PSI - Issue 84

Michele Placido Antonio Gatto et al. / Procedia Structural Integrity 84 (2026) 111–118

112

1. Introduction Rainfall-induced shallow landslides (soil slips) are frequent and often sudden natural phenomena, triggered by the rapid infiltration of intense rainfall over short time intervals. Despite their limited spatial extent, they can cause critical interactions with infrastructure, resulting in direct damage or service disruptions. The early identification of areas prone to soil slips is therefore essential for planning mitigation measures and ensuring the safety of existing or planned engineering works. To this end, both data-driven and physically based approaches can be adopted. The former, relying on statistical analyses or machine learning techniques, require large datasets and provide rapid results, but may be difficult to generalize. The latter are based on the understanding of instability mechanisms and on the geotechnical properties of soils, offering greater physical robustness and interpretability, albeit at the cost of higher application complexity, especially at large spatial scales. Among simplified physically based models, the SLIP (Shallow Landslide Instability Prediction) model represents a well-established solution, developed since the early 2000s and successfully applied at different spatial scales (Montrasio 2000; Montrasio et al. 2011; Gatto and Montrasio 2023; Montrasio et al. 2023; Gatto et al. 2023; Gatto et al. 2025). SLIP estimates slope stability as a function of the physical and mechanical properties of the shallow soil layer (1–1.5 m) and the cumulative rainfall over the preceding 30 days, adopting simplifying assumptions regarding the effects of partial saturation on soil shear strength. The model achieves a good balance between accuracy and ease of use. For spatial applications, the X-SLIP platform was developed in the MATLAB environment (Gatto and Montrasio 2023). To predict potential triggering conditions, it is necessary to employ synthetic rainfall inputs representative of the local rainfall regime, both in terms of average rainfall and intense events potentially responsible for triggering failures. The analysis of interference risk with infrastructure can then be performed by considering geometric parameters, such as the distance and orientation between unstable areas and engineering works. This study proposes a simplified methodology for predicting potential soil slip triggering. The proposed approach is based on the application of the SLIP model and the X-SLIP platform to synthetic rainfall scenarios, characterized by a “background noise” over the first 29 days and a triggering rainfall peak on the 30th day. Synthetic rainfall series are generated through statistical analysis of long-term recorded rainfall time series (up to 10–20 years) from local rain gauge stations. 2. Methodology for the assessment of potential soil slip triggering 2.1. The SLIP Model The SLIP (Shallow Landslide Instability Prediction) model is a simplified physically based model that allows the stability of shallow soil layers to be assessed under the assumption of an infinite slope configuration. The first simplification introduced in the model concerns the representation of rainfall infiltration into the soil. The progressive accumulation of infiltrated water is modelled through the formation of saturation bubbles, which are subsequently assimilated into an equivalent saturated layer with thickness m varying over time. This parameter enables the effects of rainfall occurring during the 30 days preceding the event under analysis to be accounted for in a synthetic yet effective manner. The second simplification concerns the calculation of the apparent cohesion c ψ , which is introduced into the Mohr–Coulomb failure criterion. This cohesion is a function of the initial degree of saturation S r0 and varies over time in response to rainfall inputs and to the drainage capacity of the slope, represented by a drainage coefficient k t . The parameters m and c ψ are expressed as follows: m (t)= ξ nH (1- S r 0 ) ∑ h j ω j=1 e - k t (t-t j ) (1a) c ψ = AS r 0 (1- S r 0 ) λ (1-m) α (1b) A , λ , and α are modeling parameters. A depends on soil type, for which typical values are available in the literature, whereas λ and α can be assumed equal to 0.4 and 3.4, respectively, and are considered valid for all soil types. ξ