Groundwater flow and development were characterized in four groundwater basins of the Death Valley regional flow system in Nevada and California with calibrated, groundwater-flow models. Natural groundwater discharges in the Furnace Creek, Lower Amargosa, and Saratoga Spring areas were defined and distributed consistently with a revised hydrogeologic framework. This simplified hydrogeologic framework was limited to four hydraulically unique, hydrogeologic units: (1) basin fill; (2) carbonate rocks; (3) volcanic rocks; and (4) low-permeability granitic and siliciclastic rocks. Hydrogeologic units and division of carbonate and volcanic rocks between shallow and deep were supported by results from 271 aquifer tests and specific-capacity estimates. Greater than 90 percent of field-estimated transmissivity occurred within 1,600 feet (ft) of the water table. Pumping in the study area from 1960 to 2010 averaged 46,000 acre-feet per year (acre-ft/yr), which is 80 percent of the predevelopment discharge. The central Amargosa Desert and Pahrump Valley were the two primary pumping centers and measurably affected water levels across 900 square miles in 2018.

Water levels in Devils Hole were a special focus because endangered Devils Hole pupfish (Cyprinodon diabolis) are affected by water-level declines. Pumping 42,100 acre-ft by Cappaert Enterprises, formerly Spring Meadows, Inc., caused a 2.3-ft water-level decline in Devils Hole, which temporarily reduced habitat of Devils Hole pupfish by 85 percent in 1972. If no pumping occurred, water levels in Devils Hole would have risen naturally about 1 ft between 1973 and 2018 from temporal variations in recharge. The 2.6-ft range of measured water-level changes in Devils Hole was simulated with a root-mean-square error of 0.2 ft during the 70-year period of record. Simulated water-level declines from pumping totaled 1.4 ft in 2018, with 25 and 34 percent attributed to pumping by Cappaert Enterprises and the central Amargosa Desert, respectively. Water levels in Devils Hole will decline at rates of 0.1–0.2 ft per decade if pumping from Ash Meadows groundwater basin and the central Amargosa Desert continue at current rates. Effects of future natural water-level fluctuations remain unknown.

Ash Meadows and Alkali Flat–Furnace Creek Ranch groundwater basins are hydraulically connected near well AD-4, about 5 miles south of the town of Amargosa Valley, Nevada. About 40 percent of the discharge from the Furnace Creek area is recharged in the Ash Meadows groundwater basin. Basin fill in the central Amargosa Desert hydraulically connects carbonate rocks east of well AD-4 with saturated carbonate rocks in the Funeral Range. About 7 percent of the 960,000 acre-ft pumped from Ash Meadows and Alkali Flat–Furnace Creek Ranch groundwater basins prior to 2019 was captured discharge from springs and phreatophytes. Greater than 40 percent of the 2,080,000 acre-ft pumped from Pahrump Valley between 1910 and 2019 was capture that primarily discharged from Bennetts and Manse Springs.

Simulated advective-flow distances and velocities from underground nuclear tests are within the range of advective transport calculations from tritium data and previous radionuclide transport investigations. Boundary conditions and flow rates from the regional model in this study are plausible for local-scale flow and radionuclide transport models. Simulated 165-year groundwater-flow paths do not extend into pumping areas and effects of regional pumping on advective transport are negligible.