Multi-scale evapotranspiration and evaporative stress as indicators of vegetation health and drought response – an application in central Missouri, USA

Authors: Anderson, Martha; Schreiner-Mcgraw, Adam; WOOD, J - University Of Missouri; Gao, Feng; YANG, Y - Cornell University; LIU, H - Cornell University; DUAN, W - Cornell University; MISHRA, V - University Of Alabama In Huntsville; HAIN, C - Nasa Marshall Space Flight Center; ZHAO, H - George Mason University; CHANG, J - George Mason University; Cirone, Richard

Submitted to: Remote Sensing of Environment
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 7/26/2026
Publication Date: 8/1/2026
Citation: Anderson, M.C., Schreiner-Mcgraw, A.P., Wood, J., Gao, F.N., Yang, Y., Liu, H., Duan, W., Mishra, V., Hain, C., Zhao, H., Chang, J., Cirone, R.J. 2026. Multi-scale evapotranspiration and evaporative stress as indicators of vegetation health and drought response – an application in central Missouri, USA. Remote Sensing of Environment. 190. Article e115286.. https://doi.org/10.1016/j.ecolind.2026.115286.
DOI: https://doi.org/10.1016/j.ecolind.2026.115286

Interpretive Summary: Satellite data can track how much water plants use and how stressed they are, and products at different pixel sizes offer information from field to regional scales. These products vary in detail and frequency, but fine resolution data often miss temporal continuity while coarse data miss local field level differences. Therefore, we evaluated 4 km, 500 m, and 30 m evapotranspiration and Evaporative Stress Index (ESI) products across central Missouri from 2017–2023, finding that coarse products best capture regional drought signals, while 30 m products reveal fine scale management effects and local stress. The combined multi scale approach provides complementary insights that support operational use of 30 m ESI in platforms like OpenET and highlight the value of multi sensor thermal satellite systems for agricultural drought monitoring.

Technical Abstract: Satellite estimates of evapotranspiration (ET) based on thermal infrared retrievals of land-surface temperature enable monitoring of vegetation water use and health/stress at sub-field to global scales. The information content of these products varies with pixel size, reflecting dominant drivers of land-surface fluxes at different scales as well as tradeoffs in satellite spatial vs. temporal coverage. This study evaluates the performance and utility of multi-scale ET and Evaporative Stress Index (ESI) products at 4-km (GOES), 500-m (VIIRS-M band), and 30-m (Landsat+ECOSTRESS+VIIRS-I5 band, spatially sharpened) resolutions across a heterogeneous agricultural-ecological landscape in central Missouri, USA. At 30-m, the modeling framework combines multiple Landsat-like thermal and optical datasets to improve temporal sampling at that critical agricultural management scale. Daily ET and weekly ESI are compared with eddy-covariance fluxes, phenocam observations, drought classifications, and crop condition data collected over the Central Missouri River Basin from 2017-2023, a period including significant drought events. Results show that 30-m ET and ESI capture fine-scale management and vegetation heterogeneity, while coarser products provide superior temporal continuity and stronger correspondence with regional drought. A rotation-correction methodology applied to the 30-m ESI reduces crop-rotation artifacts and improves timing and magnitude of stress detection relative to standard pixel-based approaches. Together, the multi-scale products reveal complementary drought-response signals: coarse ESI captures regional atmospheric forcing, whereas 30-m ESI resolves local drivers of stress, including management actions, post-disturbance recovery, and landcover-specific drought sensitivity. These findings support operational implementation of 30-m ESI within cloud platforms such as OpenET and highlight the growing value of multi-sensor thermal constellations for field-scale drought monitoring and agricultural decision-support.