Skip to main content
ARS Home » Southeast Area » Stoneville, Mississippi » Crop Production Systems Research » Research » Publications at this Location » Publication #401037
Research Project: Development of Productive, Profitable, and Sustainable Crop Production Systems for the Mid-South

Location: Crop Production Systems Research

Title: Mixed-species cover crop biomass estimation using planet imagery

Author
item Kharel, Tulsi
item Bhandari, Ammar
item Mubvumba, Partson
item Tyler, Heather
item Fletcher, Reginald
item Reddy, Krishna

Submitted to: Sensors
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 1/26/2023
Publication Date: 1/31/2023
Citation: Kharel, T.P., Bhandari, A.B., Mubvumba, P., Tyler, H.L., Fletcher, R.S., Reddy, K.N. 2023. Mix species cover crop biomass estimation using planet imagery. Sensors. https://doi.org/10.3390/s23031541.
DOI: https://doi.org/10.3390/s23031541

Interpretive Summary: Cover crop (CC) is promoted as one of the approaches for soil resource conservation while utilizing the land under a sustainable crop production system. Cover crop biomass is helpful for weed and pest control, soil erosion control, nutrient re-cycling, and overall soil health and crop productivity improvement. These benefits may vary based on cover crop species and their biomass. There is growing interest in the agricultural sector of using remotely sensed imagery to estimate cover crop biomass. Estimating biomass yield and benefit with remote sensing requires reflectance characteristics of each species and their mix, a task complicated by mixed species. Scientists from the USDA-ARS Crop Production Systems Research Unit, Stoneville, MS selected four on-going experiments that have different cover crop species and their mixture to evaluate feasibility of mix species biomass estimation using remotely sensed satellite imagery. High-resolution (3 m) satellite imageries throughout the cover crop season from November to April were analyzed and used to build random forest model to predict CC biomass. Results showed that reflectance bands and vegetation indices (VIs) derived from Planet imagery collected during March were more strongly correlated with biomass compared to imagery from November and April. The highest correlation was observed with near infrared (NIR) band during March. Cover crop biomass prediction was better when cover crop species/mix information was added along with satellite imagery bands and VIs on the model. This result suggest that if classification of cover crop species (or mix species) is carried out before hand, biomass prediction using random forest model will improve. The result from this study helps scientist to design appropriate study for machine learning approach, select appropriate bands and VIs from satellite imagery and choose appropriate timing of canopy sensing to estimate mix cover crop biomass.

Technical Abstract: Cover crop (CC) biomass is helpful for weed and pest control, soil erosion control, nutrient re-cycling, and overall soil health and crop productivity improvement. These benefits may vary based on cover crop species and their biomass. There is growing interest in the agricultural sector of using remotely sensed imagery to estimate cover crop biomass. Estimating biomass yield and benefit with remote sensing requires reflectance characteristics of each species and their mix, a task complicated by mixed species. Four small plot studies located at the USDA-ARS Crop Production Systems Research Unit farm, Stoneville, MS with different cereal, legume, and their mixture as fall-seeded cover crops were selected for this analysis. All four studies were in randomized complete block design with four replications. Cover crop biomass and canopy-level hyperspectral data were collected at the end of April, just before cover crop termination. High-resolution (3 m) satellite imageries were collected throughout the cover crop season from November to April in the 2021 and 2022 study cycles. Reflectance bands and vegetation indices (VIs) derived from imagery collected during March were more strongly correlated with biomass (r = 0-0.74) compared to imagery from November (r = 0.01-0.41) and April (r = 0.03 – 0.57). The highest correlation was observed with near infrared (NIR) band (r = 0.74) during March. The R2 for biomass prediction with the random forest model improved from 0.25 to 0.61 when cover crop species/mix information was added along with satellite imagery bands. More study with multiple timepoint biomass, hyper-spectral, and imageries collection is needed to choose appropriate bands and estimate the biomass of mix cover crop species.