263-4 Correlating Soil Survey Geographic (SSURGO) Database Soil Map Units Across County Line Using Terrain Attributes.



Tuesday, October 18, 2011
Henry Gonzalez Convention Center, Hall C, Street Level

Zamir Libohova, National Soil Survey Center, USDA-NRCS, Lincoln, NE, Hans Winzeler, Purdue University, Gettysburg, PA, Skye Wills, 100 Centennial Mall North, USDA-NRCS, Lincoln, NE and Phillip Owens, 915 W. State St., Purdue University, West Lafayette, IN
The US Soil Survey Geographic (SSURGO) database is provided on county by county basis (Soil Survey Division Staff, 1993). This creates a mosaic of surveys conducted at various times by various surveyors. The result is different soil series and soil map units associated with the same slope position across county lines. The US Soil Survey program has adopted the Major Land Resource Area (MLRA) concept that calls for creation of seamless soil maps across political boundaries. The objective of this research was to develop tools for automating the correlation process of soils and soil map units across county lines using terrain attributes (TA) derived from digital elevation model (DEM). The research was conducted at Dillon Creek watershed that is located in southern Indiana extending through Dubois and Orange Counties. The watershed is part of the loess mantled un-glaciated residual landscapes and two-thirds of the watershed is covered by the Apalona, Gilpin, Bartle, Tilsit, Zanesville, and Wellston soils series. The TA used in this study were multi-resolution index of valley bottom flatness (MRVBF) and ridgetop flatness (MRRTF), slope gradient, altitude above channel network (AACHN), and topographic wetness index derived from a 5m DEM using System for Automated Geoscientific Analyses (SAGA). SSURGO soil map unit polygons were used for the analysis. SSURGO map units were grouped according to slope position. The terrain attribute means from each polygon within 500m of the county line were treated as independent and random observations. The terrain attributes were analyzed with ANOVA to detect a difference between counties. No significant differences were found between Dubois and Orange County within the grouped SSURGO soil map unit. This allowed for the correlation of SSURGO soil map units across the county line. The correlated SSURGO soil map units were overall significantly different from each other (p<0.001) with respect to terrain attributes. This reflects the use of common link of topography in soil landscape models used for mapping and terrain attributes. Thus, the use of terrain attributes can facilitate the automation of the soils and soil map units correlation process between counties in the MLRA.
See more from this Division: S05 Pedology
See more from this Session: General Pedology: II (Includes Graduate Student Competition)