|Abstract or Summary
- There is reasonable expectation that climate warming will accelerate the hydrologic cycle, resulting in greater evapotranspiration (ET) and reduced groundwater recharge (R) (or stream flow). Though qualitatively intuitive, quantifying these potential shifts in water budget partitioning is a contemporary challenge in hydrology, because the linkage between ET and R is strongly mediated by rainfall periodicity, vegetation, and soil moisture dynamics. This challenge has been accentuated by the Intergovernmental Panel on Climate Change, and is now being addressed primarily through model simulations, which have outpaced experimental efforts due to the overwhelming challenge of measuring the entire water budget in systems with known boundary conditions, and under forecasted alterations in surface air temperatures. We present new data from a controlled-chamber experiment that examines the combined responses of ET, soil moisture (θ), and R to imposed temperature alterations in a Willamette Valley grassland ecosystem. Temperature treatments include an average increase of 3.5˚C, applied both symmetrically throughout the day, and asymmetrically such that daily minimum temperature is 5˚C greater than ambient and daily maximum temperature is 2˚C greater than ambient. Given the Mediterranean climate of this region, where rainfall and ET occur largely out of phase, we hypothesized that increasing surface air temperatures would accelerate and enhance plant growth and ET during the spring season, abbreviating the period when R occurs. Counter-intuitively, over a three year period we observed only modest enhancements of ET during the spring period under 3.5˚C warming. The most salient effect was observed during the 2008 water year, when average-cumulative ET was 26-44% and 32-41% greater on April 30 under symmetric and asymmetric warming scenarios, respectively, than under ambient climate conditions. Corresponding acceleration of θ depletion was also observed, although there was no immediate effect on R. The cumulative effect of accelerated ET and θ depletion on R only became evident during late spring rain events (May-June), when average R generated under ambient climate conditions was 160-190% greater than under either warming scenario, although these events accounted for less than 6% of total R in any year. Collectively, the results demonstrate that annual water budget partitioning in Willamette Valley grasslands is unaltered by a 3.5˚C increase in average air temperature. The temperature-driven enhancement of ET is modest and inconsequential for R during the short inter-storm time intervals typical during the spring. The contrasting seasonality of rainfall (and resulting R) and ET is the dominant climate feature determining annual water budget partitioning in the Willamette, and is here shown to effectively ameliorate the potential impact of a 3.5˚C warming signal on the annual water budget.