Crop Residue

The Institute for Technology Development (ITD) partnered with the U.S. Department of Agriculture’s (USDA) Natural Resource Conservation Service (NRCS) to investigate the use of remotely sensed imagery from the NASA Science Mission Directorate’s suite of Earth observing sensors to map crop residue and tillage practices. The NRCS national residue mapping program is done biannually due to the high costs involved. They are only able to survey a small number of fields, and the methodology is subjective. Using remotely sensed imagery may help to reduce the costs, as well as increase the accuracy by allowing many more fields to be used in the program.

Soil erosion and its related runoff is a serious problem in U.S. agriculture. The USDA has classified 33% of U.S. agricultural land as being highly erodible (Figure 1). The USDA-NRCS has developed initiatives to address this problem. These include the Biennial National Crop Residue Management Survey which estimates acreage by tillage practice, the Environmental Quality Incentives Program (EQUIP) which encourages and rewards conservation tillage, and the Conservation Security Program (CSP) which also encourages and rewards conservation tillage. All of these programs depend, directly or indirectly, on being able to determine the crop residue cover percentage, which implies the tillage practice, for a field. The methods include windshield surveys, line transects and photo-comparison.

Figure 1. Distribution of highly erodible cropland in the U.S.

Five counties in two states were selected as the study area. This represented corn, soybeans and winter wheat crops over a variety of soil types and growing conditions. The counties in the study sites were suggested by NRCS. In addition, NRCS and state personnel in these counties agreed to assist ITD in conducting the ground truth sampling. As shown in Figure 2, these sites include Scott and Rice Counties in Kansas; and Wabash, Huntington, and Grant Counties in Indiana.

Kansas contains a large number of acres of highly erodible cropland, while Indiana has a significant variation in the amount of crop residue coverage within its agricultural lands. The sites contain approximately 1500 fields.

Figure 2. Crop Residue Project study sites

The overall goal of this project was to demonstrate that remotely sensed imagery can estimate crop residue cover fractions over different tillage practices, soil types, and crops.

The results indicate that imagery has a greater ability to identify residue within corn fields than within soybean fields. The biomass produced by corn is substantially larger than that produced by soybeans, thus greater amounts of residue remain in the field after the crop is harvested. Overall, the greater amount of crop residue likely yields a stronger response in the SWIR bands. In addition, the substantially lower amount of biomass produced by soybeans likely contributes to more soil spectra being recorded by the sensors resulting in confounding of the cellulose-lignin detection.

For more information about this or any other past research, contact us.