Characterization and evaluation of elevation data uncertainty in water resources modeling with GIS

Grid based digital elevation models (DEM) are commonly used in water resources modeling. The quality of readily available DEM, however, varies from source to source in terms of horizontal resolution and vertical accuracy which are the two important aspects of elevation uncertainty in the modeling with raster GIS. This paper addresses the issue of elevation data uncertainty in GIS supported hydrologic simulations. The essential role of elevation data in the modeling is revealed by presenting DEM processing processes in distributed and semi-distributed hydrologic analyses. It is very difficult to examine the elevation uncertainties analytically due to complexities of the hydrologic models. An ideal approach is to assess the effect of the DEM uncertainty by applying varying resolutions or accuracies of elevation data in the modeling. Different grid sizes of DEM are used in observing DEM resolution dependence and resulting model outputs are compared to obtain a profile of its effect. Impact of DEM vertical accuracy is explored by Monte Carlo simulation with a large number of DEM realizations generated based on different levels of specified error. The approach is implemented in a case study with a topography based hydrologic model on an experimental watershed to analyze both aspects of the uncertainty. The results show that both DEM grid size and vertical accuracy could have profound effect on hydrologic modeling performance. The impact can be compensated by model calibrations due to interactions between model parameters and spatial factors. The study indicates that the DEM uncertainty can be effectively evaluated using the applied method. The work is to provide some insight into the characterization of elevation data quality and the association between topography and water resources models.