Analyzing Heavy to Extreme Atmospheric River Precipitation Events in the Lee of the Northern Sierra Nevada Using Self-Organizing Maps

Leeside (spillover) precipitation related to atmospheric rivers (ARs) is crucial to the water supply across the eastern Sierra Nevada, but heavy to extreme precipitation events can also lead to adverse impacts including river and lake flooding, debris flows, downed power lines and trees, and significant travel delays due to heavy snow. Few studies have focused on understanding the synoptic and mesoscale factors that influence spillover precipitation in the lee of mountain ranges, particularly the Sierra Nevada. It is hypothesized that the coupling of strong mid- to upper-level momentum directed perpendicular to the Sierra Nevada crest, divergent flow accompanying jet streak exit region circulations, ample low- to mid-level instability, preexisting moist convection upwind of the Sierra Nevada, and substantial low- to mid-level moisture are the key drivers of spillover precipitation during AR events. Using hourly integrated vapor transport estimates from the MERRA-2 reanalysis during the 1980–2011 cool seasons (October–March), 278 AR events were identified that affected the Upper and Lower Truckee River watersheds of the northern Sierra Nevada. To assess the spatial distribution of precipitation within the watersheds during each AR event, daily 6 km resolution precipitation estimates from Livneh (2013) were used, with a special focus on heavy to extreme events (85th–99th percentile). Using self-organizing maps (SOMs) for analysis, we explore the relationships between heavy to extreme spillover precipitation events, the antecedent synoptic and mesoscale conditions, and local vertical profiles of wind, temperature, and moisture from upper-air observations. The goal of this study is to give weather forecasters a reliable mapping tool that helps them recognize synoptic and mesoscale patterns associated with ARs capable of producing significant impacts in the lee of the Sierra Nevada. This tool would be derived from existing medium to long range numerical weather prediction model guidance and could help satisfy the increased demand for greater lead time by water managers, emergency managers, media, and the public.