Improving overland flow routing by incorporating ancillary road data into Digital Elevation Models

Abstract

Roads, ditches, and culverts influence hydrological and geomorphological processes significantly. However, most hydrological models continue to rely solely on regional digital elevation models (DEMs) to derive overland flow directions even though these DEMs have been shown to contain inadequate topographical information to effectively represent linear landscape features. This paper introduces a methodology that improves the accuracy of grid-based overland flow routing through the use of ancillary road, ditch, and culvert data. The road enforcement algorithm (REA) that was developed re-routes overland flow on either side of a road independently, thereby enforcing linear landscape features within the flow direction matrix. The overland flow patterns resulting from the implementation of the REA differ significantly from flow patterns derived using conventional GIS routing algorithms. A test application in the prairies of southern Alberta, Canada shows the REA affects intra-watershed runoff transport as well as the size and shape of DEM-derived watersheds. The flow direction matrices created with the REA can be incorporated into any grid-based hydrological model.