GIS Methodology to Assess Landslide Susceptibility: Application to a River Catchment of Central Italy

Abstract

This paper illustrates a geographic information system (GIS) supported methodology for the assessment of landslide susceptibility. The methodology involves four operational steps: survey, site analysis, macroarea analysis and susceptibility analysis. The Survey includes the production (or acquisition) of a large-scale litho-technical map, a large-scale geomorphological map, a detailed inventory of past and present landslide events, and a high resolution DTM (Digital Terrain Model. Site analysis leads to the definition of discriminating parameters (commonly, lithological and morphometric conditions necessary but not sufficient to trigger a landslide of a given type) and predisposing factors (conditions that worsen slope stability but are not sufficient to trigger a landslide of a given type in the absence of discriminating parameters). The different predisposing factors are subdivided into classes, whose intervals are established by descriptive, statistical analysis of landslide inventory data. A numerical index, based on the frequency of landslide occurrence, quantifies the contribution of each class to slope instability.Macro-area analysis includes the generation of Litho-Morphometric Units (LMU) by overlaying discriminating parameters, manual drawing of LMU envelopes (macro-areas), generation of predisposing factor maps from the spatial distribution of predisposing factors, and heuristic weighting of predisposing factor indices. Susceptibility analysis includes the generation of Homogeneous Territorial Units (HTU) by overlaying macroareas and predisposing factor maps, and the application of a susceptibility function to the different HTU. The resulting values are normalized before the generation of the landslide susceptibility maps. The methodology has been applied to the Fiumicino River catchment, located in the western side of Latium Apennine (Central Italy) between 200 and 1300 m a.s.l. and developed on Late Miocene calcarenites, sandstones with clay intercalations, and marls. The resulting landslide susceptibility maps will be employed in environmental management. They also represent the preliminary step for the assessment of landslide hazard and risk.

Geostatistical Modeling to Capture Seismic‐Shaking Patterns From Earthquake‐Induced Landslides

“…Substantial improvements have been made in landslide predictive mapping during the last two decades (Reichenbach et al, 2018). Research has progressed from simple heuristic approaches (e.g., Leoni et al, 2009), towards deterministic methods (Bout et al, 2018), multivariate statistics (e.g., Lombardo et al, 2014), and data mining techniques (e.g., Lee et al, 2018). However, the estimation target (i.e., the landslide susceptibility) has remained the same.…”

Landslide Susceptibility Mapping of Urban Areas: Logistic Regression and Sensitivity Analysis applied to Quito, Ecuador

“…The referential AUC value is shown in the dashed red line unbalanced, unstructured, and underreported landslide events data are expected to be overcome in the future, thus improving weights encoding. Nevertheless, the encoded data was processed using a theorical multicriteria assessment for LSM, as undertaken in other research works (Leoni et al 2009;Lombardo and Mai 2018), which this study tested with a LOGIT model and SA, as described above.…”

Landslide Susceptibility Mapping for Quito with Logistic Regression and Sensitivity Analysis

“…Van Dessel, van Rompaeya, & Szilassi, (2011) stress on the incidence that the quality of the input maps has on the resulting coefficients of logistic regression models applied for landslide susceptibility analysis. This research implements a so-called pseudo-quantitative method, which, initially consider weights from a multicriteria assessment of landslide susceptibility (Leoni et al, 2009;Lombardo and Mai, 2018), which later is, as commonly, migrated to a LOGIT model. This pondering is usually done according to criteria determined by experts and their knowledge of the study area.…”