Quantifying surface process dynamics during extreme events from storm characteristics and landslide inventories

Extreme precipitation events drive landsliding in many regions across the globe and are an important part of the erosional cycle and related hazards. The intensity and frequency of extreme events are likely increasing due to rising global temperatures, causing greater future threat to society and an urgent need to quantify the relationships between surface process dynamics and extreme events. In steep mountain belts, orography also plays a role in focusing precipitation and intensifying erosion. Yet, the influence of orography on the intensity-duration characteristics of extreme precipitation remains a subject of debate because we lack spatially distributed and high time-resolution gauge datasets needed to resolve convective-scale, short-duration storm events and satellite-derived precipitation products struggle to accurately resolve precipitation gradients over areas of high relief and altitude. Annual periods of monsoon-related landsliding in the Himalaya offer a natural laboratory in which to explore relationships between extreme precipitation, orography and landsliding processes. Here we scale the NASA’s Global Precipitation Measurement (GPM) IMERG 30-minute, 0.1x0.1 degree product with local rain gauge data to produce high-temporal resolution records used to characterize extreme rainfall events (EREs) in central Nepal where hundreds of shallow landslides occur each summer. Individual storms from the time series are defined using the average inter-accumulation time as a measure for the minimum dry period between storms and extreme storms are extracted from the series using a 90^th percentile threshold for each gauge station. Variability in storm characteristics is defined using paired K-means agglomerative cluster and principal component analyses to evaluate spatial patterns in storm characteristics over a 10 year period compared to annual landslide inventories. Spatial patterns emerge that suggest orography increases the intensity and frequency of storms, which in turn focuses landsliding in specific, and potentially predictable, regions along the steep windward flank of the mountain belt.