Skip to main content
ARS Home » Plains Area » Fort Collins, Colorado » Center for Agricultural Resources Research » Water Management and Systems Research » Research » Publications at this Location » Publication #420699

Research Project: Improving Crop Performance and Precision Irrigation Management in Semi-Arid Regions through Data-Driven Research, AI, and Integrated Models

Location: Water Management and Systems Research

Title: A novel surface reflectance calibration approach to improve the estimation of maize actual evapotranspiration in a semi-arid USA region

Author
item COSTA-FILHO, EDSON - South Dakota School Of Mines And Technology
item CHAVEZ, JOSE - Colorado State University
item Zhang, Huihui

Submitted to: International Journal of Remote Sensing
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 8/26/2026
Publication Date: 9/6/2026
Citation: Costa-Filho, E., Chavez, J.L., Zhang, H. 2026. A novel surface reflectance calibration approach to improve the estimation of maize actual evapotranspiration in a semi-arid USA region. International Journal of Remote Sensing. https://doi.org/10.1080/01431161.2026.2727171.
DOI: https://doi.org/10.1080/01431161.2026.2727171

Interpretive Summary: This research aimed to enhance the accuracy of crop water use (ETa) estimates, a critical factor for sustainable agriculture. By combining multiple remote sensing and machine learning techniques, we developed a novel approach to separate soil and vegetation reflectance signals. The results showed the improved reliability of ETa models. Our findings highlight the importance of considering site-specific factors such as irrigation practices, crop type, and soil properties for accurate ETa estimation using different remote sensing data. These insights can help farmers and water resource managers make more informed decisions using remote sensing tools for irrigation management.

Technical Abstract: This study aimed to develop a novel surface reflectance data calibration approach to determine better maize crop evapotranspiration predictions in agricultural settings. The remote sensing (RS) platforms used in the study included Landsat-8 (30 m pixel spatial resolution), Sentinel-2 (10 m), Planet CubeSat (3 m), handheld multispectral radiometer or MSR (1 m), and a small uncrewed aerial system or sUAS (0.03 m). A total of six RS of actual crop evapotranspiration (ETa) algorithms were considered in the study: Three reflectance-based crop coefficient (RBCC) approaches, the one-source land surface energy balance (OSEB), and two dual-source land surface energy balance (TSEB). Five machine-learning (ML) algorithms were evaluated in the study: Linear regression, regression tree, random forest, support vector machine, and Gaussian process regression. The RS calibration approach uses a dual-source pixel decomposition model for surface reflectance that accounts for the contributions of soil and vegetation to a given surface reflectance band in the visible and invisible light radiation spectrum. Two maize research sites in Greeley and Fort Collins, Colorado, USA, in 2020 and 2021, provided the grounds for data collection. The Greeley site had a subsurface drip system, while the Fort Collins site had surface irrigation (furrow). Results indicated that the best ML regression model for a given RS sensor data and RS of the ETa algorithm depended on the surface reflectance composite (plant or bare soil values). The best ML models for adjusting RS data were the regression tree and the Gaussian process regression. These ML approaches adjusted RS data based on surface reflectance composites in combination with physically based models accounting for canopy architecture and surface reflectance spectral response.