INTERNATIONAL JOINT COMMISSION
MENOMONEE RIVER
PILOT WATERSHED STUDY
SEMI-ANNUAL REPORT
COOPERATING AGENCIES
WISCONSIN DEPARTMENT OF
NATURAL RESOURCES
JOHN G, KONRAD
UNIVERSITY OF WISCONSIN SYSTEM
WATER RESOURCES CENTER
GORDON CHESTERS
SOUTHEASTERN WISCONSIN REGIONAL
PLANNING COMMISSION
KURT W, BAUER
Sponsored by
INTERNATIONAL JOINT COMMISSION
POLLUTION FROM LAND USE
ACTIVITIES REFERENCE GROUP
UNITED STATES ENVIRONMENTAL
PROTECTION AGENCY
OCTOBER 1975
-------
City of Milwaukee showing Menomonee River flood plain
U.S. Environmental Protection Agency
GLNPO Library Collection (PL-12J)
77 West Jackson Boulevard,
Chicago, IL 60604-3590
-------
TABLE OF CONTENTS
Page No.
SUMMARY SEMI ANNUAL REPORT 1
APPENDIX
A. River Monitoring Activities 7
B. Specific Land Use Studies 21
C. Land Data Management System 28
D. Atomspheric Monitoring 33
E. Remote Sensing 36
F. Proposal of Modeling Output Objectives 40
-------
SUMMARY
SEMI ANNUAL REPORT
Introduction
The International Joint Commission, through the Great Lakes Water
Quality Board, established the International Reference Group on Great Lakes
Pollution from Land Use Activities (PLUARG) to study and report the effects
of land uses on water quality. The "Task C" assignment requires the de-
tailed investigation of six major watersheds in Canada and the United States
which are representative of the full range of urban and rural land uses
found in the Great Lakes Basin. The Menomonee River watershed was selected
as the site for studying the effects of rapidly changing land use patterns
in an urban setting. A unique facet of the Menomonee River watershed stems
from the proposed plan to remove all municipal point sources of pollution
by 19763 at which time the effects of land use on water quality will arise
primarily from non-point sources. Thus, of the six major watersheds chosen
for intensive study, the Menomonee River watershed will serve as the focus
of investigations on the impact of urban land uses on water quality and
this data will be used to extrapolate the effects of urban land use on the
water quality of the Great Lakes.
The specific objectives of the Menomonee River Pilot Watershed Study
are:
1. To determine the levels and quantities of major and trace con-
stituents including, but not limited to nutrients, pesticides
and sediments reaching or moving in flow systems likely to reach
Lake Michigan.
2. To define the sources and evaluate the behavior of pollutants
for a metropolitan complex with particular emphasis on the impact
of residential and industrial, including utility facilities,
transportational, recreational, agricultural and constructional
activities associated with rapid urbanization.
3. To develop the predictive capability necessary to facilitate
extension of the findings from the Menomonee River Watershed
Study to other urban settings, leading to an eventual goal of
integrating pollution inputs from urban sources to the entire
-------
Great Lakes Basin.
This report will review the progress made towards achieving the specific
objectives of the study since the January, 1975 semi-annual report.
The river monitoring activities are currently achieving the goals of
the first objective. Scheduled weekly water quality sampling at twelve
automatic monitoring stations and three harbor area grab stations has
been underway since January, 1975 (refer to January, 1975 semi-annual
report for location of sites). Beginning in March, two grab sampling
stations were added on the Menomonee River in the long unsampled reach
between the confluence of the Menomonee and the Little Menomonee Rivers
and the 70th Street station. The parameter list for weekly scheduled moni-
toring includes nutrients, solids, hardness, alkalinity, color, conductiv-
ity, temperature, DO and pH. Two sets of samples were collected for
organic, trace metal and bacteriological analysis as part of the quarterly
sampling effort. Continuous in situ electronic sensor monitoring of pH,
DO, conductivity and temperature began in the spring at five automatic
monitoring stations. Discharge is monitored continuously at eleven of
the 12 automatic monitoring stations, and recording rain gauges are oper-
ating at seven of the stations and at Greenfield High School. Sediment
samples are collected by the U.S. Geological Survey (USGS) on a continuous
basis at the 12 stations. Event sampling was initiated in June at four
automatic stations (463001, 413011, 413010 and 413005) with the intention
of obtaining better yearly loading estimates and providing data essential
for the overland flow and river transport models. Data from the weekly
sampling activities were edited and stored on computer mass storage to
allow for reporting, statistical analysis and plotting.
The principal method used in achieving objective 2 is an inventory of
land use characteristics and their impact on water quality. A Land Data
Management System (Land DMS) was designed to store, retrieve, analyze and
display land data—in tabular or graphic form. The term "land data" as
used in the context of the Land DMS is a comprehensive concept in that it
denotes all those watershed characteristics that have aerial extent. Eight
land data types are presently coded for the entire watershed, one type is
in progress, and two are planned.
The Specific Land Use Study program is designed to investigate the
-------
impact of various land uses on water quality. Thirteen study areas, repre-
senting major homogeneous land uses in the watershed, have been identified
as storm event sampling sites for the Specific Land Use Studies. Detailed
surveys have been completed for seven of the sites and stations will be
operable by midwinter. Each station will consist of a control structure,
stilling well, stage recorder, sequential sampler and protective enclosure.
The parameter list will include those in the "core" list suggested by the
Tack C technical committee. The data will be incorporated into the overland
flow model.
An additional source of pollutants to the watershed arises from atmos-
pheric fallout. The nature and quantity of the pollutants is determined
by the types of land use in and around the watershed and the climatological
characteristics of the area. An air and rain water monitoring program is
presently being established in the watershed in support of objective 2. The
primary objective of the atmospheric monitoring project is to determine the
relative contribution of different land uses to the total weight and chemi-
cal composition of the aerosols measured over the watershed. The atmospheric
study will also attempt to quantify the mass of aerosols entering the water-
shed from rain and dry fallout deposition. Presently, seven high-volume
air samplers are being calibrated and will be mounted permanently on exis-
ting monitoring stations along with modified (Wong) rain samplers. Two
cascade impactors will be used periodically to size-fractionate aerosols.
The target date for initial sampling is January 5, 1976.
Also in support of the second objective is a recently initiated
remote sensing project. The intent of this project is to develop tech-
niques for mapping land use and hydrologically-active source areas in a
watershed from small scale color-IR imagery. The technique involves
measuring the film density on the image for the area of interest and with
the aid of a computer, interpreting and mapping the area. It is anticipated
that the automatic processing of digitized photography will accelerate the
land cover interpretation process. Statistical summaries of the aerial
extent of each land cover category can be obtained within any given area.
This data should be useful input to the hydrological modeling effort.
Density measurements on the film and ground checking of the results are
being accomplished at the University of Wisconsin-Madison, while the
-------
computer interpretation and mapping are being undertaken at The Pennsylvania
State University.
Objective 3 is being supported by an effort to develop land use water
quality models. Modeling techniques and mathematical models applicable
to the land use water quality models have been reviewed and evaluated for
possible use. A computer model—"STORM"—has been acquired to model over-
land flow and quality of runoff water. Preliminary testing of the model has
been performed on three small subwatersheds and subroutines of the model are
being developed. In addition, conceptual models of soil adsorption applic-
able to agricultural chemicals, nutrients, pesticides and other pollutants
transported in the soil are being prepared and will be added to the STORM
model. Ouput from the overland flow and quality model will be used in the
river transport model. The river transport model will account for in-stream
and benthic processes over a range of flows (i.e., a dynamic model) and
deliver the derived loadings to an estuary model. It is anticipated that
DO/BOD, sediment and nutrient/algae interactions and quantifications will
be derived from extensively modifying existing models.
A model is being developed to relate surface runoff quality to the
ultimate nonpoint sources, namely, atmospheric fallout, precipitation, land
and management practices and natural geological weathering. However, this
model requires extensive data collection and for this reason will not be
of general applicability. Thus, an attempt is being made to derive empirical
relationships describing the quality of surface runoff to land use and meteor-
ological parameters.
In their cooperative effort with the IJC Project, including coordination
with their programs supported by Section 208 funds under Public Law 92-500,
the Southeastern Wisconsin Regional Planning Commission (SEWRPC) is develop-
ing several watershed models. It is likely that an amended form of the
Hydrocomp Simulation Watershed Model—which contains both overland flow and
river transport components—will be capable of satisfying the requirements
of objective 3 of the Menomonee River Pilot Watershed Study.
In summary, the Menomonee River Pilot Watershed Study was designed to
investigate the impact of urban and urbanizing land use on water quality
and to extrapolate these effects to the entire Great Lakes Basin. The over-
all goals of the Menomonee River Pilot Watershed Study were divided into the
-------
the three aforementioned specific objectives. Since January, 1975 the
river monitoring activities in support of objective 1 have been fully
implemented and will continue to be conducted at the present level for
the duration of the project. Most of the projects designed to accomplish
the second and third objectives have been initiated recently and will
progress significantly towards meeting these objectives within the next
6 months. Eight land types already have been coded into the Land Data
Management System and progress has been made in identifying special study
sites and defining better the modeling activities.
-------
APPENDIX A
-------
RIVER MONITORING ACTIVITIES
Introduction
The objective of the river monitoring activities is to determine the
types and quantities of various water quality parameters reaching or moving
in the Menomonee River and its principal tributaries. The parameter list
includes the "corer list determined by the Task C Technical Committee of
the Pollution from Land Use Activities Reference Group (PLUARG), plus others
likely to affect the quality of Lake Michigan water. The river monitoring
activities will provide basic information about the hydrology-hydraulics
and water quality of the watershed. Interpretation and assessment of moni-
toring data will be handled by the development of land use-water quality
models.
Preliminary monitoring of the river began in February., 1973 with the
establishment of three stations by the Wisconsin Department of Natural
Resources (WDNR). Intensive sampling of the river and its major tribu-
taries began in January, 1975 after the site selection and construction
of 12 automatic monitoring stations.
Progress
Sampling frequency for baseflow water quality at the river monitoring
stations was seasonally adjusted to accomodate times of rapid changes in
discharge such as in the spring. Bi-weekly water quality samples were
obtained from the 12 automatic stations from January until the end of
March 1975, and from the end of March until the end of June samples were
obtained twice a week. Since the end of June the stations have been
sampled once a week. Beginninp in March two grab samples stations on the
Menomonee River, 413061 and 413062, were included in the regular river
sampling schedule. These two stations provide water quality data in the
long reach between stations 683001 and 413005 on the Menomonee River (refer
to pages 17 and 18 in January 1975 semi-annual report for location of
monitoring stations). Starting in March the scheduled weekly samples at
-------
the automatic monitoring stations were collected when the gauge height
permitted. A quality control program was initiated in April to compare
grab samples with samples obtained from the automatic samplers. Grab
sampling stations 413004, 413013 and 413012 in the Menomonee River harbor
area were sampled bi-weekly between January and the end of March, and once
a week from the end of March until the end of May when bi-weekly sampling
was resumed. The parameter list for the scheduled weekly sampling at all
stations included the basic nutrient scan, solids, total alkalinity9 hardness,
color, chloride, conductivity, pH, DO and temperature. Total organic carbon
was added to the list on July 28. Temperature, D03 pH and conductivity have
also been measured continuously at five river stations (413004, 413005,
683001, 413008 and 673001) equipped with an electronic sensor-type monitor.
Data from the weekly sampling activities have been edited and stored on
computer mass storage to allow for reporting and statistical analysis. The
data may be accessed through batch processing or terminal input and output.
To further assist in the interpretation of the water quality data, the stored
information is presently being programmed for mechanical plotting in terms
of concentration and loading and will be included in the April 15 report.
The above calculations will include the parameters from quarterly sampling.
Two quarterly samplings for metals, organics and bacteria at the 12
automatic stations were accomplished this year. Samples were obtained under
baseflow conditions in June and September. The total coliform, fecal coli-
form and fecal streptococci counts almost always exceeded accepted standards
by a large margin. Most of the samples collected on June llth and analyzed
for organics showed concentrations below detection limits. In a few instan-
ces, aldrin, lindane, heptachlor epoxide and methoxychlor demonstrated a
significant concentration above detection limits. Most phenol levels for
samples taken June 24 and September 3 were above detection limits. Phenol
concentrations varied considerably between two different sampling dates at
the same station. Many of the metals sampled on June 25 had concentrations
below detection limits at all the stations. Levels of copper, lead, zinc
and chromium appeared to be higher at stations below station 413007 (Appendix
A Table 1). Previous investigators have shown a significant increase in
metal loadings during runoff events. Therefore, an attempt is being made to
-------
Q) P
P O
O H
W -H
CU Pi
H
ft 0)
B
•^s
*- o
OO
• o
o
o
o
.
— CMootr —
I/ CM \/ w o oo irv
' P-
V v v/ v IA in
• «•
CM CM CM
— TTl/^V OOOCM
V V O CM
" »j^
CM CM
CM CM
CM CM
CM CM
rn
CM CM
_ "*>
•-.- 3
=Jo>
C C
en
c
CM rr\
\o e?
-------
10
collect quarterly metal samples during runoff events. A plan to sample
for organics during an event is also being investigated. A more accurate
estimate of monthly loadings of quarterly parameters will be made by
combining base flow and runoff event concentrations with the continuous
discharge data.
The stage heights that have been monitored continuously since November,
1974 at 10 of the 12 automatic monitoring stations (station 413009 was
operational in May) will be available from the U.S. Geological Survey
(USGS) in completed form later this year.
Starting in December, 1975 the USGS will provide primary discharge
data every 6 weeks with three seasonal updates as a check on the accuracy
of the primary data. The primary summer flow data is expected to be
accurate to within 10% of the corrected flow data, while primary winter
data is expected to be less accurate. Corrected flow data will be received
in discharge units on magnetic tape, which will be programmed to be compat-
ible with the stored water quality data. The above procedures will provide
more immediate estimates of loadings and interpretation of river flow con-
ditions. Due to the delay in obtaining corrected flow data the USGS will
not be able to provide sediment loading data until later this year. The
USGS has been monitoring continuously sediment concentration using the
automatic samplers .at the 12 automatic monitoring stations since March 1975.
In the future,, sediment concentrations and loading values will be provided
monthly on data two months old. During September, 1975 the USGS completed
the installation of rain gauges at seven of the automatic monitoring
stations (673001, 683002, 683001, 463001, 413011, 413007 and 413005) and
an eighth gauge at Greenfield High School in Greenfield. The availability
of rain data will be valuable for present efforts in modeling overland
flow during runoff events.
As an important part of the overland flow and river transport modeling
effort, a program was initiated in June 1975 to monitor numerous runoff
events at river stations 463001, 413011, 413010 and 413005 for the same
water quality parameters determined in scheduled weekly samplings. The
first 3 stations were chosen based on the homogeneous nature of the land
use in their drainage area. Station 413005, which was not sampled until
-------
11
the third event, provided water quality data for the Menomonee River at
a downstream location. Four runoff events (June 17, July 19, August 18
and August 20) have been sampled so far with six or more samples collected
at any one station during an event. Due to equipment failure and the uneven
distribution of rainfall throughout the Menomonee River watershed, only one
of the stations (413011) was sampled during the first 2 events (June 17 and
July 18) and two of the 4 stations (413010 and 413005) were sampled in the
last two events (August 18 and August 20). A significant increase in the
concentration of most of the parameters was observed during the runoff
events at stations 413011 and 413010 :, the average dissolved reactive phos-
phorus (DRP) and suspended solids levels at station 413011 during the month
of June were 0.10 mg/1 and 25 mg/1, respectively,, while the average concen-
tration during the runoff event were 0.20 mg/1 and 6,800 mg/1, respectively
(Appendix A Fig. 1). The increasing concentration with increasing discharge
rates during a runoff event suggest the watershed around station 413011 was
a possible nonpoint source of such parameters as dissolved nutrients and
particulate matter. Hardness, alkalinity and chloride concentrations de-
creased with increasing discharge rates at stations 413010 and 413011 for
all events, suggesting these parameters were contributed at a relatively
constant rate by a source such as groundwater and were present in relatively
low quantities in the watershed. Chloride levels are expected to increase
with increasing discharge rates if an event occurs during or soon after
road salting season.
Trends in parameter concentration at station 413005 during runoff
events differed from those observed at stations 413010 and 413011. Con-
centrations of most of the dissolved nutrient species for the August 18
event tended to increase only slightly with increasing discharge rates,
while the dissolved nutrient species levels for the August 20 event de-
creased with increasing discharge rates. The parameters measuring partic-
ulate species increased with increasing discharge water for both events.
The DRP levels decreased from 0.15 mg/1 to a low of about 0.035 mg/1 and
the suspended solids increased from a monthly average of about 18 mg/1
to a peak concentration of 880 mg/1 (Appendix A Fig. 2) for the August 20
event. The same decrease in hardness,, alkalinity and chloride with increas-
ing discharge rate3 observed at stations 413011 and 413010, also occurred
-------
33S/1 NI 39dVH3SId
12
co
c
>H
CO
H
0)
0)
•H
H
O
co
0) 4->
T3 C
C CU
0) >
ft
0) f~
> CD
•H H
4->
O «
ro r-
0) H
0)
T7 C
0) 3
O CD
CO ,C
CO -P
•H
T3 W)
C
r, -H
T)
NI S13A31
bO
^
rd
O CD
W 0)
•H ^
O O
bO
•H
<
X
swi sanos
c
0)
-------
13
a
oo
Q)
0)
H
W •
-d -p
•H G
H 0)
O >
M 0)
T3 CD
C H
m
rt
OH O
CN
0)
> -P
•H W
S §>
(0 3
CD <
k
bO
S ^ C
r~ —• rt\ . I
0) -H
O T)
W
W -P
•H 0)
T3 0)
&
" -P
0) CO
hO
ro -P
J2 O
o r-
w
•H p
CN
bO
X
•H
"d
C
a;
NI
dya
NI S13AJ1 sanos
-------
14
at station 4-13005 on August 18 and 20. The August 20 data at station
413005 suggest that the Menomonee watershed contained low quantities of
dissolved nutrients relative to the volume of runoff and river discharge.
The result was dilution of the dissolved nutrients in the river. However,
a discussion of concentration changes, large or small, in the various
parameters during a runoff event at any of the stations may lead to an
incorrect judgment on the full impact of a runoff event until the concen-
tration data is combined with the discharge data to obtain loading values.
The DRP loadings for the days of the scheduled weekly sampling station
413005 during the month of August were found to be about nine times smaller
(10 Kg/day vs 90 Kg/day) than the loadings for DRP observed during the
August 20 runoff event. The August 25 scheduled weekly sampling had a
higher DRP loading because of a storm event that had occurred prior to
sampling. The DRP loading for station 413005 was found to be about 22 times
(90Kg/day vs 4 Kg/day) greater than that for station 413010 during the same
event (Appendix A Fig. 3). Similar results were found for suspended solids
loading on August 20 at station 413005, where the loading was about 28 to
973 times greater (1 to 35 Kg/day vs 973 Kg/day) than the loading observed
for the scheduled weekly sampling (Appendix A Fig. 4).
Suspended solids loading at station 413005 was 80 times greater
973 Kg/day vs 12.5 Kg/day) than at station 413010 (Appendix A Fig. 4).
Station 413011 was monitored twice during runoff events which allows a
comparison of event loading and loading for weekly scheduled sampling,
plus loading difference between two separate events. The DRP loadings
during the June 1 and July 18 events at station 413011 were 5,400 (540 Kg/
day vs 0.10 Kg/day) and 50 (1 Kg/day vs 0.02 Kg/day) times greater than
the average loading for weekly scheduled sampling, respectively. Also the
average DRP loading for the weekly scheduled sampling was higher (0.1 Kg/
day vs 0.02 Kg/day) in June than in July (Appendix A Fig. 5). This was a
result of both a decrease in DRP levels and discharge in July. The DRP
loading for the June 11 event was 540 times greater (540 Kg/day vs 1 Kg/day)
than the July 18 event. This difference is directly correlated to the
higher average DRP levels in the June 17 event of about 0.20 mg/1 compared
to 0.05 mg/1 for the July 18 event and a water loading of 2.5 x 107 liters vs
-------
15
114
112
>•
g
| 96
£
i»—t
1 80
a.
£
i ^
§3 48
»— -*
32
16
-
•
j
RUNOF EVERT ^j
h AT 70TH ST, \
I i
-
1
s
^
1 1 \\f
I • i l 1 • i i i • %H i i
xSCHOOfWKER
CREBC
}2 16 20 24
AUGUST
28
Appendix A Fig. 3. Loadings for 70th Street and Schoonmaker
Creek during the August 20, 1975, runoff
event compared to loading for samples
obtained weekly at 70th Street.
-------
16
y/5
X>
s
140
CL x-
o ^
t-H
x 35
,>-
a
S
z 30
(•"4
2=
| 25
OO
a
1—4
£ 20
B
a
£ 15
00 J"'
00
10
5
0
s
I
s
i
s
fc
1
s
1^
RUNOFF EVENT"^ I
^/ AT 70TH ST, ^ \
k 1
"
-
•
-
•
i • j i i B i i i
^ 1
k —.
\
\
\
\
\
\
\
\
\
\
@/^
1
J
2I.J L.
48 12 16 20 24
AUGUST
'RUNOFF EVENT AT
SCHOONMAKER CREEK
Appendix A Fig. 4.
28
Suspended solids loading during August 20, 1975,
runoff event at Schoonmaker Creek and 70th Street
station and for weekly sampling at 70th Street.
-------
17
oo
AW/9* MI
QC.
CM
& 3 9 8
d 3AID\/3H 03ATOM
c
0)
o
I
M
C
•H
0)
&
O
co
d)
o •
^ w
ho
w c
bO-H
C H
5 §•
(0 ft)
o w
O, H
X
(1) 0)
•H &
T3 CO
0) T)
>
H C
O O
CO
CO T)
•H G
O rd
bO
•H
<
X
•l-i
c
a,
-------
18
3 x 106 liters,, respectively. Differences in water loading were partly
related to rainfall intensities, which were 0.5 inches in one hour for
July 18 and 0.70 inches in one and half hours for June 179 while the DPP
concentration differences could be partly influenced by seasonal changes
in the watershed. All the above loading data indicate the important con-
tribution of runoff events to total pollution from a watershed irrespective
of the parameter concentrations during a runoff event. A better under-
standing of the factors affecting runoff loadings, such as seasonal changes
and rain intensity, will be obtained by future monitoring of runoff events.
Further event monitoring might also allow the development of sufficient
pollutographs to predict storm loading based mainly on hydrographs.
In Situ Continuous Monitoring Equipment
Five of the river network sampling stations are equipped with electronic
sensor-type monitors for in situ continuous monitoring of dissolved oxygen,
pH3 conductivity and temperature. These sites are 1. Henomonee River above
27th Street at Falk Corporation (413004), 2. Menomonee River at 70th Street
Bridge (413005), 3. Menomonee River at 124th Street (Hwy. M) (683001), 4.
Little Menomonee River at Appleton Avenue (Hwy. 175) (413008) and 5.
Menomonee River at River Lane Road (Hwy. F) (673001).
Equipment partially consists of a subsurface sensor probe unit housed
in a protective semi-open sonde. The sonde is supported in a subsurface
location at mid-stream by attachment to one or more steel posts driven into
the stream bed. Attachment of the sonde to the post(s) using a ratchet
tightening nylon web belt allows for easy removal of the sonde unit for
cleaning and calibration. The sonde unit is connected by water tight
neoprene connectors and a urethane cable to monitoring and recording equip-
ment located inside the site building.
The above water equipment consists of a surface unit which allows
instantaneous reading of all parameters, a data scanner which separates
data .impulses and a strip chart recorder which records separate parameter
data on an alternate time basis. (Data collection via tape deck units in
place of the scanner and strip chart is available.) Strip chart data is
collected monthly and digitized for computer input.
-------
19
The entire continuous monitoring system is battery-powered by a 12 volt
direct current system and internal mercury batteries. This set-up provides
a reliable power supply independent of electrical line power. Electricity
is used, however, to provide heat and light to the sites and to charge
periodically the batteries.
Calibration frequency of the system is dependent on rate of probe
fouling, which in turn is dependent on general water quality. A weekly or
twice-a-week cleaning and calibration schedule is usually sufficient to
ensure accurate data.
-------
APPENDIX B
-------
21
SPECIFIC LAND USE STUDIES
Introduction
The nature of river sampling and the heterogeneity of land use in the
Menomonee River watershed precludes the use of most of the river sampling
stations as specific land use stations. Defining the character of storm-
water from homogeneous and/or predominant land use areas in a rapidly
urbanizing watershed requires the establishment of stations at the outlets
of these areas. In selecting sites9 an attempt must be made to select
areas that are representative of major land uses in the watershed. Data
generated by the stations will give us an insight into the types and amounts
of pollutants associated with land drainage. In addition, the data will be
used to support an overland flow modeling effort.
Study Sites
Presently, thirteen study areas have been identified and are listed
in Appendix B Table 1. Over the past few months., seven of the areas have
been emphasized and are listed in Appendix B Table 2 along with some
pertinent information. Two additional areas, Schoonmaker Creek and Noyes
Creek3 which are part of the river network sampling system, will be used.
Detailed surveys have been completed for the seven sites and stations will
be operable by mid-winter. We had originally expected a starting date in
late fall but have been frustrated by delays in obtaining permission to
access areas. The rationale for concentrating on seven sites rested with
the inability of the laboratory to handle a larger station load.
Equipment
Each station will consist of a control structurej stilling well,
stage recorder, sequential sampler and protective enclosure. See Appendix
B Table 3 for more details.
During rain events stage will be recorded and the sequential sampler
-------
22
Appendix B Table 1. Sites selected for the special land use studies
Land Use
Location
Airport
Railroad
Freeway, right-of-way
Freeway, interchange
Retail and services
Manufacturing3 heavy
High density
Low density
Developing
Golf course
Park
Landfill
Swamp
Transportat ion
Timmerman Airport at W. Appleton Ave, "'. :
Milwaukee
Milwaukee Road switching and maintenance
yard at W. Canal, Milwaukee
Segment of Stadium Freeway (USH 41) between
W. State St. and W. Washington Blvd.,
Milwaukee
Stadium Interchange (I-94--USH 41) near
Milwaukee County Stadium, Milwaukee
Commercial and Industrial
Brookfield Square Shopping Center at N.
Moorland, Brookfield
Allis-Chalmers at 5.70th St., West Allis
Residential
Schoonmaker Creek at W. Milwaukee Ave.,
Wauwatosa
Elm Grove at Wrayburn Rd., Elm Grove
Noyes Creek at 91st St., Milwaukee
Recreational
Currie Park Golf Course at N. Mayfair,
Wauwatosa
Greenfield Park at S. 124th, West Allis
Land Disposal
Germantown Sanitary Landfill at County Line
Rd.y Menomonee Falls
Wetland
Germantown swamp
-------
23
Appendix B Table 2. Some characteristics of selected areas and status
of obtaining permits to install sampling equipment
and control structures
Land Use
Airport
Freeway ,
interchange
Commercial,
shopping
center
Industrial,
heavy
Residential,
high
density
Residential ,
low
density
Residential,
developing
Landfill
Estimated
Location area (ha^) Lnipurviousttess (%)
Timmerman Airport at W. 131 15
Appleton, Milwaukee
Stadium Interchange 79 51
(I-91--USH
-------
Appendix B Fig. 1.
Brookfield Square Shopping Center -
looking south, drainage to the north.
Appendix B Fig. 2. Stadium Interchange - looking south,
drainage to the east.
-------
25
Appendix B Table 3. Station Equipment
Control Structure:
Stilling Well:
Stage Recorder:
Sequential Sampler:
Protective Enclosure:
H-flume for open channel
conditions - Palmer-Bowlus
for stormsewer pipes
Corrugated drain pipe
Leupold and Steven digital
tape stage recorder with
event marker
ISCO 1680 sampler with
extended timer for drawing
larger sample volumes
Precast cement sewer pipes
approximately 3f x 3 "
with a locked steel top.
Pipes can be added to
increase the height.
-------
26
will be actuated by a liquid level switch. This switch starts and stops
the sampler when the stream reaches a predetermined level. Stage record-
ing and sampling times will be determined by the hydrology of the area.
Sampling will not be proportional to flow due to the large variations in
flows and heads that will be experienced at the sampling sites.
Laboratory analysis will be handled by the Wisconsin State Laboratory
of Hygiene for Groups A, B and D parameters and the USGS for Groups E and
F. For Group C inorganic and organic3 a special laboratory will be estab-
lished at the University of Wisconsin-Madison. Negotiations are presently
underway for laboratory space and we expect to have an analytical chemist
on the staff by November 1, 1975.
-------
APPENDIX C
-------
28
LAND DATA MANAGEMENT SYSTEM
Introduction
The Land Data Management System (Land DMS) is a digital computer-
based system designed to store, retrieve, analyze and display—in tabular
or graphic form—land data. The term "land data" as used in the context
of the Land DMS is a comprehensive concept in that it denotes all those
watershed characteristics that have an areal extent. For example, land
data encompasses land use, soil type and civil division information but
does not include water quality or streamflow data.
Uses of the Land DMS
The Land DMS has two principal uses under the Menomonee River Pilot
Watershed Study:
1. Interpretation of water quality and quantity data acquired from
routine long-term monitoring activities as well as that obtained
from short term special land use studies.
2. Input to hydrologic-hydraulic-water quality models.
In addition to meeting the above two needs of the Menomonee River
Pilot Watershed Study, the Land DMS was designed so as to be consistent
with the recommendations of the Ad Hoc Data Handling and Processing Work
Group of the Task C Technical Committee.
Data Storage Unit: The Cell
The basic areal unit for storing,, retrieving, analyzing and displaying
land data is the cell. The Menomonee River watershed was subdivided into
about 35,000 cells by partitioning U.S. Public Land Survey system quarter
sections. Each of the four sides of each quarter section was equally
divided into eight parts and the grid marks on opposite sides of the
quarter section were connected resulting in 61 cells per quarter section
each having a nominal area of 1.0 hectare (2.5 acres). The use of cells
-------
29
that are partitions of quarter sections has one principal advantage: It
facilitates the geo-referencing of each cell since horizontal survey
control has been established for a large number of quarter-section corners
in the watershed using field survey methods and control is directly
transferrable., by computation, to the centers and corners of each cell.
Geo-referencing
An accurate geo-referencing procedure is required to permit computa-
tion of the area of each cell or of groups of cells and to facilitate
display in map form,, of selected land data. The corners of each cell were
accordingly referenced to the State Plane Coordinate System. Plane geom-
etry was then used to calculate., within the Land DMS, the State Plane
Coordinates of each corner and of the center of each cell. The Land DMS
can also readily convert the cell corner coordinates to latitude and longi-
tude and software is being developed to convert the coordinates to the
Universal Transverse Mercator System.
The Supporting Computer System
The digital computer system--hardware and software—needed to support
the Land DMS can be broken into four phases: the input phase ? the data
management phase, the data base phase, and the output phase. Under the
input phase, data are entered into the Land DMS on either magnetic tape,
magnetic diskettes or punched cards. The second or data management phase
is composed of a set of computer programs that perform contingency checks
on the incoming data, provide for the maintenance and updating of the
Land DMS, analyze the data and transfer it—on request—back to the user.
The analysis capability of the data management phase utilizes an overlay
technique to identify cells having specified combinations of land data
types. The third or data phase of the Land DMS is the actual storage of
the areal characteristics of each cell in a computer file, maintained on
magnetic tape or on a magnetic disc. The fourth or output phase, provides
transfer of land data from the Land DMS to the user in a variety of tabu-
lar and graphic formats. For example, land data can be ouput on a cell
-------
30
basis or aggregated by subwatershed or some other geographic area of
interest. System output can be obtained on several media including mag-
netic tape, punched cards, on-line printer, and plotter.
Land Data Contained in the Land DMS
Appendix C Table 1 summarizes the status of land data within the Land
DMS. Eight data types have been coded for the entire watershed, one type
is in progress and two types are planned. Other land data types will be
added in response to the needs of the Menomonee River Pilot Watershed
Study.
-------
31
Appendix C Table 1. Land data in the land data management system
Type of Coding
Status Dominant Percent
Data Type Completed In Progress Planned Characteristic of Cell Other
1. Civil Division x x
2. Sub-basins and x x
Subwatersheds
3. Wildlife Habitat x x
(with value ratings)
4. Woodland-Wetlands x x
(with value ratings)
5. Park and Outdoor x x
Recreation Sites
6. Floodlands x x
7. Perennial Streams x x
8. Conservancy, Flood- x
land and Related
Zoning
9. Soils x x
10. Ground Elevation x x
11. Land Use x x
-------
APPENDIX D
-------
33
ATMOSPHERIC MONITORING
Objectives
As part of the overall Menomonee River watershed study, an air
sampling network will be set up within the boundaries of the watershed.
Objectives of the research are listed below:
1. The primary objective is to determine the relative contributions
of different land uses to the total weight and chemical composition
of the aerosols measured over the watershed.
a. To characterize "background" aerosols, i.e., the particulate
matter not injected from sources within the watershed.
b. Identify sources of aerosols, i.e., land uses or point sources.
2. To attempt to quantify the mass of aerosols entering the watershed
from rain and dry deposition.
a. To correlate frequency and duration of rainfall with total and
elemental concentration of aerosols.
3. To determine size distribution of total particulate and individual
elements as a function of relative humidity.
4. To study collection efficiencies of water surfaces, snow surfaces,
leaves, etc.
Sampling and Analysis
Seven high-volume air samplers along with modified (Wong) rain samplers
will be set up on river monitoring stations 673001, 683002, 683001, 463001,
413011, 1+13007 and 413005. An additional rain sampler will be built at
Greenfield High School, Greenfield. Meteorological data will be obtained
from stations in the area operated by the Air Pollution Control Section of
the Wisconsin Department of Natural Resources. Data from on-site rain
gauges will also be used.
In addition to total suspended particulates, two cascade impactors will
be used periodically to size-fractionate the collected aerosols. Analysis
of about 25 elements will be done by neutron activation, flameless atomic
absorption, and flame atomic absorption on all air and rain samples.
-------
Rain samples will be freeze-dried prior to neutron activation analysis
(N.A.A.). The Whatman Ul cellulose filter paper from the high-volume air
samplers will be cut to size and along with the nucleopore filter paper
from the impactors will be directly analyzed by N.A.A. Prior to flameless
atomic absorption analysis, the remaining cellulose filter will be dissolved
in acid. In the case of the nucleopore filter only one of the analyses can
be performed. If flameless atomic absorption is chosen for the nucleopore
filter the filter is dissolved in chloroform.
Progress to Date
Presently the seven air samplers are being calibrated in Madison. The
next step will be to make test runs to compare collection efficiencies
between them, due to motor differences and filter paper variations. Some
equipment on order has still not arrived (Wong rain samplers, flameless
A.A. system, and silicon controlled rectifiers for voltage reduction on
the*, high volumes). The target date for initial sampling in the watershed
is January 5, 1976.
-------
-------
APPENDIX E
-------
36
REMOTE SENSING
The intent of this project is to develop techniques for mapping land
use and hydrologically active source areas in a watershed from small-scale
color-IR imagery. The technique involves measuring the film density on the
image for the area of interest and with the aid of a computer, interpret
and map the area. The tasks of density measurements on the film and ground
checking of the results are being accomplished at the University of
Wisconsin-Madison, while the computer interpretation and mapping are being
done at The Pennsylvania State University.
The imagery used by this project will be high altitude color-IR
imagery flown by the National Aeronautics and Space Administration (NASA)
(scale 1:120,000) and low altitude 70 mm imagery acquired with the aid of
the Wisconsin Department of Natural Resources (WDNR) DC-3 (scale 1:72,000).
Densities on selected portions of the imagery will be measured with an
Optronics P-1000 scanning microdensitometer. This instrument can measure
the density every 25, 50 or 100 microns in the areas of interest on the
film imagery. The imagery will be scanned through three different colored
intereference filters in order to perform a color separation of the photo-
graphic imagery. The best choice of filters to perform this color separa-
tion is under investigation. The output from the scanning densitometer
is a nine-track computer tape on which is written the film densities at
each point in the area of interest. The film densities on the computer
tape must be transformed before the data is used in the computer inter-
pretation programs. This transformation depends on the properties of the
microdensitometer,properties of the three dyes in the film, and radiometric
calibration of the roll of imagery of interest. Once this transformation
has been determined, a computer tape of the transformed film densities will
be generated and sent to The Pennsylvania State University for processing
through the data processing system of the Office for Remote Sensing of
Earth Resources.
This system is couched in a multivariate framework. Each observation,
identifiable by scan line and element number consists of a vector composed
of three response values (one each for the response in the blue, green and
-------
37
red portions of the electromagnetic spectrum). Spectral signatures will
be determined for selected training areas. A variety of classification
programs will then be used to classify the data into discrete land cover
classes.
It will be possible at this time to determine if there are any problems
with data registry. The photography has to be scanned three successive
times and it is essential that each scan exactly overlays the previous scan.
Computer techniques will probably have to be developed to align and regis-
ter the data.
The output from the computer processing will result in the form of
digital maps showing the spatial distribution of various land cover types.
Most of these maps will be in black and white grey scale. Combinations of
grey scaling and four colored ribbons may also be used to produce a colored
output map.
Several other output modes will also be used. One of these modes
would be to output the land cover types using the CalComp plotter. Another
method will be to display the results on a color monitor and to photograph
this color display.
The various types of output will be compared to the land use mapping
that was done by the Southeastern Wisconsin Regional Planning Commission.
It is anticipated that this automatic processing of digitized photog-
raphy should greatly speed up the land cover interpretation process.
Statistical summaries of areal extent of each land cover category can be
obtained within any given area. This data should be useful input to the
hydrological modeling efforts.
-------
APPENDIX F
-------
39
PROPOSAL OF MODELING OUTPUT OBJECTIVES
For greatest effectiveness, water-quality modeling must respond to the
overall project objectives within the confines of relevant model outputs.
Many previous water-quality models have focused on dissolved oxygen as a
sole parameter. Perhaps the first significant precedent for a DO model
occurred with the 1966-70 Delaware Estuary Water Quality Modeling Project.
Although this model included five physical, six mineral, four nitrogen,
three oxygen, three bacteriological, five metal and four miscellaneous
parameters, predictive use of the time-varying model was focused almost
entirely on DO, and on salinity which was considered to be a conservative
substance. The continuing narrowing of focus doubtless arose from such
statements as:
"The system developed during the research being reported here is
based on the assumption (underlinings added) that there exists a
single measurable attribute that is indicative of the river water
quality. It is also assumed that there is a minimum (or maximum)
value of this attribute beyond which it is undesirable to go. Further,
it is assumed that there is a causative relationship between the
amount of materials being introduced into the river and the downstream
water quality ..." "... therefore, the river presents a reaction
system having BOD as its major input and DO as a measure of output.
(A River as a Chemical Reactor. Penobscot Tidal Estuary Study, E. G.
Bobalek, University of Maine, 1971.)
Taken far beyond its limited context of presentation, BOD/DO fixation—
as an end rather than a means—has largely turned concern away from
eutrophication-inducing nutrients, bacterial and viral hazards, toxic
metals, organic toxicants and sediment as threats to environmental quality
and the general public health. Certainly, there is increasing need to re-
focus on basic objectives of the environmental clean-up effort (see
especially The Uncertain Search for Environmental Quality, Ackerman et al.,
The Free Press, New York, N.Y.
Lake Michigan has a long retention time and can be described as a
"sink" situation in the classical sense. Thus, DO content delivered by the
Menomonee River is of small consequence in comparison to the off-shore
-------
processes within the lake. Nutrient content, BOD9 fecal and other coliforms,
metals, organic compounds and sediments should be of primary concern. None-
theless, attention to the modeling of DO/BOD constitutes the first step
toward the modeling of the other parameters and should serve to check flow
rates and as an important guide in determining the reactions of non-
conservative substances. The determination will involve extensive litera-
ture search and application of University of Wisconsin-Madison technical
expertise developed in the International Biological Program supported simula-
tion studies on Lake Wingra. Sediment quantity and size distribution is an
important pollutional input not only in its own right, but because it acts
largely as a reservoir of potential toxicants. Nutrient inputs—not only
nitrogen and phosphorus, but carbon, silica and some trace elements—are
important in determining the extent of algae and macrophyte activity and
the degree of eutrophication, which doubtless influence the reactions of
non-conservative substances. A brief catalogue of parameters for which
modeling is important (the order of listing denotes suggested order of
priority) is presented (Appendix F Lists 1 and 2).
The Development of Data Input via STSAM
The development of a Statistical and Time-Series Analysis Model is of
first importance to the application of the collective expertise of project
personnel.
Efforts are impending to analyze long-term precipitation and runoff
records for the project vicinity using power-spectra analysis. Median
values (and other appropriate confidence levels) will be plotted within
the identified cycle periods. Scrutiny of this information allows defin-
ition of hydrologic seasons within the study area; model input and output
could then be expressed over periods of truest physical significance.
Looking toward good model output, we must examine a basic assumption ,
namely, is land use an effective descriptor of vegetation and surface
imperviousness, occurrence probability of particular parameters, and
population/urbanization characteristics, such that runoff water-quality
conforms to particular land uses. While land use does provide a good
-------
Appendix F List 1. High-priority parameters
Polychlorinated Biphenyls (PBCs)
'"'Radium 226/Alpha Radioactivity
Lead
ABarium
Surface Contaminants
Asbestos
Silica
Phthalates
DDT/DDD
Oil and Grease
Streptococci Coliform
Copper and Zinc
Manganese and Iron
Aluminum
*Suggested additions to Task C "core"
parameter list
-------
Appendix F List 2. Moderate-priority parameters
Arsenic
Selenium
^Titanium
*Zirconium
Nickel
Chromium
Vanadium
Chloride
Cadmium
Mercury
Phenolics (and Chlorophenols)
Heptachlor/Heptachlor Epoxide
v':Polynuclear Aromatic Hydrocarbons
'''Strontium
Molybdenum/Beryllium
'"'Suggested additions to Task C "core" parameter
list
-------
43
beginning for categorization, the need to minimize statistical deviations
will probably dictate that, to the extent feasible , the classification
should be expanded in three particular areas,, namely, agricultural, resi-
dential, and manufacturing areas. -,
Specific information on farm-type and class of crop within the ;'cellv'
will be determined to better describe vegetation and water quality param-
eters. Similarly, zoning classifications from planning sources might
quantify better the extents of surface imperviousness and population/
urbanization characteristics and also capture some -essence ' of the
consequent dependence of land use on vegetation and fallout-derived param-
eters. Hydrologically, it is probably acceptable to categcrically-unify
single and two-family homes. However, it would also be important to divide
the aggregate into perhaps four classifications by plot acreage, as well as
to distinguish between low and high-density multi-family units (e.g.,
garden-apartments vs. high-rises). Furthermore, either planning agency
information or an industrial agency's input would allow manufacturing
breakdown by Standard Industrial Classification (S.I.C) and further cate-
gorization as appropriate to species of pollutants. Thusly, there would
be limitation to the enormous heterogeneity of surface imperviousness and
occurrence probability of particular pollutants inherent to a single class.
e.g., manufacturing. Such category subdivision9 where needed, would likely
save the Menomonee River Pilot Watershed Study from having to echo the
lament of a 1970 San Jacinto River Basin report:
;'The results from models were rejected because of the general incom-
pleteness of the predictor data and because the available predictor data
were not directly related to water quality...1"
A Model Relating Water Quality, Vegetational Structure and Urbanization
in the San Jacinto River Basin. Jameson: D. L. Houston Univ., Texas.
Such land-use derived category expansion might require extensive expan-
sion of the special study sites for calibration purposes (Appendix F List 3).
The modest site preparation and grab sampling called for would allow the esta-
blishment of "a matrix model of the relationship between predictors and water
quality/1 which parallels the San Jacinto Basin Study and which represents
elaboration upon the ''location j ' dimension of the 'data matrix for moni-
toring work units' (Fig. 20 of the 1st IJC semi-annual report).The difficulties
-------
Appendix F List 3. Land use category expansions
For purposes of the MRWS Pilot Study it is likely that expansion of
the Land Use Categories defined by SEWRPC need expansion into some or all
of the sub-categories listed below:
Single Family Residential might be expanded to:
1. One/two family residential - 1/4 acre or less
2. " " " " - 1/2 acre or less
3. " " " " - up to 2 acres
4. " " " " - 2 acres or more
5. Mobile homes
Multi-Family Residential might be expanded to:
6. Multi-family - high rise
7. Multi-family - low rise
8. Residential Under Development, as is
9. Retail and Services (local plus regional), as is
Wholesale and Storage might be expanded to:
10. Wholesale and storage - open
11. " " " - enclosed
Manufacturing and Extracting might be expanded to:
12. Manufacturing - Food and kindred products - general
13. " - " " " " - dairy
14. " - »' " " " - beer
15. " - " " " f: - meat
16. " - Textile Mill Products
17. " - Paper and allied products - general
18. " - " " " " - de-inked pulp
19. " - Printing and publishing - general
20. " - " " " - newspaper
21. " - ChemicaJsand allied products - general
22. " - " " " !i - organic specialists
23. " - " " " " - non-organic specialists
24. " - " " " " - Pharmaceuticals
-------
25. Manufacturing - Energy materials - general
26. " - " " - petroleum products
27. " - " " - coal products
28. " - " " - nuclear isotope products
29. " - Rubber and plastic products - general
30. " - " " " " - rubber
31. " - " " " " - plastics
32. " - Leather and leather products
33. " - Primary metal industry
34-. " - Fabricated metal industry
35. " - Machinery except electrical
36. " - Electrical machinery
37. " - Transportation equipment
38. " - Miscellaneous manufacturing
39. Extractive - Quarries - active
40. " - " inactive
41. " - Mines
Transportation, Communication and Utility Facilities might be expanded to:
42. Transportation, Comm., etc. - Rail, bus and ship terminal
43. " " " - Railroad right-of-way
44. " " " - Railroad yards
45. " " " - Airports (terminal and field)
46. " " " - Local and collector streets right-of-way
47. " " " - Arterial streets £ hwy right-of-way
48. " " " - Freeway £ expressway right-of-way
49. " " " - Truck terminals
50. " " " - Off-street parking
51. " " " - Communication and utility facilities
Governmental and Institutional might be expanded to:
52. Institutional (local and regional)
53. Government (local and regional)
Park and Recreation might be expanded to:
54. Public recreation area (local and regional) - open
55. " " " " " " - enclosed
56. Private and other recreation areas - natural intensive use
57. " " " " " - artificial intensive care
-------
Croplands and Rotation Pasture might be expanded to:
58. Croplands - general
59. " - grains (including oats)
60. " - truck farm (i.e., vegetables)
61. " - corn
62. " - alfalfa and hay
63. Pasture - general
64. " - dairy
65. Barnyard
Orchards and Nurseries might be expanded to:
66. Orchards - general
67. " - stone fruits (e.g., cherries, plums, prunes, almonds,peaches)
68. " - seed fruits (e.g., apples and pears)
69. Nurseries - quantity
70. " - specialty
Fowl and Fur Farms might be expanded to:
71. Fowl (i.e., poultry)
V2. Fur
73. IeV?g, Rivers, Swamps and Canals, as is
7M-. Sw-imps, Marshes and Wetlands, as is
75. UriM7^3 Lands, as is
76. Land Fill and Dumps, as is
77. Woodlands, as is
Categories for application elsewhere in Great Lakes Basin as necessary:
15a. Tobacco manufacturers
16a. Apparel and related products
16b. Lumber and wood products
16c. Furniture and fixtures
32a. Stone, clay and glass products
37a. Instruments and related products
68a. Vineyards
An evaluation of the relative importance of the sub-categories is presently
being undertaken to allow the project to be conducted within manageable
limits.
-------
introduced by additional sampling are not to be ignored, but neither can
the experiential record of past modeling disasters. Careful choice of
sampling sites, field techniques, simplifications derived from hydrologic
season awareness and correlation data for the more sophisticated continuous
monitoring sites could provide the matrix which could markedly assist in
the solution of the modeling objectives (Appendix F Tables 1 and 2),
Overland Flow Element
Literature Review of Modeling Techniques for Overland and River Transport
Systems
Modeling techniques and mathematical models applicable to the land use-
water quality model have been reviewed and evaluated for possible use. The
following areas of the land use-water quality transformation have been
covered:
Rainfall - Runoff Relationship
Rational formula
Unit hydrograph - linear systems
Variable unit hydrograph - nonlinear systems
Kinematic wave models
Flow Routing Methods
Numerical solutions of St. Venants equations
Muskingum method
Hydrograph method
Infiltration and Groundwater Storage
Exponential decrease of infiltration
Soil moisture dependence models
Interception and Depression Storages
Evaporation and Evapotranspiration
Snow Melt and Snow Storage
Erosion and Sediment Transport
Universal soil loss equation (USLE)
Transport of cohesive sediments
Sediment saturated flow
Deposition and scour of cohesive sediments
------- An error occurred while trying to OCR this image.
------- An error occurred while trying to OCR this image.
-------
M-5
Soil Adsorption
Solubility of soil phosphate compounds
Phosphate adsorption process and isotherms
Soil water adsorption models
Over 100 references have been reviewed. A special review report cover-
ing hydrological problems of transport has been prepared. Similar reports
on erosion and sediment transport, and on soil adsorption processes will
be issued in the near future.
Testing of Existing Models
A computer model "STORM" has been acquired and preliminary testing of
the model has been performed on three small subwatersheds.
1. Schoonmaker Creek - a mostly single family residential area in the
southeastern part of the Menomonee River watershed.
2. Noyes Creek - a mixed open (park) and residential subwatershed in
the Little Menomonee River watershed.
3. Donges Bay Road in the most northern part of the watershed draining
farm and agricultural areas.
The model computes the runoff in hourly intervals by a method similar to
the Rational Formula and erosion using USLE. Other quality parameters are
related to the sediment content. The model is adequate for very rough
estimations of sediment originating from dust and dirt accumulation in urban
areas. The results using the erosion portions are unsatisfactory.
The Hydrocomp Simulation Watershed Model, based on the Stanford Water-
shed Model is now being tested by the Southeastern Wisconsin Regional
Planning Commission. Several other models have been reviewed but rejected
as not meeting the objectives of the IJC - land use-water quality modeling
task.
Model Development
Presently, subroutines of the proposed overland flow and quality model
are being developed and combined together under a working name "LANDRUN".
The model which has been formally debugged contains the following sections:
-------
a. rain editing, snow melt and snow pack computation
b. depression and interception storage
c. infiltration
d. excess rainfall - surface runoff transformation
e. erosion by the USLE
f. dust and dirt accumulation
g. sediment transport and deposition during the overland flow
The model is now being tested and delineated.
In addition, conceptual models of soil adsorption applicable to agri-
cultural chemicals, nutrients, pesticides and other pollutants transported
in the soil, are being prepared and will be added to the prepared model.
Parameters of the Land Use
Hydrological Model
Water is the primary carrier of pollution in a watershed. Thus, the
knowledge of the hydrology of the watershed, rainfall runoff transformation,
turbulence and residence times in channels, are the primary parameters to be
estimated by the models.
Overland Flow Model Hydrology
A schematic picture of a watershed hydrology model is shown in Appendix
F Fig. 1. This schematic can be translated into a complex computer model
such as the Stanford Watershed Model (Crawford, Linsley - 1966). The flow
chart of the proposed hydrologic model for the Menomonee River sub-drainage
areas is shown in Appendix F Fig. 2.
Hydrologic Parameters of the Overland Flow Model
The necessary parameters for the model are:
Major Inputs: a. rain intensity mm/hr (in hourly or shorter
intervals)
-------
RMN ( SNOW)
EVAPOTRANSPI RATION
OF AREA
TRANSPORT
(SURFACE 4 GCOUNDWATER WITHDRAWALS)
Appendix T Fig. 1. Overland flow model.
-------
RAIN (SNOW MELT)
MINI
PERVIOUS AREA
EVAPORATION
IMPERVIOUS A.REA
INTERCEPTION
STORAGE
EVAPORATION
STORAGE
EVAPOT(?AN-
&PI RATION
EVAPORATION
OVERi-ANO
STORAGE
SURFACE-
INFLOW RUNOFF
INFILTRATION
ZONE FLOW
AMP STORAGE
INTERFLOW
SURFACE RUNOFF
kMITHORAWAL TO
ANOTHER AREA
OR TO SEWAGE
INTERMEDIATE
Z.ONB FLON
AND
WITHDRAWAL
TO OTHER
AREAS OR TO
SEWAGE.
GROONDWATER
FuOKl AND
STORAGE
.GROUNDNATER( BASE)
RUMOFF
GEOLOGICAL
l
-------
b. temperature °C
c. potential evaporation (daily) mm/day
System Parameters:
a. drainage area - hectares
b. percent imperviousness
c. land use categories
d. roughness of the surface (as Manning's n) for
each category
e. curb density m/hectare for urban areas land use
categories
f. depression and interception storage for each
land use category (mm)
g. surface elevation
h. slope and flow length
Soil Data:
a. hydrologic soil group
b. porosity
c. minimum soil moisture
d. percolation rate
e. depth to water table
f. particle size distribution (% clay, silt and sand)
g. specific gravity of soil particles
Groundwater Data:
a. elevation of the aquifer base and surface
b. hydraulic conductivity
c. porosity
Demographic Data:
a. population density for each land use
Hydraulic Parameters: The River Flow Model
Channel Data:
a. identification of the reaches
b. length (meters) of the reaches
c. bottom elevation of the upper and lower end
-------
50
d. coefficients for velocity - flow relationship
e. coefficients for stage - flow relationship
f. bottom roughness (Manning's h)
Quality Components of the Model
A preliminary schematic picture of the quality subroutines is presented
in Appendix F Fig. 3.
Pollutant Cumulation and Washout in Urban Areas
The pollutant cumulation in urban areas is a function of the land use
characteristics and can also be related to air pollution and wind erosion
characteristics of the area. In the simplest form, the pollutants cumulation
rate can be expressed as
pollutant cumulation ,. ,.. , , . .e ,, -.
—. = f (land use; dust fall)
time '
which can be directly inputed or incorporated as a default value in the
model. More sophisticated models will use wind speed as a factor affecting
the pollutants cumulation rate.
The washout function will utilize the first order decay function in the
form
AP = P0 (l-e P )
where AP = amount of pollutants washed into the runoff during the
time period t
P = amount of pollutants present on the surface before the
beginning of the rain
K = a coefficient depending on the rain intensity
Model Parameters
a. Rates of pollutant cumulation in urban areas for each pollutant
to be modeled.
b. Coefficient K for particulate pollutants.
------- An error occurred while trying to OCR this image.
-------
52
Soil Loss Model
The available modeling techniques for soil loss (erosion) estimation
and particulate matter channel routing will be discussed in a separate
report.
The estimation of soil loss by erosion will be accomplished by the
Universal Soil Loss Equation. The parameters necessary for the soil loss
estimation are (in addition to those necessary for the hydrological model):
a. soil credibility factor
b. cropping management factor
c. erosion control practice factor.
The above parameters are necessary for all soils subject to erosion in the
sub-basins.
To estimate the impact of the soil particles on rivers, the information
must include:
d. soil delivery ratio.
An attempt will be made to estimate the soil delivery ratios by the model.
Sediment Transport in Streams
The model has not, as yet, been formulated. The following information
is deemed necessary for the model in addition to the parameters discussed
previously:
a. settling velocity of sediments (can be computed from particle
sizes and specific gravity)
b. parameters for the estimation of equilibrium concentration (i.e.,
concentration of the sediments at which scour - deposition)
c. plasticity limit and liquid limit of sediments.
Nutrient Model in Overland Flow - Phosphorus Model
The phosphorus model will be based on the adsorption kinetic model
given as:
-------
53
Dissolved Phase
-
at
Sorbed Phase
J9X5
= K (Se-S)
9C
- p
9S
9t
- EN
where
and
C .
S .
S
e
DL
v .
K .
EN
b .
Q°
p •
t .
x .
q -
e ~
1+bC
phosphorus concentration in the soil water phase
phosphorus sorbed on soil particles
saturated equilibrium of sorbed phosphorus
diffusion coefficient
convective downward velocity of infiltration
kinematic coefficient of adsorption
sinks and sources of phosphorus (plant uptake and
release and organic phosphorus transformation)
partition coefficient
maximal sorptivity of phosphorus for soil
soil density
time
coordinate of flow
The preliminary structure of the model is shown in Appendix F Fig.
Inputs to the Model
a. phosphorus content of dust and dirt
b. phosphorus content of atmospheric precipitation
c. fertilizers application
d. other phosphorus inputs
Model Parameters
a. soil pH
b. clay and organic carbon content
c. phosphorus loss and release by plants
-------
DUST AMD DIRT
IN URBAN AREAS
RAINFALL p
COMMERCIAL
INFILTRATION
SOLID PHASE
P
1
SOLUTION P
PLANTS
6ROUNDNATER P
SOIL LOSS
AND EROSION
SOLUBLE1 P
MICROBIAL p
PLANTS AND
RESIDUES
GEOLOGICAL
PHOSPHORUS
suRr/sc? IN
RUNOFF
ORGANIC P
PHOSPHORUS
IN GROUND
HATER RUNOFF
Appendix F Fig. 4. Phosphorus model in the overland
flow model.
-------
55
d. phosphorus adsorption isotherm and kinetic parameters
e. aluminum and iron content
Phosphorus Movement in Streams
Similarly to soil phosphorus, phosphorus in the aquatic environment is
either bound to suspended particles or dissolved. Thus, a similar adsorp-
tion sorbed phase-free phase equilibrium model may be formulated. The
dissolved phosphorus is in equilibrium with that adsorbed on suspended
particles or on the bottom deposits. The release of phosphorus from the
bottom deposits is a diffusion process taking place in the bottom boundary
layer. The final form of the model has not, as yet, been formulated.
Nitrogen Model - Soil Nitrogen
Soil nitrogen model will consist of the following components:
Nitrogen fixation by soil bacteria
Ammonia adsorption on soil particles
Nitrification
Denitrification
Deamination
Nitrogen uptake of plants
In its final form the model will be based on the simultaneous solution of
differential equations describing the above processes in the soil profile.
Pesticides and Heavy Metals Transport
Modelling of the above components will be discussed in a separate report
following evaluation of the literature survey.
Determination of the Relation of Surface Runoff Concentrations
to Land Use and Meteorological Conditions
The purpose of this project is to develop an empirical means of relating
the quality of surface runoff to land use and meteorological conditions
-------
56
rather than directly to the ultimate nonpoint sources of atmospheric fallout,
precipitation, land and road management practices and natural weathering.
The direct relation approach requires extensive monitoring of these nonpoint
sources. Although these data are being collected in the Menomonee River
Pilot Watershed Project, they are relatively rare elsewhere. Thus, it is
worthwhile to investigate other means of predicting surface runoff quality.
The method being developed involves analysis of the mass loadings of
flow during runoff events from small drainage areas where point sources are
minimal. The total mass loadings represent integrations of the inputs from
all the nonpoint sources contributing to the area plus the baseflow. Separ-
ation of baseflow and baseflow loadings from the event hydrograph and mass
loading curves, respectively, leaves the discharge and mass loading of the
surface runoff itself. The concentration of a particular constituent in
the surface runoff (CL) is then simply the mass load of that constituent
K
divided by the discharge at the same instant during the event. A dimension-
less relative concentration (C_,/C9 where C is the mean concentration of the
K
constituent in the surface runoff for a given event) can then be defined.
Studies underway on small tributaries to the Milwaukee River, immediately
east of the Menomonee River have shown that the relative concentration
shows a time distribution during an event that is related to land use.
Runoff event data from Noyes Creek, Schoonmaker Creek and the Little
Menomonee River, plus the specific land use sites in the Menomonee Watershed,
and the Milwaukee tributaries will be used to determine surface runoff con-
centrations and relative concentrations. The relation of these relative
concentrations to time during the event, and to land use and meteorological
conditions, such as rainfall intensity and duration, and antecedent rainfall,
will be analyzed statistically. This analysis will provide an empirical
relationship between the relative concentration's time distribution and
land use and rainfall conditions which will allow prediction of surface
runoff concentrations throughout the event. These concentrations can in
turn be combined with a unit hydrograph to produce mass loading curves for
the event.
-------
57
River Transport Model
The third element of the modeling, namely, the river transport model,
is to pick up the output runoff and sediment of the overland flow model and
describe resultant downstream output to the boundary of Lake Michigan. How-
ever, closer scrutiny has shown this to be a particularly complex problem.
Above the control structure which admits mainstem flow to a channelized
section commencing at 45th Street (i.e., approximately 10 blocks downstream
of Hawley Road) is conventional streamflow; however, below that point is
either potential and (in most cases) actual major influence of seiche from
the Lake. Thus, to provide output applicable to lake input quantification
the model must consist of two elements: a conventional river transport and
an estuarine transport endeavor.
A fully-dynamic model might incorporate in-stream/benthic-effect
modifications but more importantly show other benefits. Input of data
would be easier and the model could allow for better parameter quantification.
The River Transport modeling would also have to address itself to the
impending elimination of sewage treatment plant discharges. Similarly,
best application of the Army Corps of Engineers Scour and Deposit program
and the (extension of the) Stanford Sediment program is critical to the
important, bottom-transport aspects of the River Transport Model.
Estuarine focus initially involves definition of limits. The 45th
Street flow-control device on the Henomonee River, the large dam below
North Avenue on the Milwaukee River and the old railroad culverts adjacent
to Cleveland Avenue on the Kinnickkinnick River down to the harbor entrance
would probably be the most appropriate definition of estuary limits. This
would include extensive cross-correlation potential from industrial acre-
age, a flushing tunnel from the Lake to the Kinnickkinnick River, sheet
pile flow-control devices(s) at the Falk Corporation, a present monitoring
installation (413004) and three grab sample sites (413014, 413013 and
413012). Continuity of these sites plus establishment of six more grab-
sites (Milwaukee River and Kinnickkinnick River estuary-limits, sites,
prior to their entries into the harbor, at the harbor entrance and an
inharbor site) along with assessment of dam and tunnel records, applica-
tion of aerial photography, and attention to the understanding of the
-------
58
"secondary" seiche mechanisms (i.e.3 wind-generated5 as opposed to barometric
pressure) is needed to complete a river transport model. Coordination with
an upcoming University of Wisconsin-Milwaukee project to investigate seiche
in the Milwaukee Harbor will be effected.
Evaluation of a Continuous Simulation Model
The Southeastern Wisconsin Regional Planning Commission in their role
as a prime contractor under the IJC project and as a designated planning
commission for receipt of Section 208 (P.L. 92-500) funding play a unique
part in coordinating efforts on the Menomonee River Pilot Watershed Study.
In this coordinating role work was initiated on the evaluation of an exis-
ting continuous simulation model to determine if the model could—with
perhaps some modification—satisfy the third objective of the Menomonee
River Pilot Watershed Study which calls for developing the predictive
capability necessary to extend the findings of the study to other urban
settings in the Great Lakes Basin. The overall approach in this model
evaluation is to test first the effectiveness of the model in reproducing
historic hydrologic-hydraulic data and then to determine model capability
with respect to water quality phenomena. This procedure is based on the
premise that successful water quality simulation is contingent upon effec-
tive hydrologic-hydraulic modeling since runoff from the land and flow in
the streams provide the transport mechanism for water quality constituents.
Model Description
The digital computer model selected for evaluation is based on a
hydrologic-hydraulic model that originated at Stanford University in the
early 1960's and is now available—with a water quality feature—from the
consulting firm of Hydrocomp, Incorporated. The model is unique in that it
continuously simulates hydrologic-hydraulic and water quality processes for
an indefinite period of time in response to a full spectrum of meteorological
conditions. This is in contrast to the more common discrete event models
that generate hydrographs and pollutographs only for major hydrologic events.
Continuous process models are most likely to meet the needs of the pilot study
-------
59
because they permit the computation of total annual transport of potential
pollutants from a watershed to the Great Lakes for the complete spectrum
of runoff conditions and streamflows and because they facilitate statistical
analysis of the manner in which that material is delivered to the receiving
waters.
The model consists of three submodels: The Hydrologic Submodel, the
Hydraulic Submodel and the Water Quality Submodel. The principal function
of the Hydrologic Submodel is to determine the volume and temporal distrib-
ution of runoff from the land to the stream system. Meteorological data and
land data constitute the two principal types of input for operation of the
Hydrologic Submodel. The key output from the submodel consists of a con-
tinuous series of runoff quantities for each land segment in the watershed.
The function of the Hydraulic Submodel is to accept as input the runoff
from the land surface as produced by the Hydrologic Submodel, to aggregrate
it, and to route it through the stream system thereby producing a continuous
series of discharge values at predetermined locations along the surface
water system of the watershed. Input for the Hydraulic Submodel consists
of parameters describing the stream reaches and impoundment sites as well
as the output from the Hydrologic Submodel. The Water Quality Submodel
simulates the time-varying concentration, or levels, of water quality
indicators at selected points throughout the surface water system. Opera-
ting on a reach-by-reach basis, the submodel continuously determines water
quality as a function of reach inflow and outflow, dilution, and biochemical
processes. Input to the Water Quality Submodel consists of output from the
Hydrologic Submodel, channel data and diffuse and point source data. Out-
put from the submodel consists of a continuous series of water quality
levels at selected points on the watershed stream system.
Data Base Development
Data base development consists of the acquisition, verification and
coding of data needed to operate, calibrate and apply the model. The model
data base for the watershed and environs is being assembled and when com-
pleted will consist of a large, readily accessible computer file of informa-
tion subdivided into six distinct categories: meteorological data, land
-------
60
data, channel data, riverine area structure data, diffuse source data and
point source data. The meteorological data set is the largest because it
contains 35 years of semi-monthly, daily, or hourly information for seven
types of meteorological data. The data base is being developed by combin-
ing data obtained from the Southeastern Wisconsin Regional Planning
Commission with data from the U.S. National Climatic Center.
Initial Calibration Runs
Calibration consists of comparing simulation results with historic fact
and, if a significant difference occurs, making parameter adjustments so as
to tailor the model to the natural and man-made features of the watershed.
The initial calibration of the hydrologic-hydraulic portions of the model
were conducted on sub-watersheds outside of but close to the Menomonee
River watershed that were essentially spatially homogeneous with respect to
soils, slope and land use-cover and had combinations of these three key
land characteristics that were similar to those in land characteristics
that were similar to those in land segments of the Menomonee River water-
shed. Gaged subwatersheds outside of the Menomonee River watershed were
selected because the monitoring network that exists within the watershed
had not yet yielded sufficient streamflow data for calibration purposes.
The underlying objective was to use the initial calibration process to
determine land parameters for the homogeneous subwatersheds which could
in turn be applied to the Menomonee River watershed.
The initial calibration runs on the three test subwatersheds reveals
that the combination of the Hydrologic Submodel and the Hydraulic Submodel
can effectively reproduce the hydrologic-hydraulic response of a basin to
a wide variety of meteorologic inputs. A close correlation was achieved
between simulated and recorded annual and monthly runoff volumes, simulated
and recorded hydrographs for major runoff events, and simulated and recorded
annual instantaneous peak discharges. *
Work Elements Planned and in Progress
Upon successful completion of the initial stage of the calibration process,
-------
61
calibration runs were initiated in the Menomonee River watershed using lim-
ited historic stream flow records. An example of the results of that simula-
tion is presented in Appendix F Fig. 5 in the form of a graphical comparison
between simulated and recorded monthly runoff volumes. These model runs
will be supplemented with calibration runs using precipitation and stream-
flow data collected under the pilot study monitoring program. After completion
of the hydrologic-hyraulic calibration, the model will be tested against the
water quality monitoring data.
-------
7.0
6.0
CO
O 5.0
z
UJ
5
u.
u.
O
z
O
cc
Q
UJ
CO
4.0
3.0
2.0
1.0
Z
1-0 2.0 3.0 4.0 5.0
RECORDED RUNOFF VOLUME IN INCHES
6.0
70
SOURCE- SEWRPC
Appendix F Fig. 5.
Comparison of recorded and simulated monthly
runoff volumes for the Menomonee River at the
Wauwatosa Gauge, January 1, 1963 to
September 31 ,1973.
-------
U.S. Environmental Protection Agency
GLNPO Library Collection (PL-12J)
77 West Jackson Boulevard,
Chicago, IL 60604-3590
------- |