CONSTRAINING THE IMPACT OF ATMOSPHERIC RIVER STRENGTH ON CALIFORNIA'S SHALLOW LANDSLIDES USING A NEW UNIVERSAL THRESHOLD BASED ON ANOMALOUS SOIL HYDROLOGY

California (USA) receives most of its wintertime precipitation from atmospheric rivers (ARs), long filaments of atmospheric moisture transported to the Pacific coast via mid-latitude cyclones. Historical studies in the San Francisco Bay area reveal that shallow landslide events are often associated with AR-like meteorological conditions, but it is currently unclear how these relationships may hold in other regions of the state and to what extent the strength of landfalling atmospheric rivers dictates the scale of regional landslide response. Estimating regional landslide impacts remains an ongoing challenge as post-event landslide inventories typically cover a limited area relative to the true area impacted by landslides. Furthermore, rainfall thresholds developed for landslide triggering are often calibrated to specific sites and therefore hold particularly limited value for extrapolating across such a climatically diverse state. In this study, we develop a shallow landslide triggering threshold based on the local climatology of a simple tank hydrologic model applied to over 19 years of gridded rainfall data across California. We use four landslide inventories from different storm events across varying climatic regimes to show that the fifteen-year recurrence of modeled soil water at a given site, which we call A^*, represents a consistent threshold for shallow landslide triggering. We then use the distribution of A^* for nine unique landslide-inducing storm events to calculate the total landslide potential area (LPA) of above-threshold hillslopes throughout the state and compare this to atmospheric river strength via the AR scale of Ralph et al. (2019). We find that although most storms in our catalog experienced moderate to strong AR conditions, AR scale does not correlate with event LPA. Instead, monthly LPA monotonically declines from Dec./Jan. to June, following a similar pattern to observations of monthly AR arrival frequency across the U.S. west coast. As storm frequency may act to keep soil water elevated by limiting inter-event drainage and thus set antecedent moisture conditions preceding a storm arrival, this process may exert a primary control on the scale of regional landslide response to storms.