Open Source Dynamic Mapping of Slope Susceptibility to Shallow Landslides Potentially Interacting With Bridges and Viaducts by Ground-based and Remote Sensing Monitoring.
The high susceptibility to landslides of the Italian territory represents a significant risk factor for many infrastructures, especially bridges and viaducts. Landslides can be fatal for existing bridges and viaducts even when they are properly designed, built, and maintained, since the landslide-induced stresses can overcome the structural resources of such infrastructures.
In 2020, the Italian Ministry of Infrastructure introduced the "Guidelines for Risk Classification and Management, Safety Assessment, and Monitoring of Existing Bridges", which require the assessment of the Attention Class (also determined by landslide risk) of each bridge according to a multi-risk and multi-level procedure, with the aim of optimizing the resources allocated to infrastructure maintenance, based on the identified risk conditions. Therefore, a good understanding of the stability conditions of the slopes around a bridge is essential for analyzing possible landslide-bridge interactions.
The most frequent landslides in Italy are rain-induced, so it is very useful to zone the susceptibility of the territory to such instabilities. In assessing rain-induced shallow landslides susceptibility, Surface Soil Moisture (SSM) estimate is crucial, since this type of instability, which typically affects the slope in the first 2-3 m depth, often occurs in unsaturated soil conditions, frequently even in the days following rainfall, as a result of the propagation of the saturation front.
In this scenario, an innovative monitoring system for the SSM, integrating direct and remote sensing techniques, has been installed in San Carlo di Cese (Liguria, Italy), in the context of Mind The Bridge (MTB) Project, co-financed by the Liguria Regional FESR Program 2021-2027.
Within the LAMP (Landslide Monitoring and Predicting) System, in its modelling component dedicated to shallow landslides Soil Apparent Cohesion (SAC), the obtained SSM maps are the main input of a physically based 3D model implemented in QGIS for slope stability assessment, together with a Digital Terrain Model and geotechnical parameters. Typically, the analysis can be conducted over a large area (several km²) at 10 m spatial resolution in a short computational time. LAMP System processing procedures are available at github.com/LabGeomatica. At the San Carlo di Cese site, the area under consideration is the one interacting with a bridge.
Soil moisture is directly measured using low-cost capacitive sensors placed in the soil at different depths. In addition, the correlation between the reflectance of the examined surfaces and soil moisture allows the monitoring of the spatial and temporal variations of the latter variable, mapped using Sentinel-2 Multispectral images, obtained using both the Copernicus Data Space Ecosystem Portal and SADASADAM (SAme DAy SAtellite DAta Mosaics), a Python-based command line tool that automates the creation of atmosphere-corrected and cloud-free mosaics from Sentinel-2 scenes. Sentinel-2 images provide several bands (Red Edge, Broad Near Infrared, and Near Infrared) particularly suitable for the estimate of vegetation conditions, which are an indirect indicator of SSM. The obtained maps provide an estimate of SSM on natural slopes strictly referring to the first 10 cm of depth. However, this estimate can be extended to a depth of 40 cm, based on direct measurements obtained by the soil moisture sensors positioned at that depth.
SSM maps are realized according to a land cover specific (woods) Reflectance-Soil Moisture correlation developed on Mendatica (Liguria, Italy), a site prone to landslides selected within the Alcotra Project AD-VITAM (Analysis of the Vulnerability of the Mediterranean Alpine Territories to natural risks), assuming the characteristics of the area to be similar to San Carlo di Cese. The assumption was supported by the fact that the study site has a high prevalence of wooded areas and then positively verified by comparing satellite estimates with direct soil moisture measurements. The analogy with Mendatica in regards of soil characteristics was also exploited for the calibration of sensors in the field.
Both SSM and slope stability maps of the study area have been made available as a Web Map Service through a QGIS server. Currently, the available maps refer to October-December 2025 and are being updated. Through the WMS, territorial analysis concerning slope instability are included into the digital twin of the bridge, together with data acquired by the sensors installed on the bridge itself.
Since LAMP may allow for a dynamic analysis of the territorial susceptibility to landslides, its use could have consequent benefits in the protection of bridges and viaducts interacting with potential rain-induced shallow landslides, which represent the majority of Italian landslides.
