Assessment of rainfall erosivity from a convection-permitting model ensemble in an alpine transect

Alpine regions are highly susceptible to rainfall-induced soil erosivity and to its changes in a warming climate. Thanks to their high spatio-temporal resolution and to their ability to explicitly resolve convection, recently developed convection-permitting climate models (CPMs) outperform regional models in capturing intense sub-daily rainfall. Thus, these models offer a high potential for the simulation of rainfall erosivity and its projection in a warming climate.

This study evaluates the ability of a multi-model CPM ensemble to provide reliable rainfall data for assessing rainfall erosivity. This represents a fundamental step for then analyzing the near- and far- future projected changes in rainfall erosivity under climate change scenarios. The study is carried out based on data from an transect in the Italian eastern Alps which offers an ideal case study given its rainfall variability and complex topography. Data from 174 rain gauges are used to evaluate the CPM ensemble in simulating rainfall erosivity.

Preliminary results show that the CPM multi-model ensemble is able to reproduce well the mean annual rainfall erosivity in the study region. However, the CPM simulations over predict the number of erosive events by 21% on average. Also, the quality of the CPM-based simulations is shown to be strongly impacted by terrain elevation, with simulations in low land areas being characterized by an underestimation of number of erosive events and average rainfall erosivity. The bias reverses to overestimation of both number of erosive events and average erosivity with increasing the terrain elevation. Topography also strongly influences the spread of the erosivity simulations within the model ensemble.

These findings underscore the need for bias adjustments, considering topographic influences, for the investigation of projected changes in rainfall erosivity patterns.