Features of rainfall and latent heating structure simulated by two convective parameterization schemes

Abstract

Using the latest version of SAMIL (Spectral Atmosphere Model of IAP LASG) developed by LASG/IAP, we evaluate the model performance by analyzing rainfall, latent heating structure and other basic fields with two different convective parameterization schemes: Manabe Scheme and Tiedtke Scheme. Results show that convective precipitation is excessively overestimated while stratiform precipitation is underestimated by Tiedtke scheme, thus causing less stratiform rainfall proportion compared with TRMM observation. In contrast, for Manabe scheme stratiform rainfall belt is well simulated, although precipitation center near Bay of Bengal (BOB) spreads eastward and northward associated with unrealistic strong rainfall downstream of the Tibet Plateau. The simulated latent heating structure indicates that Tiedtke scheme has an advantage over Manabe scheme, as the maximum convective latent heating near middle of troposphere is well reproduced. Moreover, the stratiform latent heating structure is also well simulated by Tiedtke scheme with warming above freezing level and cooling beneath freezing level. As for Manabe scheme, the simulated maximum convective latent heating lies near 700 hPa, lower than the observation. Additionally, the warming due to stratiform latent heating extends to the whole vertical levels, which is unreasonable compared with observation. Taylor diagram further indicates that Tiedtke scheme is superior to Manabe scheme as higher correlation between model output and observation data is achieved when Tiedtke scheme is employed, especially for the temperature near 200 hPa. Finally, a possible explanation is addressed for the unrealistic stratiform rainfall by Tiedtke scheme, which is due to the neglect of detrained cloud water and cloud ice during convective process. The speculation is verified through an established sensitivity experiment.