Soils of Southeastern U.S.A.

The soils map for the southeastern region of the USA was derived from the STATSGO databases developed by USDA-NRCS for the states. The data were compiled by mapping soil associations within a given Major Land Resource Area (MLRA). Each state has a STATSGO database which can be obtained from the USDA-NRCS.

The STATSGO database for each state is a stand-alone spatial and tabular data set and was developed at a scale of 1:250,000. Although, there were attempts to use common soil association names across state boundaries, the numerical values assigned to the associations were not always the same between state boundaries. Thus, a Sharkey-Tunica-Dundee association in Arkansas did not have the same numerical identification as in Mississippi. Often, where soil associations cross state boundaries, the name of the association also changed. These changes may result in minor differences in soil association names or even a completely different name.

The soils map represented in Figure 5 is a composite of the STATSGO databases for all 13 states in the southeast United States. Each state database was brought into a new data set by transforming the projection of each spatial vector data from an Alpert's Equal Area projection to a Lambert's Conical projection, the same projection as the AVHRR data. The vectors were then converted to raster format. The numerical values representing the soil association for each state had to be unique. However, most states used a nearly sequential numbering system resulting in common numerical values for differing soil associations. This problem was solved by multiplying each states FIPS code by 1000 and adding that value to numerical value of each soil associations. Thus, each soil association had a unique numerical value among all 13 states. All the state STATSGO spatial data were then patched together resulting in one continuous coverage map for all 13 states. Since some soil associations were common to two or more states, there were duplications within both the numerical and category values in the patched file. The regional soils map also had a large range in numerical values varying from 1,001 to 51,100. Problems from duplications among states and the large range in numerical values were eliminated by reclassing the new map into numerical values ranging from 1 to 2,385 and combining identical soil associations into a single category. The new classed regional soils map was then resampled into a stand-alone map for the southeast region of the United States.