Usefulness of Satellite Soil Moisture Data for Improving an Existing Global Landslide Algorithm for Monitoring and Forecasting

Satellite soil moisture data, when integrated into an existing experimental global landslide hazard algorithm, have the potential to inform landslide analyses conducted at large spatial scales. This result can be reasonably expected, because increased piezometric head drives landslide hazards, although the relationship between changing soil pore water pressure (PWP) and satellite soil moisture retrievals is not straightforward. The degree to which changing PWP affects landslide potential depends on several other relatively static terrain parameters (e.g., topography, soils, land cover) related to landslide susceptibility. The existing landslide algorithm has been developed by combining these static terrain parameters with multi- satellite rainfall data to map global landslide susceptibility and forecast rainfall-triggered, shallow landslides. The rainfall trigger is based on rainfall intensity-duration thresholds derived from the Tropical Rainfall Measuring Mission (TRMM) data. The current algorithm does not account for prior surface conditions or the memory of the interaction between rainfall and terrain parameters, in the form of antecedent soil moisture. The objectives of this study are (1) to demonstrate a relationship between satellite soil moisture data and historical landslide events and (2) to determine the usefulness of integrating soil moisture into the existing landslide algorithm for monitoring and forecasting. Study sites include portions of the U.S. and Central America. Satellite soil moisture data used for this study include those derived from the TRMM Microwave Imager (TMI) and Aqua Advanced Microwave Scanning Radiometer-Earth Observing System (AMSR-E). These microwave retrievals are sensitive to only the soil moisture in the top several cm of the soil layer (i.e., "skin depth"). To include subsurface soil moisture in the study, the Global Land Data Assimilation System (GLDAS) modeled soil moisture data (surface and subsurface) are also used. The satellite and modeled antecedent soil moisture data are compared with rainfall time series and a recently compiled global landslide inventory database, to determine the difference in algorithm performance with and without the integration of soil moisture data. Characterizing the relationships between soil moisture and other terrain parameters and incorporating these relationships into an updated version of the algorithm have the potential to greatly improve landslide forecasting worldwide.