"U.S. Geological Survey topographic maps and similar maps published in other countries are the most fundamental sources of geographic information; many other types of maps are based on these. Yet they show only one-third of all the mileage of stream channels that occur on the ground. Reuben's method maps stream channels with over 90 percent accuracy using only Digital Elevation Models (DEMs)--a data source that is now available for the entire globe. That means that Reuben's method can be used to accurately map the majority of stream channels in the world in a straightforward automated fashion."

--Christopher Lant, professor of geography and environmental resources
 

Ed.--Resource managers and planners need the best possible picture of our world. The image above at left depicts a stream system in the Flint Hills region of east-central Kansas as represented in a USGS map--but in fact this shows only about 35 percent of the "stream miles" in the system. Many small tributaries don't even appear on the map. The image at right is a much more complete map generated by SIUC student Reuben Heine.

The September 2004 issue of the Annals of the Association of American Geographers published an article by Heine and Lant describing an improved, automated method, based on Heine's master's thesis research, to map streams and their tributaries. Heine, now a doctoral student in the environmental resources and policy program at SIUC, developed a mathematical method using logistic regression that allows stream maps to be generated rapidly by extrapolating from satellite data on land elevation. Ground truthing--checking the extent to which those maps match reality--has confirmed the accuracy of the method.

Heine is now working on better methods to identify which streams are perennial and which are ephemeral (carrying water only part of the year). He is a co-recipient of SIUC's first (2005) Outstanding Graduate Student Researcher award.

Comments: Perspectives Webmaster
Copyright © 2005, Board of Trustees, Southern Illinois University | Privacy Policy