A new paper published in Remote Sensing by Resolv, Inc. argues that precision agriculture can finally deliver on its promise if satellite imagery is reliably corrected to surface reflectance. The open-access study, “Surface Reflectance: An Image Standard to Upgrade Precision Agriculture,” benchmarks three atmospheric correction methods and demonstrates how a standard correction approach can unlock low-cost, automated crop intelligence.
Atmospheric correction is essential because light traveling through the atmosphere is distorted before reaching a satellite sensor. Without accurate correction, small clouds and shadows can be misinterpreted as crop problems, leading to false alarms and costly scouting. The paper argues that this has stalled precision agriculture’s adoption.
The Resolv team compared two mainstream tools, Sen2Cor and FORCE, against CMAC, a closed-form method developed by Resolv. Across a range of atmospheric conditions, CMAC produced precise surface reflectance estimates, while the mainstream tools showed systematic bias—over-correcting clear images and under-correcting hazy ones. According to the paper, this bias had gone undetected until now.
Reliable surface reflectance enables several applications: automated cloud and shadow removal, a crop start-date index to replace growing-degree-day scheduling, stable NDVI readings even with varying atmospheric water vapor, soil capability classification from imagery, and accurate remote crop irrigation based on greenness and reference evapotranspiration. Together, these could make precision agriculture self-funding.
The paper also proposes a tiered approach to reduce imagery costs. Tier 1 uses free Sentinel-2 data corrected to surface reflectance. Tier 2 fills gaps with commercial smallsat data, which can be resampled, verified, and billed automatically. This could create a turnkey pipeline for ordering, correcting, analyzing, and billing imagery without manual intervention, lowering service costs and increasing sales volume. Crop insurance could serve as a natural channel for such a system.
According to Resolv, reliable surface reflectance imagery can finally close the gap between remote sensing’s promises and its real-world performance in agriculture.
The paper is available at Resolv’s website, along with other peer-reviewed studies. The research was funded by the National Science Foundation SBIR program.
