Remote sensing-based energy balance for lettuce in an arid environment: influence of management scenarios on irrigation and evapotranspiration modeling

Efficient irrigation is critical for managing scarce water resources where precipitation is minimal. Field-scale irrigation is largely unaccounted for in landscape evapotranspiration models, primarily due to the unavailability of data and the lack of water balance components in energy balance-based evapotranspiration models. To overcome these challenges, we implemented a remote sensing-based energy and water balance model BAITSSS (Backward-Averaged Iterative Two-Source Surface temperature Solution) to calculate evapotranspiration (ET) and irrigation requirements of winter lettuce in the arid environment of the Lower Colorado River Basin. Predicted evapotranspiration and irrigation were compared against data from twelve eddy covariance (EC) sites for wide range of soil hydraulic properties operating between 2016 and 2020 and the applied irrigation, respectively. BAITSSS estimated evapotranspiration and irrigation based on vegetative formation, weather demand, soil hydraulic characteristics, and predefined management allowed depletion (MAD) (0.4–0.6). Ground-based weather data, Sentinel-2-based vegetation indices, and SSURGO (NRCS soil survey database) soil moisture characteristics were model inputs. The results showed mean seasonal ET from BAITSSS and EC were comparable, differing on average by about 7% based on a constant rooting depth (500 mm) and MAD of 0.5 for entire crop growth stages. Variations in daily and seasonal ET were mainly due to differences in applied and model-simulated irrigation. Seasonal values of applied and simulated irrigation closely agreed (~ 6%) in most sites, though some sites applied irrigation more effectively than others. Overall, this study provided insight into consumptive water use and field-scale irrigation practices, as well as the capabilities and limitations of model-simulated ET and irrigation.