Bias in modeling clouds
Zhang et al. discovered a 15% bias in CMIP5 modeling of cloud brightness.
Xuanze Zhang, Xiaogu Zheng, Zhian Sun, and San Luo, 2015: Trends of MSU Brightness Temperature in the Middle Troposphere Simulated by CMIP5 Models and Their Sensitivity to Cloud Liquid Water
J. Atmos. Oceanic Technol., 32, 1029–1041. doi: http://dx.doi.org/10.1175/JTECH-D-13-00250.1
the two approaches for estimating modeled MSU T2 are evaluated. For each CMIP5, it is shown that there exists a model-simulated static weighting function, such that the MSU T2 trend using the weighting function is equivalent to that calculated by RTM. The effect of modeled cloud liquid water on MSU T2 trends in CMIP5 simulations is investigated by comparing the modeled cloud liquid water vertical profile and the weighting function. Moreover, it is found that warming trends of MSU T2 for CMIP5 simulations calculated by the RTM are about 15% less than those using the two traditional static weighting functions. By comparing the model-derived weighting function with the two traditional weighting functions, the reason for the systematical biases is revealed.