Uncertainty in primary and secondary topographic attributes caused by digital elevation model spatial resolution

Abstract

Digital elevation models (DEMs) are increasingly being used to explain different soil and site-related characteristics using a variety of modelled topographic properties and their frequencies. It is well known that the grid cell size of a raster DEM significantly affects derived primary and secondary topographic attirbutes, such as elevation, slope, topographic position index, and wind exposition index. This study investigates the impact of increasing DEM spatial resolution on spatial distortion (defined by mean absolute error (MAE)) in two contrasting sites and tests two different data-obtaining procedures (average plot vs. single point). Findings detected and quantified significant error(s) caused by coarsening DEM spatial resolution. In addition, the plot average measurement procedure was more representative for any bounded area than a value extracted from a single point measurement. These findings provide important guidance for creating and using topographic attributes across a range of applications. This study recommends using DEM resolution of <15 m to derive primary and secondary topographic attributes. However, the optimal resolution is scale dependent and related with landscape; henceforth, coarser resolution may be sufficient for representing large-scale topographic features.