Maria Teresa Brunetti, Massimo Melillo, Stefano Luigi Gariano, Luca Ciabatta, Luca Brocca, Silvia Peruccacci, 2021, Regional Approaches in Forecasting Rainfall-Induced Landslides, Monitoring and Early Warning, pp. 251–255, 2021,
URL: http://www.cnr.it/prodotto/i/441287

Hydrogeological hazards now exacerbated by the ongoing climate change pose serious challenges for the safety of the population worldwide. Among the others, the landslide risk can be mitigated by setting up efficient and reliable early warning systems. To date, rainfall thresholds are one of the most used tools to forecast the possible occurrence of rainfall-induced failures in large regions. In Italy a dense rain gauge network with hourly or sub-hourly temporal resolution is available. However, in some developing countries, where ground measurements are still absent or are available at coarser (daily) temporal resolution, satellite-based rainfall estimates could be a vital alternative. For this purpose, the reliability of rainfall thresholds defined using both satellite (SB) and ground-based (GB) data and with hourly or daily temporal resolution is assessed in a study area comprising the Abruzzo, Marche and Umbria regions (AMU), central Italy. The comparison between the performance of the different products allows to test their capability in eventually can GB rainfall measurements are gathered at hourly time steps (OBS-H) from a national network and aggregated on a daily scale (OBS-D); SB rainfall estimates are retrieved from the Climate Prediction Center Morphing Technique (CMORPH, hourly resolution), and from the SM2RASC product, based on the application of SM2RAIN algorithm to ASCAT (Advanced SCATterometer) soil moisture product (daily resolution). Results show that thresholds defined with GB rainfall data perform better than those obtained using SB estimates regardless of the temporal resolution. CMORPH and SM2RASC thresholds are still able to predict landslide occurrence although with a high number of false predictions.

Data from https://intranet.cnr.it/people/