An assessment of digital elevation models and their ability to capture geomorphic and hydrologic properties at the catchment scale*

Abstract

Digital elevation model (DEM) data quality is paramount for accurate representation of the land surface and drainage network. This issue was investigated within a small agricultural catchment in the Upper Hunter Valley region of New South Wales, Australia for a DEM created by use of a Differential Global Positioning System (DGPS) and a 25 m DEM from the New South Wales Government's land mapping authority, Land and Property Information (LPI). A hierarchical scaling approach was used to investigate the effect of increasing DEM grid size on a number of geomorphic and hydrologic descriptors (i.e. area–slope relationship, cumulative area distribution, hypsometric curve, width function, Strahler stream order and stream network statistics), as well as addressing the issue of source data accuracy. Results of qualitative and quantitative assessments indicate that as DEM grid size increased, average slope gradients decreased and the drainage network became increasingly simplified. Geomorphic descriptors such as the width function, cumulative area distribution and hypsometric curve appear largely insensitive to DEM scale. The area–slope relationship loses definition in the diffusive region of the curve at large grid scales; however, the fluvial region appears largely insensitive to changes in DEM resolution. A comparison of long-term field soil moisture data with wetness indices derived from DEMs clearly demonstrates that high resolution DEM data are needed to model soil moisture distribution. A 5 m DEM was found to have the minimum resolution required for the current study site in order to accurately capture catchment geomorphology and hydrology and to model the spatial distribution of soil moisture.