Evaluation of water use efficiency model in evapotranspiration partitioning from high-frequency eddy covariance data - a comparison between plantation sites

Water use efficiency (WUE) represents a crucial parameter and a major source of error influencing the accuracy of evapotranspiration separation through flux variance similarity (FVS). However, there is a paucity of studies investigating the application of FVS for the separation of evapotranspiration across multiple plantations, thereby constraining our understanding of water use strategies in plantation forests. This study assessed the outcomes of four intercellular carbon dioxide models and an optimization model (Opt) within the FVS framework, utilizing Eddy Covariance (EC) data gathered from two broad-leaf plantations and four coniferous plantations. Additionally, we examined the relationship between WUE and solution rate, as well as the responsiveness of the proportion of transpiration to evapotranspiration (T:ET) to soil moisture. The results showed that the solution rate of the Sqrt model differed by only 1 % - 2 % from the constant Ci value (Ci) and the constant Ci/Ca ratio (K) model in plantations, excluding Pinus tabuliformis. Moreover, the Sqrt model exhibited an average overestimation of T:ET by 8 % compared to the Opt model. The solution rate of Ci, K, and Sqrt models were decreased with an increase in WUE. At a half-hour scale and daily scale, the plantation's solution rate was ranged from 39 % to 67 %, and 29 % to 76 %, respectively. Except for the Linear model, all other models effectively tracked the dynamic changes in soil moisture. This study enhanced our understanding about the distinctions and consistencies in FVS separation among five WUE models in plantations, providing a foundation for the selection of WUE models to elucidate water use strategies in plantations.