Large-scale mapping of Solonetzic soils using image analysis and global positioning systems (GPS)
Image analysis is a valuable tool that can be used to map soil properties that affect management practices in Solonetzic-Chemozemic soil landscapes. Because of the highly variable nature of these soil landscapes, only large-scale maps can provide the level of detail necessary for farm-scale planning. Image analysis of black and white aerial photographs can be used as a cost effective method to delineate soil landscape units within these soil landscapes. In this study, extensive field sampling and laboratory analysis were used to create a soil landscape model for a Brooking (Solonetzic)-Amulet (Chernozemic) soil landscape. Soil landscape units derived from this model were closely related to soil subgroups which in turn were associated with tonal variation on the scanned black and white aerial photograph. Image classification was achieved by grouping the digital values on the scanned photograph into classes based on soil landscape units. This procedure had a mapping accuracy of 71.4% at the specific sampling site pixel and 85.7% at a one pixel or 3 meter radius from the site. These levels of accuracy would indicate that this technique shows considerable promise for mapping soil landscape units within Solonetzic-Chemozemic map units.
Soils and Crops Workshop