7/29/2019 Simulating Hydrologic Budgets for Three Illinois Watersheds http://slidepdf.com/reader/full/simulating-hydrologic-budgets-for-three-illinois-watersheds 1/21 ELSEVIER Journal Journal of Hydrology 176 (1996) 57-77 Estimating hydrologic budgets for three Illinois watersheds J.G. Arnold”‘*, P.M. Allenb alJSDA-Agricultural Research Service, 808 East Blackland Road, Temple, TX 76502, USA ‘Baylor University, Department of Geology, Waco, TX 76798, USA Received 11 November 1994; revision accepted 10 April 1995 Abstract It is important to simulate the major components of the hydrologic budget to determine the impacts of proposed land management, vegetative changes, groundwater withdrawals, and reservoir management on water supply and water quality. As acquisition of field data is costly and time consuming, models have been created to test various land use practices and their concomitant effects on the hydrologic budget of watersheds. To simulate such management scenarios realistically, a model should be able to simulate the individual components of the hydrologic budget. However, most field studies at the watershed scale attempt to measure only one component (e.g. total streamflow, evapotranspiration (ET), etc.) and are not suitable for validating individual components of a comprehensive model. A field study was completed in the 1950s to estimate several major hydrologic components including surface runoff, groundwater flow, groundwater ET, ET in the soil profile, groundwater recharge, and groundwater heights from measured data from three watersheds in Illinois. These data were used to validate a multicomponent water budget model called SWAT. Comparison of measured and predicted values demonstrated that each component of the model gave reasonable output and that the interaction among components was realistic. This fact should allow more realistic appraisal of various land use management practices on a basin-wide scale. 1. Introduction Numerous models have been developed over the past several years to assist in understanding the hydrologic system. Such models provide a framework in which to analyze data and test hypotheses. Models are also used as a predictive tool to test changes in the hydrologic regime such as water yield and pollution brought on by changes in basin attributes such as land use or irrigation (Beasley et al., 1980; Arnold * Corresponding author. 0022-1694/96/$15.00 0 1996 - Elsevier Science B.V. All rights reserved SSDI 0022-1694(95)02782-3
21
Embed
Simulating Hydrologic Budgets for Three Illinois Watersheds
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
7/29/2019 Simulating Hydrologic Budgets for Three Illinois Watersheds
Estimating hydrologic budgets for three Illinois watersheds
J.G. Arnold”‘*, P.M . Allenb
alJSDA-Agricultural Research Service, 808 East Blackland Road, Temple, TX 76502, USA
‘Baylor University, Department of Geology, Waco, TX 76798, USA
Received 11 November 1994; revision accepted 10 April 199 5
Abstract
It is important to simulate the major co mponen ts of the hydrologic budget to determine the
impacts of proposed land management, vegetative changes, groundwater withdrawals, and
reservoir managem ent on water su pply and wate r quality. A s acquisition of field data is costly
and time consuming, mode ls have been create d to test various land use practices and their
concomitant effects on the hydrologic budget o f watersheds. To simulate such management
scenarios realistically, a mode l should be able to simulate the individual compon ents of the
hydrologic budget. However, most field studies at the watershed scale attempt to measure only
one componen t (e.g. total stream flow, evapotranspiration (ET) , etc.) and are not suitable for
validating individual compon ents of a comp rehensive mode l. A field study was com pleted in the1950 s to estimate several major hydrolog ic compon ents including surface runoff, groundwater
flow, groundwater ET, ET in the soil profile, groundwater recharge, and groundwater heights
from measured data from three watersheds in Illinois. These data were used to validate a
multicomponent water budget model called SWAT . Comparison of measured and predicted
values demonstrated that each component of the model gave reasonable output and that the
interaction among com ponents was realistic. This fact should allow more realistic appraisal of
various land use managem ent practices on a basin-wide scale.
1. Introduction
Numerous models have been developed over the past several years to assist in
understanding the hydrologic system. Such models provide a framework in which
to analyze data and test hypotheses. Mod els are also used as a predictive tool to test
changes in the hydrologic regime such as water yield and pollution brought on by
changes in basin attributes such as land use or irrigation (Beasley et al., 1980; Arnold
* Corresponding author.
0022-16 94/96/$15 .00 0 1996 - Elsevier Science B.V. All rights reservedSSDI 0022-1694(95)02782-3
7/29/2019 Simulating Hydrologic Budgets for Three Illinois Watersheds
58 J.G. Arnol d, P.M. A l l en / Journal of Hy drol ogy 176 (1996) 57-77
et al., 1990; Bultot et al., 1990; Refsgaard et al., 1992). Problems inherent to building
such models have been enumerated by Klemes (1986), Beven (1989) and Grayson et
al. (1992). Briefly, these may be summ arized. First, there is the presumption that
physical processes in the basin can be represented in deterministic ways and that
linking of these solutions is possible. Second, it is often assum ed that input datafrom the field are available for estimation of all model param eters or that the
model can accurately derive such parameters from manipulation of related field
data. Finally, there is the assumption that such inputs derived from selective
sampling in the field adequately represent the spatial and/or temporal variability in
the field.
Although many models have successfully synthesized one or more parameters of
the hydrologic budget (Freeze, 1972; Beven and Kirkby, 1979; Leavesley, 1983; Hata
and Anderson, 1983; Loague and Freeze, 1985; Hebbert and Smith, 1990), data
availability and calibration of such models is still an arduous task (Cary, 1991).
Such calibration is often based on assessment of the goodness of fit of the model to
gaged daily or monthly discharge for the watershed. Calibration is achieved often
through a combination of trial and error adjustments and limited optimization.
Following calibration, and based on the assessed accuracy of the predicted to
simulated discharges, many models are being used to evaluate other com ponents of
the hydrologic system such as evaporation, ground water flow and storage (Peuntes
and Atkins, 1989). Assessing the validity of these models to predict such components
of the hydrologic system is usually not done. T his is problematic as measurement of
components of the water balance alw ays involves errors. T he only components of the
water balance that are regionally observed from a number of stations are precipita-
tion, streamflow, and to a lesser extent pan evaporation. Except for a limited numb er
of experimental watersheds, soil moisture, evaporation and transpiration, waterstorage and infiltration are usually estimated from empirical formulae. Here, the
accuracy of the model depends on the input requirements and the degree to which
the structure of the model approximates the physical process. Winter (198 1) discussed
various types of errors inherent in measurement and computation of the various
components of the water balance. He found that long-term averages had less error
than short-term values. Errors in annual estimates of precipitation, streamflow a nd
evaporation ranged from 2 to 15% whereas m onthly estimates ranged from 2 to 30%
For w orst possible estimates of the error, the sum of these errors should probably be
considered. Given these inherent problems in the water balance computations, it is
still an extremely useful approach to assessing the interrelationships between the
components of the hydrologic system. The only way to begin to assess the usefulness
of a model is to test it against actual data and other independently modeled
interpretations. Grayson et al. (1992) suggested that the following procedures be
part of analyzing any model. First, the model must be tested and calibrated over a
wide variety of watersheds under a wide range of conditions. Second, b oth positive as
well as negative results should be reported, and the uncertainty surrounding the
model predictions should be discussed. Finally, the source and precision of the
input data should be presented. The purpose of this paper is to compare the water
balance o utput of the SWA T (Soil and Water A ssessment Tool) model (Arnold et al.,
7/29/2019 Simulating Hydrologic Budgets for Three Illinois Watersheds
J.G. Ar nol d, P.M. Al l en / Journal of Hy drol ogy 176 (1996) 57-77 61
2. Study area
The three watersheds are located within the till plains of central and western Illinois
(Fig. 1). The general characteristics of each watershed are summ arized in Table 1,
after Schicht and Walton (1961).
3. Historical water budget calculations
Schicht and Walton (1961) used precipitation, stream flow and groundwater level
data to ascertain groundwater recharge, runoff, and evapotranspiration for three
basins in Illinois. They used a simple balance equation that contains basic elements
of the water budget,
P = R + E T + U f A S s * A S g (1 )
where P is precipitation, R is stream flow, E T is evapotranspiration, U is subsurface
underflow, ASS is change in soil moisture, and AS g is change in groundwater storage.
Precipitation was measured for each basin with rain gage densities ranging from
17.12 krr2 per gage for Hadley Creek to 20.20 km2 per gage for Goose Creek and
27.84 km2 per gage for Panther Creek. S treamflow was monitored at each basin outlet
for the study period. Groun dwater levels were monitored with continuously recording
gages for three w ells on Goose Creek, and for five wells on Hadley and Panther
Creeks. At times as many as 16-21 wells were monitored in the last two basins to
verify the quality of the data being recorded at the continuous mon itoring wells. Soil
moisture was not measured. Evapotranspiration was solved from the water budget
equation assuming no significant change in soil moisture during the year. Subsurfaceunderflow was calculated from a modified form of the Darcy equation, Q = T I L ,
where Q is underflow (in 1day-‘), T is coefficient of transm issibility (in 1day-’ ft-‘), Z
is hydraulic gradient of the water table (in m m-l), and L is width of the cross-section
of the deposits in meters. The change in soil moisture was assumed to be zero. The
change in groundwater storage wa s estimated from the change in mean groundw ater
stage from the observation wells and the gravity yield of the wells,
ASg = AH( Yg) (2)
where A Sg is change in groundwater storage, AH is mean chang e in groundwater
stage, and Yg is gravity yield of the deposits described by the equation
Yg =P - R - E T - U
AH(3)
This equation was assumed valid for periods when soil moisture change was not
significant. It was assumed that evapotranspiration averaged 90 mm per month,
soil moisture change w as eliminated, and the equation was solved for inventory
periods during winter and early spring when the water table was rising.
The groundwater budget was stated as
P g = R g + E T g + U f A S g (4 )
7/29/2019 Simulating Hydrologic Budgets for Three Illinois Watersheds
62 J.G. Ar nol d, P.M. A l l en / Journal of Hy drol ogy 176 (1996) 57-77
where P g is groundwater recharge, R g is groundwater runoff, E T g is groundwater
evapotranspiration, U is subsurface underflow, and ASg is change in groundwater
storage. Groundwater runoff was estimated based on standard hydrograph separa-
tion techniques assuming that surface runoff is complete within 5 days after rainfall
and the reminder of fair weather flow is all groundwater runoff. R ating curves ofmean groundwater runoff corresponding to a groundwater stage were prepared for
both periods of high groundwater evapotranspiration (April-October) and low
groundwater evapotranspiration (November-M arch). The difference in ground-
water runoff between the two curves was specified as ground water evapotranspira-
tion. Recharge w as estimated from the groundwater budget equation where
groundwater runoff and evapotranspiration were determined from the mean ground-
water stage-runoff rating curves. Groundw ater storage was computed from gravity
yield and average water level declines as previously described .
Panther Creek a nd Goose Creek are each underlain by glacial till and dis-
continuous sand and gravel lenses. Such glacial deposits are marked by their great
variability in hydraulic conductivities, which may range from 1.3 x lop7 cm s-i to
3.81 x 10e5 cm s-l depending on clay content, mode of deposition, degree of
J.G. Ar nol d, P.M. Al l en / Journal of Hy drol ogy 176 (1996) 57-77 63
readily available inpu ts, (c) is comp utationally efficient to operate on large basins in a
reasonable time, and (d) is a continuous time model and is capable of simulating long
periods for computing the effects of managem ent changes.
SW AT uses a comm and structure for routing runoff and chemicals through a
watershed similar to the structure of HY MO (Williams and Hann, 1973). Com mand sare included for routing flows through streams an d reservoirs, adding flows, and
inputting measured data or point sources. Using a routing com mand language, the
model can simulate a basin subdivided into grid cells or subwatersheds. Additional
comm ands have been developed to allow measured and point source data to be input
to the model an d routed with simulated flows. Also, output data from other simula-
tion models can be input to SW AT. Using the transfer comm and, water can be
transferred from any reach or reservoir to any other reach or reservoir within the
basin. T he user can specify the fraction of flow to divert, the minim um flow remaining
in the channel or reservoir, or a daily amount to divert. The user can also apply w ater
directly to a subbasin for irrigation. Although the model operates on a daily time step
and is efficient enough to run for many years, it is intended as a long-term yield model
and is not capable of detailed, single-even t, flood routing.
5.1 . Subbasin components
The components of SWA T can be placed into eight major divisions-hydrology,
weather, sedimentation, soil temperature, crop growth, nutrients, pesticides, and
agricultural managem ent. A detailed description of the SWA T components has
been given by Arnold (1992) and Arnold et al. (1994). A brief description of the
hydrology components is presented here.
5 .2 . Su r face runo f hyd ro l ogy
Surface runoff from daily rainfall is predicted using a procedure similar to the
CRE AM S runoff model, option one (Knisel 198 0; Williams and Nicks, 19 82). Like
the CRE AM S model, runoff volume is estimated with a modification of the SCS curve
number method (USD A Soil Conservation Service, 1972). The curve number varies
non-linearly from the 1 (dry) condition at wilting point to the 3 (wet) condition at field
capacity, and approaches 100 at saturation. The SW AT model also includes a
provision for estima ting runoff from frozen soil.
Peak runoff rate predictions are based on a modification of the Rational Formula.The runoff coefficient is calculated as the ratio of runoff volume to rainfall. The
rainfall intensity during the watersh ed time of concentration is estimated for each
storm as a function of total rainfall using a stochastic technique. The watershed time
of concentration is estimated u sing M anning’s Formula considering both overland
and channel flow.
5 .3 . Pe rco la t ion
The percolation component of SW AT uses a storage routing technique to predict
7/29/2019 Simulating Hydrologic Budgets for Three Illinois Watersheds
64 J.G. A rnol d, P.M. Al l en 1 Journal of Hy drol ogy 176 (1996) 57-77
f low through each soil layer in the root zone. Do wnw ard flow occurs when field
capacity of a soil layer is exceeded if the layer below is not satu rated. The dow nward
flow rate is governed by the saturated conductivity of the soil layer. Upw ard flow may
occur when a lower layer exceeds field capacity. Movement from a lower layer to an
adjoining upper layer is regulated by the soil water to field capacity ratios of the twolayers. Percolation is also affected by soil tempe rature. If the tempera ture in a
particular layer is 0°C or below, no percolation is allowed from that layer.
5.4. Lateral subsurface flow
Lateral subsurface flow in the soil profile (O-2 m) is calculated simultaneously with
percolation. A nonlinear function of lateral flow travel time is used to simulate the
horizontal component of subsurface flow. The magnitudes of the vertical and
horizontal components are determined by a simultaneous solution of the two
governing equations.
5.5. Groundwaterflow
Groundwater flow contribution to total streamflow is simulated by creating a shallow
aquifer storage (Arno ld et al., 1993). Percolate from the bottom of the root zone is
recharge to the shallow aqu ifer. A recession constant, derived from daily streamflow
records, is used to lag flow from the aquifer to the stream. Other com ponents include
evaporation, pum ping withdraw als, and seepage to the deep aquifer.
5.6. Evapotranspiration
The model offers three options for estimating potential ET-Hargreaves
(Hargreaves and Sam ani, 1985), Priestley-Taylor (Priestley and Taylor, 1972) and
Penman -Monteith (Mon teith, 1965). The Penman-M onteith method was used in
this study and requires solar radiation, air temperature, wind speed, and relative
humidity as input. Daily values of wind speed, relative hum idity, and solar radiation
were generated from average monthly values.
The model computes evaporation from soils and plants separately, as described by
Ritchie (1 972). Potential soil water evaporation is estimated as a function of potential
ET and leaf area index (area of plant leaves relative to the soil surface area). Actual
soil water evaporation is estimated by using exponential functions of soil depth and
water content. Plant water evaporation is simulated as a linear function of potential
ET and leaf area index an d can be limited by soil water content. It is assumed that
30% of total plant uptake comes from the upper 10% of the root zone and roots can
compensate for water deficits in certain layers by using more water in layers with
adequate supplies.
5.7. Snow melt
The SWA T snow melt component is similar to that of the CRE AM S model (K nisel,
7/29/2019 Simulating Hydrologic Budgets for Three Illinois Watersheds
66 J.G. Arno ld , P.M . A l l en { Journal of Hy drol ogy 176 (1996) 57-77
evaporation, seepage from the reservoir bottom, and diversions and return flow.
There are currently three methods to estimate outflow. The first method simply
reads in measured outflow and allows the model to simulate the other com ponents
of the water balance. The second method is for small uncontrolled reservoirs, and
outflow occurs a t a specified release rate when volume exceeds the principal storage.Volume exceeding the emergency spillway is released within 1 day. For larger
manag ed reservoirs, a monthly target volume approach is used.
8. Model inputs and calibration
The SWAT hydrologic model requires input on soils (bulk density, av ailable water
capacity, sand , silt, clay, organic m atter, and saturated conductivity), land use (crop
and rotation), managem ent (tillage, irrigation, nutrient, and pesticide applications),
weather (daily precipitation, temperature, and solar radiation), channels (slope,
length, bankfull width and depth), and the shallow aquifer (specific yield, recession
constan t, and revap coefficient). Revap is defined as water that is extracted from the
shallow aquifer by deep roots or water that travels from the shallow aquifer to the soil
profile and is then lost to soil evaporation or plant root uptake (Arnold et al., 1993). A
complete list of inputs has been given by Arnold (1992).
The watersheds were subdivided to account for differences in soils and land use. No
channel (flood) routing was simulated; thus, it was assumed that all surface runoff
reached the basin outlet on the day of the runoff event. Each basin was subdivided
into three subbasins-one for pasture and woodland, one for a corn-soybean
rotation, and one for a soybean-corn rotation. As most of the cropland is in
corn-soybean rotation, half of the land is in corn one year and soybeans the next,so in any given year half of the land is in corn and the other half in soybeans.
Topographic and land use data is taken from Table 1. Upland prairie silt loams
were characterized by the Drummer soil series and timber silt loams by the Flanagan
soil series. Requ ired soil properties for each series were obtained from the Soils-5
data base (USD A Soil Conservation Service, 1992). Daily precipitation and tempera-
ture were collected from the following stations in Illinois: (1) Delan d for Goose
Creek, (2) Barry for Hadley Creek; (3) Minon k, Gridley, and Panola for Panther
Creek.
Inpu ts to the model are physically based (i.e. based on readily observed or
measured information). However, there is often considerable uncertainty in model
inputs owing to spatial variability, measurement error, etc. In this study model, inputs
were allowed to vary within a given realistic uncertainty range to calibrate to annual
measured values. For exam ple, Soils-5 properties are listed as ranges (i.e. available
water capacity might range from 0.11 to 0.13). As the exact values were unknow n, the
model was manually calibrated within the uncertainty ranges for annual streamllow,
and annual surface and groundwater contributions. The input variables used in
calibration were soil properties and the curve numb er. The curve number has
categories for good, fair, and poor hydrologic condition and was allowed to vary
within these ranges for calibration.
7/29/2019 Simulating Hydrologic Budgets for Three Illinois Watersheds
J.G. A rnol d, P.M . A l l en 1 Journal of Hy drol ogy 176 (1996) 57-77 67
9. Simulation results and discussion
Table 2 gives measured and predicted hydrologic budget com ponents for selected
years (reported by Schicht and Walton (1961)) for all three basins. Streamllow is
separated into surface runoff and groundwater flow, and ET is divided into surfaceand soil ET and groundwater ET. In SWAT, soil ET is the sum of soil evaporation
and plant root uptake from the crop root zone (approximately 2 m). Groundwater ET
is plant root uptake (trees and shrub s) from soil and rock layers below the crop root
zone or water loss that occurs as the water from the shallow aquifer re-enters the soil
zone. Groundwater recharge is the amount of water that percolates past the soil
Table 2
Com parison of hydrologic budgets for the Illinois basins
Measured
(mm)
Predicted
(mm)
Goose Creeek , 1957
Precipitation
Stream flow
Surface runoff
Groundwater flow
Evapotranspiration
Surface and soil ET
Groundwater ET
Groundwater recharge
Change in groundw ater storage
Underflow
Hadley Creek, 1957
Precipitation
Stream flow
Surface runoff
Groundwater flow
Evapotranspiration
Surface and soil ET
Groundwater ET
Groundwater recharge
Change in groundwater storage
Underflow
Pant her Creek, 1952
Precipitation
Stream flow
Surface runoff
Grotmdwater flow
Evapotranspiration
Surface and soil ET
Groundwater ET
Groundwater recharge
Change in groundwater storage
Underflow
944.4
240.8
144.3
96.5
617.2
535.9
81.3
264.2
+86.4
neg.
1009.1
353.8
305.8
48.0
626.9
604.5
22.4
98.8
+26.7
1.8
822.4
249.4
67.6
181.9
608.1
557.0
51.1
204.0
-28.9
neg.
253.5
145.1
121.2
603.0
521.6
81.4
210.0
+85.0
not simulated
366.4
300.5
65.9
634.6
612.9
21.7
88.8
+38.9
not simulated
239.0
85.6
153.4
594.9
556.1
38.8
191.1
-9.7
not simulated
7/29/2019 Simulating Hydrologic Budgets for Three Illinois Watersheds
J.G. Arnol d, P.M . A l i en / Journal of Hy drol ogy 176 (1996) 57-77 71
Table 3 shows the results of measured and simulated surface runoff, groundwater
flow, and total w ater yields for the three basins. It is important for simulation models
to produce frequency distributions that are similar to measured frequency distribu-
tions. Close agreement between means and standard deviations indicates that the
frequency distributions are similar. Generally, simulated values compare well withmeasured values considering that the basin characteristics utilized are relatively crude
for estimating model input p arameters (Table 1). A common criticism of simulation
models is that they do not simulate extremes w ell and thus underpredict standard
deviations. In this case, measured and predicted standard deviations compare well for
all flows. Regression line slopes and R2 values near unity also indicate a close relation-
ship between m easured and predicted yields. Statistics are valuable criteria but often a
graph gives considerable insight into the goodness-of-fit. Measured vs. predicted
monthly streamflow for Hadley Creek is plotted in Fig. 4. Regression lines and
lines-of-perfect-fit (1: 1) are plotted with the regression points.
Seasonal trends can easily be visualized by plotting measured and predicted
150
25
R-squared 0.94560 stddevmeas 34.35 .’
M eesM ean 17.20
25 50 75 100 125
M easured M onthly St~amflow(mm)
Fig. 4. Mea sured vs. predicted monthly total streamflow for Hadley Creek.
7/29/2019 Simulating Hydrologic Budgets for Three Illinois Watersheds
12 J.G. Ar nol d, P.M. Al l en / Journal of Hy drol ogy 176 (1996) 57-77
monthly values against time. Fig. 5 shows measured and predicted monthly stream-
flow for the Hadley Creek watershed during a 22 month period in 1954-1956 . These
graphs show if the model tends to over- or underpredict during certain seasons of the
year. Although the timing of modeled groundwater flow and volume parallels the
measured flow, there is some apparent deviation from the peak values (Fig. 6).Groundwater flow in the model is a function of three variables: (1) percolation or
recharge; (2) recession factor; (3) groundwater evaporation. Although it is assumed
that groundw ater discharge is a linear function of recharge, deviation is possible if the
aquifer behaves in a nonlinear manner as described by Rushton and Tomlinson
(1979), or if two aquifers are present in which tw o different recession factors would
have to be input as described by Riggs (1985). So, with the knowledge that a more
complex model m ay come a little closer to actual catchm ent processes, it appears
justifiable to adopt this simpler m odel ba sed on the results to date. A similar approach
has been advocated by Nathan and McM ahon (1990). Sensitivity analysis of these last
two factors will be analyzed in the future.
It is also important that the model sim ulates an nual variations in the hydrologic
components, although model inputs are static (not updated annually) during the
simulation. The Panther Creek watershed provides an excellent example of the
potential magnitude of the annual variability. In 1951, surface runoff w as greater
than groun dwater flow with a measured ratio of groundwater flow to total flow of
0.33 (Table 4 ). However in 1952, the ratio of groundwater flow to total flow was 0.73.
140
120
Bl w
Lo
18x60
B
40
20
r- Measured
---predicted
Fig. 5. Mea sured and predicted total flow by month for Hadley C reek.
7/29/2019 Simulating Hydrologic Budgets for Three Illinois Watersheds
However, it is important that the model realistically predict daily peak characteristics
over time and daily hydrograph recessions. Fig. 8 shows daily measured and predictedflows after a large runoff event (the largest in 1955 -1957) in the Goose Creek
watershed. Although this is only one large event, it does show that the model can
adequately simulate the daily hydrograph recession.
10. Sum mary and conclusions
A multicomponent water budget model (SWA T) has been tested for three
watersheds in Central Illinois. The model appears to be able to simulate all com-
ponents of the budget w ithin acceptable limits on both an annual and monthly timestep. Comp arison of the modeled results with measured water budgets allows com-
parison of the accuracy of the different components of the model. In this particular
case, it demonstrates that each component of the model gives reasonable output. This
fact should allow more realistic appraisal of various land use managem ent practices
on a basin-wide scale. It should also better pinpoint exactly how each alternative will
affect the water budget, thu s allowing for more innovative managem ent practices to
be tested a priori and their effects traced through each hydrologic component of the
watershed.
References
Arnold, J.G ., 19 92. Spatial scale variability in model development and parame terization. Ph.D.
Dissertation, Purdu e University, We st Lafayette, IN, 187 pp.
Arnold, J.G. and Williams, J.R., 1987 . Validation of SWR RB-simulator for water resources in rural
basins. J. Water Resour. Plann. M anage. ASCE, 113(2): 243-256.
Arnold, J.G., Williams, J.R., Nicks, A.D. and Sammons, N.B., 1990 . SW RRB-a Basin Scale Simulation
Model for Soil and Water Resources Management. Texas A&M University Press, College Station, TX,
255 pp.
Arnold, J.G ., Allen, P.M. and Bernhard t, G., 1993 . A comprehensive surface-groundwater flow model. J.
Hydrol., 142: 47-69.
Arnold, J.G., Srinivasan, R. and Muttiah, R.S., 19 94. Large-scale hydrologic modeling and assessment. In:
Effects of Human-Induced Changes on Hydrologic Systems. AWRA Annual Summer Symp., Jackson
Hole, WY. American Water Resources A ssociation Tech. Publ. Ser. TPS-94-3, AWRA, Bethesda, MD ,
pp. l-16.
Beasley, D.B., Huggins, L.F. and Monke, E.J., 198 0. ANSW ERS: a model for watershed planning. Trans.
ASAE, 23(4): 938-944.
Beven, K., 1989. Changin g ideas in hydrology-the case of physically based models. J. Hydrol., 105 : 157-
172.
Beven, K.J. and Kirkby, M.S., 1 979 . A physically based variable contributing area model of basin hydrol-
ogy. Hydrol. Sci. J., 24(l): 43-69.
Bultot, F., Dupriez, G.L. and G ellens, D., 1 990. Simulation of land use changes a nd impacts on water
balance-a case study for Belgium. J. Hydrol., 114: 327-348.
7/29/2019 Simulating Hydrologic Budgets for Three Illinois Watersheds
Nathan, R.J. and McMahon, T.A., 1990. Evaluation of automated techniques for base flow and recession
analyses. W ater Resour. Res., 26(7): 1465-14 73.
Norris, SE., 1963. Permeability of glacial till. US Geol. Surv. Prof. Pap., 450-E: 150-151.Peuntes, C . and Atkins, J.T., 1 989 . Simula tion of rainfall-runoff response in mined and unmined water-
sheds in coal areas of West Virginia. US Geol. S urv. Water Supply Pap., 1198, 48 pp.
Priestley, C.H .B. an d Taylor, R .J., 1 972 . Gn the assessment of surface heat flux and evaporation using
large-scale parameters. Mon. W eather Rev., 100: 81-92.
Refsgaard, J.C., Seth, S.M., Bathurst, J.C., Erlich, M., Storm, B ., Jorgensen, G.H. and Chandra, S., 1992.
Application of the SH E to catchments in India Part 1. Genera l results. J. Hydrol., 140 : l-23.
Riggs, H.C., 19 85. Streamflow characteristics. In: Developm ents in Wa ter Science No. 22. Elsevier, New
York, 249 pp.
Ritchie, J.T., 197 2. A model for predicting evaporation from a row crop with incomplete cover. Water
Resour. Res., S(5): 1204-121 3.
Rulan d, W.W ., Cherry, J.A. and Fe enstra, S., 1991. The depth of fractures and active ground water flow in
a clayey till plain Southwestern Ontario. Ground Water, 29(3): 405-417.
Rushton, K.R. and Tomlinson, L.M., 197 9. Possible mechanisms for leakag e between aquifers and rivers. J.
Hydrol., 40: 49-65.
Sangrey, D.A., Kingsley, 0. and Klaiber, J.A., 198 4. Predicting ground-w ater response to precipitation. J.
Geotech. Eng., 1 O(7): 957-975.
Schicht, R .J. and Walton, W.C ., 1 961 . Hydrologic budgets for three small watersheds in Illinois. Ill. State
Water Surv. Rep. Invest., 40, 40 pp.
USDA Soil Conservation Service, 1972. National Engineering Handbook. US Government Printing Gtlice,
J.G. Ar nol d, P.M . A l i en / Journal of Hy drol ogy 176 (1996) 57-77 71
USDA Soil Conservation Service, 1992. State Soil Geographic Database (STAT SGO ) Data Users’ Guide.
Publ. no. 1492, US Government Printing Oflice, Washington, DC,
William s, J.R., 196 9. Flood routing with variable travel time or variable storage coefficients. Trans. ASAE ,
12(l): 100-103.
Williams, J.R. and Hann, R.W., 1973. HYM O: problem oriented computer language for hydrologic
modeling. U SDA ARS-S-9, US Government Printing Office, 76 pp.Williams, J.R. and Nicks, A.D., 1982 . CREAM S hydrology model-option one. In: V.P. Singh (Editor),
Applied Modeling in Catchm ent Hydrology, Proc. Int. Symp. on Rainfall-Runoff Modeling, 18-21
May 1981, Mississippi State, MS . Water Resources Publications, Littleton, CO, pp. 69-86.
Winter, T.C., 198 1. Uncertainties in the estimating of water balanc es of lakes. Wa ter Re sour. Res., 1 7: 82-