With increasing population growth and land-use change, urban communities in the desert Southwest are progressively looking toward remote basins to supplement existing water supplies. Pending applications by Churchill County for groundwater appropriations from Dixie Valley, Nevada, a primarily undeveloped basin east of the Carson Desert, have prompted a reevaluation of the quantity of naturally discharging groundwater. The objective of this study was to develop a revised, independent estimate of groundwater discharge by evapotranspiration (ET_g) from Dixie Valley using a combination of eddy-covariance evapotranspiration (ET) measurements and multispectral satellite imagery. Mean annual ET_g was estimated during water years 2010 and 2011 at four eddy-covariance sites. Two sites were in phreatophytic shrubland dominated by greasewood, and two sites were on a playa. Estimates of total ET and ET_g were supported with vegetation cover mapping, soil physics considerations, water‑level measurements from wells, and isotopic water sourcing analyses to allow partitioning of ET_g into evaporation and transpiration components. Site-based ET_g estimates were scaled to the basin level by combining remotely sensed imagery with field reconnaissance. Enhanced vegetation index and brightness temperature data were compared with mapped vegetation cover to partition Dixie Valley into five discharging ET units and compute basin-scale ET_g. Evapotranspiration units were defined within a delineated groundwater discharge area and were partitioned as (1) playa lake, (2) playa, (3) sparse shrubland, (4) moderate-to-dense shrubland, and (5) grassland.

Groundwater ET is influenced primarily by phreatophytic vegetative cover, salinity of soil and groundwater within the playa, depth to groundwater, solar radiation, and air temperature. The annual groundwater contribution to site‑scale ET ranged from 24 to 61 percent of total ET at vegetated sites and 4 to 15 percent of total ET at playa sites. Mean annual ET_g from vegetated sites ranged from 53 millimeters (mm) (0.17 foot [ft], 7.3 percent vegetative cover) to 225 mm (0.74 ft, 24.8 percent vegetative cover). Cumulative liquid‑water fluxes in the unsaturated zone indicate that ET_g at vegetated sites was influenced primarily by plant transpiration. Binary mixing analyses of oxygen-18 isotopes in groundwater and shallow soil water indicate that plants predominantly use groundwater throughout the year. Groundwater fractions in greasewood stem water varied seasonally and ranged from 0.63 to 1.0. Mean annual playa ET_g ranged from about 11 mm (0.04 ft) at the inner playa site (near-surface volumetric water content of 37–53 percent) to about 20 mm (0.07 ft) at the outer playa site located within 2 kilometers of the playa edge (near-surface volumetric water content of 25–38 percent), but playa ET_g estimates were within the probable error (plus or minus [±] 20–23 mm; 0.06–0.08 ft). Varying playa ET_g was influenced predominantly by salinity rather than depth to groundwater. Osmotic resistance and physical impediments to ET (such as surface salt crusts and salt precipitate in the soil pore space) increased with increasing salinity toward the playa center, whereas vapor pressure decreased.

Mean annual basin-scale ET_g totaled about 28 million cubic meters (Mm³) (23,000 acre-feet [acre-ft]), and represents the sum of ET_g from all ET units. Annual groundwater ET from vegetated areas totaled about 26 Mm³ (21,000 acre-ft), and was dominated by the moderate-to-dense shrubland ET unit (54 percent), followed by sparse shrubland (37 percent) and grassland (9 percent) ET units. Senesced grasses observed in the northern most areas of the moderate-to-dense ET unit likely confounded the vegetation index and led to an overestimate of ET_g for this ET unit. Therefore, mean annual ET_g for moderate-to-dense shrubland presented here is likely an upper bound. Annual groundwater ET from the playa ET unit was 2.2 Mm³ (1,800 acre-ft), whereas groundwater ET from the playa lake ET unit was 0–0.1 Mm³ (0–100 acre-ft). Oxygen-18 and deuterium data indicate discharge from the playa center predominantly represents removal of local precipitation-derived recharge. The playa lake estimate, therefore, is considered an upper bound. Mean annual ET_g estimates for Dixie Valley are assumed to represent the pre‑development, long-term ET_g rates within the study area.