Simulated response of evapotranspiration to deforestation in Southeast Asian Massif: insights from WRF simulations using three land surface models

Southeast Asian Massif had undergone rapid deforestation over mountainous areas in the early 21^st century, bringing crucial impacts on regional climate via changing evapotranspiration. Compared with Earth System Models that may omit the fragmental mountain deforestation owing to their coarse resolutions, regional climate models are powerful tools in assessing the biophysical feedbacks of mountain deforestation due to their fine grid spacing and sophisticated structure. Within both Earth System Models and regional climate models, it is land surface schemes in the models that regulate the fluxes of heat, carbon and water vapour between land surface and the atmosphere, thus may largely affect the pattern and magnitude of how land surface change impacts climate. Here we investigate the response of evapotranspiration to deforestation in Southeast Asian Massif by employing three land surface schemes (Noah mosaic, NoahMP, and CLM) in the Weather Research and Forecasting (WRF) model. We perform the analyses by designing two sets of experiments with comparative land-cover scenarios derived from high-resolution forest cover change dataset. 75% of the land grids experienced forest cover change while the dominant land cover type had changed in only 2% grids. Theoretically, CLM scheme that considers all plant function types (PFT) should be more suitable for performing the climate feedbacks of mountain deforestation than those schemes that only use the dominant type in the grid cell such as NoahMP. However, we find that only the Noah mosaic scheme which takes the sub-grid approach is able to reasonably reproduce the response of evapotranspiration to deforestation in mosaic pixels. Both the NoahMP and CLM fail to accurately capture the evapotranspiration response of deforestation in those mosaic grids that the dominant type had not change.