We investigate whether ERA5-Land reanalysis soil moisture data, despite their 5-days publication delay, can be useful for improving the performance of relationships providing triggering conditions for landslides. Using artificial neural networks, we find that soil moisture delayed even up to 15 days allows an improvement of performance respect to precipitation-based models, therefore corroborating the potential use of ERA5-Land soil moisture for improving landslide early warning.

We propose an approach exploiting PCA to derive hydrometeorological landslide-triggering thresholds using multi-layered soil moisture data from ERA5-Land reanalysis. Comparison of thresholds based on single- and multi-layered soil moisture information provides a means to identify the significance of multi-layered data for landslide triggering in a region. In Sicily, the proposed approach yields thresholds with a higher performance than traditional precipitation-based ones (TSS = 0.71 vs. 0.50).

To mark the 20th anniversary of Natural Hazards and Earth System Sciences (NHESS), an interdisciplinary and international journal dedicated to the public discussion and open-access publication of high-quality studies and original research on natural hazards and their consequences, we highlight 11 key publications covering major subject areas of NHESS that stood out within the past 20 years.

Regional climate models (RCMs) are commonly used for high-resolution assessment of climate change impacts. This research assesses the reliability of several RCMs in a Mediterranean area (southern Italy), comparing historic climate and drought characteristics with high-density and high-quality ground-based observational datasets. We propose a general methodology and identify the more skilful models able to reproduce precipitation and temperature variability as well as drought characteristics.