Skip to main content
ARS Home » Research » Publications at this Location » Publication #242550
Title: An improved ASTER index for remote sensing of crop residue

Author
item SERBIN, GUY - Foreign Agricultural Service (FAS, USDA)
item Hunt Jr, Earle
item Daughtry, Craig
item Doraiswamy, Paul

Submitted to: Remote Sensing
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 11/5/2009
Publication Date: 11/11/2009
Citation: Serbin, G., Hunt Jr, E.R., Daughtry, C.S., Doraiswamy, P.C. 2009. An improved ASTER index for remote sensing of crop residue. Remote Sensing. 1:971-991.

Interpretive Summary: An important variable affecting the amount of organic matter in agricultural soils is the method of tillage, which can be estimated from the amount of crop residue (nonphotosynthetic vegetation) remaining in the field after harvest. Remote sensing allows for the rapid estimation of tillage method by determining crop residue cover in comparison with traditional ground-based methodology. If there is extensive residue cover, no-till was used, whereas if there is little residue cover, conventional tillage was used. The best index for estimating crop residue is the Cellulose Absorption Index, which is based on three narrow bands in the shortwave infrared. Unfortunately, these three bands are not available on operational sensors and adding more bands to satellite sensors costly. Therefore, the objective of this research was to assess other potential indices to determine the best compromise of cost and accuracy. A new index, the Shortwave Infrared Normalized Difference Residue Index (SINDRI), utilizing ASTER bands 6 and 7, is proposed for future multispectral sensors. This index was as reliable as the Cellulose Absorption Index under most conditions; the accuracy was reduced with the presence of green vegetation and some soil minerals. As ASTER is currently operating onboard the NASA Terra satellite, the technology is well established, therefore using the two ASTER bands instead of one band equivalent to Landsat band 7 would provide important information on carbon sequestration in agricultural and rangeland soils.

Technical Abstract: An important variable affecting carbon cycling in agricultural soils is the method of tillage, which can be estimated from the amount of crop residue (nonphotosynthetic vegetation) remaining in the field after harvest. Remote sensing allows for the rapid estimation of tillage method by determining crop residue cover in comparison with traditional ground-based methodology. The best index for estimating crop residue is the Cellulose Absorption Index (CAI); unfortunately, the three narrow bands in the shortwave infrared are not available on operational sensors. The addition of more bands is costly, so the goal is to determine the best compromise of cost and accuracy. A new index, the Shortwave Infrared Normalized Difference Residue Index (SINDRI), utilizing ASTER bands 6 and 7, is proposed for future multispectral sensors. SINDRI was compared against CAI and other indices at five locations in the USA on multiple dates. Results show that SINDRI performs almost as well as CAI and better than other indices. However, SINDRI is sensitive to green vegetation and can be adversely affected by wet surface conditions. These results show that a minimal upgrade from one broad band to two narrow bands would provide residue data for carbon cycle modeling.