Assessing evapotranspiration in a lettuce crop with a two-source energy balance model

Winter vegetables, including lettuce, are a significant consumptive use of water in the Lower Colorado River Basin. Precise irrigation management is needed to increase water use efficiency and reduce the negative impacts of suboptimal irrigation, including nutrient leaching, crop stress, and crop pathogens. However, lettuce has multiple features that make accurate evapotranspiration (ET) modeling difficult, including asynchronicity with meteorological evaporative demand, short growing seasons, and a shallow root zone that increases the risk of using an incorrect ET value. To improve ET modeling and understand applied irrigation effectiveness for lettuce in this region, we used an energy and water balance bio-physical model, Backward-Averaged Iterative Two-Source Surface temperature and energy balance Solution (BAITSSS) on arid farmlands in the lower Colorado River basin. The study was conducted between 2016 and 2020 at twelve eddy covariance (EC) sites in lettuce with a wide range of soil physical properties. BAITSSS was implemented using ground-based weather and irrigation data, and remote sensing-based vegetation indices (Sentinel-2). The model accuracy varied among sites, with a mean cumulative seasonal ET of similar to 3% and mean RMSE of 1.1 mm d(-1) when compared to EC. The results showed that accurate timing and amount of applied water (irrigation and precipitation) were critical to capturing ET spikes right after irrigation and tracking the continuous decrease of ET. This study highlighted the dominant factors that influence the ET of lettuce and how BAITSSS can improve ET modeling for irrigation management.