Using high-spatiotemporal thermal satellite ET retrievals to monitor water use over California vineyards of different climate, vine variety and trellis design.

• ALEXI/DisALEXI provides reliable ET estimates over range of vineyards. • Evaluation considered different varietals, climate, and trellis types. • Spatial maps highlight variability in moisture dynamics within vineyards. • Errors in LAI inputs lead to biases in modeled ET. • Spatially distributed actual ET can inform sustainable irrigation practices. Mapping the spatial variability of actual evapotranspiration (ET a) across vineyards is useful for optimizing irrigation scheduling and efficiency, leading to conservation of water resources and more sustainable wine grape production. To support efficient irrigation strategies, we investigate the utility of thermal infrared-based ET a maps over a range of vineyards located throughout California, each representing a unique local climate, trellis design, grape variety, row orientation and management practice. ET a maps are derived by combining the Disaggregated Atmosphere Land Exchange Inverse (ALEXI/DisALEXI) surface energy balance model and the Spatial Temporal Adaptive Reflectance Fusion Model (STARFM) to generate ET a estimates at high spatial (30 m) and temporal (daily) resolution. Model output is evaluated for years 2017 and 2018 over vineyard sites located in Sonoma, Sacramento, and Madera counties in California that are being monitored as part of the Grape Remote sensing Atmospheric Profile and Evapotranspiration eXperiment (GRAPEX). Overall, modeled daily ET estimates compare well with flux tower observations, with average root mean square error (RMSE), mean absolute error (MAE) and mean bias error (MBE) of 0.88 mm day−1, 0.70 mm day−1, and 0.17 mm day−1 respectively, over all four individual vineyard locations, aligning with past GRAPEX studies. Despite general agreement, record wildfires in northern California during 2018 likely resulted in positive model bias, while misrepresentation of leaf area index within a double-trellis designed canopy at the southern-most vineyard resulted in negative model bias. Spatial analysis of monthly total ET highlights the advantages of utilizing a satellite-based approach to characterize the variability in water use within and surrounding the targeted vineyards. A reliable spatial ET product at scale has the potential to improve water allocation and conservation efforts by identifying areas of uneven water use due to variations in soil texture and composition and other environmental or anthropogenic factors. [ABSTRACT FROM AUTHOR]