Investigating recent weather-induced landslides and the related weather characteristics in a test area in Campania Region (Italy) for warning purposes

The municipalities of the Amalfi and Sorrento Coasts (Campania Region, southern Italy) have historically been affected by weather-induced slope instability phenomena causing casualties and heavy economic damages, often associated with transport interruptions limiting the crucial tourist activities in the area. Due to the considerable geomorphological complexity of the area, even the simple cataloguing of the events is often challenging. Furthermore, the significant differences between the dynamics affecting the slopes (rockslides, debris flows, flowslides in pyroclastic covers) result in substantial differences in the atmospheric patterns able to trigger such events: from short-duration (up to sub-hourly scale) heavy precipitation events up to long-lasting rainfalls anticipated by particularly wet periods. Despite the significant differences among the dynamics, the warning system current operational in the Region considers three reference durations for cumulative precipitation (24, 48, and 72 hours) and it is based on three alert levels simply associated with the return time of potentially triggering precipitation (2, 5 and 10 years).

This study wants to fill this knowledge gap by investigating the main recent weather-induced slope instability phenomena in the area in relation to the recorded characteristics of the associated weather events. The investigation aims at multiple objectives.

• Comparing, over a common period, the records of different landslide catalogues available over the area - FraneItalia  (https://franeitalia.wordpress.com/), ITALICA (doi.org/10.5194/essd-15-2863-2023), Franceschini et al., 2022  (doi.org/10.1007/s10346-021-01799-y); Extreme Severe Weather Database (https://eswd.eu), Italian hydrological-geological Portal (https://idrogeo.isprambiente.it) - with the goal to verify consistency and to analyse the reasons leading to any discrepancies.
• Verifying the performance of the currently operational warning system for a set of events assumed as reliable, as they are included in more than one catalogue, taking into account the seasonality of events and the triggering precipitation patterns.

• Evaluating the capabilities of atmospheric reanalysis (ERA5land 10.24381/cds.e2161bac and CERRA 10.24381/cds.a7f3cd0be) made available by the Copernicus Climate Change Service to reconstruct the precipitation patterns that triggered the events (back-analysis). The topic is of great interest due to the high temporal resolution (1 hour) and spatial resolution (9 km for ERA5 land, 5.5 km for CERRA), which could therefore adequately cover areas where current sensor networks may be lacking or temporarily non-operational.
• Assessing if and how the soil moisture content data returned by atmospheric reanalysis can support the back-analysis and/or the forecasting of landslide events (ERA5 is updated with a delay of only 5 days with respect of present time). Given the spatial resolution of the reanalysis, they are not expected to actually reproduce local in-situ conditions, but they rather should act as proxies to evaluate the average wetness status of the slopes and then the presence of conditions predisposing to landslide triggering.

The results discussed regarding the case study of the Amalfi and Sorrento Coasts can be readily extended to other geographical and geomorphological contexts, at national and continental scale.