The influence of DEM resolution on seismic landslide hazard assessment based upon the Newmark displacement method: a case study in the loess area of Tianshui, China

The estimation uncertainties and spatial variability of input parameters are gaining more attention and have frequently been taken into account in seismic slope deformation predictions, but the influence of DEM resolution has rarely been investigated. In this paper, the impact of DEM resolution on the seismic landslide hazard is investigated through the comparison between various seismic landslide hazard assessments evaluated from multi-resolution (10, 30, and 90 m) DEMs in the loess area of Tianshui, China, using the Newmark displacement method. The results show that the 90 m relatively coarse resolution DEM underestimates the estimated Newmark displacement. The 30 and 10 m resolution DEMs basically give the similar predictions while the 10 m higher-resolution gives higher resolution at the local convex topography such as scarps and ridge crests. It suggests that the ability of Newmark method to predict landslide occurrence accurately not only depends on the assigned shear strengths, but also is controlled predominantly by the distribution of slope angles. Making accurate maps of seismic landslide hazards will require high-resolution DEM, apart from the associated ground-motion parameters and assigned shear strengths.