Effects of DEM Resolution on Surface Depression Properties and Hydrologic Connectivity

Surface digital elevation model (DEM) resolution receives increasing attention because of its importance in topographic analysis (e.g., quantification of surface depressions) and hydrologic modeling. Varied and even contrary conclusions have been obtained concerning the effects of DEM grid size on surface depression properties. Land surfaces are featured with a number of microtopography-controlled, localized areas, and their connections may significantly alter hydrologic and geomorphologic processes. Few efforts have been made to examine the effects of DEM resolution on surface depression properties and hydrologic connectivity. This study aimed to evaluate such resolution effects by two dimensionless parameters: the DEM representation scale λL (the ratio of DEM grid size to correlation length, representing horizontal resolution) and the surface roughness scale λR (the ratio of random roughness to correlation length, representing vertical topographic variability). A puddle delineation program was utilized to quantify depression properties for a variety of topographic surfaces characterized by different DEM resolutions, including small-scale surfaces with random roughness, field plots, and watershed surfaces. In particular, a puddle-to-puddle (P2P) conceptual model was used for hydrologic connectivity analysis. It was found that puddle properties depended on both dimensionless λL and λR. The λL significantly influenced the calculations of structural and functional hydrologic connectivity. Using DEMs with a coarser resolution or higher λL tended to overestimate hydrologic connectivity and simplified hydrograph for a surface with numerous small-scale depressions. The DEM resolution or dimensionless λL had significant influences on the development of functional hydrologic connectivity, especially at the early stage of the rainfall-runoff process.