hide
Free keywords:
-
Abstract:
Rainfall-triggered landslides in urban areas cause fatalities and damage globally, but landslide early warning systems operating in individual cities rely on different data sources and methods, making a comparison of thresholds challenging. In this study, we estimate the distribution of intensity-duration (I-D) and event-duration (E-D) rainfall thresholds for landslide occurrence in cities worldwide. We develop a global compilation of urban landslides and extract the coincident event rainfall from two global precipitation datasets: the hourly resolution, gauge-based Global Sub-Daily Rainfall Dataset (GSDR; Lewis et al., 2019) and the half-hourly resolution, gridded, satellite-based Global Precipitation Measurement mission (GPM-IMERG; Huffman et al., 2014). We then use Bayesian multi-level quantile regression to determine 5% I-D and E-D thresholds for each city within a single model. This approach offers an objective way to determine rainfall thresholds and their uncertainty while explicitly modeling variation between cities. Preliminary results show that I-D thresholds in urban areas are much lower than previously reported global thresholds (Guzzetti et al., 2008), and that posterior parameter estimates for the rainfall threshold slope are indistinguishable from zero in most cases. Rainfall thresholds in cities with widely varying climates, topography, and population densities are often indistinguishable from each other. At short durations (hours to days), landslide triggering rainfall has often been moderate, with high annual exceedance probabilities. At longer durations (days to weeks), landslide triggering rainfall has frequently been among the highest recorded events. We hypothesize that the processes controlling landsliding in cities are fundamentally different from those in natural settings. We surmise that anthropogenic controls such surface sealing, slope engineering, and artificial drainage play a dominant role for slope stability in urban areas, resulting in a harmonizing effect on rainfall thresholds in many cities throughout the world.
