This dissertation explores ecohydrologic interactions in western U.S. catchments using a process-based model. My research focuses on understanding how model estimates of two key components of the hydrologic budget, evapotranspiration and streamflow, are influenced by soil and vegetation physiological characteristics in three watersheds. These watersheds are located in the Oregon Cascades, California Sierra Nevada, and Colorado Rocky Mountains. Water availability in these systems is driven on a first order by annual precipitation. The majority of their precipitation is received in the winter and their forests are generally water-limited in the summer. However, they differ in the magnitude of annual precipitation received, the fraction of winter precipitation received as snow, and their seasonal energy demands. The response of these ecosystems to inter-annual climate variation is also a function of soil storage and ecophysiological characteristics.

I will present research motivated by three research questions. How do soil characteristics and climate interact to influence forest water availability? How do uncertainties in forest ecophysiology, carbon allocation strategy, and their interaction effect mature forest carbon and streamflow estimates? Finally, how do climate and carbon allocation strategy influence the rate of forest growth and streamflow recovery following disturbance? I use a physically-based model, the Regional Hydro-Ecologic Simulation System, to address these questions. Results indicate that the influence of soil storage on evapotranspiration's sensitivity to climate drivers varies across sites. In the Sierra Nevada and Cascades, low soil storage increases the sensitivity of annual ET to climate drivers. Evapotranspiration in Colorado, which is water-limited but has a summer monsoonal pulse, is not sensitive to changes in soil storage.

Estimates of forest carbon sequestration differ significantly between three carbon allocation strategies in mature (100-300) forests. Biomass estimates for leaf and fine root pools were strongly sensitive to allocation strategy and ecophysiological characteristics in the Sierra Nevada watershed. Streamflow estimates in this drier watershed are also more sensitive to vegetation ecophysiology. I show that the effect of allocation strategy effects estimates of recovery in forest LAI and streamflow more than climate variability at al three sites. This research contributes to the coupled ecosystem modeling community's understanding of key processes that influence our ability to predict water resources.