Logistic regression modelling of storm-induced shallow landsliding in time and space on natural terrain of Lantau Island, Hong Kong

Abstract

This study is aimed at developing a method for modelling the probability of landsliding using a geographical information system (GIS), rainstorm events and associated landslide distributions, with an attempt to assess where landslides are likely to occur when certain rainstorm events happen. This research focuses on the natural terrain of Lantau Island, Hong Kong. In order to derive a logistic regression model for statistical prediction of the probability of landsliding, topographic attributes derived from digital elevation model (DEM) data are combined with a variety of other digital datasets including 24 h rainfall, lithology and vegetation cover in a GIS. The model was then applied to given rainstorm events and rainfalls of a variety of periods of return time in order to predict the probability of landsliding on natural slopes in space and time. The modelling techniques described here can be used to develop effective mitigation strategies. © Springer-Verlag 2004.