Development of a toy column model and its application in testing cumulus convection parameterizations  

作　　者：Xiaocong Wang 

机构地区：[1]State Key Laboratory of Numerical Modeling for Atmospheric Sciences and Geophysical Fluid Dynamics, Institute of Atmospheric Physics, Chinese Academy of Sciences

出　　处：《Science Bulletin》2015年第15期1359-1365,共7页科学通报（英文版）

基　　金：jointly supported by the National Natural Science Foundation of China(41305102);the National Basic Research Program of China(2014CB441202,2013CB955803)

摘　　要：A single-column model is constructed based on parameterizations inherited from the Finite-volume/Spectral Atmospheric Model F/SAMIL and tested in simulations of tropical convective systems. Two representative convection schemes are compared in terms of their performances on precipitation types, individual physical tendencies, and temperature and moisture fields. The main difference between the two selected schemes is in their representation of entraining/detraining process. The Tiedtke scheme assumes bulk entrainment, while the Zhang–Mc Farlane scheme parameterizes entrainment/detrainment rates under the spectrum concept. Large-scale forcing and verification data are taken from the GATE phase III field campaign, during which abundant convective events were observed. Given the same triggering function and closure assumption, results show that entrainment/detrainment representation remains the dominant factor on the simulation of cumulus mass flux and of temperature and moisture fields. By analyzing sources and sinks of heat and moisture, this study reveals how parameterization components compensate for each other and make model results insensitive to parameterization changes in certain fields, thus suggesting the need to treat parameterizations as systems rather than individual components.A single-column model is constructed based on parameterizations inherited from the Finite-volume/Spec- tral Atmospheric Model F/SAMIL and tested in simula- tions of tropical convective systems. Two representative convection schemes are compared in terms of their per- formances on precipitation types, individual physical ten- dencies, and temperature and moisture fields. The main difference between the two selected schemes is in their representation of entraining/detraining process. The Tiedtke scheme assumes bulk entrainment, while the Zhang-McFarlane scheme parameterizes entrainment/de- trainment rates under the spectrum concept. Large-scale forcing and verification data are taken from the GATE phase III field campaign, during which abundant convec- tive events were observed. Given the same triggering function and closure assumption, results show that entrainment/detrainment representation remains the dominant factor on the simulation of cumulus mass flux and of temperature and moisture fields. By analyzing sources and sinks of heat and moisture, this study reveals how parameterization components compensate for each other and make model results insensitive to parameterization changes in certain fields, thus suggesting the need to treat parameterizations as systems rather than individual components.

关 键 词：Single-column model （SCM） TCM GATE Cumulus convection F/SAMIL 

分 类 号：P407[天文地球—大气科学及气象学] P426.5