THE INFLUENCE OF FUTURE WARMING ON THE SHALLOW LANDSLIDE CHARACTERISTICS OF AN EXTREME STORM SEQUENCE (ARKSTORM 2.0) IN CALIFORNIA

Historic, landscape-transforming events in California such as the Great Flood of 1861-1862 are thought to result from back-to-back arrivals of atmospheric river storms over a period of weeks. In addition to the severe flooding that accompanies these extreme storm sequences, referred to as “ARkStorms,” they are also predicted to trigger abundant landsliding across the state. Recently, Huang and Swain (2022) utilized downscaled meteorological models to examine the impact of climate change on the probability and scale of a future ARkStorm and showed that a changing climate not only increases the probability of an ARkStorm, but amplifies the rainfall impacts.

How then might the shallow landslide hazard, which depends on both rainfall intensity and duration, change in response to a warming-enhanced ARkStorm? Here we develop an empirical landslide hazard index, R^*_a, that combines 48-hour anomalous rainfall and a measure of antecedent soil moisture, to gauge how climate change might affect shallow-landslide potential in future extreme storm scenarios across California. We first verify the efficacy of R^*_a on a catalog of eight landslide-inducing storms across California since 2004 and find R^*_a to be a reasonable predictor of regional landslide occurrence. We then calculate the spatio-temporal evolution of R^*_a for both the historic and warming-enhanced ARkStorm scenarios of Huang and Swain (2022). We find that both the regional extent and mean value of R^*_a from the historical modeled storm sequence is on par with some single large-magnitude atmospheric river storms that have occurred in California since 2004. Whereas the duration of rainfall from a historic ARkStorm may elevate other large-scale hazards such as flooding, modeled rainfall intensities suggest that shallow landslide hazard may not increase commensurately. For the warming-enhanced ARkStorm scenario, however, two-day intensities are substantially higher across the state, yielding an average magnitude and footprint of R^*_a higher than any individual storm in our catalog. Thus, a future ARkStorm occurring in a warming California, in addition to producing substantially enhanced runoff and flooding, may also result in a commensurately extreme increase in shallow landslide hazard across the state.