ARS | Publication request: Evaluation of Spectral Indices for Estimating Crop Residue Cover and Soil Carbon

Submitted to: USDA Greenhouse Gas Symposium
Publication Type: Abstract Only
Publication Acceptance Date: November 21, 2006
Publication Date: February 5, 2007
Citation: Serbin, G., Daughtry, C.S., Hunt, E.R, Doraiswamy, P.C., Brown, D.J. 2006. Evaluation of spectral indices for estimating crop residue cover and soil carbon [abstract]. Fourth USDA Greenhouse Gas Conference: Positioning Agriculture and Forestry to Meet the Challenges of Climate Change. 2006 CDROM.

Technical Abstract: Long term use of conservation tillage practices can lead to increased soil organic carbon (SOC) compared to intensively tilled soils. However, monitoring soil tillage intensity over large areas for assessing changes in SOC is difficult. Remote sensing can potentially estimate crop residue cover, an important indicator of soil tillage intensity, for modeling soil carbon. Aerial hyperspectral images were acquired over agricultural fields in Indiana in May 2006. Crop residue cover was measured in corn and soybean fields using line-point transects, and was linearly related to the Cellulose Absorption Index (CAI) for the whole scene (r2 = 0.70). By combining information on previous season’s crop classification (USDA-NASS 2005 crop data layer product) with crop residue cover after planting in 2006, an inventory of soil tillage intensity by previous crop type was generated for the whole scene (5 km x 23.5 km). As CAI values for bare soils were variable, our next objective was to evaluate the effects of a wide range of soil types on several spectral crop residue indices, and to use these spectral indices for measuring crop residue cover in agricultural fields using aerial hyperspectral images. Reflectance spectra acquired by Brown et al. (2006, Geoderma 132 pp. 273-290) of over 3,700 soils from the US National Soil Survey Center- Soil Survey Laboratory database were analyzed. Mean CAI values for soils were -2.2±0.8 for surface soil samples (0-1.5 cm depth) and -3.5±1.5 for all samples analyzed. In contrast, values for dry crop residues were typically 5.0±1.1. Although CAI moderately correlated (r2= 0.40) with organic carbon content in surface soils (0-1.5 cm sample depth), the correlation decreased with depth suggesting that CAI is primarily sensitive to relatively undegraded cellulose and lignin and not to other forms of SOC, e.g., humic and fulvic acids.