Application of Remote Sensing Based Tillage Mapping Technique to Evaluate Water Quality Impacts of Tillage Management Decisions in Upper White River Basin

Tillage practices directly impact runoff processes, erosion and water quality in agricultural watersheds. Consequently, environment models require tillage information for water quality modeling but often this information is not available at required spatial and temporal scales. Remote sensing approach facilitates tillage mapping at a larger scale than the conventional surveying methods. Models based on remote sensing can classify contrasting tillage practices with accuracy of 80% to 92%. The objectives of this study were to use reflectance based logistic regression models to classify contrasting tillage practices in agricultural watersheds using Landsat imagery. This study is conducted in the Upper White River Basin (7,000 km²) located in central Indiana, an agricultural watershed with corn-soybean rotation. A single tillage practice can have different level of impact at different slope classes within the watershed therefore; the tillage system information is coupled with the slope classes within a watershed using geographic information system (GIS). This can identify the optimum slope class for a tillage system for minimum sediment loss. The tillage data derived are input in the Soil and Water Assessment Tool (SWAT) model to evaluate the distribution of tillage practices and their impacts on runoff, sediment, and pollutant losses.