A Bayesian regression modeling framework for identifying land surface temperature and evapotranspiration treatment effects in UAV and airborne data
Environmental Research: Water · 01.09.2026 · 2 (3) · 035002
DOI: 10.1088/3033-4942/ae774e ↗Kurz gefasst
Fachartikel in der Zeitschrift Environmental Research: Water, online erschienen am 22.06.2026. Verfasst von Logan A Ebert und 4 weiteren Autor:innen. Die Arbeit nennt 68 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: California Tomato Research Institute, California League of Food Production, California Department of Food and Agriculture.
Abstract (Originalsprache)
Remote sensing is becoming an integral part of agricultural management, not only for monitoring but also for detecting crop responses to different management strategies (e.g. deficit irrigation). High resolution spectral data collected from airborne or unmanned aerial vehicles (UAVs) can facilitate the detection of treatment responses across replicated, field scale studies. However, frequentist statistical models, like linear regressions and generalized additive models, may not be suitable for identifying treatment responses using the spectral data due to the inherent spatial autocorrelation. In this study, we used Bayesian regression modeling to evaluate the effects of four deficit irrigation treatments on land surface temperature (LST) and evapotranspiration (ET) in processing tomatoes in the San Joaquin Valley of California. To address growing water scarcity challenges for processing tomato production, this study focused on identifying responses from four deficit irrigation treatments implemented across two growing seasons. We collected LST data and modeled ET multiple times throughout both seasons using high-resolution airborne and UAV imagery to assess treatment effects at field scale. We observed significant differences in LST and ET across treatments, with higher water deficits resulting in increased LST and reduced ET, supporting the utility of Bayesian approaches for identifying field-scale treatment responses. These findings underscore the potential for combining remote sensing with spatial regression modeling to detect crop responses to different water management strategies in agriculture. By providing precise, actionable insights into how crops respond to irrigation strategies, this approach offers scalable solutions to address increasing water scarcity as well as potential for application across agriculture with high resolution spectral data.
Förderung
- California Tomato Research Institute
- California League of Food Production
- California Department of Food and Agriculture
Zitieren
APA
Ebert, L. A., Gal, A., Brooks, W., Jha, G., & Nocco, M. A. (2026). A Bayesian regression modeling framework for identifying land surface temperature and evapotranspiration treatment effects in UAV and airborne data. Environmental Research: Water, 2(3), 035002. https://doi.org/10.1088/3033-4942/ae774e
Vancouver
Ebert LA, Gal A, Brooks W, Jha G, Nocco MA. A Bayesian regression modeling framework for identifying land surface temperature and evapotranspiration treatment effects in UAV and airborne data. Environmental Research: Water. 2026;2(3):035002. doi:10.1088/3033-4942/ae774e
BibTeX
@article{Ebert2026,
author = {Ebert, Logan A and Gal, Andrew and Brooks, Wesley and Jha, Gaurav and Nocco, Mallika Arudi},
title = {A Bayesian regression modeling framework for identifying land surface temperature and evapotranspiration treatment effects in UAV and airborne data},
journal = {Environmental Research: Water},
year = {2026},
volume = {2},
number = {3},
pages = {035002},
doi = {10.1088/3033-4942/ae774e}
}
EUVIMED