Use of Bias Reduced L3SMP_E Surface Moisture Estimates in Slope Stability Analyses

This study addresses the feasibility of using bias reduced SMAP Level 3 enhanced radiometer surface moisture (L3SMP_E) surface moisture estimates to predict and use root zone soil moisture (RZSM) to detect incipient failure conditions at known landslide locations. This was accomplished by: (1) acquiring L3SMP_E data from the NASA SMAP satellite mission over 20 in situ sites in Kentucky, (2) using an ensemble Kalman filter assimilation routine to estimate and reduce biases between L3SMP_E estimates and in situ measurements, (3) using a Kentucky-specific version of the soil moisture analytical relationship (SMAR) infiltration model to estimate RZSM at known landslide sites within Kentucky, and (4) using an unsaturated soil infinite slope stability model to analyze and detect incipient failure conditions at the time of reported landslide occurrence. Four known landslides within the Commonwealth of Kentucky were analyzed during this work. It was observed that RZSM estimated using the Kentucky SMAR model performed well when compared to in situ RZSM estimates. Additionally, inclusion of SMAR RZSM estimates at the four landslide sites was seen to yield stability models that were able to detect strength weakening and incipient failure conditions that aligned well with that of reported landslide occurrences and failure modes. The intent of this study is to show that bias-reduced satellite-based moisture estimates can be used to estimate soil moisture at known landslide sites for use in stability models with success in detection of reported incipient failure conditions.