.0%       5.0%
Pocomoke                                   „   •  95.0%       5.0%
Potomac                              .           53.0%      47.0%
Wicomlco                              1.0%      99.0%
                                      90

-------
      Segment  160:  MD  rates  were  used  for  the  entire  segment  because  no
      information was available  for UV,  MD comprised the  next largest portion,
      and  MD  provided  detailed  information  on  Z  of  cropland  receiving
      fertilizer, manure, and both.  The overall  impact  should be  small since
      only 3.3Z of the segment is in Cropland  and 6.7Z is in Hay.

      Segments .170  and 175:  Rates  from  the  neighboring  segments  in  VA,
      Shenandoah Segments 190  and 200, were used with some adjustment of the 'X
      receiving manure', because it was  felt that the nature of agriculture in
      these segments  would  be more  similar to  that in  VA than in  the other
      states.   Based on animal  populations reported  by Lugbill  (1990).  the
      percentages of cropland  receiving manure  were reduced to 30Z of the values
     •in Segment 200.  The overall impact of these assumptions  should be small
      since only l.OZ of Segment  170 and 3.0Z of  Segment 175 are  in Cropland,
      while 4.0Z andr5.OX,  respectively, are in  Hay.            '   '

      Segment 200:  VA rates were used for the entire segment since only 8Z Is in
      WV where no data was available.

      Segments 180. 210. and  220: The rates for these segments were weighted
      between  two  states: PA  and MD for Segments  180  and 210, VA and  MD for
      Segment 220.   The segment  rates were calculated from the  percentages for
      each state shown in Table 4.8.  MD information on 'Z receiving fertilizer
      only, and  both  fertilizer  and manure7,  was used for all three segments
      since values were provided  for every model  segment,  and  the  values were
      reasonably consistent with the information from PA  and VA.

      Coastal 11 and Potomac (BFL): Weighting of  PA and MD rates  for Coastal 11,
     . and  VA and MD  rates  for  the  Potomac BFL segment,  was  performed.   MD
      information on percentages receiving fertilizer and/or manure was used.

      Remaining BFL Segments;  For the remaining  multi-state BFL model segments
      shovn  in Table 4.8,  MD data  was  used since  the majority of  the  model
      segment area was contained in MD and no information was available for the
      portions in DE.

Although the assumptions described above are reasonable,  and were  necessary to
complete the AGCHEM  modeling  within the project schedule, any future  efforts
should include an attempt to verify the  application rates and percentages used
especially for areas in the non-member states  of NY, WV,  and DE.

4.3.4.2  Atmospheric deposition

At the request of the NRTF,  atmospheric  deposition of nitrogen  was included in
the nutrient balance of the cropland areas simulated by AGCHEM by  adding in
deposition estimates  as part  of the nutrient  input to the land surface.   The
Phase  I Watershed  Model included  atmospheric  deposition  directly on  water
surfaces as  input  to  the stream  channel  simulation in  each model  segment.   In
Phase Tl,  these  deposition values for nitrogen were also  added  to  the nutrient
input to the cropland areas.   Although the HSPF code does  not currently contain
a  capability  for  daily vet  or  dry deposition values,  the estimated  annual
nitrogen deposition  for each  model  segment 'was  included  as an  addition to the

                                      91

-------
fertilizer/manure applications  so  that it could  be considered  in  the annual
nutrient balance.  Table 4.9 shows  the model  segment values for 1982-87 along
with the annual average; the annual average value for each segment (rounded to
whole numbers) was included in calculating the total nutrient applications to the
cropland areas (see Section 4.3.4.6  below).  The development of the atmospheric
deposition values was described  earlier in Section 3.4; the same values are used
for the  cropland areas and the  deposition onto water surfaces  in  each model
segment.

4.3.4.3  Composition of nutrient inputs

All nutrients that are input to  the  AGCHEM  model must be in one of the specific
nutrient forms  simulated.   As  discussed in Section  4.3.1,  the  nutrient forms
simulated by  AGCHEH  include N03, NH<  (adsorbed  and dissolved),  Organic N, PO^
(adsorbed and dissolved),  and Organic  P.   Thus,  all  fertilizer, manure,  and
atmospheric deposition inputs are defined as one or more of these five nutrient
forms.   Figure 4.13  shows  the  assumed forms  of each  of  these  three nutrient
sources.

For fertilizers,  information provided by  the NRTF  indicated  that  the primary
forms of nitrogen fertilizers, are  applied as urea,  ammonium  nitrate,  and so-
called 'nitrogen solutions' which are typically referred to as UAN  (urea ammonium
nitrate) solutions (Tisdale et al.,  1985).   For all three primary CBP states, UAN
solutions  dominate  the  nitrogen  fertilizer  use.    Similarly,   phosphorus
fertilizers are almost exclusively applied in some .form of phosphate material,
either alone or in a combination with other nutrients such as ammonium or potash.

Based on the above information, all phosphorus fertilizer inputs were assumed to
be in the phosphate (POJ form.  For nitrogen,  we assumed a typical 301 nitrogen
UAN solution  as defined by Tisdale et al.  (1985),  which contains a 50X/25X/25X
distribution  of nitrogen as urea/ammonium/nitrate, respectively.  Since 'urea'
is not a nutrient form simulated byCAGCHEM, all the urea nitrogen was input to
the model in the ammonium (NH4) form..  Consequently,  all fertilizer was input as
752 NH4, and  25X N03.  Tisdale et al.  (1985) discuss the soil reactions (i.e.,,
hydrolysis and biomediated decomposition)  that breakdown  urea in the soil to
ammonia;  they note  that the reactions occur rapidly,  continue .to  occur at
significant rates even at temperatures  of 2ฐ C and lower, and often completely
transform urea to ammonia within a few days  in warm moist soils. Thus, inputting
all urea as ammonia appears to be a reasonable approximation.

For atmospheric  deposition,  the loading estimates developed by  CBPO indicated
that about BOX of the nitrogen was  in  the nitrate form, with the remaining 20X
as ammonia.   Due to  the relatively rapid  nitrification that occurs  in surface
soils, all the deposition was input as nitrate  in the model. Phosphorus loading?
in atmospheric deposition were  assumed to be  insignificant.

For  manure nutrients,  nutrient composition  is much  more variable,  being  a
function of animal type,  diet, storage/handling, season, etc. . The primary issue
in  terms of  modeling is the distribution between the organic  and inorganic
nutrient components  because the inorganic  fraction is partitioned between the
adsorbed  and dissolved forms,  while  the  organic  fraction  is assumed (in the
AGCHEM model)  to be  entirely bound to soil particles and  sediment.   Also, for

                                    .92                        .

-------
TABLE 4.9  ANNUAL NITROGEN ATMOSPHERIC  DEPOSITION LOADS
Average
MODEL SEGMENT
10
20
30
40
50
60
70
80
90 *
100
110
120
140
160
170
175
180
190
200
210
220
230
235
240
250
260
265
270
280
290
300
310
330
340
ANACOSTIA
BALT HARBOR
BOHEMIA
CHESTER
CHICKAHOMINY
CHOTANK
COASTAL 1
COASTAL 11
COASTAL 4
COASTAL 5
1982
8
8
8
8
10
11
9
9
10
10
9
8
8
9
10
10
9
9
9
9
9
8
8
8
8
8
9
8
8
7
7
7
8
8
- 8
8
8
8
8
8
8
8
7
8
.95
.10
.51
.80
.92
.39
.21
.78
.27
.72
.57
.30
.61
.66
.80
.13
.97
.72
.19
.53
.33
.96
.38
.35
.59
.32
.90
.85
.03
.88
.45
.24
.91
.91
.58
.58
.58
.48
.31
.44
.48
,58
.97
.38
1983
8.75
8.77
8.50
8.81
9.89
9.82
8.87
8.67
9.52
9.55
8.34
8.10
7.87
9.45
8.55
8.98
8.61
8.28
8.38
8.11
8.11
7.94
7.50
7^47
7.50
7.43
7.71
7.34
7.13
7.07
6.63
6.56
8.17
8.17
7.83
7.83
7.70
7.66
7.57
7.56
7.60
7.77
7.02
7.63
1984
10.23
9.87
9.39
9.56
11.72
11.78
10.23
10.10
10.94
11.17
9.15
8.78
8.75
10 . 34
9.43
10.00
9.29
8.55
9.26
8.92
8.92
8.75
8.38
8.28
8.38
8.24
8.52
7.88
7.88
7.81
7.17
7.10
8.85
8.85
8.72
8.72
8.72
8.41
7.77
8.11
8.01
8.72
7.44
8.04
1985
9.15
8.91
8.77
9.01
10.15
10.22
9.35
8.87
9.92
9.75
8.40
7.96
7.93
9.52
8.94
9.51
8.87
8.21
8.91
8.44
8.44
8.14
7.63
7.60
7.70
7.56
8.04
7.33
7.20
7.13
6.62
6.62
8'. 23
8.23
7.83
7.83
7.83
7.73
7.63
7.62
. 7.66
7.83
6.88
7.36
1986
8
9
9
9
11
11
9
9
10
10
8
8
8
10
10
11
9
9
10
8
8
8
7
7
7
7
8
7
7,
7
6
6
8
8
7
7
7
7
7
7
7
7
6
7
.70
.65
.38
.22
.51
.65
.83
.15
.06
.70
.81
.10
.07
.87
.43
.67
.76
.02
.19
.78
.78
.21
.70
.47
.84
.43
.72
.47
.34
.20
.76
.62
.24
.24
.83
.77
.77
.46
.23
.36
.66
.77
.95
.29
1987
9
9
8
10
11
11
9
8
11
10
8
7
7
9
8
9
8
7
8
8
8
7
7
7
7
7
7
7
6
6
5
5
7
7
7
7
7
7
6
7
7
7
6
6
.38
.12
.58
.22
.79
.86
.69
.75
.07
.91
.21
.77
.74
.52
.21
.59
.68
.73
.58
.17
.17
.67
.50
.40
.50
.36
.43
.06
.59
.39
.95
.95
.77
.77
.56
.63
.63
.26
.35
.23
.53
.70
.42
.96
Total N
9
,9
8
9
11
11
9
9
10
10
8
8
8
9
9
9
9
8
9
8
8
8
7
7
7
7
8
7
7
7
6
6
8
8
8
8
8
7
7
7
7
8
7
7
.19
.07
.86
.27
.00
.12
.53
.22
.30
.47
.75
.17
.16
.89
.39
.98
.20
.58
.08
.66
.62
.28
.85
.76
.92
.73.
.39
.66
.36
.24
;76
.68
.36
.36
.06
.06
.04
.83
.48
.72
.82
.06
.12
.61
                       93

-------
TABLE 4.9  (continued)
                                                     Average
MODEL SEGMENT   1982  1983  1984  1985  1986  1987   Total N

                                                       7.91
                                                       7.52
                                                       6.63
                                                       6.79
                                                       7.40
                                                       7.98
                                                       6.96
                                                       6.79
                                                       7.58
                                                       7.99
                                                       8.00
                                                       7.60
                                                       7.34
                                                       7.67
                                                       7.45
                                                       7.91
                                                       7.44
                                                       7.60
                                                       7.12
COASTAL 6
COASTAL 8
COASTAL~9
ELIZABETH
GREAT WICOMICO
GUNPOWDER
JAMES
NANSEMOND
NANTICOKE
OCCOQUAN
PATAPSCO
PATUXENT
POCOMOKE
POTOMAC
RAPPAHANNOCK
SEVERN .
WICOMICO
WYE
YORK
8
8
7
7
8
8
7
7
8
8
8
8
8
8
8
B
8
8
7
.51
.34
.13
.60
.28
.54
.67
.60
.38
.54
.54
.38
.28
.41
.31
.51
.31
.38
• 94
7
7
6
6
7
7
7
6
7
7
7
7
7
7
7
7
7
7
7
.77
.40
.93
.99
.53
.80
.13
.99
.49
.80
.80
.63
.39
.67
.30
.77
.43
.36
.20
8
7
7
7
7
8
7
7
7
8
8
7
7
7
7
8
7
7
7
.38
.81
.13
.20
.81
.28
.27
.20
.84
.08
.28
.84
.74
.94
.77
.25
.77
.84
.34
7
7
6
6
7
7
6
6
7
7
7
7
6
7
7
7
7
7
6
.76
.12
.25
.31
.05
.79
.65
.31
.36
.93
.79
.36
.99
.39
.36
.69
.09
.56
.99
7
7
6
6
7
7
6
6
7
7
7
7
7
7
7
7
7
7
7
.70
.19
.38
.58
.12
.86
.92
.58
.43
.86
.86
.29
.12
.46
.16
.70
.23
.29
.06
7.37
7.26
5.98
6.05
6.59
7.60
6.12
6.05
6.96
7.74
7.74
7.09
6.52
7.13
6.82
7.57
6.82
7.16
6.18
                       94

-------
                        Composition of Nutrient  Inputs
                             Nitrogen
                                        Phosphorus
   Fertilizer
25% NO3
75% NH3/Urea
                 100% PO4
   Atmospheric
   Deposition
       100% NO3
V0
in
   Manure
  40% NH3
                   24-28%
       16-12%
  60% Org. N
18%
42%
               Volatilization
               Losses
                    Input  as
                    Organic N
                             30-34%
                          Input  as NH3
            50% PO4
50%  Org. P
                   Figure 4.13  Composition of nutrient inputs to AGCHEM model.

-------
 nitrogen,  volatile losses for the inorganic component  (primarily NH3) can be a
 significant fraction of the total manure nitrogen application.

 A variety  of articles and reports were reviewed in order to estimate a reasonable
 value of percent inorganic nitrogen  for manure  applications to cropland.   The
 following table summarizes the range  and mean of values found in the literature,
 along with the literature source and the manure type:
    Percent
    Inoranic N
 6 to 59 I
24 to 62 X
15 to 75 X
             Mean: 26X
             Mean: 33X
             Mean: 38X
         39 X

         45 X


         37 X


         55 X

30 to 48 X   Mean: 39 X
Manure
Type

Beef
Dairy
Swine

Dairy

Dairy
Dairy. liquid
Swine, liquid

Dairy, Liquid
                                                       Reference
                                                       ReddyeCal., 1979
                                                             "
                                                       Loehr, 1984

                                                       VA DSWC (Flagg, 1990
                                                        personal communication)

                                                       PA DER Manual
                                                        by R.E. Wright (1990)
                                                       Fox and Piekielek,  1990
 Although, a number of the references are general  in nature, the information from
 the  VA Department of  Soil  and  Water .Conservation,  the  PA  Department  of
 Environmental  Resources,  and  the  work  of  Fox and  Piekielek (1990) are  all
 specific to the Chesapeake Bay region.   Based on this review and with primary
 emphasis on beef and dairy wastes, a value of 40X inorganic N was assumed for all
'nitrogen  applied  as  manure.    This  assumption then  provided  the basis  for
 subsequent calculations of volatile ammonia  losses from manure applications , as
 discussed below in Section 4.3.4.4.

 The remaining 60X  organic N fraction in manure is comprised of both unstable and
 stable  organic compounds ,  with the  unstable  portion derived from urea* type
 materials.  Consequently, the unstable organic nitrogen will  become available as
 inorganic  species (primarily NHซ)  through  mineralization processes-: much more
 quickly than the stable fraction.  Based on  information from Banvel (1990), the
 Maryland Department of Agriculture  (F. Samadani, 1990, personal communication,
 and  the Virginia  Department of Soil  and Water Conservation (M.  Flagg,  1990,
 personal communication), about 30Z to 50X~6f the organic N can become available
 to the crop as  inorganic N during the first year; 30X to 40X can become available
 relatively quickly within 3 to 4 weeks.  The remaining portion of the organic N
 is either non-mineralizable or slowly mineralizes over subsequent years;  about
 20X of the organic N is in this latter  category based on the sources noted above.
 Since AGCHEM includes  only one mineralization rate for organic N. we assumed that
 30X of the organic N is  quickly converted to NH\ for  input  to the model with the
 remaining fraction added to the soil organic  N storage.  In Figure 4.13, this 30X
 is shown as the 18X of Total N  (i.e. , 30X x 6. OX Organic N - 18X NHซ) added to the
                                       96

-------
non-volatilized ammonia.

For the  phosphorus  composition in manure, very  little  useful information was
found due largely to the complex chemistry and associated analytical problems in
measuring the various organic and inorganic farms  of P. Sommers and Sutton (1980)
provide a brief overview of some of these  problems, indicating that most of the
P in animal wastes is part of relatively insoluble compounds and that much of the
organic P is not well defined and of unknown chemical  structure.   They cite work
by Townshend et al.  (1969). where swine,  beef cattle, and dairy cattle wastes were
found to have soluble P (expressed as a percentage of Total P) values of .43 X,
54 X,  and  46  X,  respectively.•  Based on this limited data,  we assumed a 50/50
split between the inorganic and organic P fractions of manure  P input to the
model, as shown in Figure 4.13.  Clearly, this split is  not quite  as critical as
the corresponding one for  N,  since PQi does  not undergo volatilization losses
during and following'application to the land.

4.3.4.4  Application/volatilization losses of manure nitrogen     '

Since AGCHEM does not account for volatilization losses of nitrogen, these losses
associated with manure application were estimated and the application rates were
reduced accordingly to subtract the volatilized fraction.  Ammonia volatilization
from manure applications is highly variable, and depends on animal waste type and
characteristics,  meteorologic   conditions  (e.g.,   wind,  air  temperature).
application procedures, etc.  Since the inorganic (primarily ammonia) fraction
is the volatile component,  various literature sources were reviewed in order to
estimate the  fraction of the  inorganic component that  would likely be lost by
volatilization.  The following summarizes the  results  obtained from this review:
      Percent NH3 Loss

  4 to 96 X loss, dairy, surface applied
   (mostly in 60 to SOX range)

  39 X   (incorporated within 24 hrs)

  over 50 X (incorporated within 24 "hrs)

  over 50 X (from surface within 24 hrs)

  over 50 X (loss within 24-48 hrs, incorp.)
Reference

Reddy et al.,1979


Stevens and Logan,  1987

Thompson et al,,  1987

Lauer et al,, 1976

Fox and Piekielek,  1990
Gasman   (1989)  reviewed  a  number  of  literature  sources  on  animal  waste
volatilization  losses associated with  collection, storage and  handling,  and
application, and found that application losses were generally in the range of 10
to 60 X  of Total N.   In a  spreadsheet for calculating animal waste application.
for croplands in Maryland, Banvel (1990) provides the user with options ranging
from  20X  loss  of  NH,  for  incorporation within  1  day,   to  SOX loss  for
incorporation within 5  days.   He also allows  an option for 100X loss (by both
volatilization .and leaching) if the manure is  not incorporated at all, or later
than 5 days.                      .                .
                                      97

-------
Information on manure application procedures were provided by the state members
of the NRTF.  In VA manure is either incorporated after 7  days or not at all.  In
PA, the practice is to incorporate within 2 days about 40Z of the time, and after
2 days about 602 of the  time.  Practices in J1D typically fell between these two
extremes: incorporation within 2 to 5 days.   In discussions with the NRTF, and
considering  the literature  information  cited above,  the following  'percent
ammonia losses' were assumed for each state:

            Pennsylvania            60 Z loss
            Maryland                65 Z loss
            Virginia                70 Z loss

These percent losses were applied to the manure application rates specified by
each  state  as  shown  in Table 4.6;  the  state-supplied rates represented the
amounts applied to the land surface, accounting for any collection, handling, and
storage  losses,  so  that application  and volatilization losses were  the  only
additional mechanisms to be considered.

4.3.4.5  Nutrient application methods and timing

Once  fertilizer and  manure  application amounts  were estimated,  application
practices and procedures were studied in order to define the appropriate timing
and placement of the nutrient applications for input to the AGCHEH model for each
model segment.  'Timing' refers to the period during the year when nutrients are
applied  for  each major  crop,  and 'placement'  refers to how the  nutrients are
applied, usually either surface application or  soil'incorporated.   The  USDA
Handbook No. 628  (USDA.  1984) provided information on usual planting dates for
the major crops; information from .the NRTF, and subsequent discussions with state
representatives, helped to further  refine typical application practices. . Table
4.10  summarizes  the protocols  used  to distribute the  estimated  nutrient
application  amounts,  in both time and space,  for both fertilizer and manure
applications  to each major  crop,  along with  assumptions  on the  atmospheric
deposition component.                .

For fertilizer applications, both surface application and soil incorporation were
considered typical procedures depending  on  whether the application is  at,  or
prior to, planting or during the  growing season as  a side-dressing.  For surface
applications, the application amount was added,  in the model, to the 1 cm surface
zone.  For soil incorporation, different assumptions were used for Conventional
Tillage  (CNT) versus Conservation Tillage (CST).   For CNT,  we  assumed that 10X
of the application would remain  in the surface zone, and 90X would be added to
the 14 cm upper zone; this is based on the assumption that incorporation would
produce  a uniform distribution  of  the chemical in the top 10 to 15 cm of the
soil.  For CST, we assumed 15X would  remain in the surface, and 85X would be in
the upper zone, derived  from assuming a linear distribution of chemical from the
surface to zero at the depth of incorporation of 10 to 15 cm.  These assumptions
were  based on reviews of literature to evaluate the impacts of  BMPs  on model
parameters (Donigian, et al., 1983).

For manure applications, we assumed a 50/50 split between the surface and upper
zones  for all  nutrient  components   of  the  animal  waste.   Although  manure


                                    .98

-------
applications are typically surf ace-applied, some incorporation does occur.  Also,
the small depth  of  the  1 cm surface  zone  relative  to the large mass of manure
applied  (typically  10-15 tons/ac) would lead  to some natural mixing of manure
with native soil below the  1 cm depth, which the model, represents as the upper
zone.  Since the AGCHEM  model assumes all organic N and P  is sediment bound, any
organics  applied to the surface  remain in the  surface  except  for  losses by
mineralization and  attached to eroded  sediment.   For these  reasons,  the 50/50
split was needed to more closely represent manure nutrient applications within
Che model framework.                 .           •

For the Hay segment, we also assumed a 50/50 split for fertilizer applications
because re-seeding would likely involve, soil incorporation of applied fertilizer
but this would occur only once every two to three years.  In the other years, the
fertilizer would be surface applied.  Since some fraction of the farmers would
be  re-seeding hay  fields every  year,  we   assumed  the 50/50 split would  be a
reasonable compromise for the complex combinations of practices that would occur
each year.                   .
                                       •
Based on discussions with the NRTF,  the  frequency of manure applications can vary
from daily spreading in winter, to weekly or monthly applications for hay, to one
or two spring or fall applications for  field crops.  Unfortunately, the current
procedures for specifying nutrient applications in AGCHEM are too cumbersome to
allow daily or  weekly applications  at  the scale of this  modeling effort;  each
application must be defined explicitly for each model segment and each year of
Che simulation,  and each specific  state variable impacted by the application.
Consequently, Che minimum frequency used in Che model is monthly applications for
Che hay  segment and bi-monthly (i.e., every two months)  for the other crops,
during Che typical  application period appropriate for each crop.

.In  general,   the application protocols  described  in Table  4.10 produced a
reasonable representation of nutrient application procedures throughout Che Bay
drainage.  Some compromises  between practices in different states were needed in
order to  maintain consistency in the model representation.  Any future studies
with the  Watershed Model by the  CBPO and/or individual states should consider
refining  the  assumptions and practices as needed to better represent local and
regional  conditions.

4.3.4.6   Calculation of model input values

Calculation of the final model input nutrient application amounts was performed
with a Lotus  spreadsheet for each model segment, following the flowchart shown
in Figure. 4.12 and Che assumptions discussed in the above  sections.  Tables 4.11
and 4.12  show portions of the spreadsheet  for Model Segment'100  (Juniaca model
segment)  to demonstrate the  format of the spreadsheet and the steps  in the final
calculations.

Table 4.11 shows the calculations of the annual application amounts for the CNT
category  within Model Segment 100; separate portions of Che spreadsheet are used
Co calculate  the rates for CST and Hay. These  portions are not shown, but they
follow che same  steps as in Table 4.11 except  that  the calculations for Hay are
simpler  since no compositing is  involved.


                                       99

-------
Table 4.10  MODEL PROCEDURES FOR TIMING AND PLACEMENT OF FERTILIZER/MANURE
            APPLICATIONS AND ATMOSPHERIC DEPOSITION
CORN:                                                    ......
      Fertilizer- SOX N and 100X P applied at planting and incorporated
                  50Z N sidedressed 5-6 weeks after planting
      Manure -    Bimonthly  (i.e.,  every  2  months)  applications,  November
                  through April, with  a  50/50  split between Surface and Upper
                  Zone                                         -

SOYBEANS;
      Fertilizer- 100X of N and P applied at planting, incorporated, about 3-4
                  weeks after corn planting dates
      Manure-     Bimonthly (i.e.,  every  2 months) applications November through
                  April,  same  as   corn;  if  amounts  are  small,  use  double
                  applications  in  November and  April (for Fall  and Spring).
                  Split 50/50 between  surface and upper zone.

          NOTE: If total application amount  is small, less than 5 pounds/ac, add
          to the corn application,  i.e.,  don't  put in as separate applications.

GRAINS;
      Fertilizer- 100X   P  and  20X   N  ^application  in  Fall  (September),
                  incorporated,  and 80 X N application  top-dressed in Spring
                  (March)
      Manure-     Bimonthly summer applications, June through September, 50/50
                  split incorporation  in Surface and Upper Zone

HAYc        '     .    ;'.'••••     .                .'"              .  .  •
      Fertilizer- Use 40/30/30 split applications between August, mid-March, and
                  June to account for different seeding times (August and March,
                  incorporated 50/50 split between Surface and Upper  zones) and
                  single  top-dressing  in June.                 •
      Manure-     Monthly application  April- through October,  with 50/50 split
                  between surface  and  upper zone.                  .

ATMOSPHERIC DEPOSITION:
      Apply all  deposition amounts as  N03 (since NH3 will  be nitrified quickly)
      and add  to die fertilizer  component, but only to the SURFACE zone.
                                      100

-------
    TA51E  4.11   EXAMPLE CALCULATION  OF FINAL NUTRIENT  APPLICATION  RATES
                                                                  SEGMENT 100
                                                          j-Modc
Crap Distribution,
State-Supplied
Application RAtes,
and AtBos. Dep.
Cropland Category
Initial Composite
Based on Crop
Distribution
State-Supplied
Fractions for
Application Rate
NH3 Volatilization
and Nutrient
Coapoaitian
Assuaptions  ,
FINAL Composite
Rates and
Crop Coaponents
CALCULATION OF NUTRIENT APPLICATION RATES -
  Date:  03/08/91
State-Supplied Fertilizer and Manure Appl. Rates and CBLO ATM. Depos.

                         State Nutrient Application Rates
I Segment Nunfcer


Corn
Corn-si I
Soybeans
Grains
Total's
CNT
CROP X
46.3X
2S.3X
2.8X
2S.6X
100.0X
CST
CROP X
54. 4X
29.9X
3.5X
12. ZX
100.0X
Fertilizer *
N
95
95
34
34

P
59
59
8
8

1 Manure
N
150
150
140
140

P
26
26
,25
25

ATM. DEP
. N
10
10
10
10

Fertilizer * 2
N
0
0
0
0

P
0
0
0
0

.-G
Conventional Tillage
                                     Fertilizer
                                    Manure        ATM.  DEP   TN
                              (less NH3 vol.  losses)
                                                                   TP
Coapocition of
Final Segacnt
Application Rates
1 — N P N P
Composite 77.7 44.5 111.8 25.7
Corn 68.0 42.2 81.6 18.6
Soybeans 1.0 0.2 3.0 0.7
Grains 8.7 2.0 27.2 6.4
Area Weighted Application Rates - CNT
N
10.0
7.2
0.3.--
2.6

Fraction of area receiving fertilizer * 1 and manure
Fraction of area receiving fertilizer * 1

Fraction of area receiving fertilizer * 2 and manure
Fraction of area receiving fertilizer * 2
— Fraction of Manure available as NH3
Fraction of Manure available as ORGN
Fraction of Manure available as P04
Fraction of Manure available as ORGP
Fraction of fertilizer available as NH3
Fraction of fertilizer avajlable as N03
Fertilizer Manure
•— N P N P
Composite 67.3 38.1 35.1 8.0
Corn ^57.8 35.9 28.6 6.5
Soybeans 0.8 0.2 1.0 0.2
Grains 8.7 2.0 5.4 1.3
Fertilizer and Manure Conposit ion of TN and TP
Fertilizer * ATMOS. Dep.
ATM.
•— NH3 N03 N03 P04 NH3
Composite 50.5 16.8 10.0 38.1 15.5
Com 43.4 14.5 7.235.9 12.8
Soybeans 0.6 0.2 0.3 0.2 0.5
Grains 6.5 2.2 2.6 2.0 2.4

0.34
0.42
0.50
0.50
0,75
0.25
ATM. DEP
N
10.0
7.2
0.3
2.6


199.5 70.2
156.8 60.9
4.2 0.9
38.5 8.4
Corn Beans Grains
0.20 0.20 0.20
0.65 0.65 0.80
0.15 0.15 0.00
0.00 0.00 0.00
X





TN TP

112.4 46.2 — ->
93.5 42,4 FINAL
2.1 0.4 SEGMENT
16.7 3.3 RATES
• '
. Manure . TN TP


ORGN P04 ORGP
19.1 4.0
15.8 3.3
0.6 0.1
3.0 0.6
4.0 112.4 46.2
3.3 93.5 42.4
0.1 2.1 0.4
0.6 16.7 3.3
                                                     101

-------
TABLE  4.12   EXAMPLE CALCULATION  OF  TIMING AND  PLACEMENT OF  NUTRIENT
                   APPLICATION
     Tiarfi* an

     Fertilize
     Applicati<
of
     Application*
                    Conventional
                       TUlaซa
                    Contanwtian
                       Till***
                    •ay land
                    Convantfonal
                       TIUaซa
                    Camera t Ian
                       TllUa*
                                         FEITILIKI TIMING AND PIACEMEHT  I SEGMENT  100K Model ScgBซU NuBtaar
CUT

ซ:HN3

II:N03

f:KX


CST

ซ:NH3

• :N03

P:P04

KAT
 Spr. Cr.
  SZ

 5.22

 3.79

 0.00

 Spr. Cr.
 SZ

 2.49

.1.81.

 0.00

  Nardi
 SZ   UZ
                                         ป:*a   o.oo o.oo

                                         •:N03   3.00 0.00

                                               '  1.20 1.20
 Plant C.
 SZ  UZ

2.17 19.51

t.30 *.SO

3.5932.32

 •tant C.
 SZ  UZ

3.83 21.70

5.49 7.23

0.34 35.93

     Juw
      SZ

     0.00

     3.00

     2.40
riant I.
SZ    UZ

0.06 O.SS

0.30 0.18

0.02 0.17

•tant 8.
SZ    UZ

0.11 0.64

0.39 0.21

0.04 0.20
                                                            SZ     UZ

                                                            0.00  0.00

                                                            4.00  6.00

                                                            1.60  1.60
S.Or. C
SZ
21.68
10.81
0.00
s.or. e
sz
25.53
12.72
6.00
Til T*
10.00 8.00
rail Cr.
SZ UZ
0.13 1.18
0.56 0.39
0.20 1.84
fall Cr.
SZ UZ
0.09 0.53
0.28 0.18
n rr
77.33 38.15


M Tป
83.23 43.49

0.150.83
Total
— Fartllitar — 	
Application.
                                                                  [sttMEIIT 100|- I
                            MANURE TIMIttG MB HACEMEUT  SECMEKT 100[- Modal tl|ซant Ma*ar
                            CUT

                            mo   '
                            OtCM

                            M4/OK*
                            CST

                            no
                            OftGN

                            •04AKG*


                            MY
            COM/SOT (3 app(.)
            Mov/Jon/Nar
             SZ   UZ

            2.21 2.21
            2.73 2.73

            O.S4 O.S6

            COtMTSOT C3 appt.)
            •ov/Jan/Nar
             SZ   UZ

            2.61 2.61
            3;22 J.22

            0.66 0.64
                               OUINS (3 appl.)
                               Mar/All/Sap
                               SZ    UZ

                               0.41 0.41
                               O.JO O.SO

                               8.11 0.11

                               CM!** (3 appl.)
                               May/M/Sap
                               SZ    UZ

                               0.19 0.19
                               0.24 0.2*

                               O.OS 0.8S
                                                     KAT (6 appl.) Nar/AtVJul/Aui/Sap/Oct
                                                      SZ  UZ

                                                    .1.281.28
                                                     1.58 1.S8

                                                     0.33 0.33
                                                                                                    Total
                                                                                                    Hanuro
                                                                                                    Appdeatie
                    n • Surfaea Soil Zona
                    UZ - Uppar Soil.Zona
                                                             102

-------
At the top of the spreadsheet, the input values for the  cropping distributions,
the state-supplied  rates  for  each crop,  and  the atmospheric  deposition amounts
are shown.  In the next section,  the initial  composite value  and the  individual
crop components are calculated based on the input  cropping distribution.  These
rates also allow for NH3 volatilization losses using the value for 'fraction of
manure available as NH3' ,  which is.shown in the middle portion of the table.  For
example,  the value  shown in Table 4.11 is 0.34; this accounts  for 60Z loss by
volatilization of the  inorganic N,  plus  the  fraction of organic N available as
NH3 (as shown in Figure 4.13 and discussed above).

The middle section of Table 4,11 shows the state-supplied fractions of each crop
(i.e., corn, beans,  grains)  which receives the various combinations of manure and
up  to two  different > fertilizer  rates;  this middle section  also  shows  the
fractions used to calculate volatilization losses and nutrient composition.  The
initial composited  crop component rates are then weighted by  the crop fractions
for the various combinations in order to calculate the final  segment  rates, and
the crop components, shown  in the second to last block near the bottom of Table
4.11.  The final block in Table 4.11 shows the composition of the final segment
rates,  i.e.,  the distribution of the final  rates  into organic and  inorganic
forms.

The components of the  final segment rates, attributable  to  each crop  is carried
through the calculations so that the protocols for timing and placement for each
crop can be represented.  In Table 4.12,. the final segment rates from Table 4.11
are then distributed in time and space for specific applications on each cropland
category.  The  top half  .of  Table  4.11  shows   the  values  for   fertilizer
applications, while the bottom half shows the manure applications for CNT, GST,
and  Hay.    The values shown  in this  table- (which are  the  same  as  in  the
spreadsheet) are the specific values input to the model and can be identified in
the model segment input sequence.                                    .

For fertilizer applications,  the values  in Table 4.12 include the NH3, N03, and
P04 amounts epplied to.the surface  zone (SZ in the  table),  and  the  upper, zone
(UZ)  for applications associated  with spring grains,  planting of corn  and
soybeans, side  dressing  of. corn,  and fall grains.   For the Hay segment, .the
fertilizer amounts associated with the three separate applications (discussed in
Table 4.10) are shown.          •.''--.    '•'.'.-

For manure  applications,  the bottom half of  Table  4.12  shows the inorganic and
organic amounts, the.number and timing of the  applications, and the placement in
the surface and upper zones.                       .   .

The  final application rates  for  all model segments for the  CNT, CST, and Hay
cropland categories are .summarized in Tables 4.13, 4.14. and 4.15, respectively.
Appendix C includes copies of the individual spreadsheets for each model segment,
which are also available from the CBPO as Lotus files.

4.3.5  AGCHEM application procedures for Model Segments

As  noted in the  introduction to Section 4.3, the  primary goal of  the AGCHEM
application in Phase II was to represent nutrient balances on the cropland areas
so  that the impacts of climate, agricultural practices,  and nutrient  management

                                      103

-------
TABLE 4.13 FINAL CALCULATED MODEL SEGMENT NUTRIENT APPLICATION RATES FOR
           CONVENTIONAL TILLAGE (Ib/ac)

MODEL          FERTILIZER          MANURE       ATMOS.  DEP.     TOTAL
SEGMENT        N        P        N        P         N         N      P
10
20
30
40
50
60
70
80
90
100
110
120
140
160
170
175
180
190
200
210
220
230
235
240
250
260
265
270
280
290
300
310
330
340
84.3
97.4
95.3
77.1
77.5
•86.3
78.7
59.4
68.9
67.3
62.6
71.5
63.4
110.9
138.9
130.5
117.5
117.9
113.7
97.5
114.1
130.7
87.0
76.8
118.0
83.8
149.2
138.6
95.6
81.1
97.2
104.8
•101.0
110.4
17.2
19.7
19.3
15.9
17.1
18.9
17.3
, 22.2
39.5
38.1
26.8
28.8
27.1
13.4
30.7
30.2
21.2
29.1
28.7
26.8
25.5
22.0
27.2
29.8
27.0
26.6
32.2
31.6
26.5
27.4
27.4
26.3
14.8
15.4
22.4
22.6
22.6
23.4
33.1
35.7
34.6
43.0
35.4
35.1
61,5
66.4
60.9
21.8
28.6
28.6
24.7
97.3
93.5
26.1
1.9
0.0
0.0
0.0
0.0
0.0
0.0
5.7
0.0
0.0
0.0
0.0
29.4
16.8
5.3
5.3
5.3
5.5
7.7
8.4
8.1
10.1
8.1
8.0
14.4
15.5
14.3
7.3
9.0
9.0
7.1
30.6
29.4
7.5
0.6
0.0
0.0
0.0
0.0
0.0
0.0
1.8
0.0
0.6
0.0
0.0
9.3
5.2
. 9
9
9
9
11
11
10
9
10
10
9
8
8
10
9
10
9
9
9
9
9
8
8
8
8
8
. .. 8
8
7
7
7
7
8
8
115.7
129.0
126.8
109 . 5
121.6
133.1
123.2
111.5
114 . 3
112.4
133.2
145.9
132.4
142.7
176.6
169.1
151.2
224.2
216.2
132.6
125.0
138.7
95.0
84.8
126.0
91.8
157.2
152.3
102.6
88.1
104.2
118.8
138.3
135.2
22.5
25.0
24.6
21.4
24.8
27.2
25.4
32.2
47.6
46.2
41.2
44.3
41.3
20.7
39.7
39.2
28.3
57.7
58.1
34.3
26.1
22.0
27.2
29.8
27.0
26.6
32.2
33.4
26.5
27.4
27.4
26.3
24.1
20.5
                                        104

-------
TABLE 4.13  FINAL CALCULATED MODEL SEGMENT NUTRIENT APPLICATION RATES FOR
            CONVENTIONAL TILLAGE (Ib/ac) (continued)
MODEL
SEGMENT
Anacostia
Baltimore
Bohemia
Chester
Chickahominy
Chop tank
Coast_l
Coast_4
Coast_5
Coast_6 '
Coast_8
Coast_9
Coast 11'
Elizabeth
Gr. Wico.
Gunpowder
James
Nansempnd
Nanticoke
Occoquan
Patapsco
Patuxent
Pocomoke
Potomac
FERTILIZER
N
96.7
109.3
94.9
85.4
89.9
72.4
74.6,
79.0
111.7
119.6
77 .5
64.3
100.9
56.1
80.0
103.7
82.4
112.5
93.9
123.5
101.4
92.8
74.2
82.7
Rappahannock 79.0
Severn
Wicomico
Wye
York
91.9
66.7
77.3
79;7
P
26.9
15.8
17.9
19.8
29.2
22.6
22.2
29.6
* 16.1
35.8
29.7 '
29.7
33.3
27.7
29.2
14.9
29.2
32.1
23.0
27.1
14.7
25.7
20.5
25.5
29.1
14.7
19.5
20.8
29.3
MANURE ATMOS. DEP. TOTAL ,
N
22.8
16.4
3 .5
11.7
0.0
24.4
10.2
0.0
10.6
3.0
0.0
0.0
72.1
0.0
0.0
25.6
0.0
0.0
127.6
0.0
16.7
7.6
137.8
2.4
0.0
27.7
130.4
21.5
0.0
P
8.1
4.8
1.0
3.9
0.0
6.7
2.9
0.0
3.5
0.9
0.0
0.0
19.2
0.0
0.0
7.9
0.0
, 0.0
36.4
0.0
5.4
2.6
40.1
0.9
6.0
7.3
37.4
6.3
0.0
N
8
8
8
. 8
7
8
8
7
8
8
8
7
8
7
7
8
7
7
8
8
8
8
7
8
7
8
7
8
7
N
127.5
133.7
106.4
105.0
96.9
104.9
92.8
86.0
130.3
130.6
85.5
71.3
181.1
63.1
87.0
137.3
89.4
119.5
229.5
131.2
126.1
108.4
219.0
93.1
86.0
127.6
204.1
106.8
86.7
P
35.0
20.6
18.9
23.7
29.2
28.8
25.1
29.6
19.6
36.7
29.7
29.7
52.6
27.7
29.2
22.8
29.2
32.1
59.3
27.1
20.2
28.3
60.5
26.5
29.1
22.1
56.9
27.2
29.3
                                        105

-------
TABLE 4.14  FINAL CALCULATED MODEL SEGMENT NUTRIENT APPLICATION RATES FOR
            CONSERVATION TILLAGE (Ib/ac)
MODEL
SEGMENT

10
20
30
40
50
60
70
80
90
100
110
120
140
160
170
175
180
190
200
210
220
230
235
240
250
260
265
270
280
290
300
310
330
340
FERTILIZER
N
96.7
104.3
102.1
86.1
90.9
95.2
86.8
65.3
74.6
73.2
71.0
77.5
70.7
115.8 .
139.7
133.0
124.5
121.6
117.6
108.3
116.5
134. 5;
87.2
75.1
121.2
82.9
149.6
142.1
96.8
80.4
98.3
107.2
106.1
115.4
P
19.5
20.9
20.5
17.5
19.8
20.7
19.0
25.4
- 44.7
43.5
28.5
30.0
28.4
16.8
30.5
29.5
25.1
28.9
28.4
28.4
25.1
22.1
26.9
30.4
26.6
25.9
32.2
31.4
26.0
27.0
26.8
25.4
15.2
15.8
MANURE
N
22.6
22.7
22.7
23.9
36.7
38.2
37.1
48.2
37.8
37.5
68.6
71.3
68.3
22.3
28.6
32,4
28.1
97.3
93.5
25.7
1.9
0.0
0.0
0.0
0.0
0.0
0.0
5.7
0.0
0.0
0.0
0.0
29.5
16.5
P
5.4
5.4
5.4
5.6
8.6
8.9
8.7
11.3
8.6
8.6
16.1
16.7
16.0
7.5
9.0
10.2
8.1
30.6
29.4
7.4
0.6
0.0
0.0
0.0
0.0
0.0
0.0
1.8
0.0
0.0
0.0
0.0
9.3
5.1
ATMOS. DEP. TOTAL
N
9
9
9
9
11
11
10
9
10
10
9
8
8
10
9
10
9
9
9
9
9
8
8
8
8
8
8
8
7
7
7
7
8
8
N
128.3
132.8
133.9
118.9
138.6
144.5
133.9
122.5
122.4
120.8
148.6
156.8
147.0
148.1
177.3
175.5
161.6
227.9
220.1
143.0
127.4
142.5
95.2
83.1
129.2
90.9
157.6
155.9
103.8
87.4
105.3
114.2
143.6
139.9
P
24.9
25.7
25.9
23.2
28.4
29.7
27.7
36.7
53.3
52.1
44.5
46.7
44.4
24.3
39.5
39.7
33.1
59.5
57.8
35.8
25.8
22.1
26.9
30.4
26.6
25.9,
32.2
33.2
26.0
27.0
26.8
25.4
24.5
20.8
                                        106

-------
TABLE 4.14  FINAL CALCULATED MODEL SEGMENT NUTRIENT APPLICATION RATES FOR
            CONSERVATION TILLAGE (Ib/ac) (continued)
MODEL
SEGMENT
Anacostia
Baltimore
Bohemia
Chester
Chicka.
Choptank . .
Coast_l
Coast_4
Coast_5
Coast_6
Coast_8
Coast_9
Coast 11
Elizabeth
Gr. Wico.
Gunpowder
James
Nansemond
Nanticoke
Occoquan
Patapsco
Patuxent
Pocomoke
Potomac
FERTILIZER
N
97.3
117.4
92.2
84.4
77.5
69.2
70.3
77.0
107.9
128.9
75.4
62.0
108.1
43.9
77.0
112.3
82.6
115.3
93.1
127.8
112.2
85.5
65.8
76.6
Rappahannock 76.2
Severn
Uicomico
Wye
York
98.7
57.0
76.1
78.0
P
28.3
16.2
17.9
21.6
29.3
24.7
24.5
30.3
- 16.1
37.1
30.4
30.2
34.4
28.1
29.9
15.4
29.5
32.6
23.0
26.5
15.8
25.1
19.4
25.3
29.6
15.1
18:3
23.4
29.8
MANURE ,
N
22.6
16.2
3.6
11.5
0.0
23.7
10.6
0.0
10.6
3.0
0.0
0.0
73.6
0.0
0.0
25.3
0.0
0.0
127.5
0.0
15.8
7.0
136.9
2.3
0.0
32.0
136.8
18.6,
0.0
P
8.0
4.8
1.0
3.8
0.0
6.5
3.0
0.0
3.5
0.9
0.0
0.0
19.6
0.0
0.0
7.8
0.0
0.0
36.4
0.0
5.1
2.4
39.6
0.9
0.0
8.2
39.6
5.5
0.0
ATMOS . DEP . TOTAL
N
8
8
8
8
: 7
8
8
7
8
8
8
7
8
7
7
8
7
7
8
8
8
8
7
8
7
8
7
8
7
N
127.9
141.7
103.8
103.9
84.5
100.9
88.9
84.0
126.4
139.9
83.4
69.0
189.7
50.9
84.0
145.6
89.6
122.3
228.7
135.8
136.0
ioo.s
, 209.7
86.9
83.2
138.6
200.8
102.7
85.0
. P
36.3
21.0
18.9
25.4
29.3
31.2
27.5
30^3
19.6
37.9
30.4
30.2
54.0
28.1
29.9
23.2
29.5
32.6
59.4
26.5
20.9
27.6
59.1
26.2
29.6
23.4
, 57.9
28.9
29.8
                                        107

-------
TABLE 4.15  FINAL CALCULATED MODEL SEGMENT NUTRIENT APPLICATION RATES
            FOR HAY (Ib/ac)
MODEL
SEGMENT
ATMOS. DEP.    TOTAL
    N        N       P
10
.20
30
40
50
60
70
80
90
100
110
120
140
160
170
175
180
190
200
210
220
230
235
240
250
260
265
270
280
290
300
310
330
340
30.0
30.0
30.0
30.0
30.0
30.0
30.0
7.5
0.0
0.0
10.0
10.0
10.0
0.0
34.2
34.2
2.1
29.3
29.3
1.5
7.2
60.0
34.7
30.0
36.0
35.5
36.0
34.6
33.4
30.2
30.0
30.0
9.0
9.0
14.5
14.5
14.5
14.5
13.0
13.0
13.0
17.5
* 8.0
8.0
13.0
13.0
13.0
23.7
28.5
28.5
18.9
24.1
24.1
11.7
9.9
21.0
29.0
25.0
30.0
29.6
30.0
28.7
27.8
25.2
25.0
25.0
3.9
3.*
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
34.2
34.2
41.0
41.0
41.0
12.1
4.0
4.0
11.6
15.1
15.1
14.4
" 1.6
0.0
0.0
0.0
0.0
0.0
0.0
0.4
0.0
0.0
0.0
0.0
24.6
13.0
0.0
0.0
0.0
0.0,
.0.0
0.0
0.0
0.0
7.8
7.8
9.6
9.6
9.6
4.1
i.a
1.8
3.5
6.7
6.7
4.3
0.5
0.0
0.0
0.0
0.0
0.0
. 0.0
0.2
0.0
0.0
0.0
0.0
7.8
4.0
9
9
9
9
11
11
10
9
10
10
9
8
8
10
9
10
9
9
9
9
9
8
8
8
8
8
8
8
7
7
7
7
8
8
39.0
39.0
39.0
39.0
41.0
41.0
40.0
16.5
44.2
44.2
60.0
59.0
59.0
22.1
47.2
48.2
22.7
53.4
53.4
24.9
17.8
68.0
42.7
38.0
44.0
43.5
44.0
42.9
40.4
37.2
37.0
37.0
41.6
30.0
14.5
14.5
14.5
14.5
13.0
13.0
13.0
17.5
15.8
15.8
22.6
22.6
22.6
27.8
30.2
30.2
22.4
30.8
30.8
16.0
10.4
21.0
29.0
25.0
30.0
29.6
30.0
28.9
27.8
25.2
25.0
25.0
11.7
7.9
                                        108

-------
TABLE 4.15  FINAL CALCULATED MODEL SEGMENT NUTRIENT APPLICATION RATES
            FOR HAY (Ib/ac) (continued)
MODEL
SEGMENT
Anacostia
Baltimore
Bohemia
Chester
Chicka.
Choptank
Coast_l
Coast_4
Coast_5
Coast_6
Coast_8
Coast_9
Coast 11
Elizabeth
Gr. Wico.
Gunpowder
James
Nansemond
Nanticoke
Patapsco
Patuxent
Pocomoke
Potomac
Occoquan
Rappa .
Severn
Wye
Wicomico
York
FERTILIZER
N
0.0
9.0
6.6
9.0
31.6
9.0
9.0
30.0
9.0
0.0
30.0
30.0
2.7
30.0
30.0
9.0
30.0
30.0
9.0
9.0
9.0
9.0
18.9
36.0
30.6
9.0
9.0
9.0
30.9
P
12.0
3.9
7.8
7.8
26.4
7.8
7.8
* 25.0
3.9
9.3
25.0
25.0
10.3
25.0s
25.0
3<.9
25.0
25.0
7.8.
3.9
3.9
7.8
12.0
30.0
25.5
3.9
7.8
7.8
25.8
MANURE
N
17 .,5
16.9
3.2
8.0
0.0
21.2
5.1
0.0
10.6
2.7
0.0
0.0
55.7
- 0.0
0.0
20.2
0.0
0.0
63.4
13.9
13.5
- 5.3
2.0
0.0
0.0
; 32.6
17.6
114.2
0.0
P
6.2
4.6
0.9
2.7
0.0
5.9
1.4
0.0
3.5
0.8
0.0
0.0
14.9
'0.0
0.0
6.3
0.0
0.0
18.0
4.5
4.6
1.5
0.8
0.0
0.0
8.0
5.2
32.9
0.0
ATMOS. DEP. TOTAL
N
8
8
8
8
7
8
8
7
8
8
8
7
8
7
7
8
7
7
8
8
8
7
8
8
7
8
8
7
7
N
25.5
33.9
17.2
25.0
38.6
38 . 2
22.1
37.0
27.6
10.7
37.0
37.0
66.3
37.0
37.0
37.2
37.0
37.0
80.4
30.9
30.5
21.3
28.9
44.0
37.6
49.6
34.6
130.2
37.9
P
18.2
8.5
8.7
10.5
26.4
13.7
9.2
25.0
7.4
10.1
25.0
25.0
25.1
25.0
25.0
10.2
25.0
25.0
25.8
8.4
8.5
9.3
12.7
30.0
25.5
12.0
13.0
40.7
25.8
                                        109

-------
alternatives on NFS  nutrient loadings and water quality could be evaluated.  The
AGCHEM modules were originally  designed for  application to areas ranging from
field-size to small-to-moderate  size watersheds on  the order of the size of a
single model segment.  For  these size  areas  when  the relevant soil, crop, and
runoff data are  available,   the AGCHEM calibration  involves  the establishment of
a reasonable simulation of  soil  nutrient storages through  adjustment of plant
uptake rates,  soil  transformations,  partition coefficients, and percolation
parameters,  followed by  evaluation  of  the  nutrient  runoff simulation (in
comparison with observed values) and parameter refinement as needed.  The general
guidance provided for nutrient calibration as  described in  Che HSPF Application
Guide (Donigian, et al.,  1983) is as follows:

      1.   Evaluate initial  soil nutrient parameters from information available
          in the literature, and include fertilizer manure  and rainfall sources
          of nutrients as input  to the model.

      2.   Calibrate  initial mineralization  rates so  that annual  amounts  of
          plant-available nutrients correspond to expected values.

      3.   Adjust leaching factors based on any data available for  a tracer such
          as chloride.

      4.   Adjust plant  uptake rates to develop  the  expected nutrient uptake
          distribution during the growing season and  the estimated total uptake
          amount expected for the crop.

      5.   Adjust nutrient partition  coefficients  based on  available soil core
          and runoff data.

      6.   Refine  the leaching,   uptake,  and  partition parameters based  on
          observed  runoff  data  and  the expected  sources  of  nutrient runoff,
         •i.e., surface, interflow, groundwater.           '

Due  to  both data  and  resource  limitations,  the  type -of detailed  AGCHEM
calibration  described by  the above steps could not  be performed in Phase II.
Even if the data were available for selected sites within the Bay  drainage, the
calibrated model parameters would still need to be  refined or adjusted to the
scale of the model  segments and to represent the  'composite crop' within each
model segment.  AGCHEM model calibrations were initially planned as part of the
Phase II effort for sites in the  Patuxent, Rappahannock (Owl Run), and Potomac
(Nomini Creek) model segments; however, schedule and  data constraints precluded
completion of these efforts.  Plans are to complete these applications in future
studies either by the CBPO,  the  individual states, or  through joint efforts.

Because  of  the  above  constraints  and limitations,  the  focus  of  the  AGCHEM
simulations was a 'regional' scale application designed to represent the expected
nutrient  balances  on  each   of  the cropland  categories.   The  literature was
reviewed  to  define typical nutrient  balances   for.  each  of  the  major  crop
categories,  and the  AGCHEM  simulations were  calibrated  to represent  these
nutrient balances while maintaining the nonpoint  nutrient  loadings  within the
range of expected values.
                      ,                  1

                                     .110

-------
4.3.5.1  Expected crop nutrient balances

Tables 4.16  and 4.17, respectively  show the typical nitrogen  and phosphorus
balances expected for  the major crops when the nutrients are applied at agronomic
rates meeting crop needs.  Although  there is a fairly large range in many of the
components of the nutrient balances, it is clear that nutrient applications and
crop uptake are the dominant portions of the balance.  This is especially true
for corn which generally comprises more than half of the cropland  (excluding hay)
for the model segments above  the  Fall Line;  below the Fall Line,  soybeans and
grains are  more extensive but  corn still comprises  about 30X  to 60Z  of the
cropland in most segments.

Moreover, current total nutrient application rates including both fertilizer and
manure, as shown  in Tables 4.5  and  4.6,  are  considerably larger than expected
crop nutrient needs. •ป Table 4.18 shows the expected mean and range of crop uptake
rates  along  with the assumed crop  yields on which the rates are  based.   The
uptake rates were initially derived  from the general literature sources shown in
the  table,   and then  adjusted through  discussions  with the  NRTF based  on
subsequent information provided by state representatives on actual crop yields
experienced  in  the individual states.

Because of the dominance of the application rates and crop uptake  on the nutrient
balance, these two components were the primary focus  of our efforts with AGCHEM
to  simulate  these balances.   The extensive  efforts  to  develop  reasonable and
realistic nutrient application rates were described above in Section 4.3.3.  The
critical importance of the application rates  cannot be over-stated; these rates
defined the source and quantity of nutrients available for crop uptake and runoff
loadings.   Based on  the  defined application  rates,  the steps in  the  AGCHEM
calibration were as follows:

       1.  Estimate  soil  nutrient   storages  and  parameters  from  available
          literature and state-supplied data.     '

       2.  Calibrate uptake,  mineralization, and immobilization (fixation) rates
          to  reflect expected nutrient balances.

       3.  Compare  simulated  runoff  loadings and concentrations  to  available
          literature data.

Initial estimates of both soil nutrients and parameters were derived from past
modeling  studies using ACCHEM  at a  variety  of  field and watershed  sites  in
Georgia  and  Michigan as  part of  the ARM model development work (Donigian and
Crawford, 1976;  Donigian et al., 1977),  and in Iowa (Donigian et al.,  1983;
Imhoff et al.,  1983), Nebraska  (Gilbert et  al., 1982),  Florida  (Nichols and
Timpe, 1985), and Tennessee (Moore etal.. 1988).  However, many of these studies
focussed on small field sites with little or  no subsurface nutrient components,
and phosphorus modeling was either not attempted or de-emphasized due to limited
data, availability.    In  fact,  the  current application  of AGCHEM  is  the most
comprehensive and  extensive  application  to  date,   requiring  more  detailed
assessment of cropland nutrient processes than any previous study.
                                      Ill

-------
          TABLE 4.16  TYPICAL NITROGEN BALANCE FOR MAJOR CROPS WHEN
                      NITROGEN IS APPLIED AT AGRONOMIC RATES (Ib/ac)
                              Corn       Soybeans        Grains        Hay

Inputs:

      Fertilizer/Manure      100-160       25-35         50-100       30-60
      Atmos. Deposition        7-10         7-10          7-10         7-10
      Mineralization          25-40        25-40         25-40        25-40
               Totals        132-210       57-85         82-150       62-110
                                     ' •   ' t.
Outputs:
      Plant Uptake           120-150       25-40*        60-90        30-55
      Surface Runoff           2-5          1-3           2-4         ,1-3
      Leaching & Subsurface   10-25        10-15          5,-15         5-15
        Runoff
      Volatilization &        15-25         5-15         .10-20        10-20
        Dinitrification

               Totals        147-205       41-73         77-129       46-91

      A Storage             -15 to +5    +16 to +12    +5 to +21    +16 to +10
* - Represents uptake from the soil, approximately 25% of total
    uptake, with 75% supplied by fixation (Tisdale et al.,'1985).
                                      112

-------
         TABLE 4.17  TYPICAL PHOSPHORUS BALANCE FOR MAJOR CROPS WHEN
                     PHOSPHORUS IS APPLIED AT AGRONOMIC RATES (Ib/ac)
                              Corn       Soybeans        Grains        Hay
Inputs:
      Fertilizer/Manure       20-40        10-30         10-30        10-30
      Atmos. Deposition        0-1          0-1           0-1          0-1
      Mineralization           2-5          2-5           2-5          2-5
               Totals         22-46        12-36         12-36        12-36
Outputs:
      Plant Uptake            20-30        12-20         12-22        12-25
      Surface Runoff           1-2          0-1           0-1          0-1
      Leaching & Subsurface    0-1          0-1           0-1          0-1
        Runoff              -                                 -

               Totals         21-33        12-22         12-24        12-27

      A Storage              +1 to -1-13     0 to +14     0 to +12      0 to +9
                                      113

-------
4.3.5.2  Initial soil nutrient storages

Soil nutrient storages required by AGCHEM include initial values for all nutrient
forms  (i.e.,, organic and inorganic) for each soil zone, including  the surface,
upper, lower and groundwater zones  as  described  in Section 4.3.1.  Although a
variety of soil nutrient data was supplied by the state representatives on  the
NRTF,  the data was generally sporadic, incomplete for selected nutrient forms,
often based on different analytical  techniques, and did not provide the vertical
spatial  definition needed  for  the model  soil zones.   In  addition,' limited
resources precluded efforts to establish different nutrient storage values  for
each model segment.  Consequently,  the nutrient storages were developed from a
few selected detailed studies, past experience with AGCHEM,  and general soils
information,  and  then  compared  with  available  state  information  to  insure
consistency.

Parker et al.  (1946) prepared national maps of Total N and P205 for the surface
foot  of  soil which showed  percentages of 0.05  to 0.19 for  these two forms,
respectively.  Cunningham (R.L.  Cunningham, 1991, personal  communication) noted
that Total N was typically in the range of 0.13 to 0.16 X for Pennsylvania.  Data
from  Fox and  Piekielek (1983)  for a variety of cropland  sites throughout
Pennsylvania were consistent with the earlier information.  Bandel  et al. (1975)
reports detailed information for  N  content at three research sites in Maryland
for  different  soil  depths.   Ibison  (1990)  reported Total  N,  Total  P.   and
Inorganic P soil concentrations for a variety of Maryland and Virginia sites in
coastal regions of the Bay as part of an effort to assess nutrient contributions
from eroding banks.                   '

The above data indicated extreme, variability  in soil nutrient concentrations at
individual  sites  as a  function  of  native  soil characteristics,  cropping
practices,  historical  fertilizer and  manure applications, etc.   Lacking  the
resources  to develop segment-specific values,  we developed .initial nutrient
storages that were consistent with the  general range of reported literature data
and state-supplied information.  Brady (1990) reports  that  the vast majority of
nitrogen (i.e., 90-952 or more)  is in the organic, form, while  for.phosphorus  the
inorganic form  tends to dominate especially in subsurface soils.  Moreover,  for
our .cropland categories, nutrient applications and  crop uptake primarily impact
the inorganic nutrient forms leading to significant variations  in  soil storages
during the  annual  simulations;  this is  consistent with  the  reported data  for
-field  sites.                             -    '

Based  on the- above information  and data sources,  we  developed the following
organic  (ON, OP) and inorganic (IN,  IP) nutrient storages which were used, with
only slight adjustments, for all  model segments:            .               .*
                                      115

-------
The  initial  nutrient  storages  shown  above  provided  a  reasonable  basis for
simulating the  regional  nutrient balances as part  of  the  AGCHEM application.
Although the resources available for this effort were inadequate  to analyze the
universe of available soil nutrient data and  develop nutrient storages specific
to  each  model  segment,  any future  efforts  to refine  or  improve  the  AGCHEM
modeling should include  a more  detailed assessment of the variation in soil
nutrients throughout the region.                                               .

4.3.5.3  Estimation of AGCHEM model parameters

In addition to the initial soil nutrient storages, the AGCHEM nutrient parameters
can be grouped  into the transformation  rates,  sorption parameters,  and plant
uptake rates.   Initial values for  most parameters were obtained from selected
literature,sources  and past applications of  AGCHEM in  Georgia and Michigan as
part of the ARM  model development work (Donigian and Crawford, 1976; Donigian et
al.,,-1977),  and applications  in Iowa  (Donigian et al., 1983;  Imhoff et al.,
1983). Nebraska (Gilbert et al., 1982),  and Tennessee (Moore et al., 1988).

Stanford and Smith (1972) and  Reddy et  al.  (1979)  provided  ranges of values for
nitrogen mineralization rates, along with temperature adjustment factors.  Reddy
et  al. (1979) and Reddy  and Patrick (1984) also discussed  expected values for
nitrification, denitrification,  and volatilization rates for soils and sediments.
Recently,   Deizman   (1989)   summarized   literature   values   on   rates   for
mineralization, nitrification, and  ammonia volatilization for soils amended with
organic wastes.  Gilliam and  Hoyt  (1987)  discuss  the  impacts  of conservation
tillage  on  nitrogen  soil processes   and  summarize   much of  the  available
literature.

To represent the sorption process,  a modified Freundlich isotherm is used in the
AGCHEM application in Phase II whereby  a permanently bound  concentration of the
chemical must  be exceeded prior  to partitioning by  the  Freundlich isotherm
parameters.  Thus, for each soil  layer three parameters are required to represent
NH4 end P04 sorption:  the permanently bound concentration,  XFIX (ppm), and the
Freundlich coefficient, Kl, and  exponent, 1/N1 (Nl is the input  parameter).  The
literature on sorption was much more limited for NHซ  than  for P04,  but  a wide
range of values was  reported based  on soil  conditions, sampling procedures, and
analytical procedures, especially for P0ซ.  Work by Ardakani and McLaren (1977),
Simon (1989), Reddy  (1989; 1990,  personal communication) , and Reddy et al. (1988)
provided values from field observations for NH4.  For the more highly sorptive
P04, a number of articles for a wide range of soils and conditions were reviewed,
including Berkheiser et al. (1984), Reddy et al. (1980),  Sharpley et al. (1981),
Mansell et al.  (1977), Fitter  and Sutton (1971), Oloya and  Logan  (1980),  Taylor
and Kunishi  (1971), and McDowell et al.  (1980).

In  spite  of the  extensive Literature  available  on soil  nutrient  processes,
information  on rate  values  that  could be  used  directly in AGCHEM without
calibration was limited.   In addition, the regional scale of the modeling and the
'composite crop' representation precluded the  direct use of literature  values
developed from specific field and plot-scale data collection  efforts for specific
crops, soils, practices, etc.    Consequently, the final AGCHEM parameter values
were determined primarily through calibration to represent the expected nutrient
                  1                             *

                                     • 117                         .

-------
TABLE 4.19. NUTRIENT TRANSFORMATION AND .SORPTION PARAMETERS USED IN THE AGCHEM
            SIMULATIONS
NITROGEN:         Surface

Mineralization
Nitrification
Dinitrif ication
Mobilization/
    Fixation

PHOSPHORUS:*

Mineralization     0:0007
Immobilization/   10. Q
    Fixation

FREUNDLICH SORPTION PARAMETERS:

NH^* Sorption     Surface
                                 Upper
                  Lower
                  G.W.
0.001
10.0
0.0
5,0/6.0
0.00015
5.0
0.0/0.005
2.0/3.0
0.00015
3.0
0.005
0.20
0.00
0.50
0.03
0.00
XFIX (ppm)
Kl
Ni

 Sorption

XFIX (ppm)
Kl
Nl
                    5.0
                    1.0
                    1.5
                   25.0
                    5,0
                    1.5
                                 0.00003
                                 2.0
Upper

 5.0
 1.0
 1.2
15.0
 5.0
 1.5
                  0.00005
                  0.10
Lower

 0.7
 0.5
 1.2
10.0
 5.0
 1.5
                  0.00
                  0.00
 G.W.

 0.3
 0:5
 1.1
12.0
 6.0
 1.5
  - Rates are in units of 'per day'
                                      119

-------
      d.  The  P parameters and  representation  of the P balance  are the  most
          uncertain aspects of  the AGCHEM  simulations.   Most previous  AGCHEM
          applications were limited to nitrogen simulations, largely because the
          nitrogen cycle  is  much better  documented and  understood than  the
          phosphorus cycle in soils.   The complexity of the. phosphorus cycle and
          POt chemistry is well documented, but poorly understood quantitatively
          (e.g.,  see Sample  et.al.  (1980),  Olsen and Khasawneh (1980),  and
          sorption references cited above).   Some  of  the limitations of  AGCHEM
          for  phosphorus  simulation  were  discussed -in   Section 4.3.2.    In
          calibrating the phosphorus balance,  it was necessary to use high plant
          uptake   rates,   relatively  low  sorption  parameters,   and   high
          immobilization/fixation rates. The  sorption parameters are consistent
          with.the low end of the range of values in the literature,  while rapid
          fixation of applied P04 fertilizer (as bound or precipitated complexes
          of iron,  aluminum,  calcium, etc.) is generally accepted.

 4.3.6   AGCHEM  Simulation Results

 The AGCHEM model provides storages and fluxes of each nutrient form for each soil
 zone  and  each  simulated  process   for  every  timestep   of   the  simulation.
 Consequently,  an extensive volume  o.f information is produced for  each  model
 segment.  Tables 4.21,  4.22,  and 4.23 summarize the simulation results for the
 Conventional Tillage, Conservation Tillage, and Hay land categories, respectively
. for model segment 100 in the Juniata Basin.   The tables provide annual totals and
 the annual  average for  the. four-year simulation period, for  the  following
 variables and  fluxes:

      Nutrient applications          .
      Runoff and components
      Sediment loading. .  •'
      N03 and Ammonia loadings and components
      Sediment N loadings (ammonia  and organic N)             ,
      P04 loadings  and  components  •
      Sediment P loadings (PQ4 and organic P)
      N and  P  plant uptake from  each  soil  zone                          -
      N and  P  mineralization        '
      Denitrification
      Ammonia  and  PO*  immobilization  '

 The results  for model segment. 100 in the Juniata Basin are  representative of the
 AGCHEM  results throughout the Bay drainage.  Appendix C.2 includes AGCHEM summary
 tables  for the three cropland categories for selected model segments within each
 subbasin.

 Figures 4.14,  4.15, and 4.16 graphically  show  the inputs, losses,  and  runoff
 losses  of N and P for the Juniata model  segment  100.   Figure 4.14  shows  the
 composition  of the nutrient inputs of N and P; Figure 4.15  shows the  composition
 and amounts  of the N and P  losses; and Figure 4.16 shows the composition of the
 nutrient runoff losses.  Finally, Table 4.24 summarizes the Total N and Total P,
 and components, in runoff losses  from selected model segments  for  relative
 comparisons.   All the  information   in these  tables and  figures are partial
 extracts of  model  results included  in the  tables  in Appendix  C.2.

                                       121

-------
TABLE 4.22  AGCHEM RESULTS FOR JUNIATA SEGMENT 10P (LT COMPOSITE CROP)
Pert. + Manure
Application
Runoff (in)
 1  Surface
   Interflow
   GW
   Total
Sed. Loaa  (t/a)
Nutrient Loss, (Ib/a)
  N03
      Surface
      Interflow
      GW
      Total
  NH3
      Surface
      Interflow
      GW
      Total
NH3 Sed. (Ib/a)
ORG. N Sed. (Ib/a)

Total N (Ib/a)

  P04
      Surface
      Interflow
      GW
      Total
P04 Sed. (Ib/a)
ORG. P Sed. (Ib/a)

Total P (Ib/a)
Plant Uptake (Ib/a)
  Nitrogen
      Surface
      tipper
      Lower
      Total
  Phosphorus
      Surface
      Upper
      Lower
      'Total
Net Mineralization  (Ib/a)
  Nitrogen
 • Phosphorus
Denitrification  (Ib/a)
NH3 IMMOB.  (Ib/a)
PO4 IMMOB.  (Ib/a)
1984
1221b N/a
531b P/a
5.98
6.67
8.15
20.80
0.61
0.31
*' 13.23
4.87
18.41
1.12
1.25
0.05
2.42
6.53E-03
2.57
23.41
0.65
0.57
0.00
1.22
2.69E-02
0.68
1.93
0.01
59.53
32.57
92.11
0.04
20.01
9.82
29.87
/a)
33.03
2.44
1.48
19.73
29.50
1985
1221b N/a
531b P/a
4.73
4.73
6.36
15.82
Q.56
0.63
8.15
3.33
12.11
2.24
0.81
0.01
3.06
7.33E-03
2.36
17.54
0.86
0.48
0.00
1.34
2.71E-02
0.63
2.00
0.04
67.48
27.78
95.30
0.04
19.37
2.56
21.97

32.26
2.13
1.20
20.77
27.42
1986
1221b N/a
531b P/a
4.26
5.78
6.59
16.63
0.29
0.43
7.66
3.25
1-1.34
1.14
0.89
0.01
2.04
3.53E-03
1.21
14.59
0.91
0.73
0.00
1.64
1.33E-02
0.32
1.97
0.06
67.58
29.21
96.85
0.07
21.14
1.89
23.10

32.54
2.23
1.16
21.46
25.56
1987
1221b N/a
531b P/a
2.18
3.93
6.48
12.59
0.12
0.05
7.94
3.08
11.07
0.38
0.41
0.01
0.80
1.28E-03
0.50
12.37
0.34
0.29
0.00 .
0.63
5.56E-03
0.13
0.77
0.03
67.61
30.49
98.13
0.05
20.94
1.79
22.78

32.52
2.19
1.16
21.26
27.47
Mean
1221b N/a
531b P/a
4.29
5.28
6.90
16.46
0.40
0.36
9.25
3.63
13.23
1.22
0.84
0.02
2.08
4.67E-03
1.66
16.98
0.69
0.52
0.00
1.21
1.82E-02
0.44
1.67
0.03
65.55
30.01
95.60
0.05
20.37
4.02
24.43

32.59
2.25
1.25
20.81
27.49
                                        123

-------
       Plant Nutrient Inputs  (Ib/ac)
           Juniata Segment  100
              Conventional Tillage
          Fertilizer
            114
       Mineralization
           30
Fertilizer
  48

Mineralization
    i
              Nitrogen   Phosphorus
             Conservation Tillage
          Fertilizer
            122
       Mineralization
          33  .
Fertilizer
  53   '

Mineralization
                                      1
              Nitrogen   Phosphorus
                       Hay
          Fertilizer
            44

       Mineralization
          30
Fertilizer
  16

Mineralization
    2
              Nitrogen   Phosphorus
Figure 4.14  Plant nutrient inputs -  Juniata Segment 100.

                         125

-------
          Nutrient  Runoff  Losses  (Ib/ac)
                     Conventional Tillage
                      Juniata Segment 100
              NH4-N
             2.3 11.7%
          Organic N
          2.4 12.2%,
Organic P
0.67 34.9%
                           NO3-N
                          14:9 76.0%
                                             PO4-P (Ads.)
                                              0.03 1.6%
                  Nitrogen
             PO4-P (Dlss.)
              t22 63.5%
      Phosphorus
                     Conservation Tillage
                      Juniata Segment 100
               NH4-N
              2.112.4%
           Organic N
           17 10.0%,
                          N03-N
                         13.2 77.6%
                  Nitrogen
             PO4-P (Diss.)
              1272.3%
      Phosphorus
                           \Hay
                      Juniata Segment 100
            Organic N
            0.619.0%
                         NO3-N
                        5.9 86.6%
                   Nitrogen
   Organic P
   0.16 20.8%
                                             PO4-P (Ads.)
                                             " 0.011.3%
           PO4-P(DiSS.)
            0.6 77.9%
      Phosphorus
          1984-1987 Annual Mean Values
Figure  4.16   Nutrient runoff'losses  - Juniata Segment 100,

                                127

-------
Review of  the  AGCHEM model  results  presented  in these  tables and figures, and
included in Appendix C, indicate the following:

      a. 'The nutrient applications  of fertilizer and manure  and the associated
          plant  uptake of  N  and P  dominate   the  input  and  output; portions,
          respectively, of  the nutrient balance of the croplands.

      b.  Mineralization is a significant source of plant available N, but not
          a significant source for plant available P in these simulations.

      c.  Plant uptake is a  direct function of the nutrient application rates on
          each segment and  land use category.   Annual uptake of N is typically
          702  to  902 of .the application, while  for P  the uptake is typically
        .  about 502  of the  application amount.  However,  there is considerable
          variation from segment to segment due to variations  in the application
          rates.

      d.  Uptake amounts were generally less than calculated  'composite target'
          amounts  when the  segment  application rates were similar  to or less
          than the targets, and  greater than  the targets when the application
          rates were greater  than the targets by a significant  amount,   this
         - simply indicated that on individual  segments higher application rates
          than expected  (or required by the  crop) produced  higher  yields and
          uptake  than was used  in  our calculation of the  ''composite'  target
          uptake.

      e.  Plant   available  N   is   generally   reduced  by  the  processes  of
          denitrification,  NH3  volatilization,  fixation, immobilization,  and
          leaching;  plant  available  P is reduced  primarily by  fixation and.
           immobilization-  Since NH3 volatilization and N  and P fixation are not
    .      specifically  modeled,  .immobilization was  used  as  the  primary
          mechanisms   for  losses   of  plant   available   nutrients.     HH3
           immobilization was typically in the  range of 152 to  302 of application
          amounts,  while P04 immobilization  was typically  about 502 of the
      .    application.  Denitrification was assumed to be small.

      f.  The  calibration process focused on  maintaining plant uptake amounts
          close  to  target  levels,   representing reasonable  losses  of  plant
          available nutrients, and keeping N03, NH3, and P0^ runoff losses in the
           range of expected values.                     T

      g.  Total N  and Total P runoff losses are generally in  the  range  of 10Z to
          .302,  and 52  to  152 of  application amounts, respectively.  N03 is the
           largest  component of Total N losses, followed by soluble NH3, Organic
          N and sorbed NH3.  For Total P, soluble P04 is  the largest, followed
          by  Organic P and sorbed  P0ซ.   The  loadings of sediment associated
           organics,  and sorbed NH3 and P04 are direct functions of the calibrated
           sediment loading rate, which  includes the impact of the assumed
           delivery ratio of 152.  Consequently, the sediment associated loadings
           are  considerably  less than  expected  edge-of-field  values due  to
           deposition processes  affecting delivery  at  the scale  of the  model
           segments.             .

                                      129

-------
Table 4.25  SUMMARY OF OBSERVED CONCENTRATIONS FROM SELECTED FIELD MONITORING STUDIES (Means and Ranges,  mg/l)
NH4-N
Mean Range
N03-N
Mean
Range
Total
Mean
N
Range
Ortho-P
Mean Range
Total P TN/TP ratio
Mean Range
Owl Run Watershed (In Segment 230) • 7.8 ซq. ml.
OOA
OOB
ooc
000
Nomlni
ONI
QN2
- 4.5
- 0.2
- 1.8
- 1.3
•q.
•q-
•q.
•q.
ml.
ml.
Ml.
nl.
Cr. Watershed
- 5.8
- 0.9
•q.
aq.
mi.
ml.
2.44
0.61
0.32
0.43
(In Potomac
0.33
0.30
0.098-28.9
0.0-2.6
0.0-1.2
0.0-1.35
Baain, Below
0.01-1.48
0.0-1.58
2.90
1.61
, 1.17
2.60
0.0-9.78
0.07-6.5
0.0-3.9
0,0-7.75
Fall Line) - 6.7 aq.
0.63
1.15
0.11-1.86
0.12-2.09
12.60
7.92
6.29
8.40
mi.
5.75
7.85
1.7-89.5
2.4-33.1
0.0-15.2
0.0-28.9

1.80-15.1
2.46-39.7
0.68 0.02-4.9
0.55 0.05-1,38
0.04 0.0-.15
0.26 0.0-1.56

0.03 0.0-0.14
0.03 0.0-0.16
2.96 0.29-17.6 4.3
1.48 0.29-2.95 5.4
0.94 0.0-16.2 6.7
1.60 0.0-9.4 . 5.3

0.91 0.05-12.9 6.3
1.57 0.10-16.7 5.0
Jug Bridge, Monocacy River (in Segment 210) : 817 sq. ml.                                   .




   Base Flow data          0.37  0.05*0.98      2.100.05-3.57      3.23  0.65-5.37       0.23  0.08-0.52      0.44  0.14-0.88      7.3




   Storm Event             0.25  0.11-0.65      1.77  0.05-3.10      4.32  0.95-9.1         0.16  0.01-.44       1.10  0.06-1.9       3.9







Patuxent Watershed - Field Sites (Segment 340)           -




   Station H01 - 5 ac.     0.09  0.0-0.26       0.88  0.07-5.3       3.87  0.77-18.8       0.63  0.01-1.50      1.21  0.30-5.2       3.2




   Station CA1 - 8 ac.     0.07  0.02*0.24      0.33  0.02-0.70      7.49  1.60-45.0       0.13  0.03-0.40      1.54  0.40-3.6       4.9

-------
Ul
OJ
      Table 4.27  SUMMARY OF PATUXENT (SEGMENT 340) SIMULATED CONCENTRATIONS by ACCHEM (Means and Ranges, mg/l)
                                   NH4-N                N03-N              Total N                 P04-P     -       Total P       TN/TP ratio
                                                  -                                                               •                     '
                              Mean    Range        Mean    'Range       Mean     Range      '  Mean     Range       Mean    Range
      CONVENTIONAL TILLAGE
Surface & Interflow      0.38  0.04-2.51      2.58  0.0-17.60      9.96  1.83-37.10      0.41  0.12-1.90      2.340.35-10.80       4.3


Total                    0.02  0.01-2.33      1.41  0.05-14.60     1.60  0.50-36.10     0.009  0.00-1.54     0.0560.0-10.50       28.6




CONSERVATION TILLAGE


Surface t Interflow      0.45  0.06-4.74      2.30  0.01-16.40    10.30  1.68-35.10      0.41  0.15-2.61      2.42  0.35-6.31       4.3


Total                    0.02  0.02-4.25      1.26  0.07-14.90     1.41  0.44-31.60     0.007  0.0-2.12      0.042  0.0-6.03       33.6




HAY                          •


Surface & Interflow      0.13  0.01-1.53      0.35  0.0-1.16       5.38  0.31-14.20      0.26  0.04-1.20      1.57  0.11-3.96       3.4
                                                                              ^

Total                    0.01  0.0-1.42       0.99  0.06-2.96      1.07  0.45-13.40     0.004  0.0-1.03      0.023  0.0-3.72       46.5




Notes:    1. Surface I Interflow concentrations are calculated as daily load/dally flow for these components

             when the dally flow (surface and Interflow only) exceeds 0.1 in.


          2. Total concentrations are based on both storm (I.e. surface and Interflow) and

              baseflow values for the entire simulation period

-------
The  extreme values  of  the  ratio,  in the  range  of  100 to  500,   for  total
concentrations  in the  Rappahannock  (Segment 230)  in Table  4.26  are due to
extremely low mean TP concentrations  as reflected in the observed water quality
data.  Since baseflow will have higher N and lower P than storm runoff, the TN/TP
ratio  will  increase as  baseflow and subsurface  contributions  increase  as
reflected in the AGCHEM results.

4.3.7  AGCHEM Conclusions and Recommendations

The AGCHEM application to the cropland areas  of the Chesapeake Bay drainage has
satisfied the objectives outlined at  the beginning of Section  4.3, namely to (1)
represent nutrient balances for  the major  cropland categories in  each  model
segment,  (2) allow  investigation of  sensitivity to  climate variations  and
agricultural  practices,  and  (3) provide a  mechanism  to project  impacts of
nutrient management  alternatives for agricultural cropland.   The  results  are
consistent  with   expected   nutrient  balances,   observed ranges   of  runoff
concentrations, and  expected ranges  of nutrient loadings from  cropland  areas
(discussed below  in  Section 4.4).      •   •          •

The AGCHEM approach  to modeling  the  nutrient balances on agricultural cropland
Is exactly the type of approach recommended by the Chesapeake Bay Honpoint Source
Evaluation Panel  in  their report to the U.S.  EPA and the GBP Executive Council.
In  March  1990,  the EPA  Administrator  convened  the  Panel to  "assess  the
effectiveness of current efforts to reduce nonpoint source  loadings of nutrients
entering the  Bay  system".   They reviewed and evaluated the effectiveness of a
wide range of nonpoint source programs,  including program design, implementation,
budgets, modeling and assessment methods, and research efforts.  One of the key
recommendations of the Panel related to new approaches to nutrient control was
as follows:     •                        ;

       "The  Panel  recommends that  the  Bay' jurisdictions and  the federal
       agencies  develop  a mass  balance accounting system,  where nutrient
       loadings  are balanced by the nutrients  removed from  the.system plus'
       those  which are  introduced and stored" (Chesapeake Bay  Nonpoint
       Source Evaluation  Panel,  1990).            ,

The AGCHEM application provides an initial assessment of this  'mass balance' for
cropland  areas,  a tool for assisting in the 1991  Re-evaluation of  the  40X
nutrient reduction goal of the  Bay Agreement, and a framework  for future efforts
on nutrient management based on the  mass balance approach. . Recommendations to
improve the use and  application of AGCHEM by CBP include  the following:

       a.  A detailed review and refinement of the current  fertilizer and manure
          application  rates, and -changes  expected under nutrient  management
          alternatives  should  be initiated  since  these rates are  the starting
          point for  nutrient reduction from cropland. Plant uptake amounts under
          reduced nutrient application rates will need to.be investigated to
          ensure  that crop needs  are  met;  plant  uptake rates may  need  to be
          adjusted.

       b.  Detailed site applications of AGCHEM should be diligently pursued by
                                      /

                                      135

-------
In the Phase I,  the .first  steps  in the nonpoint  calibration effort involved a
review and evaluation of nonpoint loading rates associated with individual land
uses and nonpoint parameters  used in the original  Watershed Model.  The goal was
to define the expected range  of loading rates from the available literature, as
a basis  for  evaluating  and calibrating the  model predicted  loading rates, and
determine if any  changes or adjustments to the original nonpoint parameters could
be justified.  The State representatives on the  CBP Nonpoint Source Workgroup
provided data summaries of monitoring projects and  studies  conducted in their
respective regions to supplement  the efforts of the  CBPO and AQUA TERRA, on this
task.  Table 4.29, developed  in Phase I, provides  a brief summary of the results
of this effort, with ranges of loading rates for individual nutrient forms for
the major land use categories  in  the Watershed Model.

Unfortunately, the available data was not sufficiently extensive or detailed to
allow delineation of'different loading rates for different portions of the Bay
drainage area. As shown in Table  4.29 and in the  original literature, the rates
are quite variable with  cropland  showing the greatest variability and forest the
least variation.  For urban  pervious  and  impervious areas,  the average annual
National Urban Runoff Program (NURP) loads,  as complied by Schueler (1987) were
used  to  supplement the  information in Table 4.29  and guide  the  calibration
adjustments.  In  summary, as shown in Table 4.29,  the  model predicted rates could
fall within a relatively large  range and still be  consistent with the literature
data.         -            .             •         - •     ^ •"'•    '

In fact,  the second part of  this effort involved an assessment of consistency
between the original model  rates  and the literature.   In order to assess whether
refinements  to the  potency factors and washoff  parameters  were Justified,  we
compared the model predicted  loading rates (which are based on the input potency
factors and washoff parameters) with the loading- ranges shown in Table 4.29.  Our
reasoning was that if the model loading rates were consistent with the literature
values,  we  did not have   a  basis for changing  the parameters  except through
calibration.  The comparison showed that the model rates were  consistent with the
literature   values,  partly  due  to   the  wide  range  in  the  literature.
Consequently, no changes in potency factors or washoff rates were made initially
until the calibration indicated  that changes were justified.

However, the subsurface concentrations used in the original Watershed Model were
based on calibration at instrearn sites during low flow periods.  Thus, the sane
concentrations were used for all  land uses.  In the  Phase  I work,  we developed
different subsurface concentrations for each land .use for  consistency with the
hydrologic/land  use  variability   and  to better  represent  effects of  land use
changes.  Initial subsurface concentrations  were estimated  for NH3,  N03, P04,
BOD, and 00 from  a variety of sources including past HSPF applications within the
Bay  drainage,  data from sites within the  Patuxeat  and Monocacy rivers,   and
selected literature review articles.   Generally, the highest pollutant values
were  estimated for  cropland areas, with conservation  tillage  having somewhat
higher values  than conventional  tillage.  Urban subsurface concentrations were
the  next highest, followed  by pasture, with forest areas having  the lowest.
concentrations.   Subsurface 00 values were assumed  to be the'same for all land
uses, and the  same  seasonal  (i.e. monthly)  variations  as used in the original
model were  imposed.  These initial values were then adjusted as needed during


                                      137

-------
the calibration process based on the observed data and model predictions for the
calibration sites throughout the drainage area.

In Phase II,  Table 4.30 was developed by the CBPO from further analyses of field
studies specific to the .Chesapeake Bay drainage area; many of the studies were
provided by members  of  the NRTF.  Note that a number of the entries  in Table 4.30
indicate 'no data'  due to  insufficient  information for  selected land uses and
individual nutrient species.  Table 4.31 from Beaulac (1980) and Reckhow et al
(1980)  show  the range and median  of  expected  loading  rates  (or  export
coefficients) only for Total N and Total P based on nation-wide data from field
studies.  Typically, Total N and Total P are better defined than the individual
components  of  NH3,  PO*. and organics;  N03 loading  rates are  usually  better
defined than the other  species since it is the most commonly measured nutrient
form.               ,

Tables 4.32, 4.33, and 4.34 are tabulations of the simulated loading rates from
the Phase II simulations for each nutrient  form, plus BOD, for each year of the
simulation, along with the  mean annual value, for model segments 60,  180, and
280, respectively.  These three segments are the largest  segments in each of the
three major basins  • Susquehanna, Potomac, and James.  Similar tabulations are
provided for additional model  segments  in Appendix C. .  Note that Total  N and
Total P  are  calculated from the other nutrient forms and BOD  as described in
Section 5.0 and  the table footnote.

Table 4.35 summarizes the mean annual rates  for  17 model  segments'throughout the
Bay  drainage, area,  including  both  above  and below Fall Line  segments;  the
detailed tabulations  for these model segments  are included in Appendix C.I.
Table 4.3.6 provides a capsule summary of the mean and ranges of loading rates for
the  segments  lisced in Table 4.35; although this summary is not based  on all
model  segments  (segment  tabulations are being  currently compiled), ' it does
provide a good indication of the simulated mean and range for each land use and
selected nutrient forms.  Comparing the mean annual rates in Table 4.36 with the
expected means and ranges from Tables 4.29  through 4.31  indicates the following
general conclusions:

      a.  Generally, the simulated annual loading rates  are within the range of
          expected  values  shown in  Table  4.29  with some deviations.   Annual
          rates  for P0ป from forest and pasture, and NH4 from forest occasionally
          tend  to be  toward the  lower  end of the  defined  range.   Annual P04
          rates  from the cropland areas are somewhat higher than the defined
          range.

      b.  For non-cropland  categories., the Total  N and Total' P simulated values
          compare favorably with both the expected means and ranges.

      c.  For  the  cropland categories  of Conventional  Tillage,  Conservation
          Tillage,  and  Hay, the  Total  N and  Total P simulated values are
          generally close  to the mean lumped cropland data;  the Hay values are
          generally less than the mean, while the tillage categories are usually
          greater than the  mean but  well within the observed range.
                                      139

-------
TABLE 4.31  NUTRIENT EXPORT COEFFICIENTS FROM BEAULAC (1980) AND
            RECKHOW ET AL. (1980)
                                  Phosphorus
Land Use
Forest
Pasture
Cropland ,
Urban
Land Use
Forest
Pasture
Cropland
Urban
Minimum

0.017
0.12
0.09
0.17
Minimum
'.
1.23
1.32
1.96
1.32
Median

0.18
0.72
0.89
0.98
Nitrogen
Median

2.32
4.63
8.03
4.91
Maximum

0.74
4.37
1.96
5.56
Maximum

5.58
27.52
71.0
34.3
  Source:  Nutrient export coefficients from (Beaulac (1980) and
           Reckhow et al. (1980).
                              141

-------
TABLE 4.33  WATERSHED MODEL SEGMENT 180 ANNUAL LOADING RATES (LB/AC/YR)
LAND USE
FOREST
CONV.TILL
CONSER. TILL
PASTURE
URBAN-PERV.
HAY
URBAN-IMPV.
MANURE SEC.
YR
84
85
66
87
AVERAGE
84
85
86
87
AVERAGE
84
85
86
87
AVERAGE
84
85
86
87
AVERAGE
84
85
86
87
AVERAGE
84
85
86
87
AVERAGE
84
85
86
87
AVERAGE
84
85
86
87
AVERAGE
NH4
0.072
0.053
0.046
0.059
0.057
2.73
3.05
1.38
2.79
2.49
3.82
2:08
1.42
2.98
2.58
0.10
0.07
0.06
0.08
0.08
0.31
0.20
0.16
0.22
0.22
0.42
0.13
0.15
0.30
0.25
1.42
1.34
1.36
1.39
1.34
243.0
198.2
170.7
202.0
203.5
K03
4.12
3.41
2.99
3.47
3.50
22.71
12.74
9.15
13.70
14.58
21.85
11.20
7.06
11.63
12.94
6.08
5.16
4.40 '
5.10
5.19
7.61
6.23
5.20
6.19
6.31
4.68
1.72
1,78
2.18
2.59
3.04
2.85
2.89
2.95
2.93
60.7
49.5
42.7
50.5
50.9
P04
0.012
0.005
0.004
0.009
0.008
1.46
1.29
0.92
0.97
1.16
2.06
1.36
1.00
1.16
1.40
0.06
0.04
0.03
0.04
0.04
0.25
0.14
0.11
0.17
0.17
2.14
0.69
0.88
1.49
1.30
0.66
0.62
0.63
0.64
0.63
60.7
49.5
42.7
50.5
50.9
1 BOD
7.91
2.81
2.93
5.09
4.69
400.00
86.00
84.90
174.00
186.23
181.00
44.80
31.10
80.20
84.28
48.00
29.80
25.00
35.30
34.53
54.90
29.50
22.40
36.60
35.85
203.00
42.70
27.40
89.10
86.25
86.50
81.00
82.30
84.00
83.45
4,251.6
3.468.5
2.987.8
3.535.7
3.560.9
ORC-N





5.86
1.20
1.47
2.44
2.74
4.66
0.89
0.67
1.78
2.00










1.97
0.36
0.27
0.79
0.85





1.822.1
1,486.5
1,280.5
1.515.3
1.526.1
ORG-P





1.61
0.33
0.41
0.67
0.76
1.23
0.24
0.18
0.47
0.53







1


0.53
C.W
0.07
0.21
0.23





364.4
297.3
256.1
303.1
305.2
TN •
4.61
3.61
3.19
3.80
3.80
37.66
18.36
13.35
.21.70
22.77
33.21
14.88
9.64
17.67
18.85
8.72
6.81
5.79
7.05
7.09
10.83
7.99
6.55
8.35
8.43
10.30
2.89
2.63
4.69
5.13
.04
.48
.61
.79
.73
2,125.8
1.734.2
1,493.9
1,767.9
1.780.5
TP •
0.072
0.026
0.026
0.048
0.043
3.98
1.82
1.52
2.04
2.34
3.70
1.70
1.25
1.81
2.12
0.42
0.27
0.22
0.31
0.31
0.67
0.37
0.28
0.45
0.44
3.13
0.88
1.01
1.90
1.73
1.31
1.23
1.25
1.28
1.27
425.2
346.8
298.8
353.6
3S6.1
* For all land uses, except CUT. CST. HAY and MANURE  land uses, TN and TP are calculated by:
TN * NH3 * N03 + 0.053*800               TP ซ P04 * 0.0076*600

For CNT. CST, and HAY land uses, TN and TP are calculated by:
TN ซ NH3 ป N03 + ORGN * 0.3*0.053*800    TP • P04 ป ORGP * 0.3*0.0076*800

For the MANURE land use, TN and TP are calculated by:
TN * NH3 + N03 + ORGN                    TP ซ P04 * ORGP
                                                     143

-------
TABLE 4.35   SUMMARY OF  AVERAGE ANNUAL LOADING RATES FOR EACH LAND USE IN SELECTED
             MODEL  SEGMENTS  (LB/AC/YR)                 '
SEGMENT
SEGMENT 10
E. Branch Susquehama





SEGMENT 30
E. Branch Susquehanna

' -



SEGMENT 40
E. Branch Susquehanna


.. /


SEGMENT 50
U. Branch Susquehanna





SEGMENT 60
•U. Branch Susquehanna
1 .• '




SEGMENT 110
Lower Susquehanna





SEGMENT 160 .
N. Branch Potomac





LANOUSE
Conv. Till
Cons. Till
Hay
Pasture
Manure Seg
Forest
Urban
Conv. Till'
Cons. Till
Hay
Pasture
Manure Seg
Forest?
Urban
Conv. Till
Cons. Till
Hay
Pasture
Manure Seg .
forest
Urban
Conv. Till
Cons. Till
Hay
Pasture
Manure Seg
Forest
Urban
Conv. Till
Cons. Till
Hay
Pasture
Manure Seg
Forest:
Urban
Conv. Till
Cons. Till
Hay
Pasture
Manure Seg
Forest
Urban
Conv. Till
Cons. Till
Hay
Pasture
Manure Seg
Forest
Urban
NM4
2.39
2.12
0.76
0.08
212
0.04
0.59
2.31
2.83
1.02
0.09
255
0.05
1.07
3.21
3.15
1.13
0.09
287
0.05
1.34
6.95
8.77
1.41
0.11
301
0.10
0.65
.4.93
5.14
1.33
0.10
261
0.09
0.72
2.08
2.39
0.49
0.21
221
0.09
0.95
4.64
3.34
0.31
0.11
240
0.06
0.66
N03
13.80
12.10
8.84
3.00
53
1.91
4.53
13.00
12.80
10.00
5.89
64
5.10
7.17
12.60
11.20
8.58
5.88
72
6.02
7.03
22.00
20.50
15.00
4.26
75
3.00
5.97
20.80
19.40
14.20
5.16
65
4.62
7.74
11.60
11.90
4.79
7.60
55
6.04
8.26
13.20
9.83
3.83
6.98
60
3.61 '
5.58
P04
0.95
0.93
1.10
0.02
53
0.01
0.31
0.96
1.07
1.17
0.02
64
0.01
0.38
1.12
0.96
1.65
0.02
72
0.01
0.66
1.38
1.38
1.30
0.02
75
0.01
0.26
1.19
1.22
1.06
0.02
65
0.01
0.33
1.31
1.37
0.75
0.01
55
0.01
0.36
0.96
1.09
2.29
0.08
60
0.01
0.38 •
ORG-N
3.91
2.48
1.20
1.06
1.593
0.27
3.28
3.49
2.37
1.08
1.39
1.915
0.34
3.56
5.79
3.35
1.29
1.88
2,155
0.43
4.13
4.25
3.43
1.29
1.06
2.257
0.35
3.88
4.62
3.98
1.36
1.62
1.962
. 0.33
3.64
18.02
9.71
5.92
3.39
1.657
0.35
5.99
6.76
4.03
1.93
2.26
1.799
0.24
3.72
ORG-P
0.76
0.49
0.23
> 0.15
319
0.04
0.42
0.69
0.46
0.21
0.20
383
0.05
0.50
1.14
0.70
0.25
0.27
431
0.06
0.58
0.79
0.51
0.23
0.15
452
0.05
0.56-
0.95
0.58
0.26
0.23
393
0.05
0.52
3.06
1.73
0.97
0.49
.332
0.04
0.85
1.16
0.70
0.33
0.32
360
0.04
0.54
TN
20.10
16.70
10.80
4.U
1,858
2.22
8.40
16.80
18.00
12.10
7.37
2.234
5.49
11.80
21.60
17.70X
11.00
7.85
2,514
6.50
12.50
33.20
32.70
17.70
5.43
2,633
3,45
10.50
30.55
28.52
16.89
6.88
2,287
5.04
12.10
31.70
24.00
11.20
11.20
1.933
6.48
15.20
24.60
17.20
6.07
9.35
2.099 ,
3.91
9.96
TP
1.71
1.42
1.33
0.17
372
0.05
0.73
1.65
1.53
1.38
0.22
447
0.06
0.88
2.26
1.66
1.90
0.29
503
0.07
1.24
2.17
1.89
1.53
0.17
527
0.06
0.82
2.14
1.80
1.32
0.25
458
0.06
0.85
4.37
3.10
1.72
0,50
387
0.05
1.21
2.12
1.79
2.62
0.40
420
0.05
0.92
                                                      145

-------
TABLE 4.35   (Continued)
SEGMENT
SEGMENT 310
Appomattox
SEGMENT 340
Patuxent
SEGMENT 380
Chester
SEGMENT 400
Choptank
SEGMENT 430
Pocomoke
SEGMENT 450
Lower Susquehanna
 SEGMENT  500
 Lower Patuxent
LANDUSE
Conv. Till
Cons. Till
Hay
Pasture
Manure Seg
Forest
Urban
Conv. Till
Cons. Till
Hay
Pasture
Manure Seg
Forest
Urban"
Conv. Till
Cons. Till
Hay
Pasture
Manure Seg
Forest
Urban
Conv. Till
Cons. Till
Hay
Pasture
Manure Seg
Forest
Urban
Conv. Till
Cons. Till
Hay
Pasture
Manure Seg
Forest
Urban
Conv. Till
Cons. Till
Hay
Pasture
Manure Seg
Forest
Urban
Conv. Till
Cons. Till
Hay
Pasture .
Manure Seg
Forest

Urban
NH4
1.90
1.68
0.57
0.05
242
0.05
0.69
1.55
1.22
0.11
0.07
216
0.04
0.64
1.59
1.14
0.25
0.07
190
0.04
0.45
1.88
1.18
0.30
0.07
190
0.04
0.48
3.20
2.19
0.16
0.07
190
0.04
0.55
1.55
1.16
0.21
0.08
212
0.05
0.68 .
1.15
0.71
0.13
0.08
211
0.05

0.58
H03
12.50
11.50
4.33
1.51
61
0.68
2.35
10.20
8.69
2.47
2.64
54
1.33
4.72
11.20
9.92
3.59
4.73
48
2.20
4.69
11.90
10.20
3.80
4.69
48
2.28
4.83
17.80
16.60
2.94
4.54
.*•
2.30
*•ป
11.00
10.20
3.10
5.55
53
2.35
5.25
8.27
6.42
2.57
4.95
53
2.35
I
5.00
P04
0.93
0.89
1.01
0.03
61
0.01
0.30
0.63
0.46
0.15
0.03
, 54
0.01
0.28
0.42
0.35
0.18
0.04
48
0.01
0.22
0.53
0.44
0.22
0.04
48
0.003
0.23
1.34
•1.22
0.21
0.04
48
0.004
0.26
0.89
0.63
0.35
0.03
:53
0.003
0.31
0.56
0.39
0.15
0.04
53
0.003

0.27
ORG-N
8.60
5.02
2.82
2.31
1,815
0.28
2.93
6.75
ซ.59
2.18
0.53
1.623
0.22
3.98
4.71
3.14
1.60
1.04
1.425
0.26 .
2.11
3.82
2.52
0.92
0.95
V*25
0.10
2.20
3.80
2.41
0.74
1.74
1,425
0.13
2.13
3.75
2.94
1.71
0.89
1,594
0.10
2.31
9.58
6.07
4.91
1.25
1,583
0.14

2.29
ORG-P
1.62
0.99
0,54
0.33
363
0.04
0.42
1.72
1.15
0.53
0.07
325
0.03
0.57
0.93
0.62
0.29
0.15
285
0.03
0.30
0.68
0.46
0.16
0.13
285
0.02
0.31
0.68
0.44
0.13
0.25
285
0.02
0.30
0.86
0.65
0.31
0.13
319
0.02
0.34
,2.33
1.44
0.89
0.18
316
0.02

0.33
TN
23.00
18.20
7.72
3.87
'2,118
1 .01
5.97
18.50
14.50
4.76
3.24
1,893
1.59
9.34
17.50
14.20
5.44
5.84
1,663
2.50
7.25
17.60
13.90
5.02
5.71
1,663
2.42
7.51
24.80
21.20
3.84
6.35
1.663
2.47
7.42
16.30 •
14.30
5.02
6.52
1.859
2.50
8.24
19.00
13.20
7.61
6.28
1,847
2.54

7.87
TP
2.55
1.88
1.55
0.36
424
0.05
0.72
2.35
1.61
0.68
0.10
379
0.04
0.85
1.35
0.97
0.47
0.19
333
0.04
0-52
1.21
0.90
0.38
0.17
333
,0.02
0.54
2.02
1.66
0.34
0.29
333
0.02
0.56
1.75
1.28
0.66
0.16
372
0.02
0.65
2.89
1.83
1.04
0.22
369
0.02

0.60
                                                     147

-------
TABLE 4.36  OVERALL SIMULATED MEAN AND RANGE of LOADING RATES FROM SELECTED
            SEGMENTS LISTED IN TABLE 4.35 (LB/AC/YR)


                    NH3-N          N03-N        P04-P              TN            TP

FOREST
  Mean               0.06           2.77          0.01            3.09          0.04
  Range           0.04-0.10      0.55-6.04    0.003-0.03       0.81-6.50    0.02-0.09

CONVENTIONAL TILL
  Mean               2.57          13.19          0.95           22.04          2.18
  Range           0.90-6.95    8.27-22.00     0.42-2.17   ,  15.20-33.20    1.21.4.37
                      ^             "                '   .

CONSERVATION TILL
  Mean               2.24          11.69          0.91           17.90          1.68
  Range           0.71-8.77    6.42-20.50     0.35-1.97      11.0-32.70    0.90-3.10
PASTURE
Mean
Range
URBAN*
Mean
Range
HAY
Mean
Range
MANURE
Mean.
Range
0.10
0.05-0.25
0.65 !.
0. 21-1. 34
0.51
0.11-1.41
230
190-301
4.70
1.01-8.53
5.49
1.27-9.73
6.13
2.47-15.00
57
-48-75
0.06
0.01-0.28
0.33
0.16-0.66
0.87
0.15-2.29
57
48-75
6,47
2.28-11.20
9.27
3.43-15.20
8.72
3.84-17.70
2010
1663-2633
0.29
0.10-0.91
0.78
0.42-1.24
1.26
0.34-2.62
402
333-527
*  - Includes area weighted totals from both pervious and impervious segments..
                                         149

-------
                                  SECTION  5.0

                 RIVER TRANSPORT AND TRANSFORMATION SUBMODEL
This section discusses the HSPF modules used to simulate the instream transpo'rt
and water quality processes considered in  the Phase  II Watershed Model.   Each
Above Fall Line  (AFL) Watershed Model segment is associated with a corresponding
river  channel/reservoir  which  receives  nonpoint,   point,   and  atmospheric
deposition loadings.  These inputs  of flow, heat,  sediment,  and inorganic and
organic  nutrients  then  undergo  the various  instream  processes,  including
phytoplankton growth,  and are  transported downstream  toward the Bay.  Table 4.2
lists the water quality constituents simulated in Phase II.

A major  focus of Phase  II was to  include a more detailed  representation of
instream  nutrient processes,  particularly with  respect to  sediment-nutrient
interactions. With  the exception of a simple organic N and P settling algorithm,
previous versions of the  Watershed Model  have not considered sediment transport
and  the  associated  sediment-nutrient  interactions  that  are  critical  to  an
accurate  representation  of  the water  quality  of  the .rivers  in the  basin.
Consequently, the Phase  II effort  included significant modifications  to the
nutrient  simulation routines  in  HSPF  to  allow  inorganic  N  (ammonium)  and P
(orthophosphate) to adsorb to sediment and undergo settling,  resuspension, and
transport.  Also, the denitrification algorithm was modified  to provide a more
flexible representation of nitrate losses via this process.

The methods  used to simulate instream processes in run-of-the-river and reservoir
reaches are described in the following subsections.  Additional details of the
model equations and algorithms are contained in the HSPF Users Manual (Johanson
et at, 1984).

5.1   OVERVIEW OFTHE INSTREAM MODEL

Within  the  HSPF  program, the  RCHRES  module  sections are  used  to  simulate
hydraulics,  sediment transport, water temperature, and water  quality processes
that result in delivery of flow and loadings to the Bay.  The structure chart of
the RCHRES module sections is  shown  in Figure 5.1.  Each module section consists
of a group of subroutines that perform the  individual process simulations.  For
example the HYDR module performs the flow/hydraulic simulation that-drives the
sediment transport and pollutant advection. The water quality processes in the
Watershed Model are performed by the RQUAL module, which is futther subdivided
into submodules for performing  simulation  of different water  quality processes
as shown in  Figure 5.2.  All processes in  the Watershed Model are simulated using
a one hour  time step.
                                      151

-------
                         RQUAL
OXRX
  Simulate
  primary
  dissolved
  c-xygen and
  BOD
  reactions
                           Simulate
                           constituents
                           involved in
                           biochemical
                           transform-
                           ations
NUTRX
  Simulate
  primary
  inorganic
  nitrogon and
  .phosphorus
  reactions
PLANK
  Simulate
  behavior of
  plankton and
  associated
  reactions
PHCARB
  Simulate pH
  and inorganic-1
  carbon
  reactions

   Figure 5.2  Structure chart of the RQUAL section of the  RCHRES module.

                                153

-------
Geological Survey (Arnibruster, 1977)  and the Susquehanna River Basin Commission
(Jackson, 1979).                                     ,

5.2.2  Water Temperature'

Water temperature is simulated in the Watershed Model because of its fundamental
effect upon water quality processes.   Dissolved oxygen saturation level and the
rates of  BOD decay, nitrification, and phytoplankton growth  are  all directly
affected by water temperature.

Instream water temperature processes  are simulated by the HTRCH module, shown in
Figure 5.3.  Changes in heat content  of a reach result primarily from advective
transport, runoff and rainfall entering the reach, and heat transfer processes
across the air-water interface.  The advective transport is  simulated by treating
water temperature as*a thermal concentration.   This allows use of the standard
method  within  the  model  for computing  advective  transport  of  dissolved
constituents.               •                      •

The  mechanisms affecting heat transfer  across  the  air-water interface  are
absorption  of  solar (shortwave)   radiation,  absorption/emission  of  infrared
(longwave) radiation, conduction-convection, and evaporation.   Each mechanism is
evaluated separately, and the net transfer is the sum of the individual effects.
Solar radiation absorption  is computed by the  following equation:
      . -,                    -                                       *
      QSR    -  0.97 * CFSAEX * SOLRAD * 10.                             (5.1)

where:
      QSR    -  radiation  absorbed  (kcal/mz/hr)
      0.97   -  fraction of  incident radiation  absorbed
      CFSAEX -  fraction .of: gage radiation that  is incident to the surface
      SOLRAD -  solar radiation (langleys/hr)
      10.    -  conversion from langleys to kcal/m2

The net infrared radiation transfer,  considering both atmospheric radiation and
back radiation  from the water surface,  is dependent on temperature  and cloud
cover.  It is computed using  the following equation:                     .
where:
      QB     - SIGMA * (TWKELV* - KATRAD * 10ซ * CLDFAC * TAKELV*)        (5.2)

     • '
      QB     - net transport of infrared radiation (kcal/m2/hr)
      SIGMA  - Stefan-Boltzmann constant multiplied by 0.97 (kcal/m2/hr/K*)
      TWKELV - water temperature ('K)                                  .      '
      KATRAD - atmospheric longwave radiation coefficient
      CLDFAC - cloud factor - 1.6 + (0.0017 * CLOUD2)
      CLOUD  - cloud cover (tenths)
      TAKELV - air temperature corrected for elevation (*K)

Air temperature is corrected for  elevation by application of a lapse rate to the
gage temperature.  If it is raining,  the  lapse rate is 0.00194 C/ft; otherwise,
the lapse rate is selected from a table of  24 hourly values ranging from 0.0019
to 0.0028 C/ft.     .

                                      155

-------
Conductive-convective heat transfers  are  caused  by  the temperature difference
between the air and the water; the net transfer is computed as follows:

      QH     - CFPRES * KCOND * 1
-------
Since sediment migration characteristics and adsorptive properties of sediment
vary with particle size,  SEDTRN considers three size fractions (sand, silt, and
clay), each with  its  own set of  parameters.  The silt and  clay fractions are
categorized as cohesive sediments,  and are  simulated differently from the sand
or noncohesive fraction.  The module maintains storages  of each size fraction in
the bed and in suspension, and computes scour and deposition fluxes between these
storages.  It also computes a bed depth based on  sediment  porosities and width
and length of the bed.  SEDTRN assumes a single bed layer,  so that all sediment
is available for scour; burial and armoring are not considered.

Scour  and  deposition of cohesive  sediments depends on the bed  shear stress
computed in'the HYDR module, and is based on equations developed by Krone (1962)
and Partheniades (1962).  Whenever bed shear stress in the reach is  less than the
user-defined critical shear stress  for deposition, sediment settles to the bed.
Conversely, whenever  the  shear is  greater  than the critical shear  stress for
scour, erosion of sediment  from the bed occurs. The following two equations are
used  for each  size  fraction  to  compute   the  deposition  and  scour  fluxes,
respectively.

      DEPCNC - CONG * (1.0 - EXP(-W/AVDEPM) * (1.0 -  TAU/TAUCD))         (5.5)

where:
      DEPCNC - concentration of suspended sediment lost to deposition (mg/L/hr)
      CONC   - concentration of suspended sediment at start of Interval (mg/L)
      V      - settling velocity of size fraction (m/hr)
      AVDEPM - average depth of reach (m)                                  -
      TAU    - bed shear stress (kg/m2)
      TAUCD  - critical shear stress for deposition (kg/m2)


      SCRCNC -  M/AVDEPM * 1000 * (TAU/TAUCS - 1.0)                      (5.6)

where:                        .               .
      SCRCNC - concentration of suspended sediment added- to .
               suspension by scour (mg/L/hr)
      M      -credibility coefficient (kg/mVhr)
      1000   - conversion from kg/m3 to mg/L
      TAUCS  -critical shear stress for scour (kg/m2)

The principal  calibration parameters for cohesive sediments are  the settling
velocity (W), credibility coefficient (M), and the critical shear stresses (TAUCD
and TAUCS).  The  critical  stresses  affect  the timing of deposition and scour,
while V and M are useful in adjusting the magnitudes.  In  the Watershed Model,
initial values of TAUCD and .TAUCS were  determined from time series plots of the
simulated bed shear stresses in each reach.   These adjustments were made so that
occurrence of scour and deposition  was  restricted to extremely  high  and low
stress periods, respectively. As an example,  based on the plot of shear stress
shown in Figure 5.5, the  TAUCS and TAUCD values for REACH 290 in the James River
were set to the following values:
                                      159

-------
                     Critical Shear Stresses  for REACH  290
                                   (lbs/ft2)

                           DEPOSITION     SCOUR

                  SILT        0.030       0.137
                  CLAY        0.029       0.135

Values of M and KSER  (the soil erosion parameter for pervious areas) were Chen
calibrated by comparing simulated and observed suspended sediment concentrations
during, scour events and periods of high loading.

Sand transport  in SEDTRN  is simulated by one of three optional methods.   The
Toffaleti and Colby  methods  are  detailed formulations  primarily applicable to
wide rivers, while the user-specified power function (of velocity) is simpler and
more generally applicable.  The power function method is used in the Watershed
Model because  of the variety of  channel types encountered in  the  basin,  and
because  of  the  greater degree of  control over the sand simulation  that this
option provides.,

Sand scour and deposition in SEDTRN depends on the amount of material the flow
is. capable of carrying.  If the amount of  sand being transported is less than the
flow can carry for the hydrodyriamic conditions of the reach,  sand will be scoured.
from the bed.  Conversely, deposition occurs if the sand in suspension exceeds
the flow's capacity.  The potential sandload or sand carrying capacity for the
power function method is computed directly as:                           '

      PSAND  - KSAND  * AVVELEEXI>sm>                                        (5.7)

where:                •           '
      PSAND  - sand carrying capacity of the flow (mg/L)
      KSAND  - coefficient in the sandload equation             •
      EXPSND— exponent in the sandload equation
      AWELE - average velocity of the flow (ft/s).

5.2.4  Water Quality  Processes

Water quality simulation in the Watershed Model is  performed by the RQUAL module
which consists  of four  interacting submodules.  Three  of these suboodules are
used in  the Watershed Model: 1) OXRX, which simulates oxygen and BOD dynamics;
2) NUTRX, which simulates  inorganic nitrogen and  phosphorus  reactions;  and 3)
PLANK, which handles  phytoplankton and other organic materials.  These submodules
are executed sequentially in the order listed, and each submodule requires the
execution of those above it.  For example execution of NUTRX requires that OXRX
be executed, and PLANK requires execution of both OXRX and NUTRX.  OXRX,execution
does not require  any  other submodule.

In this  section,  overviews of the nitrogen and phosphorus cycles are described
first,  and  then the individual processes  simulated  in  each submodule  are
described under  the  corresponding section below.   Table 5.1 provides a summary
of all of the water quality constituents and processes simulated in the Watershed
Model along with  the  submodules  that handle them.

                                      161

-------
The complete schematic of instream  nitrogen  cycling  in the Watershed Model is
shown in Figure 5.6.  The distinct forms of nitrogen are nitrate (N03), dissolved
ammonia (NK3),  particulate ammonia (ads-NH3),  degradable organic N (BOD), living
organic N (phytoplankton),  and dead refractory organic N (ORN).  The principal
changes made as part of Phase II are the addition of the particulate ammonia and
a more flexible algorithm for denitrification of N03.

The parallel schematic for phosphorus  is  shown in Figure 5.7.  As for nitrogen,
improvements made  to the Phase  II  model are the addition  of the  particulate
orthophosphate constituent and its associated processes sorption/desorption and
settling/resuspension.
5.2.4.1  Dissolved oxygen and BOD dynamics
The primary  reaction's  of dissolved oxygen and BOD simulated in the Watershed
Model are performed by OXRX, while  processes  simulated  in the NUTRX and PLANK
submodules also affect oxygen and BOD concentrations.  Figures 5.8 and 5.9 are
the flow diagrams for oxygen and BOD; the processes simulated in the Watershed
Model are delineated by dashed lines.

Reaeration is  computed as  the  product of a coefficient  dependent on hydraulic
conditions,  and a driving  force  which  is  the  difference between  the actual
dissolved  oxygen  concentration  and  the  saturated value   for  the  current
conditions.  The general expression is:

     . DELDOX - KOREA * (SATDO - DOXS)                                    (5.8)
                     i                                     ,
where:
      DELDOX - change  in oxygen concentration over the interval (mg/L)  .
      KOREA  - reaeration coefficient
      SATDO  - oxygen  saturation level at the current water temperature (mg/L)
     DOXS    -  oxygen concentration at start of interval (mg/L)

The saturation value is based on temperature as follows:

      SATDO  - 14.652-+ TW*(-.4102 + TW*(.00799 - .7777E-4*TW)) * CFPRES (5.9)

where:
      TW     - water temperature (*C)
      CFPRES - ratio of atmospheric pressure to pressure at sea level

Estimation of  the reaeration coefficient for free flowing  river channels uses a
modified version  of Churchill's empirical relationship of  depth and velocity
which has the  following form: .

      KOREA  - 0.726 * AVVELE0-'*9 * 1.024(TW ' "'VAVDEPE1-673              (5.10)
where:
      AWELE -  average velocity  (ft/s)
      AVDEPE -  Average depth  (ft)
                                      163

-------
  WATER  COLUMN
                                      uptake
               adsorption/
                derorptlon
       PO4-P
     (adsorbed)
                      LIVING ORGANIC
                             P
                       (phytoplankton)
  PO4-P
(dU(Olvod)
                                                                           death/
                                                                           reซp.
           deposition/
              • oour
3-
                                                                   BOO
                                                                  dtoay
                                                                    death/
                                                                  respiration
OEORAOABLE
 ORQANIC P
   (BOO)
                               • •tiling
REFRACTORY
 ORQANIC P
   (ORP)
                        • •ttllng
                                • •tiling
1       PO4-P      i
     (adsorbed)   |
I       Infinite
1     •ouroe/ซlnk   i
                           Infinite
                            • Ink
                   Infinite
                     • Ink
                 I    I
                 I    I
                                                                                                                 1
   Infinite
    • Ink
                                                  I  _  __  J
                                                  :	  I
  SEDIMENT
                       Figure 5.7   Insteam phosphorus dynamics .in the Watershed Model.

-------
           IBOD
           Inflow
             to
          RCHRES
O\
                                    phy to-
                                   pi an K ton
                                    death
                                          storage  of BOD
                               settllnd
benthlc
release
                                      outflow
                                     'through
                                      exit N
 BOO
decay
                Figure  5.9  Flow diagram  of BOD in the OXRX section of the RCHRES  module.

-------
     TW     - water temperature (*C)
     TAM    - total dissolved ammonia-N. concentration (mg/L)

This process  is dependent upon a  suitable  supply of dissolved  oxygen;  i.e.,
nitrification is set to zero if the DO concentration  is below 2 mg/L.  In HSPF,
the amount of oxygen used  during nitrification is 4.33  mg oxygen per mg NH3-N
oxidized completely to N03-N.

Denitrification is the reduction of nitrate  by facultative anaerobic bacteria
such as Pseudomonas,  Micrococcus,  and Bacillus.  These bacteria can use N03 for
respiration in  the same  manner that oxygen  is  used  under aerobic conditions.
Facultative organisms use oxygen until the environment becomes nearly or totally
anaerobic, and then switch over to  N03 as their oxygen source.  The end product
of denitrification is assumed to be nitrogen gas.

Oenitrification does not occur unless the dissolved oxygen concentration is below
a  user-specified  threshold  value  (DENOXT) .     If  that  situation  occurs,
denitrification  is assumed  to be  a  first-order process based  on the  N03
concentration.  The  amount of denitrification  is calculated by  the following
equation:

     DENN03 - KN0320 *  (TCDEN**(TW-20))  * N03                            (5.13)

where:
     DENN03 - amount of N03 denitrified  (mg N/L per hour)
     KN0320 - N03  denitrification rate coefficient at 20*C (/hr)
     TCDEN  - temperature correction coefficient for denitrification
     N03    - nitrate concentration (mg N/L)

NUTRX  simulates  the  exchange  of  ammonium  between  the  dissolved state  and
adsorption on suspended sediment,   i.e., adsorption/desorption is not simulated
In the bed sediment.  The sorbents considered  are sand, silt, and. clay, which are
simulated in section SEDTRN.  The adsorption/desorption process for each sediment
fraction  is represented  with an  equilibrium   linear  isotherm   (K^) which  is
described as follows :

     ADSNH3 - DISNH3  *  KDNH3                                             (5.14)

where :                          -   .         .
     ADSNH3 - equilibrium concentration of adsorbed ammonia on sediment
                fraction J  (mg/Tcg)
     DISNH3 - equilibrium concentration of dissolved ammonia (mg/L)
     KDNH3  - adsorption parameter (or Kj)  (L/kg)
The advective processes for ammonia adsorbed to a sediment size fraction are:

       1.     Inflow to the RCHRES of ammonia attached to suspended sediment.
                      4
       2.     Migration of  ammonia from  suspension to the  bed as a  result of
             deposition of the sediment to which the ammonia is adsorbed.
                                     ,169

-------
TALGRH,  no  growth occurs.    Between  TALGRL and TALGRM,  the  correction factor
increases linearly from 0  to 1,  and  between TALGRM and TALGRH, the correction
factor is 1.  In the Watershed Model,  the high temperature threshold is 95*F, and
the optimum temperature (TALGRM)  is 86 *F.  The low temperature threshold was set
to 26.7*F for most of the basin, except  for James and-other Virginia tributaries,
where lower values (> -10*F) are heeded  to produce sufficient algal growth during
winter.

Algae depend on uptake of orthophosphorus to provide the phosphorus necessary for
growth.  In situations  where phosphorus is  the  limiting factor, the growth rate
depends on .both phosphate and nitrate  as follows:

                    MALGRT * P04  * N03
      CROP   -	:	                             (5.15)
               (P04  + CMMP) * (NO, -I-  CMMNP)

where:                                                     .
      CROP   - unit growth rate based  on P-limitation  (/hr)
      MALGRT - temperature-corrected maximum growth rate (/hr)
      CMMP   - phosphate Michaelis-Menten constant for P-limited growth (mg/L)
      CMMNP  - nitrate Michaelis-Menten constant for P-limited growth  (mg/L)

Nitrogen, which is also essential to algal  growth,  is obtained from the total
pool  of  dissolved NH3 and  N03 in the reach.  The ratio of N03 uptake to total
Inorganic N (N03 + NH3) uptake is governed by a preference factor (ALNPR) which
ranges from 0.25  to 0.35 in the  Watershed  Model.   The nitrogen-limited growth
rate  is calculated by the following equation:

                MALGRT * (NH3 + N03)
      GRON   -	—                                   (5.16)
     '            CMMN + (NH3 > NOj)

where:
      GRON   - unit growth rate based  on N-limitation  (/hr)
      CMMN   - Michaelis-Menten  constant for N-limited growth (mg/L)

During each time .step, the model  computes the amount of radiation available to
support  algal  growth by applying light absorption rates to the incident light
available at  the  surface,  and assuming that all phytoplankton are at the mid-
depth  of the  reach.   The  overall  light extinction  is the  sum of the  base
extinction coefficient, a chlorophyll A factor,  and a sediment factor.   Finally,
the light-limited growth rate is  computed as:                  .

                MALGRT * LIGHT
      GROL   - ——	                                         (5.17)
                 CMMLT + LIGHT                         •    •.    -

where:
      GROL   - unit growth rate based  on light-limitation (/hr)
      CMMLT. - Michaelis-Menten  constant for light-limited growth (Ly/min)
      LIGHT  - light intensity available to algae (Ly/min)


                                      171

-------
parameter used  in the Watershed Model "to adjust algal losses from the system.

.5.2.4.4  Organic nitrogen, organic phosphorus, and total organic carbon

Dead  refractory organic matter is  simulated as three state .variables  in  the
model.   ORN, ORP, and  ORC  are the nitrogen,  phosphorus,  and carbon species,
respectively.   These constituents undergo advection, they  are generated as a
result of phytoplankton death (see above), and  they are lost from the system  via
settling.   They are.constituents in the respective  total  organic  quantities
(TORN,  TOR?, TORC)  which  are also  made  up  of contributions  from BOD  and
phytoplankton.   It  is  these simulated  total organic  state  variables that  are
compared directly with the "observed" organics concentrations (organic N, organic
P,  and  TOG).    The following   expressions,  based  on  the   Watershed  Model
stoichiometry,  define the total organic state  variables:
                    *
      TORN    -  ORN + 0.0863 *  PHYTO + 0.0529 * BOD                      (5.19)
    •  TORP  .  -  ORP + 0.0123 *  PHYTO + 0.00756  * BOD                     (5.20)
      TORC    -  ORC + 0.49 * PHYTO + 0.301 * BOD     .                    (5.21)

where:
      PHYTO  -  phytoplankton concentration (mg/L as biomass)
      BOD     -  biochemical oxygen demand (mg/L as oxygen)

In  the  Phase II Model,  total organic carbon (TOC)  replaces-BOD as  the primary
indicator of carbon species.  This change was implemented because of  the greater
availability of observed TOC data for  direct  comparison with the  simulation.
However, in order to simulate TOC,  it was necessary to  characterize the loadings
of  carbon since the  Phase I model  did not account for the  total  carbon loadings.
Previously,  loading of nonliving carbon species  occurred entirely through the  BOD
loading.  In the Phase  II model,  refractory organic carbon (ORC) loadings from
point and nonpoint sources are explicitly included by assuming that ORC loads  are
proportional to the existing BOD loads.  Therefore, the BOD loads  are multiplied
by  a  factor and input to the system as  refractory organic carbon.

5.2.4.5  Review and  adjustment of instream water quality parameters        .

The chemical and biological parameters  were.reviewed  and adjusted in Phase  II.
Some of the parameter values were atypical of  the ranges used  in surface water
modeling.   This may have been  due to the magnification or reduction of certain
parameters  in order  to  compensate for phenomenon that are not  simulated by  the
model.  For example, in the initial, versions of the model,  temperature endpoints
of  the smoothing functions were set at  atypical levels to permit phytoplankton
growth  in all  four  seasons.   In  addition,  in previous  versions of the model
-maximal algal growth rate may have been set artificially high to account for both
free-floating and benthic algae as well, as adsorption-sedimentation mechanisms
for reducing inorganic  P and N during instream transport.

A series of steps, following the hierarchy of model sections were executed during
the process of  adjusting the water quality parameters. For instance, hydraulics
simulation  was finalized before temperature was  adjusted and the dissolved oxygen
simulation  was  completed  before   proceeding  to  the  adjustment of inorganic
nutrients.

                                      173

-------
TABLE 5.2   INSTREAM WATER QUALITY PARAMETERS
Parameter Description
 Value
Unit   Ref.
TCBOD     Temperature Coefficient
          for BOD decay
TCGINV    Temperature Coefficient
          for gas invasion
TCNIT     Temperature Coefficient
          for nitrification
TALGRH    Temperature above which
          algal growth ceases
TALGRL    Temperature below which
          algal growth ceases
TALGRM    Temperature below which
          algal growth is retarded
CMMLT     Half-saturation constant
          for light
CMMN      Half-saturation constant
          for nitrate
CMHNP     N03 Concentration for   .
       .   Phosphorus limiting growth
CMMP      P04 Concentration for
          Phosphorus limiting growth
KBOD20    BOD Decay rate at 20 *C
KODSET    BOD settling rate
KNH320    Nitrification rate at *C
MALGR     Maximal algal growth rate
EXTB      Base light extinction
          coefficient
ALR20     Algae respiration rate at
          20 *C
ALDH ,    High algal.death rate
PHYSET    Phytoplankton settling rate
REFSET    Dead refractory settling
          rate     ,  .  '
RATCLP    Chlorophyll-a to P ratio
          in biomass
ALNPR     Fraction of N03 as nitrogen
          source for algal growth
 1.047      N/A    1,2

 1.024      N/A    1,2

 1.045      N/A    1,2

 95.0       *F      2

 26.7       *F      2

 86.0       *F      2

 0.040  (ly/min)   1,2

 0.025    (mg/L)   1,2

0.0001**  (mg/L)   1,2

 0.005    (mg/L)   1,2

 0.004    (1/hr)   1,2
 0.021*** (ft/hr)  1,2
 0.004    (1/hr)   1,2
 0.142    (1/hr)
 0.575    (I/ft)
        1,2
         *
 0.005    (1/Hr)   1,2

 0.003    (1/hr)   1,2
 0.027*** (ft/hr)   1
 0.021*** (ft/hr)   1

  0.68      N/A    1,2

  0.35      N/A     1
1. Reference 1 - Bowie et al., 1985
2. Reference 2 • Potomac Model, Thomann and  Fitzpatrick, 1982
* calculated from tidal fresh monitoring data.
** Set at a value as to make the parameter inoperative.
*** Adjusted by calibration for each reach.
                                  175

-------
TABLE 5.3  CHANNEL REACH INFORMATION


LENGTH
SUBBASIN REACH
REPRESENTED

SUSQUEHANNA
EAST BRANCH
20
30
40
WEST BRANCH
60
70
JUNIATA
100
LOWER
110
CONOWINGO
140 ,
PATDXENT
340
POTOMAC
UPPER
170
175
SHENANDOAH
200
LOWER
210
220
RAPPAHANNOCK
YORK
MATTAPONI
240
PAMUNKEY
260
JAMES
270
280
290
APPOMATTOX
310

10
184.4
,94.9
68.4
50
88.2
42.1
90
111.9
80
24.2
120
14.0
330
21.0

160
126.6
43.3
190
57.8 .
180
64.5
41.0


235
35.7
250
44.5
265
114.0
107.3
39.0
300
13.8
TRIBUTARY
(MI) AREA (MI2!

108.0
4991
2440
1492
91.0
4262
1330
27.0
2418
50.6
1922
19.6
295
12.0
215
,
92.3
1431
1257
139.0
1406
78.6
971
955
230

27.6
338
16.3
737
12.0
2917
2989
512
108 .4
158

2647
S
S .
S
1428
S
S
937
S
2286
S
743
L
133
S

1349
S
S
1623
S
2507
S
S
73.0

257
S
334
S
348
S
S
S
1194
L
) TYPE* PRIMARY TRIBUTARY

S EAST BRANCH
EAST BRANCH
EAST BRANCH
EAST BRANCH
S WEST BRANCH
WEST BRANCH
WEST BRANCH
L RAYSTOWN RESERVOIR
, JUNIATA RIVER
S SUSQUEHANNA RIVER
SUSQUEHANNA RIVER
L CONOWINGO RESERVOIR
CONOWINGO RESERVOIR
L PATUXENT RESERVOIRS
PATUXENT RIVER

S N. BRANCH POTOMAC
S. BRANCH POTOMAC
POTOMAC RIVER
S SHENANDOAH RIVER
SHENANDOAH RIVER
S POTOMAC RIVER
' MONOCACY RIVER
POTOMAC RIVER
1604 S RAPPAHANNOCK RIVER

S MATTAPONI RIVER
MATTAPONI RIVER
L LAKE ANNA
PAMUNKEY RIVER
L JAMES RIVER
JAMES RIVER
JAMES RIVER
JAMES RIVER
S APPOMATTOX RIVER
LAKE CHESDIN

*  S  - River  reach
   L  - Lake/reservoir reach
                                  177

-------
                                  SECTION 6.0

                          HYDROLOGY CALIBRATION RESULTS


 this  section  discusses  the  results  of the  hydrology  calibration,  including
 discussion of tasks performed in both Phase I and Phase II.  Much  of the  text
 from  the Phase I  report  is retained in order  to provide a  complete  description
 of the hydrology calibration effort; however,  only the results from Phase II are
 presented here  since they  supersede  the Phase  I  results  for  the  1984-87
 simulation period.   Readers  interested in the  1974-78  hydrology results  are
 referred to the Phase I  report (Donigian et al,  1990).

 6.1 OVERVIEW AND  STEPS IN THE HYDROLOGIC CALIBRATION PROCESS

 the purpose and justification for the Phase I hydrology re-calibration was to
 include  snow simulation.,  evaluate any changes resulting from the revised land use
 data  and addition of the animal waste  segments,  and incorporate any  changes
 necessitated by  allowing different hydrologic  parameters for  each land  use
 category.   In Phase  I,  the  1974-78  period was  chosen  for  the  primary  re-
 calibration effort because the existing data base covered this period and the
 model had been previously verified for the same period.   In effect,  the goal of
 the effort  was to evaluate  the  cumulative impact of all  the changes to  the
 hydrologic parameters, and then perform the parameter adjustments needed to 're-
 calibrate'  the model.   The model was then run  on  the  1984-85 period  and the
 results  were analyzed,  evaluated,  and provided the basis for the  recommended
.focus of the hydrology calibration tasks in Phase II.

 As discussed in Section  1.3,  the  Phase I recommendations included extending the
 simulation and calibration through  1988 in order to provide a more  realistic
 evaluation of the Watershed Model' s ability to represent the hydrology during the
 critical 1984-88  period.   This period was used as a basis for both the Bay model
 and  the  40Z Chesapeake  Bay Agreement.   Additional Phase  I recommendations
 included evaluation  of  the  snow simulation for 1984-88,  investigation of the
 Conowingo Reservoir operation and simulation,  and further  investigation of the
 simulation of the November 1985  storm.  Due to limitations on data  availability
 for 1988 at the time the meteorologic was  needed,  the simulction could only be
 extended through 1987.

 The steps performed in the combined Phase  I/II efforts were  as follows:

       a.  Identify  watershed areas needing snow simulation.

       b.  Perform snow calibration for 1974-78 to observed snow  depth data.

       c.  Evaluate hydrologic parameters for each land  use category.

                                       178                         '

-------
TABLE 6.1  ELEVATION, TEMPERATURE AND SNOW DEPTHS FOR WATERSHED MODEL LAND SEGMENTS
Sub
RCH

10(1)
20(1)
30(1)
40(2)
50(1)
60(1)
70(2)
^80(2)
90(2)
100(2)
110(2)
120(2)
140(3)
160(1)
170(2)
175(2)
180(2)
190C3)
200(3)
210(2)
220(3)
230(4)
250(4)
260(4)
265(4)
270(4)
280(4)
290(4)
300(4)
basin
Elev
(feet)
3300-650
3300-650
2000-650
1640-650
3300-650
3300-650
3300-500
1640-330
3300-650
3300-300
1640-330
1000-500
1000-
2400-650
3300-650
1640-650
2000-400
3600-650
1640-400
2000-300
2000-
1000-
1000-
.1000-
3300-1640
3300-650
1000-
500 -
500 -
Average*
Elev
(feet)
1860
1512
1368
930
1980
1260
524
6UO
, 4
- 772
469
,510
40
1693
' 1868
1288
547
1880
460
710
910
500
: 220
320
2579
1060
916
164
.164
C
Normal S
Temp
27.3
25.9
27.2
27.3
26.4
28.9
30.7
32.3
30.9
• 31.5
33.3
32.2
34.4
. 31.4
34.7
33.3
33.4
35.4
34.5
33.3
35.8
34.5
37.9
38.7
33.9
40.0
38.5
39.9
39.4
ECEMBER
now on
Max
10.3
9.7
8.5
4.4
6.0
6.9
3.3
1.9

2.2
1.7
3.2
1.5
4.7
3.2
.9
.6
.2
.0
.8
.8
.2
.3
.6
2.3
T
T
T
T

Ground*
Min
T
T
T
T
0.0
0.0
0.0
0.0

T
0.0
0.0
0.0
0.0
0.0
0.0
0.0
o.o
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
ป
Normal S
Temp
22.5
20.8
22.5
23:3
22.1
24.4
26.2
28.3
26.9
27.5
29.3
28.5
30.4
27.8
31.3
29.5
26.6
32.1
30.8
29.4
31.4
31.2
34.7
35.8
30.8
35.3
35.4
36.6
37.0
IANUARY
inow on
Max
13.5
14.8
12.8
5.5
11,9
10.5
8.1
6.7

6.2
5.7
8.1
4.0
11.3
.5
.1
.9
.3
.1
.3
.3
4.9
4.0
5.5
6.2
4.2
3.8
3.2
3.2

Ground*
Min
3.3
3.1
3.1
T
1.3
1.1
T
T

T
T
0.0
0.0
1.0
T
T
0.0
0.0
0.0
T
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
1
Normal !
Temp
23.5
22.0
, 23.8
25.0
23.2
25.7
28.2
30.3
28.5
29.2
31.2
30.5
32.2
29.7
33.3
31.2
31.8
34.1
32.8
31.6
33.6
33.0
36i4
38.1.
32.8
37.8
37.4
38.9
39.2
'EBRUARY
Inow on
Max
13.9
15.7
14.9
6.4
15.2
11.7
7.4
7.7

7.0
6.1
9.6
6.9
12.9
9.0
8.6
7.1
6.5
4.9
9.2
5.9
4.7
2.9
4.0
5.4
4.3
3.9
2.7
2.7 .

Ground*
Min
4.8
4.3
4.1
T
3.1
2.6
1.1
1.6

T
T
T
0.0
2.2
T
1.5
,T
T
T
T
T
Tr
0.0
0.0
T
0.0
0.0
o.o
0.0
* - Average of show measuring stations elevation (ft) in that Subbasin (see attached Notes).
* - Average maximum snow on ground and average Bininum snow on ground (see attached Notes).
T - Trace of snow.        '                                                         /      .

Segment Categories:

(D-Temperature below 30 (F) and max snow on ground above 10 (in) for two or nore winter  months.
(2)-Temperature below 32 (F) and max snow on ground below 10 (in) and above 6 (in) for two or
    •tore winter months.
(3)-Temperature below.32 (F) and max snow on ground below 10 (in) and above 6 (in) for one month.
(4)-Temperature above 32 (F) and/or max snow on ground below 6 (in) for two or more months.

•NOTE* For subbasins 40, 110 and 200 the snow gages are located at the lower end of the elevation
       range. So, an exception is made to include these subbasins under the next highest  category.
                                              480

-------
extreme, segments under Category //4 do not require  snow simulation due to above
freezing temperatures and minimal observed snowfall.

With regard to  the  'marginal'  segments under Categories  #2 and #3,  the actual
snow depth records were  reviewed (as  discussed under step c. above) to assess
both the maximum depths achieved and the  typical duration of snow cover during
the winter months.  Based  on this assessment, segments  under Category #3 were
eliminated from contention because maximum depths rarely exceeded 10 to 12 inches
and usually lasted only a few days. Similarly, under Category |2, segments  110,
120, 140, 180, and 210 were eliminated due to depth and duration of snow cover;
depths rarely exceeded -12 inches and cover occasionally lasted up to two weeks
at the most.             '                     -

However, segments 40,^70, 80, 90, 100,  170. and 175, under Category 02 frequently
showed snow cover lasting  a month or  more,  and'depths often exceeded 12 to J.8
inches.  Consequently, we felt  that these segments would have significant snow
depths,  and  subsequent storage and carry-over  of moisture  from one month to.
another; in our opinion, this justified the need for snow simulation.
                   *                                                    •     '  i
Based  on these analyses,  the  dividing line for snow simulation  (as shown in
Figure 6.1) occurs about midway through the Lower Susquehanna, at the northern
boundary of the Lower Potomac, and at the southern boundary of the Upper Potomac.
Since  each segment is  simulated with one set of input and  parameters,  the
dividing  line follows segment  boundaries.   To add or  delete areas for  snow
simulation, an entire  segment would need to  be added or deleted.   Additional
areas  can be added for snow simulation  in future  efforts  to those currently
simulated, if justified.   However,  the overall hydrologic calibration results
(discussed below) appear to  confirm the snow  segment selection.

6.2.2  Snow Calibration and  Simulation'Results

In both Phase I and Phase II, snow calibration was performed on Segment 20 in the
East Branch of the Susquehanna,  Segment 50 in the West Branch of the Susquehanna,
and Segment 160 in the Upper Potomac. these segments were chosen to specifically
span the  geographic extent of the snow simulation region and include segments
with some snow depth measurements;  see Figure 6.1 for  the segment locations
within the snow region.  Since  snow depth is  highly variable between locations
within  a  segment, we included as   many  observations  as  possible  in  the.
calibration.

The primary goals of the calibration were to capture the  general time period of
snow cover,  the magnitude  cf the depth of the snow pack, and the general time
frame for disappearance or melt of the pack in the  spring.   Parameters relating
to  the  'catch' of the precipitation gage (SNOVCF),  the  magnitude of  the melt
components (CCFACT), the snow density/moisture content (RDCSN), and the ground-
associated melt (MGMELT)  were  adjusted in Phase I to obtain the best overall
agreement  between the  observed and simulated timeseries and frequency of snow
depth values.

The Phase  I  report  included the calibrated  frequency  curves  for snow depth for
Segments  20,  50,  and 160,  and the daily observed and  simulated snow depths for
Segment 20 for each winter period from 1974-78 starting in January 1974. Both the

                                       182

-------
00
ou

27

24


21


g 18
& 15
O

^ป
o
^ 12


9

6



V

i i i i i i i i i i i i

-

	 QFrMI

	 rnoii
	 t/Ur\ 1 L
RIKJPH
DIINon
-•


- '

i
"1
j
j
i i i i i i i i i i i
r.NT20
AND
^MTON





•
?, '
i i i i




ii
|1
/(•'•



i
•


*!
: • •

il

fป
; rl
, Y
'
-------
   25.0

   22.5

   20.0

   17.5

g 15.0
•ฃ
& 12.5
00
- o
  i/j
        10.0
         7.5
         5.0
         2.5
               i  i  i  I  r
                                   TI  i  i  i  i i  i  i  n  i  i i  iii  i  n n  i  r
           0
                                 SEGMENT 160
                                 SAVAGE RIVER DAM
                                        . Ill 1 I
                                                           n II I  I  I  I
              J FMA MJ J ASONDU F M A.M J J A S 0 N DIJ  F M A M J J A S0 N Old  FMAMJ J A SOND
                       1984          I         1985         I         1986         I         1987
                                              SNOW DEPTH - SEGMENT 160
                       Figure 6.4  Simulated and observed snow  depth for Model  Segment 160, 1984-87.

-------
         TABLE 6.2  LAND USE COVER VALUES USED IN THE WATERSHED MODEL
         .           FOR NON-CROPLAND AREAS
         JAN   FEE   MAR   APR   MAY   JUN   JUL   AUG   SEP   OCT   NOV   DEC



FOREST  0.97  0.97  0.97  0.97  0.97  0.97  0.97  0.97  0.97  0.97  0.97  0.97

PASTURE 0.90  0.90  0.90  0.91  0.91  0.91  0.93  0.93  0.93  0.91  0.90  0.90
N '                   -       .
URBAN   0.90  0.90  0.90  0.91  0.93  0.93  0.93  0.93  0.93  0.91  0.90  0.90
hydrologic simulation,  the monthly rainfall interception values in the Model were
adjusted to be consistent with the  seasonality and magnitude of the COVER value
for each land use.                                                •

Values for SLSUR and KRER were provided by CBPO based on soils information for
all  segments  in  the  watershed.   Slope  values  for SLSUR  were  derived  from
knowledge of the  land  use  categories typically associated with different soil
classifications.  Values for KRER were estimated by the 'credibility factor' in
the Universal Soil Loss Equation based on tabulations from the SCS.

LSUR values were estimated with. respect to corresponding slope values for each
land use; areas with high slopes were assigned .low values  of LSUR  and low slopes
were assigned higher values, based on experience with the model *

NSUR is the Manning's roughness factor for overland flow. The following values
assigned  to  each  land category  were  initially  developed  in  Phase I  and
subsequently revised in Phase  II based on information  in KnLsel et al (I960),
Engman (1986), and Donigian et  al (1983):
               ,            .            .     i  -'.''.•"'.'
            Forest                               0.35
            Pasture                              0.15
            Urban Pervious                       0.10
            Conventional Tillage          0.05-0.06
            Conservation Tillage          0.16 - 0.185
            Urban Impervious                     0.03                   '      ,

INFILT values  are primarily based on calibration.    For  each land segment, we
imposed a distribution of values based on experience that  had  the  highest values
for  forest  areas,  the  lowest values for cropland,  and intermediate values for
urban pervious  and pasture.   During  the  calibration effort,  this ordering of
values was maintained with individual segment values adjusted as dictated by the
calibration process.
                                      188

-------
     TABLE 6.3  WATERSHED MODEL HYDROLOGY AND WATER QUALITY CALIBRATION STATIONS
 Model  Sta.  #  Station Name
                             Status  Hydro.
 SUSQUEHANNA RIVER BASIN

 20   01515000 Sus.  R,  nr Waverly,  NY
.40   01540500 Sus.  R.  nr Danville,  PA
 70   01553500 WB Sus.  R.  @ Lewisburg,  PA
 100  01567000 Juniata  R.  @ Newport,  PA
 80   01570500 Sus.  R/@ Harrisburg.PA
 110 01576000 Sus.  R.
 140 01578310 Sus.  R.

 POTOMAC RIVER BASIN
        @ Marietta, PA
        @ Conowingo, MD
 JAMES  RIVER BASIN                 '

 280 02035000 James R.  @ Cartersville,  VA
 290 02037500 James R.  nr Richmond,  VA
 270 02025500 James R.  (ง Holcombs Rock, VA
 OTHER SITES

 310 02041650
 220 01646500
 180 01638500
 210 01643000
 230 Q16680QO
 240 01674500
 260 01673000
 340 01594440
 330 01592500
Appomattox R. (ง Matoaca, VA
Pot. R. @ Lit. Falls Dam, DC
Pot. R. @ Point of Rocks, MD
Monocacy R. @ Jug Bridge, MD
Rappahannock R. nr Fred., VA
Mattaponi R. nr Beulah., VA
Pamunkey R. nr Hanover, VA
Patuxent R. nr Bowie, MD
Patux.. R. nr Laurel, MD
P
C
C
C
C
P
C
 175 01613000 Potomac R.  @ Hancock,  MD      C
 220 06146580 Pot.  R. @ Chain Bridge,  DC    C
 200 01636500 Shenandoah R. @ Millville, W  C
 180 01618000 Potomac R.  @ Shepardstown, WV  P
                               C
                               P
                               C
C/P
P
C
P
C
C
C
C ,
P
X
X
X
X
X
X
X
                                      X
                                      X
                                      X
                                      X
       X
       X
       X
X
X
X
X
X
X
X
X
X
               "Water Cont. 'Co'nt.
                Qual. Temp. Sed.
                                               X
                                               X
                                               X
                                               X
                .X
                X
                X
X
X
X
X
        XX
        X    X
                 X(Buchanan)
        X
        X
X
X
 Proposed Stations for Hydrology and Water Quality Phase I Recalibration
 C - calibration station;    P - potential calibration stations.
                                         190

-------
        TABLE 6.*   (continued)
        Model    USGS
        Seg.     Station
        Number  Nunfcer
Station
                 1974-78
Mean Ota.  Mean Sim.       ft*     R*
Annual     Annual         Ave.   Ave.
flow (In)  Flo* (In)     Dally  Monthly
                                                     1984-87
                                     Mean Ota.    Mean Sin.      R2       R*
                                     Annual       Annual       Ave.    Ave.
                                     FloM (In)    Flou (in)    Dally   Monthly
        260     01673000


        310 „    02041650


        340     01594440
Pamunkey R. 8
Hanover, VA.

Appomattqx R. 8
Matoaca, VA.

Patux. R. nr.
Bowie, ND.
11.48
11.43
0.3658    0.9392
14.26       14.10      0.7170    0.8323


13.31       13.88      0.7328    0.8153


11.3*4       11.75      0.5928    0.7729
vO

-------
The end result of these differences and inconsistencies between the two periods
was to preclude the use of the 1984-85 as an effective period for verification
of the hydrology  simulation in Phase I.  Since the evaporation difference was the
major concern, the only calibration efforts performed for  the 1984-85 simulation
were constant adjustments  to  the input evapotranspiration timeseries.  That is,
we adjusted a single number, usually in the range of 0.8 to 0.95, multiplying the
input data  to obtain improvement in annual  flow volumes.  These were the only
adjustments during the Phase I calibration for the 1984-85 period.

As reported in the Phase I report, .the results  for 1984-85 were not nearly as
close to observed data values as in the earlier 1974-78 period.  Although mean
annual flow volumes were calibrated with  the evaporation  adjustment to maintain.
close agreement,  both the daily and monthly R2 values  were considerably lower at
most stations.   In addition,  the frequency  curves generally show much greater
deviations, such as the lower peak flows and the higher base flows  than observed.
However,  the daily  flow  comparison  still  showed good   tracking between the
simulated and observed values, except for selected peak flow events and during
winter snowmelt periods.

Due  to the  need for • better flow  simulation  for  the  1984-85 period,  our
recommendations for additional efforts in Phase II were presented in Section 1.3
and are briefly listed below:

      a.  Extend  the data  b*ae  and simulation  through 1988  or  1989,  and
          calibrate as needed to improve the flow simulation.

      b.  Obtain  snow depth data for  the 1984-89 period  to check and evaluate
          the snow simulation (AS discussed above).

      c.  Re-evaluate and  investigate the Conovlngo  Reservoir  simulation for
          1984-89.

      d.  Investigate the storm of November 1985  in the Upper Potomac since it
          was not well simulated,  in  terms of peak flow,  in the Phase I effort.

In Phase II, all  these tasks were performed, except that  the meteorologic data
base could  only be extended through 1987 since the 1988 data was hot available
at  that time.   With the extended data base,  the model  required  only minor
adjustments to the evaporation for most  subbasins to produce the final results
shown in Table 6.4.  This, in effect, was a pseudo•verification of the Phase I
calibration, since the seme model parameters were used except for minor changes.
to the  evaporation multiplier.  For  the  smaller subbasins and portions of the
Susquehanna Basin, additional minor adjustments  .were  needed to the infiltration
(INFILT) and groundwater  recession (AGURC)  parameters to  improve the seasonal
flows and frequency curves.                 •

Figures 6.5 through  6.12  show the flow frequency and daily timeseries curves,
respectively, for the four major basin stations: Susquehanna River at Harrisburg,
Susquehanna River at Conowingo Reservoir, Potomac River at Washington, D.C., and
James River at Cartersville.   Complete results for all simulation years and all
                                      194

-------
          LOWER SUSQUEHANNA RIVER AT SEG.  80
                                 FLOW (CFS)
                        RED DASHED: SIM.. BLUE SOLID: DBS.
s
I
M
 500000-
 400000 •
 300000 -
a
n .
d
0
B
S
  200000 -
  100000 -
     0-
1 I  I I r~1I I I  I T
                                                                T"TT"T
       00000000000 00 00 00000000 000 0000000 00000000000000 00
       1111111111111111111111111111111111111111111111111
       JFMAMJJASONDJFMAMJJASONDJFMAMJJASONDJFMAMJJASONDJ
       AEAPAUUUECOEAEAPAUUUECOEAE A P A U U U E C 0 E A E A P A U U U E C 0 E A
       N B R B Y N L G P T V C N B R R Y N L G P T V C N B R R Y N LGPTVCNBRRYNLGPTVCN
       8888888888888888888888888888888888888888888888888
       4444444444445555555555556666666666667777777777778

          ..'''•    :   :    '       -  'DATE   '-.•'.'•
            FMyure 6.0 .Simulated and obacirved daily flow for llarrisbury, 1>A, Suymunt 80,

-------
S
I
M
 500000-
 400000 -
 300000 -
0
B
S
  200000 -
  100000-
             SUSQUEHANNA RIVER AT CONOWINGO
                                  FLOW (CFS)
                         RED DASHED: SIM..  BLUE SOLID: OBS.
                       I i  i I  i i
I i
i i  i i
i i  i
ill i i
       000000000000000000 000 00000000. 00000000000000000000
       i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i
       JFMAMJJASONDJFMAMJJASONDJFMAMJJASONDJFMAMJJASONDJ
       AEAPAUUUECOE A E A PAUUUECOEAEAPAUUUECOEAEAPAUUUECOEA
       NBRRYNLGPTYCNBRRYNLGPTVCNBRRYNLGPTVCNBRRYNLGPTVCN
       888888888888888888888888888 8888 888888888888888888
       4444444 44 44455555 55 5555566666 666 6 6 667 7 7 77 77777778

                 •-    '•     •/'  '•  "  '  ."  DATE   '•'    •     ••••-.
         . Pitjin•(•. 6.0  Simulated and obnorvod dally J.low for Cunowingo. Ruuu-rvul r, Ml), Mi-yini-nt 14O, 1UU1-UV.

-------
                LOWER POTOMAC RIVER AT SEG. 220
                                  FLOW (CFS)
                        RED DASHED: SIM.. BLUE SOLID:
                                                   OBS.
   300000 -
  S
  I
  M
   200000-
ง
  0
  B
  S
100000-
       0-
                                                                     I I  III I
     T~~I n i  T~T i  riiii n  i i i  i i i  i i  i n i i  i i i  i i  TT~T~T n i i  i
     ooooooo ooooooo ooooooooo  ooooooooooo oooo ooooooooooo
     1111111111111111111111111111111111111111111111111
     JFMAMJJASONDJFMAMJJA80NDJFMAMJJASONDJFMAMJJASONDJ
     AEAPAUUUECOEAEAPAUUUECOEAEAPAUUUECOEAEAPAUUUECOEA
     NBRRYNLOPTVCNBRRYN L 6 P T VCNBRRYNLGPTVCNBRRYNLGPTVCN
     88888868888888888688888888688888 688 88888-888888888
     44444444444455555 555 55556666666666667777777777778
            Figure 6.10
                                      DATE

                 Simulated and observed daily flow for Potomac River, Washington, DC, Segment 220,
                 1QP4-87.

-------
                         JAMES RIVER AT SEG. 280
                                        FLOW  (CFS)
                              RED DASHED; SIM.. BLUE SOLID: DBS.
   300000-
ro
to
 S
 I
 M
 a
 n
   200000-
 0
 B
 S
100000-
       0-
           iinFIiiir
           oooooooooo
           i
           F
           E
           B R R
         8
         4 4
         1111111
         M A M J J A S
         A P A U U U E
             N L G P
   Tiiiiiniiirnir
   000000000000000
   111111111111111
   DJFMAMJJASONDJF
                                               I
                  i  i r
                                             00000000000000000000000
  1
  N
C 0
   E A E A P A U U U E
T V C N B R R Y N L G P TV
111111111111
M A M J J A S-0 N D J F
APAUUUECOEAE
RRYNLGPTV C N B
  1
  A
1 1
J A
A P A U U U
R R Y N L G
1 1
0 N
C 0
  V
1
J
A
N
       888888888888888888868888888888888888888888888888
         44444444445555555555556666666666667777777777778
                                            DATE   :
         Figure 6.-12  Simulated and observed' daily flow for James River, Cartersville, VA, Segment 280, 1984-87.

-------
      f.  In  Phase  II,   the  Conowingo Reservoir  hydraulic  simulation  was
          investigated and adjusted to include  a minimum summer release rate of
          5000 cfs and minor adjustments to the  weekly outflow coefficients (see
          Section  5.2.1.1).   Both the  daily  timeseries  (Figure 6.8)  and the
          frequency  (Figure  6.7)  results  show  .very  good agreement  between
          simulated and observed  values,  with  some deviation- in the  low flow
          simulation since these flows are controlled entirely by the operating
          procedures of the reservoir.

      g.  Seasonal R2 values calculated by CBPO are shown  in Table 6.5 and in
          Appendix A.  The seasons were defined as follows:

                  Season 1   > January - February
                  Season 2    March • May
                  Season 3    June - September               .
                  Season 4    October - December

          The  results  in Table   6.5  show  that  the  winter  (Season  1)  is
          consistently simulated worst  than the  other  three seasons,  and that
          the summer and fall  (Seasons 3 and 4) are usually the best,  at least
          for  the  three major basins.   For most  of  the smaller  basins,  the
          spring (Season 2) Is often the best simulated.

Overall,  the  Phase  II  hydrology   calibration  results   are   a   very  good
representation of the Chesapeake Bay basin hydrology and provide a  sound basis
for nonpoint load generation and delivery of pollutant loads to the Bay.  Ongoing
updating of the meteorologic, diversion, and land use information is recommended
to allow for complete verification of the hydrology using the  1988-91 period.
                                      204

-------
                                  SECTION 7.0

                       WATER QUALITY CALIBRATION RESULTS


This section discusses the results of the water quality calibration, including
discussion of tasks performed  in  both Phase I  and Phase  II.   Much of the text
from the Phase I report is retained in order to provide a complete description
of the water quality calibration effort; however,  only the results from Phase II
are presented here since they supersede  the  Phase I results for  the 1984-87
simulation period.  Readers  interested  in the  preliminary  1984-85 results are
referred to the Phase I report (Donigian et al.,  1990).

7.1  OVERVIEW AND STEPS IN THE WATER QUALITY CALIBRATION PROCESS

In  both Phase  I  and  Phase  II,  water  quality  calibration  was  begun  after
completion  of  the hydrology simulation."  However,  in Phase I the  land use
revisions were also important  for the water quality  modeling, and in Phase II
additional input updates were required for point sources, atmospheric deposition,
and animal  waste  contributions.   Although the water quality portions of the
Watershed Model had been calibrated in the  original NVPDC effort,  all  these
revisions to the input for 1984-87 required re-assessment and re-calibration of
the water quality parameters.              .    •
                             x         "       -   "         ' -.     *
Whereas flow modeling deals with a single constituent - water quantity -  and a
single  primary  source - .precipitation,  water quality modeling must consider
numerous constituents, various forms or species,  and multiple sources.   Thus,
nutrient modeling must consider various  forms of nitrogen and phosphorus,  their
interactions and transformations, and other parameters or constituents (e.g., DO,
BOD, water temperature, sediment) that affect these interactions.  In the current
version of the Watershed Model,  the sources considered include point (municipal
and industrial)  discharges,  watershed land uses,  atmospheric deposition, and
animal waste contributions.  Nutrient contributions from all these sources are
estimated,  hydrologic  transport .processes are superimposed,  and  then  water
quality modeling is performed to  allow adjustments in parameters and sources as
part of the calibration process.

Water quality calibration is an iterative process;  the model predictions are the
integrated result of all the assumptions used  in developing the model input and
in  representing  the modeled processes.    Differences  in  model predictions and
observations require the model user to re-evaluate, these assumptions, in terms
of both the estimated model input and model parameters, and consider the accuracy
and  uncertainty  in the  observations.    At  the  current  time,  water  quality
calibration is more an art, than a science, especially for integrated simulations
of nonpoint, point, and instream water quality.


                                  '206

-------
      2.  Calibrate sediment delivery from each land use by adjusting the KSER
          parameter.

      3.' Develop initial values of instrearn parameters  (critical shear stresses
          for scour and  deposition) based on  simulated shear stresses in each
          channel reach.

      4.  Adjust KSER and instrearn parameters  to calibrate sediment at selected
          sites where observed data are available.

The sediment loading calibration was based on  "expected sediment loading rates"
(in units of tons/acre/year) for pervious land  areas.  Expected loading rates for
pasture,  cropland,  and forest  areas  were  developed for each model segment by
applying the USLE (using county soil data from  the 1982 NRI survey), and assuming
a  uniform  delivery "ratio  of  0.15  for all  areas.   The  loading  rates  for
conservation tillage cropland were assumed to be 60X of the conventional tillage
rates,  and values  for hay areas were  estimated  by adding  one third  of  the
difference  between  pasture and conservation  tillage cropland to  the pasture
rates.  A constant value  of 0.16 tons/acre/year* was  assumed for all urban areas
in the watershed, based on NURF data.  The resulting expected sediment loading
rates for all model segments are compiled in Appendix C.3.

The  sediment erosion model,  which  is  described  in Section 4.1.1,  involves
sediment detachment/attachment from the soil matrix and transport of the detached
sediment.    Parameters for  the sediment detachment (by rainfall  impact)  and
attachment  were  developed  (in Phase  I) from  soil  data, while calibration was
largely  performed by adjusting the parameter,  KSER,  which is the coefficient in
the transport equation.   The sediment loading for all segments was calibrated so
that the average of the simulated annual sediment loads for 1984-87 approached
the expected sediment loading rates.  Appendix C.3  contains the final calibrated
average  annual loading rates and the  cbrresponding KSER values by model segment
and land use.       ซ

The instream sediment simulation (described in Section 5.2.3) considers loading,
scour,  deposition,  and advective transport of three sediment size fractions,
.sand, silt, and clay.   Since the loading model considers a single sediment size
fraction, the total  load was apportioned into three size  fractions using constant
factors  estimated from soil composition data  and different delivery ratios for
the three sizes.      .                        .                      J

Initial  values of the ins tr earn scour/deposition parameters for silt and clay were
developed by examination of daily time  series plots of simulated velocity and
shear stress in each reach.  The sand transport parameters were set so that sand
always  settles to the bed,  except at high flow velocities.   The critical shear
stress  for  scour  (TAUCS) was set so that channel scour would occur only during
extreme  flow events, while the critical shear  stress for deposition (TAUCD) was
set so that silt and clay deposition would occur-only at very low flaws.  The net
effect  is that the  channels are not .eroded over time, and loadings of silt and
clay  generally remain suspended, while sand deposits.
                                      208

-------
   2000


   1800

   1600

   1400

 J 1200
o
uu
1000


 800

 600

 400

 200


   0
         i i  n  I i  i  i  i ii  ii ii  ii i  n  r
                       SIMULATED
                       OBSERVED
i  i  i i  i  i i  i  i i  i  i  i  i i  i  i i  i  i  i i i
       FMAMJ J ASONDU FMAMJ J ASOND
                 1984
r
                                                  J FMA.MJ.J ASONDU FMAMJ  J.ASOND
                                                       1986
                      r
1987
                       1985
                       POTOMAC RIVER AT CHAIN BRIDGE
                    .SUSPENDED SEDIMENT  (MG/L)
Figure 7.1  Suspended  sediment  concentration - Susquehanna TUver at Conowinqo.

-------
       800

       720

       640

       560

    5 480

   .tE 400
NJ ••  I . I

    ft 320

       240

       160

        80

         0
 1IIIIIIFiIIIII
                                     i  i  i  n  r  i  i  i  11  ii-  i i  ii r
/LJ'I
                   SIMULATED
                   OBSERVED
1,
                                        I
J FMA.'MJ J A SOND
         1984
  J F MA M J J  A SO ND
           1985
                                      J F MA M J .J  A S 0 N D
                                                                                   JFMAMJJASOND
                                               1986                    1987
                           JUNIATA RIVER AT NEWPORT
                      -     SUSPENDED SEDIMENT  (MG/L)
Figure 7.3  Suspended sediment concentration -  Juniata  River at Newport.

-------
UJ
          I I   I  I  III  II  II  I  II  I  I  I  I  I  III  II  I  I  I  I  I  I  I  I  III  I  II II I  I   II
                           SIMULATED
                           OBSERVED
                              i  i  i    ii  ii
        J F MA MJ J A SON D
                  1984
JFMAMJJASOND
   -      1985
J F MA M J J A.S 0 NDIJ FMA M J J A S 0 N D
         1986
1987
                                             JUNIATA RIVER AT NEWPORT
                                             SUSPENDED SEDIMENT LOAD  (TONS/DAY)
                           Figure 7.5  Suspended srdiinont load - Junint-.n Rlvo'r nซ Newport.

-------
The calibration  involved numerous model runs and  iterations  at each selected
calibration site in order to maintain parameter consistency within and among the
basins.  In the sections below we present and discuss the final Phase. II results
of the comparisons  performed in steps  d and e above for the concentration and
Fall Line loads.  In Phase I,  the  greatest  effort was devoted  to  the  three Fall
Line stations  due to the need  to provide  preliminary Basin  loads  to the Bay
model.   In Phase II,  calibration  was performed  on  the Fall Line  stations,
numerous subbasins,  and  the minor tributaries, comprising 13 separate calibration
sites.   Due  to the large number  of plots  produced by  this  effort,  the model
results presented below will  focus primarily on the Fall  Line sites, but the
discussion will include issues relevant to all sites. Complete results for all
water quality calibration sites are provided in Appendix B.

7.3.1  Instream Water Quality Concentrations   •                               -

Once nonpoint and point source loading rates were deemed to be reasonable, the
ins tr earn water quality  calibration focused on adjustments to selected ins t ream
parameters  to  improve  agreement  with observed  concentrations.   The primary
parameters  of concern  were the  settling  rates  for  phytoplankton  and BOO,
nitrification  rates,   phytoplankton   growth  rates,   algal   nutrient  uptake
parameters, and partition coefficients and bed concentrations  for NH4 and P04.
Final model parameters  were discussed in Section 5.2.4.

The figures  provided at the  end  of this section include the  water quality
simulation results for four major stations in the basin, including the  three Fall
Line Stations on 'the SusqUehanna,  Potomac,  and James rivers.  The. four stations
and corresponding figures are as follows:              •  • i

.  .  .  Model  ..     .'.."'.'.        .   '        .  . •   .'•.•'.''''.
                  Station                         -   Fiures
      .80    Susquehanna R. @ Harrisburg, PA           7.6 - 7.17
      140   Susquehanna R. @ Conowingo, MD            7.18 - 7.28
      220   Potomac R. @ Washington, D.C.             7.29-7.39
      280   James R. @ Carters villa, VA               7.40-7.51

In  these  figures, daily  simulated and observed values are provided  for the
following water quality constituents for each station in the order shown below:

          Water temperature (ฐC)
          Dissolved oxygen (mg/1)
          Suspended sediment (mg/1)
          Nitrate plus nitrite  (mg/1)
          Ammonia, dissolved (mg/1)
          Particulate nitrogen  (mg/1)
          Total nitrogen  (mg/1)
          Phosphate, dissolved  (mg/1)
          Particulate phosphorus  (mg/1)
          Total phosphorus (mg/1)
          Total organic carbon  (mg/1)
          Chlorophyll a (ug/1)
                                      216

-------
    timeseries with discrete, individual observations.

c.  The Phase II results are a  significant  improvement  over  the Phase I
    results due to the water quality parameter re-evaluation, the AGCHEM
    nutrient loading simulations, and the additional calibration effort.
    Although some problems remain, the results 'show improved simulation of
    seasonal variations in  nitrate  and Chi a, closer representation of
    individual nutrient species, and better agreement with most observed
    concentrations and total loads.

d.  Large deviations are still  seen  in  the  individual nutrient forms at
    some sites, such as P04  in portions of the Susquehanna and inorganic
    nutrients in the minor tributaries.  Chi a observations were limited
    at many calibration sites,  thus  restricting the  calibration process.

e.  the Potomac simulation results are  the best  of  all  the major basins
    and are  generally a good  to very  good representation  of nutrient
    concentrations throughout the Bay basin.  This could be partially due
    to the fact that the Potomac data is more extensive than in any of the
    other  basins,   allowing  a  better  representation  of  the  expected
    variability in the observations  and a better calibration.  The total
    nutrient  concentrations,  and  the  individual   species,  are  well
    simulated throughout the entire period.  Concentration peaks do exceed
    individual  observations,  especially for NH3 and P04,  but  they are
    generally within the range of the observed data; ,

f.  The Susquehanna simulation shows  good agreement between simulated and
    observed  concentrations,  but  this  was  accomplished,  only  after
    intensive calibration effort of  the  entire basin,  with particular
    focus on the reservoir simulation.  Generally, Total N and component
    concentrations  at  Conowingo are .somewhat under•simulated.   PO^ and
    Total  P  concentrations  are  somewhat  over-simulated throughout the
    Susquehanna, possibly due  to interactions  with sediment  and metal
    oxides, and increased sorption under low pH conditions.

    USGS data for pH show extremely low values for two main stem stations
    in the East and Vest Branch subbasins,  with mean  monthly values in the
    3.0  to 4.0 range.   Detenbeck  and  Brezonik   (1991)  show  greatly
    increased P binding by lake  sediments for pH in the range of 4.5 to
    6.0, with greatly decreased diffusive P fluxes under these conditions.
    Consequently,   increased   sorption  and  retention,   and  possibly
    precipitation of PO4 from the overlying water may be a reason for most
    of the observed concentrations being close to detections  limits.

    Clearly,  further  investigation  is warranted  and  should  focus  on
    instream processes in the Upper and Lower basins  for both nitrogen and
    phosphorus delivered to  the Conowingo Reservoir, and processes within
    the  reservoir.   This  should be  performed  in  conjunction with  a
    thorough loadings analysis  (see Section 7.4) to  confirm the relative
    loadings contributions from all sources.  Further calibration of the
    reservoir water quality  is  needed to  improve the sediment and Chi a
    simulations,  which may also assist  in improving  the  phosphorus

                                218

-------
          the TOG  simulation.  the  simulation results for  TOC are generally
          quite  good .for  most sites,  and are similar  in  appearance, to the
          organic N and organic P simulations.

7.3.2  Fall Line Load Simulation and Regression Comparison

The Phase I water quality calibration included comparison of  simulated nutrient
loads at  the  Fall Line  stations' with regression loads developed by the U.S.
Geological  Survey  (L.   Zynjuki  R.  Summers,  and  T.  Cohn,  1989,,  personal
communication) and HydroQual, Inc.  (1989),  TheU.S.G.S. regression was  used for
all stations  except the Rappahannock,  James,  Mattaponi,  and Pamunkey rivers,
where the HydroQual regression was used.  In Phase II, regression loads for 1984-
86 were provided by the U.S.  Army Corps of Engineers so that the same regression
loads would be  used^by both the  Watershed Model  and  the  Bay model.   In both
phases,   the  calibration  effort  considered  both  the  available  observed
concentrations (as discussed above) and the regression loads  whenever parameter
adjustments were proposed.  If parameter changes would improve one comparison at
the expense of the other, we placed greater emphasis and priority on the actual
.observed  concentrations.

Tables 7.2 through 7.5 show a comparison of annual  simulated and regression Fall
Line loads for the three  major basins, and the Patuxent. along with their ratios
for each  year, 1984 through  1986, and for  the  entire  three-year period.

Review  of  the  comparisons   in  the  tables  indicates  the  following  general
conclusions and recommendations:   •      '  '
              •                       •               '
    :  a.  For Total N  and  Total  P,  the  model  and  regression  annual loads
          generally  agree within about 10-202;  for  the  Potomac, the maximum
          difference is closer to 25Z. Problems with Organic  P and, to a lesser
          . extent Organic N,  in the Potomac and James in 1985  (partly  due to the
          November  1985  storm) lead  to  greater differences for  that year.

      b.  In .the Potomac  and to some extent the Patuxent,  the closeness of the
          concentration   comparison is  also  reflected in the Fall  Line load
          ratios.   For  both  these  rivers,  extensive data  was available for
         . developing the  regression .and performing  the model calibration.

      c.  Nitrate  generally . shows  the  closest   agreement   between  the  two
          estimates,- at  least   for the  three  major  Fall  Line  stations.
          Interestingly,  nitrate  dominates the Total N and  all other nutrient
          loads  in  terms  of .total mass,  especially  for   the  Susquehanna,
          Rappahannock,  and the Patuxent,  and  to a  lesser  extent,  the  Potomac.
          In  the   James,   Organic   N   dominates  the Total  N  load  and  is
          significantly  greater than the nitrate load.

      d.  The differences between the two  estimates are generally much greater
          for the minor  tributaries for most nutrient forms (with the exception
          of  the Patuxent),  than for the major Above Fall Line basins.   For the
          minor tributaries  and the James,  significantly less data was available.
          for developing the regression. Consequently, the .uncertainty of the
          regression is  greater for  these  basins.
                                      220

-------
TABLE 7.2        Susqueharma River Fall Line Loads: Comparison of Regression and
                 Model Estimates - (Millions of Pounds)
Constituent Yc
N03 1
1
I
1
84-M
NH3 1
1
1
1
•4-U
Org H (

1
1
•4-M
Tot N 1
1
1
1
•4-tl
P04 1
1
1
1
•4 -ft
Org P


'
•**
Tot P



•4-
TOC



•4-
Regression . Model
ar Estimate Estimate
M 123 128
6 80 74
16 108 W
17 - - . 77
Ntan 104 . 99
* 11.0 14.6
5 7.3 7.4
I* 9.5 9;4
17 • 7.3
, Haan 9.3 10.5
14 60 52
B 33 27
ft 46 36
ff - . 29
. aปan 46 38
ft 194 194
B 120 108
ft 163 13*
ff • 113
kftMn 159 ' U7
ft 1.6 1.0
B 0.9 0.8
ft 1.0 0.8
ff • 0.
kซaซป 1.2 0.
7.9 8.
3.1 4.
5.5 5.
4.
•MM 5.5 6.
9.5 9.
4.1 S.
6.5 6.
$.
•MR 6.7 7.
417 634
209 235
299 336
263
•Mn 308 402
Model
Regr.
1
0
0

0
1
1
0

1
0
0
0

0
1
0
0

0
0
0
0

0
1
1
1

1
1
1
1

1
1
1
1

1
Est. /
Est.
.04
.93
.87
.
.95
.33
.01
.99
.
.13
.86
.81
.78
.
.82
.00
.90'
.85
.
.93
.63
.89
.80
.
.74
.08'
.39
.05
.
.13
.00
.24
.02
-
.05
.52
.13
.12
' .
.30
To compare modปl aattwtaa with regression estimates, the following model  outputs Mere used:

1.       N03  • WOt ซ ซS                                .
2.       NH4  • Oiuolvad ซH4 * Adsorbed MH4
3.       Organic •• ซ6-K * (0.08631 • PHYTO)* (0.05295  • BOO)
4.       Total Ultrotan • 1 * 2 * 3
5.       P04  ป Dissolved P04
6.       Organic • • ORC-P + (0.01233 * PHYTO) ป (0.00756  * BOO)  * Adsorbed P04
7.       Total Photpnorous = 5*6
8.       Total Organic Carbon = ORG-C ป (0.49 • PHTTO)  * (0.3006  • BOO)
                                           222

-------
TABLE 7.4        James River  Fall Line Loads: Comparison of Regression and
                 Model Estimates  -  (Millions of Pounds)
Constituent
N03




NH3




Org N ,




Tot N




P04


.\

Org P




Tot P



• . :
TOC

- .


Regression
Year Estimate
84
85
86
87
84-86 Mean
84
85
86
87
84-86 Mean
84
85
86
87
84-86 Mean
84
85
86
87
84-86 Mean
84
85
86
87
84-86 Mean
84
. 85
86
87
84-86 Mean
: 84 - •" '
, -. 85 '
86
87
84-86 Mean
84
85
86
'87
84-86 Mean
6.2
. 4.4
2.3
.
4.3
0.73
0.64
0.36
.
0.58
7.9
5.6
3.2
.
5.6
15
12
6
-
11
1.2
1.2
1.0
. . •
1.1
2.7
5.2
0.7
•
2.9
3.9
6.4
1.7
•
4.0
130
165
44
•
113
Model
Estimate
7.8
4.9
3.4
7.0
5.4
0.57
1.08
0.60
1.10
0.75
6.9
7.8
1.3
11.8
5.3
15.3
13.8
5.3
19.9
11.5
1.1
1.6
0.4
1.5
1.0
1.5
1.4
0.3
2.0
1.1
2.6
3.0
0.7
3.5
2.1
101
169
17
231
96
Model Est. /. ,
Regr. Est.
1.26
•1.11
1.48
.
1.25
0.78
1.69
1.67
• -
1.30
0.88
1.40
0.40
. .
0.96
1.02
1.15
0.88
.
1.04
0.92
1.33
0.40
.
0.91
0.56
-0.27
0.43
•
0.37
0.67
0.47
0.41
; . .
0.53
0.78
1.02
0.38

0.85
To compare model estimates with regression estimates, the following model outputs were used:

1.     .  N03 * N02 ซ  N03
2.       NH4 * Dissolved  NH4 + Adsorbed NH4
3.       Organic N •  ORC-M + (0.08631 * PHTTO) + (0.05295 * BOO)
4.       Total Nitrogen * 1 + 2 * 3
5.       P04 ซ Dissolved  P04
6.       Organic P ซ- ORG-P + (0.01233 * PHYTO) * (0.00756 • BOO) * Adsorbed P04
7.       Total Phosphorous ซ 5 * 6
8.       Total Organic Carbon ซ ORG-C * (0.49 * PHYTO) * (0.3006 • BOO)
                                            224

-------
            LOWER SUSQUEHANNA RIVER AT SEG, 80
                               WATER TEMPERATURE  (C)
                          RED DASHED:  SIM.. BLUE STARS: DBS.
   40
 s
 I
 M
en
a
n 20-
d •
o
B :
s *
"
ซ

' "Si H
: 'f ',V
," 'ปซ*
Kl' i* M
• ?• %' '
' • -VW • ' * '
i ^ i
J J . I
1 h '"
' il |i "
I j 2/
           I I
                   T I I  I I T1TTT
i~n
                                        i  i ri i i  i i i  r*ri" i i  i i .r'TTT
0000000000000000000000000000000000000000000000000
i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i  i i i i i  i i i  i i i i i  i i i i i i
JFMAMJJASO NO J F M A M J J A SONDJFMAMJJASONDJFMAMJJASONDJ
AEAPAUUUECOE A E APAUUUECOEAEAPAUUUECOEAEAPAUUUECOEA
NBRRYNLGPTVCNBRRYNLGPTVCNBRRYNLQPtVCNBRRYNLGPTVCN
8888888888888888688888888888888888888888888888886
4444444444445555555555556666666666 66 7777777777778
                                DATE
Figure 7.6 Simulated -:ind observed water temperature (ฐC)
                                          Susquehanna River at Harrisburq, PA, 1984-87.

-------
       LOWER  SUSQUEHANNA RIVER AT  SEG.  80
       .                TOTAL SUSPENDED SOLIDS  (MG/L)
                 RED DASHED:  SIM. TOTAL. BLUE BARS: DBS. TOTAL
700 -
•
600-
S
1 500-
M
*

400-
a
r*
.
d 300-

0
S 20ฐ"

'.
100-

0-










• ••. • '•. ' :'••.-.
1 ' ./ ':'•:•

'it1 . ' ' •' . •
!'|!
i,il . • •. .
!'

iซ< 'i
1 * l • 'l
ijt ซ !'
i|i n 'f. .
iซi n - i| . •
1 1 n •!•'.•"' • 1
! nl ii i*
: ป5 'i ij
1 n* h n
* iih - i| l'
* ll'i *i M
n MM iiir n
^ i\l \ f\ซ !\fซ i'v'i Lii • J.'ซ
\j\ ^ 'f/'v \M' ^__iniซ ^ jj* V^jfvP'^A^V ^
II 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 11 1
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
JFMAMJJASONDJF MA M J. J A S 0 N
AEAPAUUUECOEAEAPAUUUECO
NBRRYNLGPTVCNB R R Y N L Q .P T -V
88888888888888888888888


..'."!':
!!
n :
• n
n
ii
n
n
X
II
II .
II
II
II
II
II
II
II
II
J II 1
ff !'•••'
8 L' • J i
S ' ! I
ii * ! 5i i

V./*!ปX"%!ป^ซ ^j
i i iii i i.i i i
0000000000
DJFMAMJJAS
EAEAPAUUUE
C N B R R Y N L G P
8888888888
5666666666
•

'' '". • ' "

• • . . '• . • .


.


i i

' • ป " •
! i! .
'
J j! ซ.; | .
j :! , J \.
i }t',( ^ฃ jl j!
LI* \4* ^'MS-Ai^f vซ*^*i i*vi_*'^
1 1 1 1 1 1 1 1 1 1 1 1 1
0000000000000
ONDJFMAMJJASO
COEAEAPAUUUEC
TVCNBRRYNLGPT
8 8 8 8 8 8 8888 8 88
6667777777777










.,





,
f '', iซ
i,1^*!^'1
1 1 1
00 0
14 4
1 1
N D J
0 E A
V C' N
,888
778
                                DATE
Figure 7..n  Simulated and observed suspended sediment (mg/1), Susquehanna River at llarrisburg, 1'A, 1VM/1-H7.

-------
       LOWER SUSQUEHANNA RIVER AT SEG. 80
                        NH3 CONCENTRATIONS  (M6/L)
                     RED DASHED: SIM.. BLUE STARS: OBS.
EST
1.1-
1.0-
0.9-
0.8-
0.7-
0.6-
.
0.5-

0.4-

0.3-
0.2-
0.1-
0.0-






.
* . ' •-


*l
j
1
j

i 11
S, t i i
1 i
\f J > . l^'-.s'Mi'

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
000000000000000
J F M AM JJASONDJFM
AEAPAUUUECOEAEA


- ' *
• i- . . ' .
• ' • ' . i • -

. ' '• • ' j

II 1
n
11 i
n S
n i
ii *
1 !l i I 1 ! I
| n j , i i T
MtJW*m TR ป"• •ซ* ^m ป" . *
\ \ i i i i 1 \ i \ i i i i i i i i ii i i i i i i i i i i i i
000000000 000 0000 00 0000 000000 00 00
AMJJASONDJFM AMJJASONDJFM A MJJASON
P A U U U EC 0 E A E A P AUUUECOEAEA PA U U U EC 0












„_
K

1 l
o o
D J
E A
N B R R Y N L 6 PTVCNBRRYNL.6PTVCNBRRYNLGPT V C NBRRYNL6PTVCN
8888888888888688888888688888888888888888888888888
44444444444455555555555566666666666677777777777 7 8

                               DATE

 Figure 7.10 Simulated and observed ammonia (mg/1), Susquehanna River at Harrisburg, PA, 1984-87.

-------
       LOWER SUSQUEHANNA RIVER AT SEG. 80
                       TOTAL N CONCENTRATIONS (M6/L)
                     RED DASHED: SIM.. BLUE STARS: OBS.
EST
5:

4-



3-
o
S3 '.
2-



1-

0-











'.••"•
t • ' ~ •
I M • '
il* it * ป
f*\ 1 1 p
rU Ml * M
TO! iv "V/H-^i
r VJ ^ 1 \ • ' ' *f j^l M/1 ^ il
'i\* M ง !/ป ^J
, "*" \ป/N v *

• '••••'• ' .' : *
.••-••--
I 1 1 1 1 , 1 1 1 1 1 1 1 1 1 1 1 1 1
000000000000000000
JFMAMJJASONDJFMA M J
AEAPAUUUECOEAEAPAU
N B R R Y N L 0 P T V C N B. R R Y N
888888888 868 6 8 88 88



"• " :".'••' '*'•• '• . :• '
'.'••"'••' ••'..-,-

•".'' .- . . ;'. ' •_••/•-.
n
• . - . . l| 	
. . • *ป ll w
M ^ [| ^
f ;V* V •."*f''1 ^ 1* 1 ^W
^ \ *^C L • if ' '\l • * 1
NH, W ' f •' * ^m*!/^!'
^1 ' • . .
' ". 'v " /. : "*• • '

i i i i ii i i i t i i i i ii ii
000000000000000000
JASONDJFMAMJJASOND
UUECOEAEAPAUUUECOE
L6PTVCNBRRYNL6PTVC
868888888888888886
555555666866666666


*

t -
*
' * •'"..

** 8 * ซ i
m/4 K r[ i J )l
II ll • \ IJ II .1
Jjft/jfi "iw Jl*l* d
ป" • iffflWjvk j
j/ ~ • 1 ?ซ P
? -M\j
VI
• m

1 1 I 1 i i > i i
000000000
JFMAMJJAS
AEAPAUUUE
N B R R Y N L 6 P
888888886









•I,
i^X4/ *V

JK



till
0000
0 N D J
C 0 E A
T V C N
8888

                                DATE
          .        .                              ...
Figure 7.12 Simulated and observed total nitrogen (mg/1), Susquehanna River at llarrisburg, I'A, I9H.4-H/.

-------
          LOWER SUSQUEHANNA RIVER AT SEG.  80
                        PARTICULATE-P CONCENTRATIONS (MG/L)
                        RED DASHED:  SIM.. BLUE STARS: DBS.
EST
0.36-
0.34-
0.32-

0.30-
0.28-
0.26-
0.24-;
0 . 22 -
0.20-
O.lB'-j
0.16-
0.14-
0.12-^
0.10-
0.08-
0.06-
0.04-
0.02

o.oo:



.


•
1
j
j
!
j
i
/
t
i
^



\ i
0 0
1 1










•
i't\
yilW
\
• *
* m


i i i
OOOl
i i i ,










l'
II




1
3
1

ซ '
1 .




i

'-';-•
' . 1 i
1 'j'i 8
MI ;v ' ซ
lU f' \ป i
Vl/ "! /
" W
m x
' '. • * ' • • '


i i i i i i i i
00000 00 0
11111111
1
7



.'•'''
1 '

1 1 •
1 I
1 I
,,,Pii,,lj
>-fy \M
A**
X M * * J
M --..JK • gf
W ' ' '' • . m

1 J l.| 1 1 II 1 II 1 1
00 00 O 000 00000
11 11 i i i i i i 111
[









K
,v
;
.K
K



1
0
1

• '• ' ' W
m
>


1

1
1
1
H i i
ii* s h
ijwj/v rrsfci
JW^MW*
„ ;-. ซซ V^
M v~y w^
* • s J^
)({)){ )K "^
K

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
0000 000 000000 OC

1


1
I
I
1

2
III I
v3 1-
s',-v •
v i'V
ป 5K K * *
*


1 1 1 1 1 1
(000000
L 1 1 1 1 1 1
                             .
   JFMAMJJASONDJFMAMJJASONDJFMAMJJASONDJFMAMJJASONDJ
   AEAPAUUUECOEAEAPAUUUECOEAEAPAUUUECOEAEAPAUUUECOEA
   NBRRYNL6PTVCNBRRYNLQPTVCNBRRYNL6PTVCNBRRYNL6P TV C N
   88 BBSS 8688888888686888 6888888888 8 88888868888688 88
   444444444444 55 555555555 56 6 6 666 6666667777777777778

          '   ••          '   :.'•     '    'DATE ••.;_'.-.
    Figure 7.14  Simulated and observed particulate. phosphorus (mg/1),  Susquehanna River at llarrisburg, I'A,

-------
M
Ul
                                                      I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I
                JFMAMJJASOND
                          1984
JFMAMJJAS ON 0
         1985
JFMAMJJASOND
         1986
|J F M A M J J A S 0 N P

         1987
                                                      HARRISBURG TOC (MG/L)
              Figure 7.16 .  Simulated and observed total organic carbon, (mg/1), Susquehanna River at Harrisburg,  PA,
                           1984-07.

-------
              SUSQUEHANNA RIVER AT CONOWINGO
                               WATER TEMPERATURE (C)
                          RED DASHED: SIM,. BLUE SOLID: DBS.
   40-
  S
  I
  M
  30-
CO
oo
a
n
d

0
B
S
   •20-
    10-
                                                            \/
    I  I J
    0  00
    111
    J  F M
    A  E A
    N  B R
    888
    444
I  I I  I l  I I l
00000  0 0 0
11111111
AMJJASON
                      I  I I  I. I I  I I  I II I  I .1  II  I I I  I. I  I II I  FJ. TIT I II I  I f  I I
                      000000000 0*0 000000000000000000000000000
                      i  i i  i i i i i  i i i i i i  i i  i i i i i  i i  i i  i i i i i i i i i  i i  i i
                      DJFMAMJJASONDJFMAMJJASONDJFMAMJJASONDJ
          PAUUUECOEAEAPAUUUECOEAEAPAUUUECOEAEAPAUUUECOEA
          RYNLBPtVC NBRRYNL6P TV CNBRRY.MLBPTV. CNBR'RY N L 6 P T V C N
          8866888888686888668868888868888888888888888888
          4444444445555555555556666666666667777777777778
                                        DATE
      Figure 7.18  Simulated and observed water temperature (ฐC), Susquehanna River at Conowlnpo, MD, 1984-87.

-------
          SUSQUEHANNA RIVER AT CONOWINGO
                        N023 CONCENTRATIONS (MG/L)
                     RED DASHED: SIM..  BLUE STARS: DBS.
EST
B:

A
3
M
^^
+•*
o '
2


1

0-










• • ' • • -
* ' • • . . . .
/" ill * ft
f\ /! ! l\.
/L/ i. A f. * "IV"
Vvwป!! • /wMv
/ if ?• } ป* *;ซi/ 1 r U _
n * 1 ป *— I ซ 1 • ••^fc *•
'' * - ^IVN % A sr - v\
V'^v/ V V
• ' ; .• . -"* . v
i i i i i i i i i i i i i i i i i i i i
0 00 0 0 00 0 000 0 00 00 000 0
14444444444444444444
1111111111111111111
JFMAMJJASONDJFMAMJJA
AEAPAUUUECO E A E A P A U U U
NBRRYNL6PTVCNBRRYNL6
88866886688888888888

' • . - *
'*



t. JL ft ' *
m 'ป *% !l *
• I '\ ' * I
•ป^N ^ 1 \ . fl*\A w \ * *r*
Xw \ /T.i rj. \ 1 J* V V l-rซ W V # J*v
. f^ }i v'V% vS w /? ^i Aซ v * ^ *
• rf • i - Jt ^tr.^. r/
* / • 1 A • /~ * * VM */ *
-\x-v; V!
•^ \/
1 1 i i 1 i i 1 i 1.1 i i i.i i i i i i 1 i i i ii ii i
0 0 000000000000000000000000000
S 0 N D J FMAMJJASONDJFMAMJJASONDJ
E C 0 E A E A P A U U U E C 0 E A E A P A U U U E C 0 E A
P'T'V C N B R R Y N L 6PTVCNBRRYNLGPTVCN
88888888888888686888868668868
4 A A A A 4 A A A A A A S S S S S S B ฃ S S S B 6 6 6 6 6 6 6 6 6 6 6 6 7 7 7 7 7 7 7 7 7 7 7 7 B


             • • '   . ' ' .  .   '•• .-• -  DATE  •    •'•;-'

Figure 7.20 Simulated and observed nitrate (nig/I), SuH(|iielinnnu Klvcr at Conowliigo, Ml), I984-H7.

-------
             SUSQUEHANNA RIVER AT CONOWINGO
                         PARTICULATE-N CONCENTRATIONS (MQ/L)
                         RED DASHED: SIM.. BLUE STARS: OBS.
EST

  4H
  a-
  1-
        x
                                                         x


   ooo oooooooooo ooo oo oo oooooo od ooooooooooooooooooo
   11111111111111111111111111 111111111111111111111
   JFMAMJJA30NDJFMAMJJASONDJFMAMJJASONDJFMAMJJASON
   A ฃ A PA UUUECOEAEAPAUUUECOEAEAPAUUUECOEAEAPAUUUECO
   N B R R Y N L 9 P T V C N B R R Y N L Q P T V C N B R R Y NLGPTVCNBRRYNLGPTV
   88686886888888668888886886888688688888888888888
   44444444444455555555555866666666666677777777777

           .     . • ' .     .     '••'..  DATE .              '           '
    Figure 7.22  Simulated and observed pnrtlculnte nitrogen (mg/l), Susquohnnnn Rfvcr ;it Conowln^n, Ml), lซ)H/i--H/,
0 0
1 1
D J
E A
C N
8 8
7 B

-------
          SUSQUEHANNA RIVER AT CONOWINGO
                          P04 CONCENTRATIONS  (M6/L)
                      RED DASHED: SIM., BLUE STARS:  OBS.
  EST
 0.16-
 0.15 -
 0,14-
 0.13-
 0.12
 0.11
 0.10-j
 0.09
;: 0.08
 0.07-
 0.06-
 0.05-
 0.04-
 Q.03-
 0.02-
 0.01
 0.00
                 .*'
                     * *
                        II
         '   4
\ป   *xW, x xx
-------
              SUSQUEHANNA RIVER AT  CONOWINGO
                            TOTAL P CONCENTRATIONS (MG/L)
                          RED DASHED: SIM..  BLUE STARS: DBS.
EST
1.5-
1.4-
1.3-
1.2-
1.1-
i.o-
0.9-
0.8-
0.7-
0.6-
0,5-
0.4-
0.3-
0.2-
0.1-
0.0-

















X
-'!' A
ป' '"•"*
i i i i
0000
J F M A
A E A P
N B R R
8888
/ - . ' • . " ' . ' - • . *
.''•'' '"'.'•'• • ' '
*






• • '
i t .-.'....." . •
II ' ' ' ' ' •• • ' " ' ' '• '
~ *
1 'I 1 T * |
\jV\ j ' * \\ fป ^ i \ '\ ^.A iiซr * /v . M
ii i i i i i. i i i. i i i r i •! 1. i 1.1 i ii i i ii i i i i i i i i ii i ii i i i
00 0 0 0 00 000 0 00 0 000 00 0 0 0 0 0000 0000000000000000
MJJASONDJFMA M J J A SON 0 J FMAMJJASONDJFMAMJJASON
AUUUECOEA E A P A UUUECOEAEAP A U U UECOEAEAPAUUUECO
YNL6PTVCNBRRYNLOPTVCNBRRYNLGPTVCNBRRYNLGPTV
8888888888888 88 B 888 88888888888888888888 88 88








•



,1.
****
1 T
0 0
14
a
D J
E A
C N
8 8
                                      DATE

    Figure 7.26 Simulated and observed total phosphorus (tng/1), Susquehnnna River at Conowingo, MD, 1984-87.

-------
             SUSQUEHANNA RIVER AT  CONOWINGO
                           CHL_> CONCENTRATIONS  (UQ/L)
                        RED DASHED: SIM.. BLUE STARS: DBS.
EST
110-
100 -
90 -

80-
'
70 -
60-
to „ -'
ซ 50- 1. J
K ,n
"• A,i%
*:./ 4
20- , |f
10- j If
1
o- •**
i i i i
0 0 0 0
J F M A
A E A P
N B R R
8888


. •• ' .. • './V:'. ' -..':.
fv'l
P,
1
1
* 1
* 1 '
l( l\ 1 lซc
JWvji jc'^
/ '- ^rfij/ 1 riW| !
' n/ i! !
w U l| '
f h .
* * * IK ' MW JK '
. JK * *.-..• . ' .
i i i ii lit II i i.i i i i i i i
00000000000 000 000 0 0
MJJASONDJFMAMJ JASON
AUUUECOEAEAPAUUUECO
Y N L Q P T V C N BRRYNLOP TV
8888880888888888888

^
' ' ' • '••'• . .'•' ••'
1
' \ '••'•..
•I
I M

| '. J\'> i
i i 11 ll L I* ^ 1 1 1 ^
jlซ ^ 1*1* • VJ- ll*l *l/ Fib I* K ^w
* fl A ^V II r 1 MI 5l * *
A f 'lu'if. ^ j/*- ^ l/ป
•V " 'f , J/V .:•!.;
.' ' ' •' .- ' ' • ••• : •' "
** ••'.'.'•,. . •
i i i i i i i i i i i i i i i i t i i i i i i i i i
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DJFMAMJJASONDJFMAMJJASONDJ
E A E A P A U U U EC 0 E A E A P A U U U E C 0 E A
CNBRRYNLQPTVCNBRRYNLQPTVCN
868 88 8888888 888 888 88 8888 88

                                   DATE
    Figure 7.28  Simulated and observed chlorophyll A .(ug/1) , Susquehanna River at Conowingo, MD, 1984-87.

-------
                LOWER POTOMAC  RIVER AT SEG. 220
                          DISSOLVED OXYGEN CONCENTRATIONS (M6/L)
                           RED DASHED:  SIM.. BLUE STARS:  OBS.
N)
Ln
o
 M
  17-



  15-

  14

  13-

  12-

  11-

  10-
d 8
  7

0 6
B e
    2

    1

    0
                          M


                          M
                        I  I
                             I I
I I  I I
I I  I I
I I
II I
                                                                    I I
I  I I  I I  III I I  IT

00000000000 00 000 000 0 0000000 000 0000000 000 000000000
111  11111111111111 111 111111111 1111111 .111111111 1111
JFMAMJJASONDJFMAMJJASO NO JFMAMJJASONDJFMAMJJASONDJ
AEAPAUUUECOEAEAPAUUUECOEAEAPAUUUE C 0 E AEAPAUUUECOEA
NBRRYNL&PTVCNBRRYNL6PTVCNBRRYNL OPT V C N B R R Y NLGPTVCN
8  88  8 88 8 88 88 8 8 B 88 B 88 8888888888888888 88 888888888888
4444444444445555555555556666666666667777777777778


               '•      '     '-.''.    " DATE      • '    ' '     '   •
Figure 7.30  Simulated and observed dissolved oxygen (my/1), Potomac River  at Wushinlun, U.C., J'J01-U7.

-------
                LOWER POTOMAC RIVER AT SEG. 220
                                NH3 CONCENTRATIONS  (MG/L)
                            RED DASHED: SIM..  BLUE STARS: DBS.
10
EST
0.8-
•
0.7-
0.6-
0.5-
0.4-
•
0.3-
0.2-
0.1-
0.0-












•

\ ซ•(!•
1*1
•J*Uikli '
!Hfw1
•MMrt
iiWflH."ir
III 1
0000
J F M A
A E A P
N B R R









L
MM*
i i
0 0
M J
A U
Y N

••'.'..• .' ' "-•••• ' • ' -
• • -
•'.'•• • '
•---'...' '•'". /. •'••'.-' : '
• <; . • • •
. • "~- • •
1 ••''.'- . ' . ' • ' •" . •
' J
••••*' I - | |'| '
i ir Vi ••'•/'•"•• -i/i v.j
1 1 I 1 • -b" iS'l^ H fc. *l 1 lll'T it Ll i
kLdkAA^ Md^ปlfe*vuiii&
i i i i i i i i i i i i i i i i 1 1 1 i i i i i i i i i i i i i i i i i
000000 0 000 000 00000000000000000000000
JASONDJFMAMJJAS ON D J F M A M J J A SO NO J F M A M J
UUECOEAEAPAUUUECOEAEAPAUUUECOEAEAPAU
LGPTVCNBRRYNLGPTVCNBRRYNLGPTVCNBRRYN






.


II
x 1 H
I '1 * ,
4 n*./i
/ 'USilf
S(Jlf*5KTK THfffX^K
1 1 1 1 1 I 1
0000000
J A S 0 N D J
U U E C 0 E A
L 6 P T V C N
8888
           8 88888888888888888 8 BBS 888 88 888888888888888868

     4444444444445555555555556666666666667777777777778


                       -  ; '. ';'••  '  •'•/•'  DATE      '  •  .  :. •

      I'igure 7.32 Simulated and ohscrved ammonia (mg/l), I'utomac Kiver at Washington, DC, 1984-87.

-------
          LOWER POTOMAC RIVER AT  SEG.  220
                        TOTAL N CONCENTRATIONS (MG/L)
                      RED DASHED: SIM.. BLUE STARS: DBS.
EST
9-
8-
7-
6-
5-
K
* 4-
3-
2-
1-

0-







• • ' . •
-''''••••ซ ':.•••



ป . . ''*•_.
* - ' • '
Sl JK
I ll w M !|l
4fiซ-j^i V* ( * N 'i' w*n
^i^j<^ fcj i ijf^" V \?
. v • . y ' . .' ^ *^*jr*
"... . ' ••• . ..••...
. - : . • ' •
i i i i i .1 i i i i i i 1 i i i i i i i i .1 i i i i i i
0000000000000000000000000000
JFMAMJJASONDJFMAMJJASONDJFMA
AEAPAUUUECOEAEAPAUUUECOEAEAP
NBRRYNL6P TV C N B fl R Y N L 0 P T V C N B R fl
8888888888888888888888888888



'" - : • . • * •





1 .1 i ! •
•I ' n^^ ^' Ivnii i(W n
|l jj >< gw J"l L F "1?
\r n ซซw * • • || TJ^ Ir^yKl^^ M/ i^J V jiciK
• i •• g__ . j ^y ™ m j(f ^ f^tj/I ^* n

* - ' • •

i ii i ii i i i i i ii i i i lit
00000000000000000 0.0
MJJASONDJFMAMJJASON
AUUUECOEAE A P A U U U E CO
Y N L 6 P T VCNBRRYNLGPTV
8888 88 88888 88888888









m
'*•



1 r
0 0
1j
l
0 J
E A
C N'
8 8
7 8
       • .         -. •         ••••..' -.DATE  '  •  '  .      - . -

Figure 7.3A Simulated and observed total nitrogen (mg/1), Potomac River at Washington, DC, 198/^-87.

-------
               LOWER POTOMAC RIVER AT SEG.  220
                          PARTICULATE-P CONCENTRATIONS (MQ/L)
                           RED DASHED: SIM.. BLUE STARS:  DBS.
EST
  3
  2-
Ln
      K
   000 0000 0000000 0000 00 00000 000 000 00000000 0000000000
   ill 11111 111111 11 i Ji i i iiiiiiiiiiiiiiiiiiiiiiiii 11 i i
ONDJFMAMJJASONDJFMAMJJASOND
  0 E A E A PA UUUECOEAEAPAUUUECOE
                                                         JFMAMJJASONOJ
                                                         AEAPAUUUECOEA
                  V C N B R R Y N L 0 P T V C N B R R Y N L Q P T V C N B R R Y N L 6 P T V C N
J F M A M J J A S
AEAPAUUUEC
N B R R Y N L G P T
6868688888886888888688888868688688888888888888888
4444444444445555555555556666666666667777777777778

 '•  '    .      •  '•'.''   ••'•:'•'   .DATE  .      '     '   .      ..-'•;
 Figure 7.36  Simulated and observed particulate phosphorus (mg/1), Potomac River at Washington, DC, I98/I-H7.

-------
M
in
03
   60

   54

   48

   42

   36

3, 30

   24
o
o
              12

               6

               0
                   \  n  i  i  i  i  i  i  i  i  i  i n FI I  rn  n  i  i  n  i i  i  i  i  i  i  i  i  i  r i  i  i  i  i  i i i i i
                        —	     SIMULATED
                        *      *'   OBSERVED
        i  i  i  i  i  i  i  i  i  i  i  i  i i  i  i  i  i  i  i  i- i  ill  ii i  i i
                 J F MAMJ J A SON D
                          1984
                              JFMAMJJASONDIJFMAMJJASOND
                                       1985          I         1986
                                         LOWER POTOMAC TOC (MG/L)
                                                                             JFMA'MJJASOND.
                                                                                      1987
                 Figure  7.38  Simulated and observed total organic  Carbon  (mg/1), Potomac River at WnshintjUm, I) c
                             1984-87.

-------
                          JAMES RIVER AT  SEG.  280
                                   -  WATER TEMPERATURE (G)
                               RED DASHED:  SIM.ป BLUE STARS:  DBS.
   .40 H
N)
a\
o
  n
  0
  B
  S
    30 '-]
    10-
     0-
       i i  i  i i  i T~T  r i  i  i r~i  i i  i  i i  i r~i  i i  r~r '  i  i i  i i  i  i r~i  i i  rn i  i  i i  r'T'i  r r
       0 0  0 000 00  0 0 00 0 0 0 0.0 00  00 00 00 000.00000000  0000000 000000
       1 1  1  1 1  1 1 1  1 1 l' 1 1 1  1 1  1  1 11 1 1.1 1  11 11  1 11111111111111111 1  1 1
      : J F  M A M  J J A  S 0 N D \J F M A  M J J  A S !0  N D  J F M A M J  J A S  0 N D  J F  M A M  J J A  S 0 N  D J
       AEAPAUUUECOEAEAPAUUUECOEAEAPAUUUECOEAEAPAUUUECOEA
       NBRRYNLGPTVC N 8 R R  Y N L  G P T  V C  N B R R Y N  L G P  T V C  N B  R R Y  N L G  P T V  C N
       86686668868886888686886 666 66666668886866888886888
       444444 44  4444555555555555  66 6.6 666666667777777777778

            .        ' . ':  ..        .        '   DATE  ;     .'..--.  .       ...
       Fiyure 7.40 "Simulated and observed water temperature (ฐC)ป James River  at Cartersvilie,  VA,  11JM4-H7.

-------
                 JAMES RIVER AT SEG.  280
                        TOJAL SUSPENDED SOLIDS  (MG/L)
                 RED DASHED: SIM. TOTAL. BLUE BARS: DBS. TOTAL
1800 -
1700-
16.00-
1500 -
S .1400-
I 1300-
M 1200-
• iiop-
a 1000^
n 500 •
d 800-
700-
0
Q 60ฐ-
B
g 500 -
. 400-

300 -
200 -
100-
0'
-








• ''.'.' . . . • - ;
" * p

. ' • '• _






. .. . i
> • ' • ' i
• i
" " ' ' • i
i
~ • • ' • - i
.' . ' i
•''•'" 1
•i ^
'. '••-.' . i
• ' .1
•'''•!•" - • • i
m ซ" ' • • l
i . • 1
.ฃ''••' , • ..'•''" '
<, fejjfc ^ jjl^AJL Jw.. ปw . -A* J
*', ** •'• *..* ^
ii i i i i ii i i i ii i.i.i i i i iii i
0 0 000 0 0 0000000000000000
J F M A M J J A S 0 N D J F M A M J J A S 0 N
AEAPAUUUECOEAEAPAUUUECO
NBRRYNLGPTVCNBRRYNLGPTV
888888888888888888888 88

. • ' ' • - .. . •

• * - .
• , * •


".'••".• ' - ' ' • . • '


i - • - . .
•
i • . .
i • •
i -. • • . '
• • •
f
- •' • ' ' • • .
' *•
i • . •:•' ' •. •- :.'..*
m . • .
i . . . . . . • ,
i i
• i
i
5 • ''. • ••• ' ' '•
\t . .. j- \.\
-
iii i i i ii i ii i i i iiii i ii i
0000000000000000000000
DJFMAMJJASONDJFMAMJJAS
EAEAPAUUUECOEAEAPAUUUE
CNBRRYNLGPTVCNBRRYNLGP
6888888888888883888388
566666 6666 .6 6 6 777 777777















i

r.
i
i
i
i
Lx..

iiii
0000
0 N D J
C 0 E A
T V C N
8888
7778
Figure 7.42 Simulated and observed suspended sediment (mg/1), James River at Cartersville, VA, 1904-07.

-------
                  JAMES RIVER AT  SEG. 280
                   '        NH3 Concentrations  (mg/1)
                       RED DASHED: SIM..  BLUE .STARS:  DBS.
0.7-
0.6-
s •
1 0.5-
M
*

0.4-
a
n

d O.Y
0
I P-a-


0.1-


0.0-













•
. t\
1*1
: •





i
I
ซ•'
i
I'll f i
• *VI 9f ' *ซ• 1

'• ' ' '• -' -•• ' .-'..- ' '•'•• ' ..-.- '-'''.''






• '. •••'''•'', '."


m • ' •'
''„••••'.' ' • ' " " • • •'


• '. . " 'I
"•. .-.-•....••• ..in
••-.-.••.: • - n . •
. . 1 M
•.„•.- 1 '1
* II
..M
/i I ' V
M, /! , 111 1
' J 1
i ^'i . j\ v I I in] ''' "MI 1 I '!• • ' r i * i !
!! if V V1 A /! lii ง ! j}yrlLl/ N1! 1 1 ! jl,1 L / "Vi i n 1 1
'#r* V W4/ 'AA^y "Ir'w * * WAซ-d'V^ ; ^/Jitt i'1 "1


1 1 III II 1 III 1 1 1 1 1 1 1 1 1 1 II 1 1 1 1 1 1 1 1 II 1 1 1 1 II 1 III
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 6 6 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
J F M A M J J AS ONDJFMAMJJASON D J FMAMJ JASON DJFMAMJJ
AEAPAUUUECOEAEAPAUUUECOEAEAPAUUUECOEAEAPAUU
NBRRYNLQPTVCNBRRYN L 6 P T V C.N B R R Y N L 6 P T V C N B R R Y N L











1
i

- ji
;i
i> ' i*
,j' IK 'A.
"

i i i
000
A S 0
U E C
G P T


-








|,
Af.
4
' V
'(
ป


1 1 1
000
N D J
0 E A
V.C'N
8686888888888886888888888866808888888888888888888

4444444444445555555555556666666666667777777777778
          • •  •   '    . .   .    .-  ' •  •   'DATE-  •'-•••'   •-..-   . '.
Figure 7.A4- Simulated and observed ammonia  (mg/1), James River at Cartersville , VA,
                                                              1984-87.

-------
                         JAMES RIVER AT SEG.  280
                                Total N Concentrations  (mg/1)
                              RED DASHED: SIM..  BLUE STARS: DBS.
NJ
4-
s - •
I 3'
M
•
a
n 2-
d
0
B
s i-


0-








• : • • . • . /
' ' "'.*•••

* • . ' .

•ป• 'ซ'••• i i : '!
fm^^AJjl^^
* '*••.••'• x *
• *.-•. '

1 1 II 1 1 1 -1 1 II 1 1 1 1 1 1 1 1 II 1 1 1 1 II 1 1 II 1 1
000000000000000000000 00 0 0 00 0 000 0 0
JFMAMJJASONDJFMAMJJASONDJFMAMJJAS
AEAPAUUUECOEAEAPAUUUECOEAE A P A U U U E
NBRRYNLGPTVCNBRRYNL6PTVCNBRRYNLGP
88 888888 88 8888 888 88 8888 88 888 8 8 88 8
444 4 444444 44535 5 55555 S 5 5666 6 66666
*
*•

f
'

ii
!\
J W

*

1 1 1 1 1
00000
0 N D J F
C 0 E A E
T V C N B
88888
66677



•- -


M • "' i
' ''•' S -' • •'!
•Ait j
\j*
m

i i i i i i i
0000000
M A M J J A S
A P A U U U E
R H i Y N L 6 P
8888888
7777777






. • .
1 .
1A
^ X



Till
0000
0 N 0 J
C 0 E A
T V C N
8888
7778
                                          DATE

      Figure 7.46  Simulated and observed total nitrogen (mg/1), James River at Cartersvllle, VA, 1984-87.

-------
                        JAMES RIVER AT SEG.  280
                             Particulate-P Concentrations  (mg/1)
                             RED DASHED:  SIM., BLUE STARS:  DBS.
s
I
M
a
B
S
1.5-
1.4-
1.3-

1.1-

0.9-
0.8-
CT.7-
0.6-

0.4
0.3
0.2
0.1
0.0
                                         I
                                                    U

       \  i i  i  r i  i  i7}  r i  i  i i  i i  i  i i  i  i i  i i  i  i i  i i  i  i i  i i  i  i TJi  i  m  i i  n i  r
     0000000000.000  000 00  000.0000000000000 000000000  0000 00
     1111,11 111111111  1111111 1111  111 11111111111 111111  111
     J F  M A M  J d A S 0 N O J  F M  A M J  J A S 0 N  D J F  M A  M J d ASONDdFMAMddASONOd
     A E  A P A  U U U E C 0 E A  E A  P A U  U U E CO  E A E  A P  A U U U E  C 0 E A E A P A U U  U E CO E  A
     NBRRYNLGPTVCNBRRYNLGPTVCNBR R  Y N LGPTVCNBRRYNLGPTVCN
     88888888 8 888 888B 88888888 88  8 88888888 88 8 888 8 8888888
     444444444 4 4 455555 5.  555 5 55666666 66666 6,7 777777777778
            '. .          •            :        DATE. •
     Figure 7.48' Simulated and observed particulate phosphorus (mg/1), Jnmes River at CartersvilIc, VA,

-------
to
~J
o
O
O
80

72

64

56

48

40

32

24

•16.

 .8

  0
                      i  r
                     i  i  i  i  i  i  i  i  i  i  i  i  rn  i  i  i  ri i  i  i  i  i  i  i  i  i  i  i  i i  i i i \ TIT  T  T
                         ——    SIMULATED
                         *      x   • OBSERVED
                    F MA M J J A S O.N 0
                           1984
                               J FMA MJ J A SO NO
                                         1985
                                                     J  FMAMJJASONDIJ FMAMdJASOND
                                                              1986
r
1987
                                                        JAMES RIVER TOC (MG/L)
      Figure-  7.SO   Simulntud and observed total oryunic carbon  (my/i) ,  .hi
                    1984-87.
                                                                                       -n R.ivt, VA,

-------
                  PATUXENT RIVER AT SEG. 340
                           Total N Concentrations  (mg/1)
                         RED  DASHED: SIM.. BLUE STARS:  DBS.
14-


s
I
M
•
a
n

0
B
8
•



13-
12-
11-
.
lo-
g-
s'
.
7-
6-
5-
•
4-
3-
2-
1-

IT I  ป I  I
                                    TT 1
  0000000 00' 000 O 00000000 OOOO 00000
  lllllllllliilillilillillllllll
  JFMAMJJASONDJFHAMJJASONDJFMAMJ
  AEAPAUUUECOEAEAPAUUUECOEAEAPAU
  M B R R Y N L 6 P T V C N B R R Y N L G P T V C N B R R Y N
  888888886888688888888688868888
  4 4 A A 4 A A A A 4 4 4 6 5 5 6 S 5 8 5 8 B B 5 6 6 66 66

                                     DATE
*  i i  i i  n n i i  i ~~\ r r t i  i r
bbooooooooooooooooo
                                 i i
                          i i
                                 JASONOJFHAMJJASONOJ
                                 UUECOEAEAPAUUUECOEA
                                 L 9 P T V C N B R R Y N L 8 P T V C N
                                 8888886888688888888
                                 6 6 6 6 6 6 7 7 7 7 7 7 7 7 7 7 7 78
   Figure 7.52  Simulated and observed total nitrogen (m/'/l), Pntuxi'nt Rlyor at Ihiwli:, Ml), I9H/I-H/.

-------
                                  SECTION 8.0

                                  REFERENCES
Anderson, E.A.  1968.  Development and Testing of Snow Pack Energy Balance
      Equations.  Water Resour. Res. 4(l):19-37.

Anderson, E.A., and ft.H. Crawford.  1964.  The Synthesis of Continuous
      Snownelt Runoff-Hydrographs on a Digital Computer.  Department of Civil
      Engineering, Stanford University.  Stanford, CA.  Technical Report No.
      36.  103 pp.

Ardakani, M.S. and A.D. McLaren.  1977.  Absence of Local Equilibrium During
      Ammonium Transport in a Soil Column.  Soil Sci. Soc. Am. J., 41:877-879.

Armbruster, J.T.  1977.  Flow Routing in the Susquehanna River Basin:  Fart I
      - Effects of Rays town Lake on the Low-Flow Frequency Characteristics of
      the Juniata and Lower Susquehanna Rivers, Pennsylvania,  U.S. Geological
      Survey Vater Resources Investigations 77-12.                       '   '

Bandel, V.A., S. Dzienia, G. Stanford, and J.O. Legg.  1975.  N Behavior Under
      No-Till vs. Conventional Corn Culture.  I. First-Year Results Using
      UnlabeledN Fertilizer.  Agronomy Journal, 67:782-786.

Bandel, V.A. 1965.  Plant Food Removal by Field Crops.  Fact Sheet 98.
      University of Maryland Extension Service.  Baltimore, MD.

Bandel, V.A.  1990.  Instructions for Using MANURE9.WK1:  An Enhanced
      Spreadsheet for Calculating Application Rates of Animal Manure 
-------
Donigian, A.S.,  Jr., and N.H. Crawford.  1976a.  Modeling Pesticides and
      Nutrients on Agricultural Lands.  Environmental Research Laboratory,
      Athens, GA.  EPA 600/2-7-76-043.  317 p.

Donigian, A.S.,  Jr., and N.H. Crawford.  1976b.  Modeling Nonpoint Pollution
      from the Land Surface.  Environmental Research Laboratory, Athens, GA.
      EPA 600/3-76-083.  280 p.

Donigian, A.S. Jr., D.C. Beyerleln, H.H. Davis, Jr., and N.H, Crawford.  1977.
      Agricultural Runoff Management (ARM) Model Version II:  Refinement and
      Testing.  Environmental Research Laboratory, Athens, GA.  EPA 600/3-77-
      098.  294 pp.

Donigian, A.S., Jr., and H.H. Davis, Jr.  1978.  User's Manual for
      Agricultural Runoff Management (ARM) Model, U.S. Environmental
    .  Protection Agency, EPA-600/3-78-080.

Donigian, A.S,, Jr., J.C. Imhoff. and B.R. Bicknell.  1983.  Modeling Water
      Quality and the Effects of Agricultural Best Management Practices in
      Four Mile Creek, Iowa, EPA Contract No. 68-03-2895, U.S. EPA
      Environmental Research Laboratory, Athens, GA, (PB-83-250183).
                                 " *                             -•'"'.      /
Donigian, Jr., A.S., B.R. Bicknell, and J.L. Kittle, Jr.  1986  Conversion of
      the Chesapeake Bay Basin Model to HSPF Operation.  Prepared by AQUA
      TERRA Consultants for Computer Sciences Corporation, Annapolis, MD, and
      U.S. EPA Chesapeake Bay Program, Annapolis, MD.

Donigian, Jr., A.S., B.R. Bicknell, L.C. Linker, J. Hannawald, C.H. Chang, and
      R. Reynolds.  1990.  Watershed Model Application to Calculate Bay
      Nutrient Loadings:  Preliminary Phase I Findings and Recommendations.
      Report Prepared for U.S. EPA Chesapeake Bay Program, Annapolis, MD.

Edwards, W.M., L.B. Owens, and R.K. White.  1983.  Managing Runoff from a
      Small, Paved Beef Feedict.  J. Environ. Qual. 12(2):281-286,

Edwards, W.M., L,B. Owens, R.K. White, and N.R. Fausey.  1985.  Effects of a
      Settling Basin and Tiled Infiltration 3ed on Runoff from a Paved
      Feedlot.  In:  Agricultural Waste Utilization and Management, American
      Society of Agricultural Engineers, St. Joseph, MI.

Engman,  E.T.  1986.  .Roughness Coefficients for Routing Surface Runoff.  ASCE
      J. Irr. and Drain. Eng.  112(l):39-53.

Fitter,  A.H. and C.D. Sutton.  1975.  The Use of the Freundlich Isotherm for
      Soil Phosphate Sorption Data.  J. Soil Sci. 34:902-907.

Fox, R.H. and W.P. Piekielek.  1990.  Long-Term Nitrogen Fertilization of
      Continuous Manured and Non-manured Corn and Corn in a 3-year Alfalfa, 3-
      year Corn Rotation:  The First Six Years, 1982-87.  Agronomy Series
      Number  110. College of Agriculture, Pennsylvania State University.
                                      276

-------
Imhoff, J.C.,  B.R. Bicknell, and A.S. Donigian, Jr.  1983.  Preliminary
      Application of HSPF to the Iowa River Basin to Model Water Quality and
      the Effects of Agricultural Best Management Practices, Office of
    •  Research and Development, U.S., Environmental Protection Agency,'Contract
      No. 68-03-2895, (PB-83-250399).

Jackson, D.  1979.  Hydrologic Studies of Effects of Conowingo Dam.          . •
      Susquehanna River Basin Commission, SRBC Tech. Report No. 1, Harrisburg,
      PA.

Jaworski, N. A. and L. C. Linker.  1991.  Uncertainties in Nitrogen Mass
      Loadings in Coastal Watersheds.  Proceedings from 1990 Chesapeake Bay
      Research Conference, In press.

J chanson, R.C., J.C. Imhoff, J.L. Kittle, Jr., and A.S. Donigian, Jr.  1984.
      Hydrological Simulation Program-FORTRAN (HSPF):  User's Manual for the
      Release 8.0.  EPA•600/3-14•066.  U.S. Environmental Protection Agency,
     • Athens,  GA.

Knisel, W.  (ed).  1980.  CREAKS:  A Field Scale Model for Chemicals, Runoff,
      and Erosion from Agricultural Management Systems.  U.S. Department of
      Agriculture.  Conservation Research Report No. 26.  640 pp.

Kohler, M.A.,  T.J. Nordenaen. and W.E. Fox.  19SS.  Evaporation from pans
      and lakes.  Research Paper No. 38. Weather Bureau.

Krone, R.B.  1962.  Fluve Studies of the Transport of Sediment in Estuarial
      Shoaling Processes.  Hydraulic Engineering Laboratory and Sanitary
      Engineering Research Laboratory, University of California, Berkeley, CA.

Lauer, D.A., D.R. Bouldln. and S.D. Klausner.  1976.  Ammonia Volatilization
      from  Dairy Spread on the Soil Surface.  J. Environ. Qual. 5:134-141.

Leohr, R.C.  1984.  Pollution Control for Agriculture.  2nd Edition.  Academic
      Press, Inc.. Orlando. PL. 467 p.                                  -

Lightner, J.W.. D.B. Meiujel. and C.L. Rhykerd.  1990.  Ammonia Volatilization
      from  Nitrogen Fertiliser Surface Applied to Orchardgrass Sod. Soil Sci.
      Soc.  Am. J. 54:1471-1412.

Lugbill, J.  1990.  Potomac River Basin Nutrient Inventory.  Metropolitan
      Washington Council of Governments, Washington, DC.

Hansel1, R.S., H.M. Sell*. P. Kanachanasut, J.M. Davidson, and J.G.A. Fiskell.
      1977.  Experimental and Simulated Transport of Phosphorus Through Sandy
      Soils. Water Resour. Res. 13(1):189-194.

Maryland Power Plant Research Program.  1988.  Acid Deposition in Maryland:
      The Status of Knowledge in 1987. Annapolis, MD.

•Maryland Power Plant Research Program.  1989.  Acid Deposition in Maryland:
      Summary of Results Through 1988. Annapolis, MD.      .

                                      278

-------
Olsen, S.R. and F.E. Khasawneh.  1980..  Use and Limitations of Physical-
      Chemical Criteria for Assessing the Status of Phosphorus in Soils.
      Chapter 14.  In:  The Role of Phosphorus in Aericulture. F.E. Kwasawneh,
      E.G. Sample, and E.J. Kamprath (eds.).  American Society of Agronomy,
      Madison, WI.  pp 361-410.

Onstad, C.A., and G.R. Foster.  1975.  Erosion Modeling on a Watershed.
      Trans. Am. Soc. Agric. Eng.  18(2):288-292.

Partheniades, E.  1962.  A Study of Erosion and Deposition of Cohesive Soils
      in Salt Water.  Ph.D. Thesis, University of California, Berkeley, CA.

Penman, H.L.  1948 i  Natural Evaporation from Open Water, Bare Soil, and
      Grass.  Proc. Royal Soc. London, Ser. A 193(1032):120-145.
                    . ^                '                             '
Phillips,  R.L., and M.R. Overcash.  1976.  Economics and Technology for
      Controlling Dairy Feedlot Runoff. Paper No. 76-4032. ASAE, St. Joseph,
      MI.            .           .    ..        •'..'•'  . •'  -

Reckhow, K.M., M.N., Beaulac. and J.T. Simpson. 1980. Modeling Phosphorus .
      Loading and Lake Response Under Uncertainty: A Manual and Compilation of
      Export Coefficients. U.S. Environmental Protection Agency,
      EPA  440/5-08-011.
                                   \            ซ •                          •

Reddy, K.R.  1989.  Unpublished results. University of FL, Gainesville, FL.

Reddy, K.R., R.E. Jessup, and  P.S.C. Rao.  1988.  Nitrogen Dynamics in a
      Eutrophic Lake Sediment.  Hydrobiologia, 159:177-188/

Reddy, K.R.  and W.H. Patrick.  1984.  Nitrogen Transformations and Loss in
      Flooded Soils and Sediments.  CRC Critical Reviews in Environmental
      Control, 13(4):273-309.

Reddy, K.R., M.R. Overcash, R. Khaleel, and P.W. Westerman.  1980.  Phosphorus
      Adsorption-Desorption Characteristics of Two Siols Utilized for Disposal
      of Animal Wastes.  J. Environ. Qual. 9(l):86-92.

Reddy, K.R., R. Khaleel, M.R.  Overcash, and P.W. Westerman.  1979a.  A
      Nonpoint Source Model for Land Areas Receiving Animal Wastes:  Ammonia
      Volatilization.  Trans.  ASAE.  22(6):1398-1405..                  .

Reddy, K.R.., R. Khaleel, M.R.  Overcash, and P.W. Westerman.  1979b;  A
      Nonpoint Source Model for Land Areas Receiving Animal Wastes:  I.
      Mineralization of Organic Nitrogen.  Trans. ASAE. 22(4):863-872.
                                    /                                .  '  •
Sample,  E.G., R.J.  Soper,  and G.J. Racz.   1980.  Reactions of Phosphate
       Fertilizers in Soils.  Chapter 11.   In:  The Role of Phosphorus  in
      Agriculture.  F.E. Kwasawneh,  E.G. Sample,  and E.J. Kamprath  (eds.).
      American Society of Agronomy,  Madison, WI.  pp. 263-310.
                                      -280

-------
U.S. Army Corps of Engineers.  1956.  Snow Hydrology, Summary Report of the
      Snow Investigations.  North Pacific Division.  Portland OR.  437 pp.

U.S. Army Corps of Engineers.  1988.  Water Needs Assessment of the
      Susquehanna River Basin. Engineering report prepared for the U.S. Army,
      Baltimore, MD.

U.S. Bureau of the Census.  1980.  1978 Census of Agriculture, Volume 1,
      Geographic Area Series, U.S. Department of-Commerce.

U.S. Bureau of the Census.  1984.  1982 Census of Agriculture, Volume 1,
      Geographic Area Series, U.S. Department of Commerce.

U.S. Department of Agriculture.  1978.  Estimating U.S. Livestock and Poultry
      Manure and Nutrient Production, Report No. ESCS-12, Economics.
      Statistics, and Cooperatives Service.

U.S. Department of Agriculture.  1975.  Animal Vaste Management Field Manual,
      Soil Conservation Service, U.S. Government Printing Office.

U.S. Department of Agriculture, Soil Interpretations Records (SCS-SOI-5 data)
      file; and the National Resources Inventory (NRI), Soil Conservation
      Service.            -   .                               .         .

U.S. Department of Agriculture.  1984.  Usual Planting and Harvesting dates
      for U.S. Field Crops.  Agricultural Handbook No. 628, Statistical
      Reporting Service, U.S. Department of Agriculture, Washington, DC.

U.S. Environmental Protection Agency.  1979.  Animal Vaste Utilization oh
      Cropland and Pastureland:  A manual for Evaluating Agronomic and
      Environmental Effects. 'Report No. EPA 600/2-79-059, Office of Research
      and Development.                                             '

U.S. Environmental Protection Agency.  1982.  Chesapeake Bav Program Technical
      Studies: A Synthesis. Washington. D.C.

U.S. Environmental Protection Agency.  1983.  Chesapeake Bay: A Framework for
      Action. Region 3, Philadelphia, PA;

U..S. Forest Service.  1982.  Forest Statistics for New York, 1980, Resource
      Bulletin NE-71, U.S. Department of Agriculture, Forest Service,
      Broomall, PA.

U.S. Forest Service.  1980.  Forest Statistics for Pennsylvania, 1980,
      Resource Bulletin NE-65, U.S. Department of Agriculture, Forest Service,
      Broomall, PA.                                                  .   .

U.S. Forest Service.  1978.  Forest Resources of West Virginia, Resource
      Bulletin NE-56, U.S. -Department of Agriculture, Forest Service,
      Broomall, PA.
                                      282

-------