IoT monitoring and reanalysis data of soil moisture and rainfall for landslide warning: a test case

Shallow rainfall-induced landslides are triggered by intense or prolonged rainfall. Warning models employed within territorial landslides early warning systems (Te-LEWS) are typically based on rainfall thresholds expressed in terms of cumulative rainfall or average intensity with respect to the duration of the rainfall event, completely neglecting antecedent conditions. However, recent studies demonstrated that introducing, directly or by means of models, the effects of antecedent soil moisture content in empirical thresholds can improve the performance of the warning models.

This preliminary study focuses on the definition of a pilot monitoring site that produces rainfall and soil moisture data measured by an Internet of Things (IoT) monitoring network and by the use of analogous reanalysis products (i.e., ERA5-Land dataset). The activities are being developed in the context of the Horizon Europe project “The HuT: The Human-Tech Nexus - Building a Safe Haven to cope with Climate Extremes”. The final aim is to use IoT monitoring of rain and soil moisture, combined with reanalysis data, to improve, at municipal level, the territorial warning procedures already existing and operational at regional level.

The test site has been installed within the Campus of the University of Salerno in Fisciano, Campania region (Italy) since February 2023. The pilot site has been instrumented with sensors monitoring soil moisture from different providers and with a weather station; the sensors have been installed at different depths and with different procedures. The collected data were analyzed and processed using various data analysis algorithms, with the aim of: i) establishing correlations between the local weather conditions and the hydrologic soil response; ii) make a comparison between the data collected from different providers and in different local conditions.

Establishing these relationships allowed to evaluate the peculiarities and reliability of the different sensors and to identify the best configuration for future in-situ installations. More generally, this study highlights the importance of developing a monitoring network based on diffuse low-cost sensors and a proper real-time data transmission, analysis and processing, in order to provide further knowledge to system managers of territorial warning systems in the analysis of the monitoring data, and thus support for their decisions before and during extreme weather conditions.