Application of remotely sensed land-use information to improve estimates of streamflow characteristics

Land-use data derived from high-altitude photography and satellite imagery are presented for 49 basins in Delaware, and eastern Maryland and Virginia. Based on 1:100,000 scale maps from high-altitude photography, basin land cover was extracted at the generalized Level I and the more detailed Level II classification categories. Level I land-use data summaries were prepared for 46 of the basins using the 1:250,000 scale maps derived from Landsat imagery. Land cover in the basins ranged from 93.9 percent urban at Little Falls Branch near Bethesda, Maryland, to 96.2 percent agricultural at Morgan Creek near Kennedyville, Maryland. Applying multiple regression techniques to a network of gaging stations monitoring runoff from 39 of the basins, it was demonstrated that land-use data from high-altitude photography provides an effective means of significantly improving estimates of streamflow. Forty streamflow-characteristics equations incorporating remotely sensed land-use information, were compared with a control set of equations using map derived land cover. (Woodard-USGS)