Table 6. Summary of Lotus Simulation of Cypress Creek
Above San Antonio: 1979.
A B C D E F G H I J K
Precipitation
Cypress Saint Weighted
Creek Leo Mean Evap
inches inches inches inches in
TOTAL 61.06 66.89 63.975 60.07
MEAN
MAX
MIN
STANDARD DEVIATION
KEY ASSUMPTIONS:
1.ET=(ETMAXETRATE*(GRDELWTEL)) E
Cale
ET Streamflow
ches cfs inches
45.1 10051 7.87
0.12 27.54
0.42 205.7
0 0
3.07 44.51
GRDEL
ETMAX
ETRATE
2.DRAINAGE AREA, SQUARE MILES =
3.STORAGE COEF. =
4.DISCHARGEHEAD RELATIONSHIP = K*H^B
5.ELEVATION FOR WHICH BASE FLOW = 0, FEET
6.LEAKANCE, IN/AR
7.WEIGHTS OF RAIN GAGES CYPRESS CRK
ST LEO
CONTINUITY CHECK: dS/dT = (PETQL)
dS/dT=
CALC
MEAS
63.975 IN.
45.169 IN.
7.886 IN.
8.200 IN.
2.719 IN.
2.611 IN.
PERCENT CONTRIBUTIONS
INFLOWS
P= 100%
OUTFLOWS
ET= 73.74%
Q= 12.87%
L= 13.39%
25
$'5 026
Meas
Flow
cfs
10140
27.78
500
0
58.47
Meas
Head
inches
68.48
72.71
65.64
1.77
Cale
Head
inches
68.39
74.89
62.01
2.84
74
1
0.04
47.4
0.085
3.3
2
67
8.2
0.5
0.5
located approximately 60 miles from the study area. In HSPF, the
mean monthly evaporation was used. Evapotranspiration is
estimated as a function of evaporation and water table elevation,
i.e.,
ET = (ETmaxn*(HgrdHwt))*E .................... (13)
where ET = evapotranspiration, inches of water, ETmax = maximum
ET rate, inches per day, n = reduction in ET rate per unit
decrease in head, Hgrd = ground elevation, feet, Hwt = water
table elevation, feet, and E = pan evaporation rate, inches per
day. ETmax, n, and Hgrd are used as calibration parameters.
The discharge in Cypress Creek is approximated by fitting a
power function of the form:
Q = a*Hb ......................................... (14)
where Q = estimated flow in cfs, H = estimated water table eleva
tion at well 4, feet, and a,b = parameters. The calculated flow
in cfs is converted to inches/day over the catchment. Pumpage is
represented as a change in leakance. The model was calibrated
with water level data from well 4, located just north of the
Cypress Creek Wellfield.
The model estimates the daily flow and stage for the year.
The power of the spreadsheet approach is apparent during the
calibration. The key assumptions and calibration parameters are
shown at the bottom of the table. All of the equations in the
table are expressed in terms of the calibration parameters.
Thus, the new result is calculated immediately once the revised
parameter estimates are inserted. The final estimates are shown
26
9a17
in the table. All of these values fall within the expected
range. The most sensitive assumption is the weighting on the
rain gages.
The results are quite good. The estimated mean flow of 27.5
cfs agrees closely with the measured flow of 27.8 cfs.
Similarly, the estimated mean stage of 68.4 feet is very close to
the measured stage of 68.5 feet, and the continuity check is very
good. The simulated vs. measured flows and stages are shown in
Figures 4 and 5 respectively. The fit is very good. The most
important loss term is ET. It is over three times as large as
the runoff, the next largest item in the water budget. The model
was calibrated to fit well for the normal range of flows. Thus,
simulated high flows may differ significantly from the measured
high flows.
This model is able to track the movement of the water table
below the streambed; this was a limitation of HSPF. Of course,
this model can be criticized for not explicitly tracking some
parts of the hydrologic cycle. The potentiometric head is not
analyzed. However, the analysis of the well data indicated
clearly that the potentiometric head is highly correlated with
the water table. Also, the unsaturated zone is not analyzed
explicitly. However, since no good information about the
relationship between the unsaturated zone and runoff and ET
exists, its effect can be included in the other terms of the
water budget.
Overall, the results for both the flows and heads at San
27
5 C9 T
