- This event has passed.
William Rudisill Dissertation Defense
June 6 @ 1:00 pm - 3:00 pm MDT
From the River Bed to the Cloud Tops: Evaluation and Applications of Regional Climate Modeling in Mountain Watersheds
Mountain ranges are vital “water towers” of the world and are uniquely threatened by anthropogenic climate change. At the same time, the paucity of observing networks limits our understanding of hydrometeorological processes in these water-resource critical regions, including the Western United States. In the past decade, non-hydrostatic, convection-permitting (~1-4km horizontal resolution) regional climate models (RCMs) have emerged as a promising tool for both reconstructing regional scale mountain hydroclimates and forecasting the impacts of climate and landuse/landcover perturbations on watersheds and water resources. Still, many challenges remain. To date, computational limitations have generally precluded many RCM studies to a handful of individual years, limiting the characterization of model uncertainties/biases and thus the interpretation of model outputs. Moreover, validating spatial precipitation fields from RCMs remains a challenge, as gridded precipitation datasets are uncertain in locations far from observing stations. Consequently, further validations of RCMs require leveraging diverse or indirect sources of hydrologic information. In this dissertation, I develop three studies to meet these challenges. First, I examine the fidelity of coupled hydrologic-model/RCM for reconstructing streamflow in four water resource significant, snow-dominated basins in the Boise River basin, and explore soil-moisture atmosphere feedbacks. In the second, I develop a long-term RCM simulation (1987-2020) in the Upper Colorado River basin and evaluate precipitation fields using a novel Bayesian inference strategy that leverages streamflow observations, parsimonious hydrologic model structures, and airborne LIDAR measurements of snow surfaces. Finally, I examine orographic precipitation sensitivities to three different cloud-microphysical parameterization schemes within the RCM by leveraging the aforementioned LIDAR datasets. Together, the results from this dissertation demonstrate the utility of multi-decadal, convection-permitting RCMs for interrogating mountain hydroclimate, as evidenced by high correlations/low biases between RCM fields and NRCS Snotel observations, statistically reconstructed precipitation, observed runoff, and snow remote sensing products. Still, key process uncertainties are found. This dissertation underscores the multi-disciplinary methods required to evaluate hydrologic fluxes in mountain regions, provides essential benchmarking metrics for applying RCM outputs to water-resource applications, and builds foundation for future studies of climatic perturbations on the water cycle, both in the Western US and in mountain environments globally.
Advisor: Alejandro Flores
Co-Advisors: HP Marshall, James McNamara, Rosemary Carroll, Alejandro Flores