Historical Trends in ET in the Western U.S. from Large Scale Hydrologic Simulations (Invited Presentation)

Projections of evapotranspiration (ET) for future climate conditions often point to increasing potential evapotranspiration (PET) due in part to wide spread warming simulated by global and regional climate models. These projections are uncertain, however, because other mitigating factors such as changes in solar radiation or humidity are generally not considered. Furthermore, in the western U.S., where evaporation is often constrained by water availability rather than energy limitations, relatively uncertain projections of precipitation change may dominate the response of actual evapotranspiration (AET). Finally, direct measurements of AET are rare, which presents formidable obstacles to conventional observational studies. Modeling of historical trends in AET presents an alternate approach, and also facilitates sensitivity analysis of the separate effects of temperature (T) and precipitation (P) on AET. Both T and P have increased historically in the west over most of the 20th century. Simulated trends in AET over the west as a whole from 1916-2003 are strongly influenced by increasing trends in P, which systematically increase AET in water limited areas. These effects are enhanced by increasing T in some areas. AET in cells with substantial snow accumulation in the current climate show a pronounced T influence on the timing of warm season AET, particularly in the latter half of the 20th century when T trends are large. AET increases from April-June, and decreases from July-Sept, and these effects are dominated by T and the effects on snow. The increasing trends in warm-season P observed for the 20th century for the Pacific Northwest stand out in contrast to the global climate model projections, which mostly point to drier summers. If in fact summers become wetter in the future, an increase in warm season ET and a reduction in the runoff ratio would be the expected outcomes.