Improved surface layer simulation using refined vertical resolution in the GFDL Atmospheric General Circulation Model

Surface layer (SL) variables (e.g., 2-m temperature [T-2] and 10-m wind [U-10]) are diagnosed by applying the flux-profile relationships based on Monin-Obukhov similarity theory to the lowest model height (LMH). This assumes that the LMH is in the SL, which is approximately the bottom 10% of the boundary layer, but atmospheric general circulation models rarely satisfy this in stable boundary layers (SBLs). To assess errors in the diagnostic variables due to the LMH solely linked to the diagnostic algorithm, offline tests of the flux-profile relationships are performed with LMH from a few meters to 60 m for three SBL regimes: weakly stable, very stable, and transition stability regimes. The results show that T-2 and U-10 are underestimated by O(0.1-1 degrees C) and O(0.1-1 m/s), respectively, if the LMH is higher than the SL height. The stronger the SL stability is, the larger the temperature biases are. The negative wind biases increase with the surface stress. Based on these findings, we analyze the impacts of the LMH on the climatologies of the diagnostic parameters in the Geophysical Fluid Dynamics Laboratory (GFDL) Atmosphere Model 4.0/Land Model 4.0. The results show reduced negative biases in T-2 and U-10 by lowering the LMH. The decrease of the overall bias over land is mainly due to the sensitivity of the diagnostic method to the LMH in SBLs, as shown in the offline tests. The overall increase in T-2 and U-10 over the oceans results from the increase in the actual near-surface temperature and wind rather than from the diagnostic method.