Search for a command to run...
Abstract Elevated soil moisture and heavy precipitation contribute to landslides worldwide. These environmental variables are now being resolved with satellites at spatiotemporal scales that could offer new perspectives on the development of landslide warning systems. However, the application of these data to hydrometeorological thresholds (which account for antecedent soil moisture and rainfall) first needs to be evaluated with respect to proven, direct measurement‐based thresholds that use rain gages and in situ soil moisture sensors. Here we compare ground‐based hydrologic data to overlapping satellite‐based data before, during, and after a recent season of widespread shallow landsliding in the San Francisco Bay Area (California, USA). We then explore how the remotely sensed information could be used to empirically define hypothetical thresholds for shallow landsliding. We find that the ground‐based thresholds developed with a single monitoring station show superior performance because the in situ soil saturation data better reflect the gravity‐dominated subsurface flow conditions that are characteristic of hillslopes during the rainy season. Although the satellite‐based thresholds can identify most of the landslide days, they include a greater number of false alarms due to overestimates of soil moisture between major storm events. To avoid the type of false alarms that are characteristic of our satellite‐based thresholds, further postprocessing of the near‐surface hydrologic response data should be integrated into satellite‐based model outputs to better reflect gravity‐dominated drainage. Our results encourage further deployment of ground stations in landslide‐prone terrain and cautious exploration of satellite‐based hydrometeorological thresholds where in situ networks are nonexistent.