Comparison of ground-water flow model particle-tracking results and isotopic data in the Mojave River ground-water basin, southern California, USA

Abstract

Flow-path and time-of-travel results for the Mojave River ground-water basin, southern California, calculated using the ground-water flow model modflow and particle-tracking model modpath were similar to flow path and time-of-travel interpretations derived from delta-deuterium and carbon-14 data. Model and isotopic data both show short flow paths and young ground-water ages throughout the floodplain aquifer along most the Mojave River. Longer flow paths and older ground-water ages as great as 10,000 years before present were measured and simulated in the floodplain aquifer near the Mojave Valley. Model and isotopic data also show movement of water between the floodplain and regional aquifer and subsequent discharge of water from the river to dry lakes in some areas. It was not possible to simulate the isotopic composition of ground-water in the regional aquifer away from the front of the San Gabriel and San Bernardino Mountains—because recharge in these areas does not occur under the present-day climatic conditions used for calibration of the model.

Introduction

The study area is the Mojave River ground-water basin in the western part of the Mojave Desert about 120 km east of Los Angeles, California (Fig. 1). The Mojave River ground-water basin consists of unconsolidated alluvial deposits defined by the bedrock-basin-fill contact. In general, ground-water within this area drains toward the Mojave River or topographically closed basins within the area. The area is arid with cold winters and hot, dry summers. Population of the area is growing rapidly, and ground-water pumped from the underlying aquifers is the only dependable source of water supply. In recent years, ground-water pumping increased with increased population, and water levels in the aquifers have declined (Stamos and Predmore, 1995, Mendez and Christensen, 1997, Stamos et al., 2001).

In 1990, the city of Barstow and the Southern California Water Company filed a complaint alleging that the cumulative ground-water pumping upstream from Barstow (Fig. 1) overdrafted the Mojave River ground-water basin (Mojave Basin Area Watermaster, 1996). In 1993, more than 200 parties stipulated to a ‘Physical Solution’ to the basin's water-supply problems. The purpose of the Physical Solution was (1) to ensure that downstream producers are not adversely affected by upstream water use, (2) to raise money to purchase supplemental water for the area, and (3) to encourage local water conservation (Stamos et al., 2001). To gain a better understanding of the relation between surface flows in the Mojave River and ground-water in the floodplain aquifer and the surrounding and underlying regional aquifer, the US Geological Survey (Stamos et al., 2001) updated a ground-water flow model of the area originally developed by Hardt (1971). The updated model was intended as a management tool to guide the implementation of the Physical Solution to the region's water-supply problems.

The updated model reflects geologic and hydrologic information collected as part of test drilling between 1992 and 1998 at 84 sites in the study area (Huff et al., 2003), and the model was rigorously calibrated to extensive historical and present-day hydrologic, pumpage, and water-level data sets (Stamos et al., 2001). Isotopic data collected in the study area were used to determine if the flow paths and travel times simulated by the ground-water flow model were reasonable.

The idea that ground-water flow models developed on the basis of traditional hydrologic data could be improved through the use of isotopic data has long been recognized (Campana and Simpson, 1984, Adar and Neuman, 1988). The stable isotopes of oxygen and hydrogen have been extensively used to identify sources of recharge, delineate aquifers, and delineate flow paths within aquifers (Verhagen et al., 1991, Adar and Leibundgut, 1995). Carbon-14 has received considerable attention from model developers because of the potential for estimating hydrologic properties over large areas on the basis of carbon-14 data (Phillips et al., 1989).

Phillips et al. (1989) used carbon-14 dating as a basis for estimating aquifer transmissivity as part of numerical model development in the San Juan Basin of New Mexico. Robertson (1992) used carbon-14 data to estimate ground-water ages, direction of ground-water flow, and ground-water velocity in the San Pedro basin in Arizona prior to flow-model development. Kalin (1994) used carbon-14 data to estimate ground-water ages to support numerical model development in the Tucson basin, Arizona. In recent years, more formal techniques have been developed to link model results to isotopic data. Zhu (2000) used a linked transport and flow model to adjust hydraulic-conductivity values in a model of the N aquifer in the Black Mesa basin, Arizona. CH₂M-Hill, Inc (2001) used a combination of numerical particle-tracking techniques and carbon-14 data to estimate aquifer porosity in the East Bay Plain, near San Francisco, California.

In most ground-water studies in the southwestern United States, application of combined modeling and isotopic techniques has been restricted to large regional flow systems where ground-water has relatively simple, well-defined flow paths. This approach has not been extensively applied to complex flow systems where there are interactions between regional ground-water flow systems and highly focused recharge along rivers.

The purpose of this study was to compare steady-state model particle-tracker results with interpretations of the source, movement, and age of ground-water estimated from isotopic data. The scope of the study included linking the particle-tracking program modpath (Pollock, 1994) to output from an existing US Geological Survey simulation of ground-water flow in the Mojave River basin (Stamos et al., 2001) to calculate the direction of particles of ground-water and their travel times. modpath is a three-dimensional particle-tracking program designed for use with output from ground-water flow simulations using modflow (McDonald and Harbaugh, 1988). Results from modpath were compared with delta deuterium tritium, and carbon-14 data collected in the study area.