Integrating the Maximum Entropy Production model and airborne imagery for understorey evapotranspiration mapping

While extensive research has focused on evapotranspiration (ET) from land surface, the spatial distributions of ET of the woodland and forest understorey remain poorly understood. This study developed a method for estimating spatially distributed understorey ET by integrating the Maximum Entropy Production model with airborne thermal imagery. Validation against ground-truth estimation showed good model performance (R² = 0.93, RMSE = 0.03 mm/h), confirming its efficacy across different land cover types, including open and understorey areas. The results revealed significant spatial heterogeneity in understorey ET with varying vegetation cover and topographic attributes, and distinct responses to wetting events. This method provides a new tool for estimating the important understorey water consumption in forests and woodlands, contributing to assessing ecosystem water use efficiency and improving water resource and vegetation management strategies.