Rainwater infiltrating process revealed from the self-potential signatures, insight for landslide monitoring

Understanding the physical process of soil imbibition and water flow in the porous media in depth is significant in assessing the risk of forming landslides. Volumetric soil moisture sensors can be used to measure water content variations in situ. However, it has a spatial gap due to the limited number of installed sensors. On the other hand, geophysics can provide integrated measurements that can be spatially resolved. Among existing geophysical methods, Self-Potential (SP) is a method of choice to monitor water flow. Indeed, pore-water flows can generate the electrical streaming current based on the electrokinetic mechanism. This electrokinetic cross-coupling process is not only sensitive to the water flow but also depends on water content variations. This study relies on a soil-column experiment by artificially imposing rainfall to examine if the electrical SP could indicate the water infiltrating process. Combined with the observed data, our results indicate the water infiltrating stages can be characterized by the extracted SP signatures under a comprehensive numerical model. As a passive hydrogeophysical method, the capacity of SP to capture the characteristics of the spatio-temporal variations of water fluxes and soil-water conditions can offer early warning information of rainfall-induced landslides.