Modifying the United States National Hydrography Dataset to improve data quality for ecological models

The US National Hydrography Dataset (NHD) is the primary spatial hydrography data source for research and conservation of stream ecosystems in the United States due to its national coverage, high resolution, and delivery of high-volume attribute data. However, despite its wide use, numerous topological errors limit the utility of the NHD for the critical needs of ecological and environmental modeling when network connectivity is of interest. In this paper we demonstrate an automated process that integrates a set of GIS tools to identify, track, and correct four major types of topological errors in the entire NHD network (at the scale of 1:100,000) of the conterminous United States, thereby quantifying and summarizing the type, magnitude, and spatial distribution of topological errors in the NHD. Topological errors occurred on approximately 20% of the flowline features in the NHD dataset, with pseudo nodes being the dominant type of error. In a case study of New River Basin of North Carolina, Virginia, and West Virginia USA, after correcting the topology of the NHD network the number of streams in the database designated as first order decreased by 14%, and the average length of stream between confluences (i.e., inter-confluence stream segments) increased by slightly over half of a kilometer (0.53 km). The topology-corrected NHD dataset should particularly facilitate large-scale (e.g., national or regional) ecological and environmental applications requiring topologically sound stream networks.