Comparison of Land Use Area Estimates from Three Different Data Sources for the Upper Mississippi River Basin

This study presents the results of comparing land use estimates between three different data sets for the Upper Mississippi River Basin (UMRB). The comparisons were performed between the U.S. Department of Agriculture (USDA) Natural Resource Conservation Service (NRCS) National Resource Inventory (NRI), the U.S. Geological Survey (USGS) National Land Cover Data (NLCD) database, and a combined USDA National Agricultural Statistics Service (NASS) Agricultural Census - NLCD dataset created to support applications of the Hydrologic Unit Model for the U.S. (HUMUS). The comparison was performed for 1992 versions of the datasets because that was the only consistent year available among all three data sources. The results show that differences in land use area estimates increased as comparisons shifted from the entire UMRB to smaller 4- and 8-digit watershed regions (as expected). However, the area estimates for the major land use categories remained generally consistent among all three data sets across each level of spatial comparison. Differences in specific crop and grass/forage land use categories were magnified with increasing refinement of the spatial unit of comparison, especially for close-grown crops, pasture, and alfalfa/hayland. The NLCD close-grown crop area estimates appear very weak relative to the NRI and HUMUS, and the lack of specific crop land use estimates limits its viability for UMRB agricultural-based modeling scenarios. However, the NLCD is a key source of non-agricultural land use data for HUMUS and supplemental wetland land use area estimates for the NRI. We conclude that comparisons between more recent versions of the data sets (i.e., 1997 NRI, 1997 or 2002 Agricultural Census, and 2001 NLCD) would not result in significant additional insights and that the 1997 NRI is a viable land use data source for current CARD UMRB water quality modeling studies. However, adoption of other land use data such as USDA-NASS remote sensing data should be investigated.