Improved Crop Residue Cover Estimates from Satelite Images by Coupling Residue and Water Spectral Indices

Crop residues protect the soil against erosion, improve runoff water quality and determine C sequestration. Thus, the capability to assess crop residue cover can improve predictions of the impact of agricultural practices. Our objective was to develop a method to alleviate the adverse effect of variable moisture conditions on crop residue estimates from satellite imagery. Fields with uneven and with uniform water distribution were identified in satellite images (WorldView-3) from Maryland (USA). The results showed that moisture correction of spectral bands based on a water index reduced the root mean square error of most common residue indices, NDTI (Normalized Difference Tillage Index) and SINDRI (Shortwave Infrared Normalized Difference Residue Index). If bands are available, crop residue estimation should be based on SINDRI. If only Landsat or Sentinel-2 satellites are available, crop residues estimated combining NDTI with a water index could alleviate the adverse effect of variable moisture conditions.