Skip to main content
ARS Home » Northeast Area » Beltsville, Maryland (BARC) » Beltsville Agricultural Research Center » Hydrology and Remote Sensing Laboratory » Research » Publications at this Location » Publication #406056
Research Project: Enhancing Agricultural Management and Conservation Practices by Advancing Measurement Techniques and Improving Modeling Across Scales

Location: Hydrology and Remote Sensing Laboratory

Title: Comparison of five spectral indices and six imagery classification techniques for assessment of crop residue cover using four years of Landsat imagery

Author
item Stern, Alan
item Daughtry, Craig
item Hunt Jr, Earle
item Gao, Feng
item HIVELY, W - Us Geological Survey (USGS)

Submitted to: Remote Sensing
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 9/14/2023
Publication Date: 9/19/2023
Citation: Stern, A.J., Daughtry, C.S., Hunt Jr, E.R., Gao, F.N., Hively, W.D. 2023. Comparison of five spectral indices and six imagery classification techniques for assessment of crop residue cover using four years of Landsat imagery. Remote Sensing. Remote Sens. 2023, 15(18), 4596. https://doi.org/10.3390/rs15184596.
DOI: https://doi.org/10.3390/rs15184596

Interpretive Summary: Crop residue cover helps reduce erosion, increase soil organic carbon, and improve water quality. Estimating residue cover using remote sensing has been studied, yet it is still challenging due to the similarity of soil and crop residue signals at visible and near-infrared wavelengths. This study compares the accuracy of crop residue mapping using five spectral indices and six imagery classifications over four years in central Iowa. We found that crop residue cover and soil tillage intensity can be mapped using the selected indices and classification methods. However, the accuracies obtained from spectral indices and classification methods vary across different years, and the root mean square error is a better indicator of accuracy than the coefficient of determination. The study demonstrates that remote sensing provides a viable means to map crop residue cover on a large scale, thereby supporting agroecosystem monitoring.

Technical Abstract: Determining residue cover on agricultural land is an important task. Residue cover helps reduce soil erosion and helps sequester carbon. Many studies have used either spectral indices or classification methods to map residue cover using satellite imagery. Unfortunately, most of these studies use only a few indices or classification methods and generally only study an area for a single year. This manuscript investigates five spectral indices and six classification methods over four years, to determine if a single spectral index or classification method performs consistently better than the others. A second objective is to determine whether using the R2 from the relationship between residue cover and a spectral index is a reasonable substitute for calculating accuracy. Field visits were conducted for each of the years studied and used to create the correlations with the spectral indices and as ground truth for the classification methods. It was found that there isn’t a spectral index/classification method that is consistently better than all the others. Classification methods tended to be more accurate in 2011, 2013 while spectral indices tended to be more accurate in 2015, 2018. For the second objective, it was found that R2 is not a great indicator of accuracy. RMSE is a better indicator of accuracy than R2, however simply calculating accuracy would be the best of all.