Evaluating evapotranspiration for six sites in Benton, Spokane, and Yakima counties, Washington, May 1990 to September 1992

This report evaluates evapotranspiration for six sites in Benton, Spokane, and Yakima Counties, Washington. Three sites were located on the Arid Lands Ecology Reserve in Benton County: one at a full-canopy grassland in Snively Basin (Snively Basin site), one at a sparse-canopy grassland adjacent to two weighing lysimeters (grass lysimeter site), and one at a sagebrush grassland adjacent to two weighing lysimeters (sage lysimeter site). Two sites were located on the Turnbull National Wildlife Refuge in Spokane County: one at a full-canopy grassland in a meadow (Turnbull meadow site), the other a full-canopy grassland near a marsh (Turnbull marsh site). The last site was located in a sagebrush grassland in the Black Rock Valley in Yakima County (Black Rock Valley site). The periods of study at the six sites varied, ranging from 5 months at the Black Rock Valley site to more than 2 years at the Snively Basin, grass lysimeter, and sage lysimeter sites. The periods of study were May 1990 to September 1992 for the Snively Basin, grass lysimeter, and sage lysimeter sites; May 1991 to September 1992 for the Turnbull meadow site; May 1991 to April 1992 for the Turnbull marsh site; and March to September 1992 for the Black Rock Valley site. Evapotranspiration and energy-budget fluxes were estimated for the Snively Basin site, the Turnbull meadow site, and the Black Rock Valley site using the Bowen-ratio and Penman-Monteith methods. Daily evapotranspiration for the Snively Basin site was also estimated using a deep-percolation model for the Columbia Basin. The Bowen-ratio method and weighing lysimeters were used at the grass and sage lysimeter sites. The Penman-Monteith method was used at the Turnbull marsh site. Daily evapotranspiration at the sites ranged from under 0.2 millimeter during very dry or cold periods to over 4\x11millimeters after heavy rainfall or during periods of peak transpiration. At all sites, peak evapotranspiration occurred in spring, coinciding with plant growth, and the lowest evapotranspiration occurred in late summer and winter, coinciding with plant dormancy and extremely hot or cold temperatures. Water budgets for the Snively Basin, grass lysimeter, and sage lysimeter sites were based on estimates of precipitation, evapotranspiration, and surface runoff. Surface runoff was estimated at zero for all sites. For the Snively Basin site, 1991 and 1992 water budgets using Penman-Monteith evapotranspiration estimates agreed within 1 percent of the annual budgets computed using deep-percolation model estimates; daily estimates of evapotranspiration by the two methods varied considerably, however. For the Snively Basin site, 100 percent of the precipitation became ET in 1991, and in 1992, about 91 percent of the precipitation became evapotranspiration. Water budgets based on weighing lysimeter data at the grass and sage lysimeter sites agreed within 1 percent of each other for 1991 and within 5 percent of each other for 1992. For 1991, 10 percent of the precipitation became ET at both lysimeter sites. For 1992, 94 to 98 percent of the precipitation became ET at the grass lysimeter site while 98 to 9 percent of the precipitation became ET at the sage lysimeter site. Though there were uncertainties in the methods used, recharge estimates for the Snively Basin and grass and sage lysimeter sites were of the same order of magnitude. The Penman-Monteith method (which incorporated Bowen-ratio measurements), deep- percolation model, and weighing lysimeters indicated that no recharge to subsurface systems (soil profile, unsaturated zone, and ground water) occurred in 1991 and that, in 1992, recharge to subsurface systems was probably less than 10 percent of the annual precipitation at the Snively Basin and grass and sage lysimeter sites.