Landslide hazard monitoring by combining geophysical and remote sensing data.

Landslides are a major natural hazard, threatening communities and infrastructure worldwide. The mitigation of these hazards relies on the understanding of their causes and triggering processes, directly depending on soil properties, land use, and their variations over time. Firstly, we propose a new approach combining geophysical and remote sensing data into hydro-geomechanical modeling to provide a robust estimate of the probability of failure of slopes endangering surrounding structures, with a focus on an urban area. We performed a sensitivity analysis of the main parameters of the hydro-geomechanical model, which highlighted strong sensitivity to variations in soil thickness and cohesion. Based on those results, we use seismic noise measurements to assess soil thickness around our study site and remote sensing data to assess the vegetation cover, which impacts the cohesion. Our results highlight that relatively thick soil layers (above 2 m) have up to 4 times higher probability of failure. The presence of tall vegetation has a significant effect on soil cohesion, especially when the soil layer is relatively thin. The addition of vegetation cover showed a drastic reduction in the probability of failure when the soil thickness is less than 5 m. Secondly, we used those results to locate an area highly prone to sliding and endangering a bridge. We monitored this area using passive seismic and low-cost tiltmeter landslide mechanisms to better define the precursors of landslide activation. The combination of the two monitoring methods provided an accurate description of a small reactivation that occurred during a heavy rainfall event after a 7-month drought. Seismic monitoring provided a means of tracking changes in soil properties and the tiltmeter provided accurate displacement rates. Eventually, these developments will enable us to provide an accurate hazard assessment and landslide early warning.