The influence of DEM spatial resolution on landslide susceptibility mapping in the Baxie River basin, NW China

摘要

The selection of an appropriate map resolution is highly important for landslide susceptibility assessment. No consistent objective criteria, however, are currently applied to the choice of map resolution. This research, in conjunction with slope units, explores the effect of digital elevation model (DEM) resolution on susceptibility modelling using three statistical models (frequency ratio, index of entropy, and weight of evidence). Seven different spatial resolutions (30, 40, 50, 60, 70, 80, and 90 m) and three statistical models are investigated. For each resolution, we compare the performance of the three models using area under curve (AUC) analysis. The results show that, independent of the statistical models, the best performances are produced at 70 m DEM resolution. This highlights that finer resolutions do not necessarily lead to higher predictive accuracy in landslide susceptibility mapping. Rather, the frequency ratio model seems to be optimal for the coarser resolutions (i.e. 70, 80, and 90 m).