Energy and evapotranspiration partitioning over a humid region orchard: Field measurements and partitioning model comparisons

Abstract

Understanding energy and water flux cycling and quantifying their components are essential steps to elucidate the hydrological process and sustainable water resources management in agricultural ecosystems. The challenge of partitioning energy and evapotranspiration (ET) components (canopy transpiration Tc and surface evaporation Es) was addressed by the combination of eddy covariance (EC) technique, micro-lysimeters, and sap flow methods in a humid region kiwifruit orchard of Southwest China. The variations of energy and water flux and their controls were assessed, then the measured ET and its component were compared with the simulated values by Priestley-Taylor Jet Propulsion Laboratory model (PT-JPL), Shuttleworth-Wallace (S-W), and FAO-56 dual crop coefficient models (Dual-FAO). The results showed that latent heat flux (LE) dominated the consumption of net radiation energy (Rn), with the averaged ratio (EF = LE/Rn) above 0.70 during the growing seasons. More radiation energy is released in the form of LE with increased surface moisture and vegetation cover (total leaf area index, LAIt). Tc accounted for 41%–46% ET. The increase of canopy leaf area index (LAIc) strongly affected ET partitioning through the early growing seasons, while atmospheric evaporation demand dominated the Tc/ET and even caused the energy to be redistributed since high atmospheric evaporation demand caused a decrease of Tc while an increase of Es during the active growing seasons (LAIc > 2.5 m2 m−2). All the ET models performed well in estimating ET (R2 = 0.84–0.92) and its components (Tc, R2 = 0.70–0.74; Es, R2 = 0.63–0.69), allowing individual estimation of water interception by the canopy is one of the merits of the PT-JPL model for ET simulation in this humid site, with R2 of 0.92, RMSE of 0.55 mm d−1, while the S-W model performed slightly better in ET component simulations (Tc, R2 = 0.74, RMSE = 0.35 mm d−1; Es, R2 = 0.69, RMSE = 0.38 mm d−1). The results of our study may have much referential significance for eco-hydrological process analysis and water flux estimation in regions with similar environmental conditions.