Numerical simulation of ground-water flow in the central part of the western San Joaquin Valley, California

Abstract

The occurrence of selenium in agricultural drain water in the central part of the western San Joaquin Valley, California, has focused concern on strategies for managing shallow, saline ground water. To assess alternatives to agricultural drains, a three-dimensional, finite-difference numerical model of the regional groundwater flow system was developed. This report documents the mathematical approach used to model the flow system, the data base on which the model is based, and the methods used to calibrate the model. The 550-square-mile study area includes parts of the Panoche Creek alluvial fan and parts of the Little Panoche Creek and Cantua Creek alluvial fans. The model simulates transient flow in the semiconfined and confined zones above and below the Corcoran Clay Member of the Tulare Formation of Pleistocene age. The model incorporates areally distributed ground-water recharge, areally and vertically distributed pumping, regional-collector drains in the Wesdands Water District (operative from 1980 to 1985), on-farm drains in parts of the Panoche, Broadview, and Firebaugh Water Districts, and bare-soil evaporation (which occurs if the water table is within 7 feet of land surface). The model also incorporates texture-based estimates of hydraulic conductivity, where texture is defined as the fraction of coarse-grained deposits present in a given subsurface interval. The numerical model was developed using hydrologic data from 1972 to 1988. Most of the parameters incorporated into the model were evaluated independently of the model, including system geometry, the distribution of texture, the altitudes of the water table and potentiometric surface of the confined zone in 1972 (initial condition), the hydraulic conductivity of coarse-grained deposits derived from the Coast Ranges, the hydraulic conductivity of coarse-grained deposits derived from the Sierra Nevada, specific storage, recharge, pumping, and parameters needed to incorporate drains and bare-soil evaporation. Four parameters were calibration variables: the hydraulic conductivity of fine-grained deposits in the semiconfined zone, the hydraulic conductivity of the Corcoran Clay Member, specific yield, and the transmissivity of the confined zone. The model was calibrated in two phases. In the first phase, a steady-state model of the ground-water flow system in 1984 was used to constrain the relation between the hydraulic conductivity of fine-grained deposits in the semiconfined zone and the hydraulic conductivity of the Corcoran Clay Member, thus reducing the number of independent variables from four to three. In the second phase of calibration, the change in altitude of the water table from 1972 to 1984, the change in altitude of the potentiometric surface of the confined zone from 1972 to 1984, and the number of model cells subject to bare-soil evaporation from 1972 to 1988 were used to evaluate the remaining three variables. The calibrated model reproduces the average change in water-table altitude (1972-84) to within 0.4 foot (average measured change 11.5 feet) and the average change in confined zone head (1972- 84) to within 19 feet (average measured change 120 feet). The simulated time-series record of the total number of model cells subject to bare-soil evaporation (each cell is 1 mile square) is within the range of the measured data. The measured values are at a minimum in October and a maximum in July. The October values ranged from 103 in 1972 to 132 in 1984 (the drains were closed in 1985) to 151 in 1988. The July values ranged from 144 in 1973 to 198 in 1984, to 204 in 1988. The simulated values ranged from 103 in 1972 to 161 in 1984, to 208 in 1988.