Evaluation of Cloud and Heating Rate Profiles in Eight GCMs Using A-train Satellite Observations

Abstract:

Clouds strongly interact with radiation and modulate the amount of energy reflected, emitted and absorbed by the Earth system. The redistribution of energy within the troposphere has implications for climate prediction, as it impacts the large-scale circulation, the convection and precipitation. In this study, we take advantage of two modeling experiments and A-train satellite observations to characterize and evaluate the vertical distribution of clouds in eight GCMs and their link with their radiative heating rate profiles.

While the overall pattern of modeled zonal cloud profiles is quite good (r=0.92 for the multi-model mean), we show two main systematic biases: a positive bias in cloud frequency above 8 km (up to 10%), particularly in the tropics; and a negative bias in cloud frequency below 3 km (up to -10%), with a maximum over the stratocumulus cloud regions. Both biases are persistent regardless of the cloud regime.

Using satellite-based estimates of observed radiative heating rate profiles, we show that the global net cooling rate between 50°S and 50°N is too warm in the models (-0.81 K.d^-1 in models vs. -1.01 K.d^-1 in observations). The excess cloudiness in the high troposphere increases the absorption of solar radiation and absorbs more infrared radiation in the atmospheric column. On the other hand, the lack of clouds in the lower troposphere reduces its infrared cooling and increases the absorption of solar radiation by water vapor.

The representation of clouds in GCMs remains challenging. Yet, reducing the cloud biases would lead to an improvement of the heating rate profiles, which in turn, would help improving other aspects of the simulations such as the dynamics.