Max Planck Institute for Meteorology
Hamburg, Germany
Clouds are a very important, yet poorly modeled element in the climate system. There are many potential cloud feedbacks, including those related to cloud cover, height, water content, phase change, and droplet concentration and size distribution. As a prerequisite to studying the cloud feedback issue, this research reports on the simulation and validation of cloud radiative forcing under present climate conditions using the ECHAM general circulation model and ERBE top-of-atmosphere radiative fluxes.
Model Parameterizations
The ECHAM model carries prognostic cloud water and includes parameterizations of most of the main microphysical cloud processes including droplet coalescence, ice sedimentation, and the evaporation, melting, and freezing of precipitation. Ice and liquid water are partitioned according to temperature.
Cloud droplet optical properties are parameterized as follows: Single scattering properties are calculated from Mie theory as a function of wavelength and effective radius. These values are then averaged over the spectral intervals of the broad-band radiation model used in the GCM and expressed as functions of the effective radius using a least-square method. Ice crystal optical properties are parameterized in the same way but the asymmetry factor is adjusted from Mie theory values (spherical particles) to values consistent with data from FIRE Cirrus. This accounts for the non-sphericity of ice crystals.
Effective radius of droplets depends on cloud liquid water, droplet concentration, and a spectral shape constant, which has different values in maritime and continental environments. Currently, the model does not predict droplet concentrations so fixed values are used for maritime and continental clouds. It is planned to introduce a sulfur cycle to predict concentrations of sulfate aerosols. These would then be used to estimate droplet concentration using empirical relationships between sulfate aerosol concentration and cloud droplet concentration. Ice crystal effective radius is parameterized from ice water content according to Heymsfield.
Results
The clear sky, top of the atmosphere (TOA) fluxes agree well with ERBE satellite data and the model has been tuned to reconcile with global mean ERBE cloud radiative forcing. The geographical distribution of longwave cloud forcing in the model shows good agreement with ERBE observations, but there are problems with the shortwave cloud forcing. The model produces insufficient stratiform clouds off the coasts of California and South America, and too much convective cloud in the tropical west Pacific and Indian Ocean. Such regional problems tend to introduce global errors into the model because they bias the tuning process.
As a further means of validation, the model was run with tropical sea surface temperature (SST) anomalies in the equatorial Pacific (Figure 14.1). The model does well in replicating the associated longwave cloud forcing anomalies under these conditions (Figure 14.2). The correct pattern of shortwave cloud forcing is reproduced by the model, but the amplitude is too large by about 20W/m2 (Figure 14.3).
