USGS Home
Contact USGS
Search USGS

U.S. GEOLOGICAL SURVEY
Scientific Investigations Report 2008–5135

Back to Table of Contents

Evapotranspiration Measurements

An example time series of component-scale evapotranspiration fluxes (wolfberry, creosote bush, shadescale, burrobush, and bare soil) and net radiation (15-min intervals; Johnson and others, 2007) measured on May 2, 2005, is shown in figure 3A. Using plant crown measurements (Rc) from appendix A, bare-soil evaporation was subtracted from chamber measurements of ET, yielding transpiration for each species. Cloud cover on May 2, 2005, alternated between partly cloudy and clear. The mean root-zone (0.15–1.0-m depth) soil moisture measured with a neutron probe was 0.06 ± 0.01 m³/m³ and was twice the soil moisture (0.03 m³/m³; Johnson and others, 2007) measured in near surface soils (0–0.08-m depth) using a water-content reflectometer. Net radiation (R_n) and component-scale ET fluxes followed similar patterns reflecting the diurnal solar cycle.

In addition, substantial fluctuations in 15-min R_n measurements occurred during partly cloudy periods (fig. 3A). Sudden increasing and decreasing R_n values (caused by intermittent cloud cover) were reflected in burrobush, creosote bush, and shadscale chamber flux measurements (for example, R_n period beginning at 12:00 and 13:15 PST). Although, some variable fluctuations in R_n did not correspond to fluctuating ET fluxes (14:30–15:45), most variability affecting short (about 2 min) chamber measurements during intermittent cloud cover was captured by 15-min R_n measurements. Unlike the flux patterns for burrobush, creosote bush, and shadescale, wolfberry and bare-soil chamber fluxes were less sensitive to fluctuations in R_n(fig. 3A), indicating that measurements were likely collected when overhead skies were clear. In addition to R_n, variability in component-scale ET fluxes measured in May 2005 and during additional quarters (appendix B) occasionally corresponded with lesser factors affecting ET variability (such as air temperature, wind speed, and vapor pressure; data not shown). Erratic chamber fluctuations that did not correspond with measureable weather factors were likely from random noise in the chamber method (Stannard and Weltz, 2006).

Component-scale ET fluxes for vegetation typically were greater than for bare soil and varied by a factor of 1–3 between species on May 2, 2005 (fig. 3A) and by as much as a factor of 4 for all periods measured (appendix B). Creosote bush, which had an Rc that was, on average, a factor of 3.5 greater than shadescale (2003–05; appendixes A and B), exhibited the greatest component-scale ET fluxes during summer, autumn, and winter months (except Oct. 2004 and Jan. 2005 when vegetative fluxes were negligible). During the spring when all four plant species were measured, component-scale ET fluxes from wolfberry (which had a high leaf density and the largest Rc) typically were greatest, followed by burrobush and creosote bush (fig. 3A).

An example time-series of landscape-scale ET fluxes (combination of wolfberry, burrobush, creosote bush, shadescale, and bare soil) measured on May 2, 2005, is shown in figure 3B. Bare-soil evaporation dominated the total water flux at this scale, despite drier surface soil (twice as dry as the aggregate root zone). Expansive bare-soil cover across the landscape (Fc_s= 0.90) magnified lower bare-soil fluxes and reduced greater plant fluxes measured on a component scale. Plant transpiration from wolfberry and burrobush, which dominated ET on the component scale (fig. 3A), was minimized when extrapolated to the landscape scale (Fc = 0.01 in spring 2005; appendix C). At the landscape scale (Fc = 0.06), transpiration from creosote bush was greater than from wolfberry and burrobush despite having a lower component-scale transpiration than these plant species. On average, transpiration on May 2, 2005, accounted for about 40 percent of total landscape-scale ET. Variation in transpiration fluxes among species at the landscape-scale (a factor of 2–17; fig. 3B) was substantially greater than the variation for component-scale fluxes (a factor of 1–3; fig. 3A).

Transpiration accounted for about 30 percent of landscape-scale ET on average and bare-soil evaporation for about 70 percent over all periods measured. Landscape-scale fluxes measured for all periods are given in appendix C. For all measurement periods, bare-soil evaporation was the predominant contributor of moisture loss to the atmosphere as a result of its large landscape cover (Fc = 0.94 on average). The relative rank of transpiration contributions varied among creosote bush and the three drought deciduous species. Due to its abundance across the landscape (80 percent of vegetative cover), creosote bush accounted for about 90 percent of transpiration on average and ranged from about 50 to 100 percent of transpiration for all measurement periods.

Back to Table of Contents