Remote Sensing Crop residue Cover and Soil Tillage Intensity
Craig
S. T.
Daughtry, USDA-ARS Hydrology and Remote Sensing Lab, cdaughtry@hydrolab.arsusda.gov
(Presenting)
Paul
C.
Doraiswamy, USDA-ARS Hydrology and Remote Sensing Lab, pdoraisw@hydrolab.arsusda.gov
E.
Raymond
Hunt, USDA-ARS Hydrology and Remote Sensing Lab, erhunt@hydrolab.arsusda.gov
Current methods of quantifying crop residue cover are inadequate for characterizing the spatial variability of residue cover within fields or across large regions. Our objectives were to evaluate several spectral indices for measuring crop residue cover using satellite hyperspectral data and to categorize soil tillage intensity in agricultural fields. Hyperion imaging spectrometer data were acquired over agricultural fields in central Iowa in May 2004 and 2005. Crop residue cover was measured in corn and soybean fields using line-point transects. Crop residue cover was linearly related to the cellulose absorption index (CAI) with coefficients of determination (r2) of 0.85 in 2004 and 0.69 in 2005. Three tillage intensity classes, corresponding to intensive (<15% residue cover), reduced (15-30% cover), and conservation (>30% cover) tillage, were correctly identified in 63-68% of fields. Classification accuracy increased to 68-82% for two classes, corresponding to conventional (intensive + reduced) and conservation tillage. Inventories of soil tillage intensity by previous crop type were generated for the whole Hyperion scene in each year. Regional surveys of soil management practices that affect soil conservation and soil C dynamics are possible using advanced multispectral or hyperspectral imaging systems.