TWI computations and topographic analysis of depression-dominated surfaces

Abstract

The topographic wetness index (TWI) has been widely used for determining the potential of each digital elevation model (DEM) grid to develop a saturated condition, which allows for the investigation of topographic control on the hydrologic response of a watershed. Many studies have evaluated TWI, its components, and the impacts of DEM resolution on its computation. However, the majority of the studies are concerned with typical dendritic watersheds, and the effectiveness of TWI computations for depression-dominated areas has been rarely evaluated. The objectives of this study are (1) to develop a modified TWI computation procedure for depression-dominated areas, (2) to examine the differences between the new and existing TWI computation procedures using different DEMs, and (3) to assess the impact of DEM resolution on the new TWI procedure. In particular, a bathymetry survey was conducted for a study area in the Prairie Pothole Region (PPR), and the DEM representing the actual surface topography was created. The statistical analyses of TWI highlighted a two-hump pattern for the depression-dominated surface, whereas a one-hump pattern was observed for the dendritic surface. It was observed that depressional DEM grids accounted for higher values of TWI than other grids. It was demonstrated that a filled DEM led to misleading quantity and distribution of TWI for depression-dominated landscapes. The modified TWI computation procedure proposed in this study can also be applied to other depression-dominated areas.