INVESTIGATION OF REMOTE SENSING TECHNOLOGY FOR LAND COVER
ACQUISITION FOR UPPER MISSISSIPPI RIVER SYSTEM FLOODPLAIN.

Cynthia Berlin1, Peter Joria3, Larry Robinson3, Jennifer Dieck2,3, Heidi Langrehr4, and Holly
Ries2,3.  
1Department of Geography and Earth Science, 2Department of Biology, University of Wisconsin
- La Crosse, WI 54601  3U.S. Geological Survey, Upper Midwest Environmental Sciences
Center, La Crosse, WI 54603  4Wisconsin Department of Natural Resources, Onalaska, WI
54650.

The purpose of this study is to evaluate current remote sensing technologies for their utility in
developing a land cover/use (LCU) database for the Upper Mississippi River System (UMRS)
floodplain.  An LCU for the entire system, derived from 1:15,000 scale color-IR aerial
photography acquired in 1989, is nearing completion.  It would be time- and cost-prohibitive to
update this dataset using the same scale aerial photography and the same level of classification
detail.  A less complex classification scheme has already been developed through discussions
with data users.  Smaller scale photography or high-resolution satellite images may further
expedite the classification process.  The objective of this phase of the project is to evaluate
1:24,000 scale aerial photography and 4-meter IKONOS satellite images for their utility in
developing an LCU database for the UMRS floodplain.

Multispectral IKONOS images, two sets of color-IR aerial photography (1:15,000 and 1:24,000
scale) and ground-based data for three study sites were acquired in the late summer and early fall
of 2000.  Both sets of aerial photography were interpreted to 27 land cover classes.  Portions of
the 1:15,000 scale interpretation were used as training data to help classify the IKONOS images. 
Other areas of the 1:15,000 scale interpretation will be used to assess the accuracy of both the
classified IKONOS images and the interpreted 1:24,000 scale photography.  The IKONOS
classifications and the interpreted 1:24,000 scale photography will be compared on the basis of
accuracy and ease of production.  Results of the project will help determine an improved
methodology for updating the land cover classifications for the entire UMRS. 
 
Keywords: IKONOS, remote sensing, Mississippi River, land cover classification, floodplains