Prediction of seasonal variation of in-situ hydrologic behavior using an analytical transient infiltration model

Abstract Rainfall-induced landslides pose serious threats to civil infrastructure and human life. Stability in a hillslope environment is a function of the hydrologic behavior. The variations in hydrologic behavior are driven by variations in climatological events such as rainfall and evapotranspiration. Thus, because the variations in climatological events are seasonal, prediction of seasonal variation in hydrologic behavior is critical for the prediction of landslides. However, most prediction methods only consider changes to hillslope hydrologic behavior due to rainfall and exclude the contribution of evapotranspiration. This approach does not allow for a complete analysis of the hydrologic behavior during seasonal cycles wetting and drying. Including evapotranspiration constitutes a significant improvement for early warning systems that are based on real-time seasonal hydrologic events. This study presents the development and implementation of an analytical transient infiltration model to predict seasonal variation of soil hydrologic behavior during a complete cycle of a season. The model was applied to three landslide sites in Kentucky. In-situ measurements of volumetric water-content and soil suction allowed for evaluation of seasonal soil moisture and suction fluctuations. Both rainfall and evapotranspiration were considered within a framework that facilitated the prediction of soil suction and volumetric water-content with transient surface flux. In addition, this model only requires unsaturated soil parameters based on the drying season to predict soil hydrologic behavior in the wetting season. The predicted soil hydrologic behavior can be applied directly to a limit equilibrium equation to estimate seasonal variations and the stability of a slope. The practical application of this study is the prediction of seasonal variation of hydrologic data for any site once calibrated, which will support a more realistic assessment of landslide hazards.