On the Extraction of Channel Networks from Digital Elevation Data

David G. Tarboton, Utah State University
Rafael L. Bras
Ignacio Rodriguez-Iturbe


Channel networks with artibtrary drainage density or resolution can be extracted from digital elevation data. However, for digital elevation data derived networks to be useful they have to be extracted at the correct length scale or drainage density. Here we suggest a criterion for determining the appropriate drainage density at which to extract networks from digital elevation data. The criterion is basically to extract the highest resolution (highest drainage density) network that satisfies scaling laws that have traditionally been found to hold for channel networks. Procedures that use this criterion are presented and tested on 21 digital elevation data sets well distributed throughout the U.S.