Spatial and statistical differences between 1:250,000- and 1:24,000-scale digital soil databases

Abstract

In this study, 1:250,000- and 1:24,000-scale digital soil databases for Kansas were compared statistically and spatially. The soil attributes used in this comparison were soil permeability, percent clay, and hydrologic group. Results indicate that, although the two databases were correlated, the potential exists for substantial site-specific variability between them. The largest differences between the two databases typically are in and along the stream networks. With distance away from the stream networks, the mean differences generally stabilize. The results also indicate the possibility of systematic bias between the two databases that varies with landscape position. For applications using mean soil attribute values, the two soil databases usually yield similar values for the three soil attributes analyzed, especially for areas of 25 km2 (9.7 mi2) or larger. However, for applications where more detailed information on soil variability and the spatial pattern of soil properties within the landscape is required, such as for studies focused on small areas or areas in and along stream networks, the two soil databases are sufficiently different such that using one or the other may result in substantially different results.