UC San Diego

Western U.S. residential, municipal, and agricultural water supplies depend on cool season storms. The majority of these storms are landfalling atmospheric rivers which travel in the troposphere from the tropical Pacific Ocean. Sudden changes in mesoscale features or storm ingredients can dramatically alter impacts affecting hydrologic processes and communities. Although storms provide vital water supplies, they also induce deleterious flooding, landslide, wind, and snow-related disasters. Precipitation phase and the type and severity of storm impacts vary depending on the altitude at which frozen hydrometeors melt relative to ground elevations. Thus, this rain-snow transition altitude, or atmospheric snow level, and its intrastorm vertical variations are key in determining storm benefits and hazards.

While there has been a considerable amount of recent progress in snow level research, there remain challenges regarding substantial changes in snow level observed during high-impact storms. The primary goal of this dissertation, therefore, is to provide a robust methodology to define, catalogue, and describe extreme intrastorm changes in snow levels during California storms. This work considers 10 vertically-oriented radar locations over six recent cool seasons, defining an extreme snow level change as a one-hour change with a magnitude of at least 400 meters. The provided dissertation identifies 134 and 113 extreme rises and falls, respectively, finding strong associations with periods of enhanced water vapor transport including atmospheric rivers. Additionally, this research designs data quality filters that reduce spurious snow level changes. This work also defines and identifies distinct events, termed semicontinuous snow level events, in which to compute hourly changes.

Further, this dissertation provides statistical, spatial, and temporal descriptions of semicontinuous snow level events, intrastorm snow level changes, and extreme changes. Results indicate an elevated number of extremes existed at northern sites, during December-March, and during anomalously wet cool seasons. Key findings reveal, for 60-100% of extreme changes at each radar, an atmospheric river occurred within the six hours preceding or following each extreme. Ultimately, this dissertation provides methodologies, findings, and a catalogue of extremes beneficial to future investigations of intrastorm snow level changes, atmospheric mechanisms controlling these variations, and implications for hazard prediction and preparedness.