WSP05388 - Laserfiche WebLink

Home Browse Search

-,. " ,---'-'-'----- ---,---------- -- , The method chosen to calibrate the model was to run the model from 1920 'to 1980 and to try to match the water-table contours in 1947 and 1978 and fit, water-level hydrographs with at least 20 years of record. The initial water, table for the model run was assumed to be the preirrigation steady-state con-, dition--before any pumping occurred. This water-table configuration was com-: puted by running the model with the computed boundary fluxes and no pumping" and solving for the position and flow of Beaver Creek. The simulation was, :run to steady-state so that the constant flux out of the north end of the, modeled area plus the streamflow in Beaver Creek equaled the boundary-flux: , inflo",. The model was run for the GO-year period starting at this initial, no- pumping condition. Estimating the distribution of pumping both spatially and temporally \."as a part of the calibration process. The model allo\'led pumping to occur only in those nodes where wells occur, as del ineated from the well- permit list. In the first calibration attempt, the total pumping was evenly 'distributed to each of these nodes. However, it is known that historic irri- gation was more concentrated at the southern end of the valley and increased ,in the northern end at a later time. Thus, a function was built into the model which weighted the distribution of the total pumping to the southern end prior to 1953 and toward the northern end after that date. Also programed into the model was the feature to decrease pumping at a node proportionately as the saturated thickness decreased from its initial value. 'Finally, a driving function for total annual pumpage from all wells in the entire valley was input to the model. This driving function was generated to correspond to the curve in figure 23 until 1955. An abrupt change in pumping was attributed to excessive drawdowns caused in part by the extreme drought conditions from 1953 to 1956. Since that time irrigation and pumping reportedly have de- cl ined steadily (Robert Samples, Water Commissioner, oral commun., 1978). The pumping rates computed by the model as a result of these different functions are shown on figure 24. The driving function assumed average precipitation every year to smooth out the vagaries in pumping estimates due to fluctuating precipitation (fig. 22). Other than the adjustments to the pumping functions and values, the only' 'significant change made to the input data during calibration was the distri-, bution of hydraulic conductivity. The 1947 water-table simulation closely matched actual conditions, but the 1978 water-table simulation was too high 'in the western part of the model where the sand dunes occur. Because there: was no evidence to suggest an unidentified \1ithdrawal in that area, and the: area in question was too far from the boundary for modifications in the :boundary flux to help, the only useful modification to the data was to in-; :crease the hydraulic-conductivity values in that area to about 400 ft/d. I . I , I I -' 47