By

Musselman, Keith N. 1 ; Henn, Brian 2 ; Lestak, Leanne 3 ; Ralph, Marty 4 ;Molotch, Noah P 5

1 INSTAAR
2 Scripps, UC San Diego
3 INSTAAR
4 Scripps, UC San Diego
5 INSTAAR

The spillway failures at Oroville Dam in February 2017 were a major public safety crisis, resulting in mass evacuations, risk of catastrophic inundation, and repair costs nearing $1B. While the spillway failures have received significant attention, the storms and runoff events that filled Lake Oroville and enhanced the crisis have not been explored. Over February 10-12 as the spilllways failed, Feather River inflow to Lake Oroville reached 185,000 cfs, their highest value since the January 1997 flood of record. The high flows resulted from a sequence of strong atmospheric river (AR) events making landfall in Northern California over February 3-12. These ARs maintained intense onshore moisture transport from the Pacific concurrent with high rain-snow levels, resulting in multi-day precipitation totals over the Feather River watershed exceeding 500 mm, falling as rain upon a deep antecedent snowpack from colder ARs in January. We evaluate the degree to which the duration of AR conditions and their rain-snow levels were historically anomalous. Using in situ snow observations of the rain-snow level and precipitation and a distributed snow water equivalent reanalysis dataset, we illustrate the role of the rain-on-snow ARs in modulating streamflow generation. The role of snow is strongly elevation-dependent, with lower-elevation snowmelt contributing to rain-driven runoff, while deeper, higher-elevation snowpack likely entrained a significant portion of the rain and limited runoff. Given observed trends and projections showing warming and strengthening of wintertime AR events, the February 2017 event may be indicative of enhanced future flood risk in the region.