Crop residue estimation using multiband reflectance

Soil erosion is very dependent on the protection effect of residue cover. We investigated hew remote sensing could be used to estimate the crop residue cover fraction. Laboratory reflectance measurements performed over corn residues in various moisture level show that, for any two wave bands, a residue line similar to that for bare soils can be defined. However it depends on the aging state of the residues, particularly in the near-infrared and middle-infrared domains. Field reflectance measurements performed over a range of crop cover fraction and soil backgrounds shore that a linear mixing model accurately describes the reflectance of the soil residue complex. An algorithm called CRIM (crop residue index multiband) is proposed. It is based on the linear mixing model of the soil-residue complex and on soil and residue lines. It can be used with any set of wave bands. Comparison with field measurements performed in Thematic Mapper (TM) bands TM2, TM3, TM3, TM4, and TM5 shows reasonably good estimates of the crop residue cover fraction compared with previously proposed indices (Normalized Difference Index and Soil Adjusted Corn Residue Index). The GRIM is a linear algorithm that is resistant to any linear transforms of the reflectance. It can therefore be used with ground-level reflectance, satellite radiance, or digital counts and should not strongly depend on view or sun geometry or on the spatial scale. The conditions for the GRIM computation and limits of its use are discussed. (C)Elsevier Science Inc., 1997.