Improving the understanding of the spatiotemporal variability of hydrometeorology across the Sierra Nevada using a novel remote sensing reanalysis approach
While large populations worldwide depend on water derived from the seasonal snowpack, a detailed picture of the spatiotemporal variability of snowfall and snow water equivalent (SWE) across high-elevation mountain ranges remains a knowledge gap in understanding the hydrologic cycle. Previous studies relying on point-scale in situ measurements often yielded spatially incomplete characterizations of montane snow accumulation processes (e.g. orographic snowfall). These limitations were overcome in this dissertation by using a novel, high-resolution distributed snow reanalysis over Sierra Nevada, USA from 1985-2015. Across the 20 basins examined, over 50% of the integrated cumulative snowfall (CS) accumulated rapidly in less than or equal to six days or three snowstorms, on average, and the largest snowstorms yielded an average 27% of the seasonal CS. Results suggest that misrepresentation of a single snowstorm could lead to significant biases in CS. The hydroclimatology of the Sierra Nevada was found to be driven by extremes as manifested in the high inter-annual variability of its seasonally-integrated CS, 4.4-41.3 km3, over the record. Seasonal orographic CS gradients were shown to be highly variable, ranging from over 15 cm SWE/100 m to under 1 cm/100 m. Hence, the seasonal/elevational distribution of water storage can greatly vary with the western Sierra Nevada experiencing about twice as much orographic enhancement during wet years as in dry years. Among the largest winter snowstorms, moisture-rich atmospheric rivers (ARs) significantly contribute to the seasonal CS. Using both satellite-based integrated water vapor and reanalysis-based integrated vapor transport methods, AR-derived CS was found to be more orographically enhanced than non-AR derived CS above ~2200 m in the western Sierra Nevada; however, the understanding of the AR-derived CS distribution and enhancement is tightly coupled to the AR detection method applied. ARs were shown to contribute from ~33-56% of the seasonal CS, on average from 1998-2015, depending on the AR detection method utilized. Overall, more robust characterizations of the spatiotemporal variability and climatology of snowfall distributions, atmospheric drivers of snowfall, and accumulation rates than previously existed were provided. The resulting insight could be used for improving water resources management and hydrologic analysis as well as evaluating climate model snowpack estimates and improving their representation of subgrid snow processes (e.g. orographic snowfall).