Development of a Groundwater Flow Model of Pahrump Valley, Nye County, Nevada and Inyo County, California for Basin-Scale Water Resource Management

Abstract

Pahrump Valley lies approximately 70 kilometers west of Las Vegas straddling both Nye County, NV and Inyo County, CA. As Nevada's most heavily allocated groundwater basin, Pahrump Valley has seen its population increase exponentially over the past 30 years (pop. ~39,000). The sole source of water within the valley is the underlying basin fill and carbonate aquifers, which provide water for irrigation, domestic, commercial and public uses. Previous studies estimate that the sustainable basin yield of the valley is 64,166 m3/d (19,000 ac-ft) (Harrill, 1986). Data obtained from the Nevada Division of Water Resources (NVDWR) demonstrate that annual pumping has continuously exceeded this sustainable basin yield estimate for over 50 years. Managing these vital groundwater resources requires determining the future impacts of both current and projected pumping rates. A three-dimensional, basin-scale groundwater flow model of Pahrump Valley was constructed to serve as a tool for evaluating alternative management scenarios. The model incorporates geologic data from the USGS Death Valley Regional Flow System Model, pumpage inventories from 1913 to 2003 for over 10,000 domestic, irrigation, and public wells, and water levels from 143 monitoring wells. Recharge from the Spring Mountains is based on an estimate of 87, 806 m3/d (26,000 ac-ft/yr) from Harrill (1986), and was implemented in the model as a constant flux boundary. The model simulates an annual evapotranspiration rate of 42,214 m3/d (12,500 ac-ft/yr) based on phreatophyte density on the valley floor and plant-specific water use coefficients. The model was calibrated in transient mode according to water level data from 1946 and 2003. The calibration process consisted of both manual calibration to northeast-southwest trending transects across the valley and automatic calibration to a global objective function using the inverse parameter estimation program, PEST (Doherty, 2004). Calibration of the model involved varying hydraulic parameters until an acceptable match between simulated head and observed head levels was achieved.Improving the accuracy of the model required simplifying the geology within the model domain, creating specified hydraulic conductivity (K) zones in the underlying carbonate aquifer, incorporating hydraulic conductivity decay with depth, optimizing recharge contributions of watersheds in the Spring Mountains, and using a pilot points method to further optimize simulated head levels in the basin fill. The calibrated model adequately simulates head measurements (RMSE= 9.7 m, relative error of 5%), and is precise at estimating drawdown at most locations. Suggested future work includes converting the model to an unconfined flow solution to enhance the reliability of simulated drawdown predictions. Once this is accomplished, the model can be used to predict future impacts from current and future pumping scenarios, optimize well placement and extraction rates within the valley, and serve as a numerical framework for determining future land subsidence.