Remote Sensing of Soil Tillage Intensity In Central Iowa.

Crop residues on the soil surface decrease soil erosion, increase soil organic matter, improve soil quality, and reduce the amount of nutrients and pesticides that reaches stream and rivers. Crop residue cover is often used to classify soil tillage intensity and assess the extent of conservation tillage practices and bio-fuel harvesting. The NRCS standard technique for measuring crop residue cover is the line-point transect method which is time-consuming, prone to errors, and not suitable for watershed scale studies.

Our objective was to assess crop residue cover and soil tillage intensity in the South Fork watershed, a USDA Conservation Effects Assessment Project (CEAP) watershed.  Cropland accounts for 88% of the 788 km² watershed and corn and soybeans are grown on 99% of the cropland.  With relatively broad band multispectral sensors, such as Landsat TM, SPOT, or AWiFS, crop residues can be brighter or darker than soils depending on soil type, crop type, moisture content, and residue age.  Acceptable classification accuracies for 2-3 tillage classes were often possible, but required timely, scene-specific surface reference data for training.  With hyperspectral data, physically-based spectral indices that detect absorption features associated with cellulose and lignin were linearly related to crop residue cover.  These indices were robust and required minimal surface reference data for mapping soil tillage intensity across agricultural landscapes.  Unfortunately, current satellite hyperspectral systems were not capable of imaging the entire watershed in a timely manner.  Stratified sampling protocols were developed that used the limited hyperspectral images to provide reliable data to train classifiers of multispectral images.  Watershed and regional surveys of soil management practices that affect soil and water quality are possible.