Multi-source climate data to investigate the nexus between climate anomalies and landslides in high-mountain regions in the Alps

As the effects of climate change become more and more evident in high-mountain regions, investigating the relation between climatic anomalies and geomorphic hazards becomes increasingly critical to predict the risk associated with such hazards and to develop reliable models. To do so, researchers often adopt statistical-based methods to detect climate anomalies at multiple time scales associated with the occurrence of different types of landslides. However, the in-situ observations commonly adopted for such studies may misrepresent some of these climatic variables in high mountain areas, sometimes leading to questionable results. For example, no daily precipitation anomaly is often reported for events such as debris flows, which are mainly triggered by short-duration precipitation. Additionally, collecting and quality-controlling in-situ observations is an extremely time-consuming task that prevents wide-scale applications of these methods. In this work, we exploit a consolidated statistical-based method to compare the results obtained from carefully controlled in-situ observations with the ones obtained from freely available quasi-global gridded datasets of (a) daily temperature observations from ENSEMBLES OBServation (E-OBS) and (b) half-hourly precipitation estimates from the Integrated Multi-Satellite Retrievals from GPM (IMERG). We focus on an extended database of 483 geomorphic hazards, including landslides, rockfalls and debris flows, occurred across the Italian Alps in the period 2000-2020. Our results show that the integrated use of open and free products is beneficial in different ways. Statistical tests indicate that E-OBS gridded temperature anomalies as well as multi-day IMERG precipitation anomalies provide as much information as in-situ observations, and can thus be used as easily available surrogates. More importantly, thanks to the ability of satellites to measure precipitation at the triggering locations, IMERG proved able to detect daily precipitation anomalies for many debris flows events for which in-situ data reported no precipitation. Examining the sub-daily variability of the triggering precipitation, we show that the anomalies missed by in-situ observations tend to be associated with events with high temporal, and hence spatial, variability such as the convective storms that usually trigger debris flows in the Alps. The use of quasi-global open datasets in place of in-situ observations can greatly speed-up the data retrieval and even provides an added value over in-situ observations. These results represent an important step ahead in the analysis of climate anomalies related to geomorphic hazards in high mountainous regions as they open the way to more accurate wide-scale applications.