Deforestation reshapes land-surface energy-flux partitioning

Abstract Land-use and land-cover change significantly modify local land-surface characteristics water/energy exchanges, which can lead to atmospheric circulation regional climate changes. In particular, deforestation accounts for a large portion of global land-use changes, transforms forests into other land cover types, such as croplands grazing lands. Many previous efforts have focused on observing modeling land–atmosphere–water/energy fluxes investigate land–atmosphere coupling induced by deforestation. However, interpreting land–atmosphere–water/energy-flux responses is often complicated the concurrent impacts from shifts in properties versus background forcings. this study, we used 29 paired FLUXNET sites, improve understanding how deforested surfaces drive changes surface-energy-flux partitioning. Each sites included an intact forested non-forested site that had similar climate. We employed transfer entropy, method based information theory, diagnose directional controls between variables, identify nonlinear cause–effect relationships. Transfer entropy powerful tool detective causal relationships asynchronous systems. The eddy covariance flux measurements showed consistent strong flows vegetation activity (gross primary productivity (GPP)) physical (e.g. shortwave radiation, air temperature) evaporative fraction (EF) over both surfaces. More importantly, transfers precipitation, GPP EF were reduced at compared sites. then applied these observationally constrained metrics benchmarks evaluate Energy Exascale Earth System Model (E3SM) model (ELM). ELM predicted relatively well, but underpredicted factors EF, indicating deficiencies describing state surface fluxes. Moreover, energy partitioning due not detected model. highlight need benchmarking simulated surface-energy corresponding against those observations, our predictability reshapes