Testing DRAINMOD-FOREST for predicting evapotranspiration in a mid-rotation pine plantation
Evapotranspiration (ET) is a key component of the hydrologic cycle in terrestrial ecosystems and accurate description of ET processes is essential for developing reliable ecohydrological models. This study investigated the accuracy of ET prediction by the DRAINMOD-FOREST after its calibration/validation for predicting commonly measured hydrological variables. The model was tested by conducting an eight year simulation of drainage and shallow groundwater dynamics in a managed mid-rotation loblolly pine (Pinus taeda L.) plantation located in the coastal plain of North Carolina, USA. Modeled daily ET rates were compared to those measured in the field using the eddy covariance technique. In addition, the wavelet transform and coherence analysis were used to compare ET predictions and measurements on the time–frequency domain. Results showed that DRAINMOD-FOREST accurately predicted annual and monthly ET after a successful calibration and validation using measured drainage rates and water table depth. The model under predicted ET on an annual basis by 2%, while the Nash–Sutcliffe coefficient of model predictions on a monthly basis was 0.78. Results from wavelet transform and coherence analysis demonstrated that the model reasonably captured the high power spectra of ET at an annual scale with significantly high model-data coherency. These results suggested that the calibrated DRAINMOD-FOREST collectively captured key factors and mechanisms controlling ET dynamics in the drained pine plantation. However, the global power spectrum revealed that the model over predicted the power spectrum of ET at an annual scale, suggesting the model may have under predicted canopy conductance during nongrowing seasons. In addition, this study also suggested that DRAINMOD-FOREST did not properly capture the seasonal dynamics of ET under extreme drought conditions with deeper water table depths. These results suggested further refinement to parameters, particularly vegetation related, and structures of DRAINMOD-FOREST to achieve better agreement between ET predictions and measurements in the time–frequency domain.