Mi SciELO
Servicios Personalizados
Articulo
Indicadores
Citado por SciELO
Links relacionados
Similares en SciELO
Compartir
Revista Universidad y Sociedad
versión On-line ISSN 2218-3620
Resumen
CHIRINOS ESCOBAR, Marcio Josué; RUIZ ALVAREZ, Maynor Alberto y CARDONA, Alex Javier. Statistical model for predicting landslides generated by precipitation in western Honduras. Universidad y Sociedad [online]. 2023, vol.15, n.2, pp. 246-255. Epub 30-Abr-2023. ISSN 2218-3620.
This paper presents a statistical model to predict the occurrence of rainfall-generated landslides in western Honduras. The model has been created with a landslide inventory based on a review of Google Earth satellite images for the period from 1998 to 2020. To perform the regression analysis, the MARS (Multivariate Adaptive Regression Splines) methodology has been used, which allows automating the construction of prediction models, selecting relevant variables, transforming the predictor variables, treating missing values and foreseeing overadjustments by means of a self-analysis. As a result, the map of susceptibility to rain-induced landslides is obtained, which has a value between 0 and 1, which has been reclassified into 5 susceptibility categories: low, medium-low, medium, medium-high and high.
Palabras clave : Landslides; Terrain-derivad; MARS; Prediction model; Susceptibility.