WSP05388 - Laserfiche WebLink

Home Browse Search

I Calibration of the Beaver Creek Valley Model 'J Calibration of a digital ground-water model consists of entering known ~nd estimated input data, running the model for some historic period of Istress, comparing modeled water levels to measured water levels, and then ad- ljusting certain input data until the modeled and measured water levels ,com- :pare within some acceptable limit of error. Typically, the boundary fluxes 'and historic "tresses are known and most adjustments of input data involve it he storage coefficient and the transmissivity. In this study, however, due ;to a lack of data, boundary fluxes and historic stresses also were adjusted Id' l'b' I ur,ng ca I ration. i I Boundary fluxes were the first variables estimated for the Beaver Creek :valley model. These boundary fluxes are the only source of "'ater into the model, representing the flow of water through the alluvial valley from the 'south and from the sand dune areas to the east and west. The recharge to the alluvial aquifer from the occasional flooding of Beaver Creek and recharge :from intense precipi tation on the val ley floor were ignored for the average 'conditions being modeled. To determine the order of magnitude for estimated :inflow, the possible recharge from the sand hill area was computed. If 3 in. per year of precipitation were recharged--2.5 in. per year of recharge was computed for Frenchman Creek basin, Nebraska, less than 100 mi east of this site, where precipitation is about 75 percent greater but only one-third of 'the basin is sand dunes (Lappala, 1978)--over the approximately 200 mi2 of 'sand dunes, a total of about 44 ft3/s would be avai lable to enter the tl"O illluvial valleys. I I The steady-state flux across the boundaries and into the stream channel was computed for the water-table configurations in 1947 and 1978. A uniform hydraulic-conductivity value was used for these computations. The initial simulation, using a hydraulic conductivity of 500 ft/d, resulted in a bound- ary flux and streamflow values that were much too large. Although no dis- charge data were available for Beaver Creek, 25 ft3/s was estimated to be the maximum possible average annual flow in 1947 and there was no flow in 1978. Pumpage in 1947 was reported to total about 12,000 acre-ft, which converts to about 16 ft3/s, and in 1978 pumping was assumed to be somewhat greater. On the basis of these values, the estimate of hydraulic conductivity was ad- Justed to about 90 ft/d. Using the water-table values of,1947, the flux at each of the individual nodes along the boundary was computed; the sum was 21 ft3/s. The values estimated using the 1978 water-table configuration were thought to be more accurate. This is because there was no flow in Beaver Creek, so the hydrologic system is less complex and the hydraulic-head data are affected by fewer external conditions. Also, the change in water levels was smaller in 1978 than in 1947, so the hydrologic system was closer to a steady-state condition. The sum of the fluxes using the 1978 water-table values was 23 ft3/s. ! i L_ " 46 -----.:-~.