Submitted to: Soil and Water Conservation Society
Publication Type: Abstract Only
Publication Acceptance Date: March 30, 2007
Publication Date: July 23, 2007
Citation: Serbin, G., Daughtry, C.S., Hunt, E.R., Doraiswamy, P.C., Brown, D.J. 2007. Hyperspectral remote sensing estimation of crop residue cover: Soil mineralogy, surface conditions, and their effects [abstract]. Soil and Water Conservation Society. 2007 CDROM. Technical Abstract: Conservation tillage practices can enhance soil organic carbon content (SOC), improve soil structure, and reduce erosion. However, direct assessment of tillage practice for monitoring SOC change over large regions is difficult. Remote sensing of crop residue cover (CRC) can help assess tillage practices, and thus, model changes in SOC. Our objectives were to evaluate several spectral indices from hyperspectral imagery of Iowa and Indiana during May of 2004, 2005, and May 2006 for measuring CRC and assess surface moisture and mineralogy effects on those indices. CRC was measured in selected corn and soybean fields using line-point transects. CRC was linearly correlated (r2=0.70 ~ 0.85) to the cellulose absorption index (CAI), which measures the relative intensity of cellulose and lignin absorptions near 2100 nm. Soil type and mineralogy effects were evaluated using reflectance spectra libraries of the USGS Spectroscopy Laboratory and the US National Soil Survey Center. Reflectance spectra of crop residues and selected soils at various moisture contents were measured in the laboratory (400-2500 nm wavelength range). Results show that all dry crop residues had positive CAI values, but all dry soils and rocks had negative CAI values. However, certain minerals, when present in significant quantities, could bias CAI either way. Moisture in CRC significantly attenuated CAI but was accountable. This suggests that local adjustment of remotely sensed CRC estimates may be required where soil composition or moisture change significantly. Nevertheless, regional surveys of CRC and soil tillage practices that affect SOC dynamics may be feasible using hyperspectral imaging systems.