AEPA
United States
Environmental Protection
Agency
Environmental Research
Laboratory
Athens GA 30613
Research and Development
EPA-600/S3-82-029 August 1982
Project Summary
M
Application of Water Quality
Models to a Small Forested
Watershed: I. The
Nondesignated 208 Area
Screening Model
J. Hesson and J. K. Robertson
The report presents an evaluation of
the application of the Water Quality
Assessment Methodology (Water
Quality Assessment: A Screening
Method for Nondesignated 208 Areas,
EPA-600/9-77-023) to a forested
watershed on portions of the U.S.
Military Academy Reservation and the
Harriman Section of The Palisades
Interstate Park in Orange County,
N.Y. As part of the calibration and
verification process, field data for
water quality, hydrologic, and meteo-
rological parameters were collected.
The report details the selection of
sampling sites, instruments, techni-
ques, and analytical methods used in the
data collection. Parameter selection
for model use is explained.
This Project Summary was developed
by EPA's Environmental Research
Laboratory, Athens, GA, to announce
key findings of the research project
that is fully documented in a separate
report of the same title (see Project
Report ordering information at back).
Introduction
Basin planning requires a set of
analysis procedures that can provide an
assessment of the current state of the
environment and a means of predicting
the effectiveness of alternative pollution
control strategies. In 1977, the Environ-
mental Research Laboratory published
Water Quality Assessment: A Screening
Method for Nondesignated 208 Areas
(EPA-600/9-77-023), which contains a
set of consistent analysis methods that
accomplish these tasks. The assessment
procedure, called the Water Quality
Assessment Methodology (WQAM), is
directed toward local and state govern-
ment planners who must interpret
technical information from many sources
and recommend the most prudent
course of action that will maximize the
environmental benefits to the community
and minimize the cost of implementation.
An integral part of the WQAM develop-
ment process is the calibration and
verification of the model on actual
watershed. This report evaluates its
use in cl aracterizing wasteloads and
water quality in small forested water-
sheds in New York.
The West Point Study Area comprises
3247 acres of watershed draining to the
dam on Popolopen Lake (Figure 1).
Elevations range from 678 ft at Popolopen
Lake to 1401 ft along the northwest
margin of the basin. Soils on the hills
are shallow with zero to 18-24 inches
overlying bedrock. Lowland soils are
deeper, up to 6 feet.
Summer temperatures at West Point
average 74 degrees Fahrenheit, but
short hot spells in the nineties are
-------
Black Rock
N
,'' Military
/ Reservation t / M
f /£""""
Forest .' .
/ ~v\Ğ
Popolopen
.' Lake
Watershed Area
5.75 sq. miles
1490 hectares
4*£/. .-Palisades /
/ .' Interstate/
Park I
/ Lake / '6/
/ Georgiana ' it m
t .'/iĞ-
/ . **5
Figure 1.
JStatute Mile
... 'Watershed Boundary
USMA Reservation Boundary
Roads
The West Point study area showing the watershed boundary and
wetland locations.
common. Winters are moderately cold,
with temperatures averaging just below
the freezing point. It is not unusual to
have rain throughout the winter.
Snowpack typically comes and goes
throughout the winter. The mean
annual precipitation at West Point is47
inches, distributed evenly through the
year.
Drainage on the watershed is a
modified trellis pattern influenced by
the lineation in the underlying rocks and
faults. Six lakes (ponds) and 21 wetland
complexes affect the flow of water on
the watershed. Five of the six lakes are
manmade; Bull Pond, the deepest lake,
is natural. Three of the five manmade
lakes (Summit, Popolopen, Beaver) have
depths greater than the dam height and
thus must have existed as small ponds
or wetlands prior to impoundment.
The streams in the area have cut to
bedrock in most cases. Channels are
strewn with boulders and stones.
Streams flash to high flow after storms
because of the impervious bedrock
material close to the surface. There is
little overland flow under the forest
canopy. A great deal of interflow takes
place at the soil-bedrock interface.
The entire watershed tributary to
Popolopen Lake was modeled for the
study. Individual sub-areas are shown
in Figure 2. The ground surfaces of all
areas were modeled individually for
nonpoint wasteloads of sediment,
nitrogen, phosphorus, and organic
matter. In addition, composite loads
from areas 1 through 4 and for trv
entire study area were predicted. Trv
estimated wasteloads of surroundini
and tributary areas were used as .input
to model three lakes for sedimentation
eutrophication, and thermal stratification
The data collection network on th<
West Point Study Area was designed t<
gather data for the EPA-supportec
model testing program and to serve as
the basis for a future program dealinc
with the hydrology and nutrient budgets
of wetlands. The study was designed tc
obtain:
1) input to a simulation model.
2) data against which the model
prediction will be compared during
the calibration of the model to the
West Point Study Area.
3) data against which the prediction
from the calibrated model will be
compared to verify that the model
is providing valid predictions.
4) data to support the wetland studies.
Data needs are in the following areas.
1) Hydrology - stream flow, evapora-
tion, precipitation, etc.
2) Meteorology - temperature, dew
point, radiation, wind speed, etc.
3) Physical characteristics of the
watershed - area, slope, soil
depth, soil types and extent, cover,
aspect, etc.
4) Water Quality - both chemical and
biological.
5) Rate constants for reactions.
The report elaborates on the instrument
and site selection criteria used. Pa-
rameter selection and sources used to
select parameters for modeling are
presented in the report.
Two basic problems may have restricted
the application of the WQAM to the
West Point Study Area: 1) the study area
is almost totally forested with no
agricultural land; 2) the assumption that
the majority of the nutrient and organic
loads are tied to sediment may not hold
for watersheds having relatively low
sediment yields. Little work has been
done on applying the Universal Soil Loss
Equation to forested lands and thus
input parameters are less clearly
defined.
Based on the loading functions
provided in the WQAM documentation,
loading functions estimates were made
of the average annual loads of sediment,
nitrogen, phosphorus and organic
matter contributed from each of eight
sub-areas of the West Point Study Area.
These estimated loads for the entire
study area (sub-areas 1 through 8) and
for the four southern sub-areas as a
-------
whole (sub-areas 1 through 4) are
presented in Table 1.
In order to make the WQAM work on
the West Point Study Area, we have in
several instances, turned to the Water
Resources Evaluation of Non-Point
Silvicultural Sources (WRENSS) model
(EPA-600/8-80-012) and borrowed
pieces to patch up difficulties in the
WQAM. Which is the better model, the
patched-up WQAM or the WRENSS
model? The WRENSS model was devel-
oped for forested areas and probably
should be used for large forested
watersheds in preference to the WQAM.
WQAM was meant to identify problem
areas in large regions, portions of which
may contain forests. It should contain
procedures to allow modeling of forested
areas or incorporate the WRENSS
model procedures for this purpose. A
comparison of the ease of use of the two
procedures and comparison with field
data are needed to decide whether the
WRENSS model or the patched-up
WQAM is the better alternative.
The impoundment models were
applied without modification. The
thermal profile predictions for lakes of
various depths and residence times can
be said to be validated for The West
Point Study Area, Figure 3, although we
could have used better guidance in the
procedures on how to decide whether
our lakes were well mixed. These
procedures could easily be applied with
confidence to other lakes in the Hudson
Highlands. For existing impoundments,
however, the data requirements for
their use are such that we think many
modelers would be prevented from
using the prediction, finding it easier
and cheaper to measure a thermal
profile than to obtain data necessary to
compute lake volume and residence
time. For planned impoundments, the
designers would know volumes and
easily predict residence time.
The sedimentation rate prediction is
easily applied but is only as good as the
nonpoint source sediment yield predic-
tions. When the best predictive method
for forested areas is determined, then a
closer look at the sediment accumulation
prediction can be made.
The impoundment eutrophication
predictions in the WQAM will indicate a
eutrophication problem only in phos-
phorus-limited situations. The proce-
dures should be expanded to allow
prediction in nitrogen limiting situations
as well as those in which neither
phosphorus or nitrogen controls. The
stream models in the WQAM docu-
N
Popolopen
Lake
Lake
Georgiana * m
*
Barnes
Lake
Scale
1000 500 0
1000
""* Meters
1/2
Statute Mile
.Summit
1 M Lake
Watershed Boundary
_ _ Sub-area Boundary
Figure 2. Sub-area basins within the West Point study area.
Table 1. Summary of Predicted Nonpoint Loads from the West Point Study Are9
Using the Water Quality Assessment Methodology Entire Study
Southern Area Area
(sub-areas 1-4) (sub-areas 1-8)
Sediment, tons/year
Sediment, tons/ acre-year
Nitrogen, Ib/year
Nitrogen, lb/ acre-year
Phosphorus, Ib/year
Phosphorus, Ib/acre-year
Organic Matter, Ib/year
Organic Matter, Ib/acre-year
621
0.66
5.818
6.1
1,305
1.38
93,150
98.5
2.115
0.65
19,839
6.1
4,442
1.37
317,250
97,7
-------
May Jun
Oil 0
8-
re-
24
8-
16
24
8-
16
24
0 o20 30 0 1'0 2\) 30 0 1'Q 10 30 0 1'0 2'0 30
Jul _ Aug Sep _ Oct
8-
16
24
24
o lo 2050 o ;'o io so o/'o 2'o 30 o i'o 2'o 30
Nov
Dec
8-
16-
24
0 I'O 20 30 0 1'0 2D 30
Temp, °C
From Burlington, VT
Max Mixing
From Burlington, VT
Min Mixing
Actual Bull Pond
Profiles May '77-Apr '78
Figure 3. Comparison of Bull Pond thermal profiles with model estimates assuming
varying degrees of mixing.
mentation have provision for this and
perhaps could be expanded to lakes.
The dissolved oxygen model gives a
reasonable approximation for DO levels
during the mid-summer maximum
stress period. Assuming Ks = 0 yields a
DO prediction higher than observed and
can lead one to conclude, as we did, that
the benthic DO demand prior to strati-
fication is zero. This is not reasonable
for every lake, and our test at Bull Pond
yielded a difference between actual and
predicted DO levels of almost 1.0 mg/l.
In lakes with higher levels of decaying
organic matter the difference would be
greater. Table V-10 in the WQAM
documentation should be used in
preference to Equation 17-13 for
assigning an LM value. The assumption,
based on observations at Bull Pond, that
ks = 0 appears justified in other parts of
the model. Again this works well for Bull
Pond, but in other lakes with a higher
settleable BOD rate, it will lead to a
prediction error.
Conclusions
1. Wasteloading from nonpoint sources
on the West Point Study Area or any
other steeply sloped area cannot be
modeled by the WQAM without extension
or modification of the algorithm for
assigning a value to the topographic
factor, LS.
2. Procedures for use of the erosion
control practice factor, P, for forested
areas need to be clarified in the WQAM.
3. The formulation for deriving the
sediment delivery ratio. So in the
WQAM, yields sediment loading values
higher than we are comfortable with.
The sediment delivery index in the
WRENSS model derived from eight
forest parameters seems better suited
to forested areas.
4. Lake eutrophication predictions in
the WQAM only work for phosphorus-
limited conditions. The model needs to
be extended to other situations.
5. The lake thermal profiles model in
the WQAM, when applied to the West
Point Study Area, gives a reasonable
prediction of actual field conditions.
6. The lake dissolved oxygen prediction
in the WQAM has been verified for one
lake in the West Point Study Area.
7. Application of the WQAM to forested
watersheds should be done with caution
until the methodology is thoroughly
tested and verified.
Recommendations
1. That further data collection be
conducted on the West Point Study Area
to enable verification of the wasteloading
prediction from nonpoint sources based
on the WQAM.
2. That WQAM model procedures be
revised to give guidance to the user with
forested terrain to model on how to set
the erosion control practice factor, P, in
the universal soil loss equation.
3. That the procedures from the
WRENSS model for calculation of the
topographic factor, LS, and sediment
delivery index, SDI, be added to the
WQAM as replacements for LS and Sa
for steeply sloped forested areas only.
4. That the WRENSS model be applied
to the West Point Study Area and
verified and that the revised WQAM
(see 3 above) be verified on the same
watershed to determine which model
gives the better prediction of water
quality from a forested watershed.
5. That for the average annual rainfall
factor, R, maps presented in the USDA
Agriculture Handbook 537 be substituted
for the generalized version presented in
the WQAM documentation.
-------
J. Hesson and J. K. Robertson are with the Science Research Laboratory, U.S.
Military Academy, West Point. NY 10996.
T. O. Barnwell andj. W. Falco are the EPA Project Officers (see below).
The complete report, entitled "Application of Water Quality Models to a Small
Forested Watershed: I. The Nondesignated 208 Area Screening Model,"
(Order No. PB 82-242 520; Cost: $ 12.00, subject to change) will be available
only from;
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Environmental Research Laboratory
U.S. Environmental Protection Agency
Athens, GA 30613
1982-559-092/0479
-------
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Postage and
Fees Paid
Environmental
Protection
Agency
EPA 335
Official Business
Penalty for Private Use $300
0000329
------- |