Using Eddy Covariance Tower Clusters To Evaluate Biogeophysical Impacts Of Land Cover In The Community Land Model (CLM)

Abstract:

The Community Land Model (CLM) land surface model has been used widely to evaluate biogeophysical responses to land cover and land use change. Here, we compare surface attributes collected from eddy covariance towers clusters to uncoupled point CLM (PTCLM) simulations. The tower clusters collect surface energy fluxes over adjacent forested and deforested land surface types located within 10-km of each other in temperate eastern North America. Summer surface albedo is very well simulated over cropland, C-3 grassland, and broadleaf deciduous temperate forests. In winter, modeled snow cover persists longer in spring than at the tower sites, resulting in higher average winter and spring albedo. Latent heat does not vary significantly among the three tower sites. PTCLM underestimates forest latent heat and overestimates cropland and grassland summer latent heat.

We evaluate temperature differences between forested and deforested sites due to changes in surface albedo, energy redistribution due to changes in surface roughness, and energy redistribution due to changes in latent and sensible heat partitioning (e.g., Bowen ratio). Surprisingly, temperature differences resulting from radiative forcing due to changes in surface albedo are relatively minor at the tower sites and generally too high in PTCLM. We conclude that the increased surface roughness of forests contributes strongly to nocturnal cooling over deforested tower sites in winter and daytime warming in summer. The importance of biogeophysical coupling between the land surface and atmosphere on energy redistribution due to surface roughness is explored using high-resolution (28-km) Variable Resolution Community Earth System Model (VR-CESM) simulations.