A DIAGNOSTIC STUDY OF FEEDBACK MECHANISM IN GREENHOUSE EFFECTS SIMULATED BY NCAR CCM1

This paper describes a diagnostic study of the feedback mechanism in greenhouse effects of increased CO2 and other trace gases(CH4,N2O and CFCs),simulated by general circulation model.The study is based on two sensitivity experiments for doubled CO2 and the inclusion of other trace gases,respectively,using version one of the community climate model(CCM1)developed at the National Centre for Atmospheric Research.A one-dimensional(1-D)and a two-dimensional(2-D)radiative-convective models are used to diagnose the feedback effect.It shows that the feedback factors in global and annual mean conditions are in the sequence of surface albedo,water vapor amount,water vapor distribution,cloud height,critical lapse rate and cloud cover,while in zonal and annual mean conditions in the tropical region the above sequence does not change except the two water vapor terms being the largest feedback components.Among the feedback components,the total water vapor feedback is the largest(about 50%).The diagnosis also gives a very small feedback of either the cloud cover or the lapse rate,which is substantially different from the 1-D feedback analysis by Hansen et al.(1984).The small lapse rate feedback is considered to be partly caused by the convective adjustment scheme adopted by CCM1 model.The feedback effect for doubled CO2 is very different from that of the addition of other trace gases because of their different vertical distributions of radiative forcing although the non-feedback responses of surface air temperature for both cases are almost the same.For instance,the larger forcing at surface by the addition of other trace gases can cause stronger surface albedo feedback than by doubled CO2.Besides,because of the negative forcing of doubled CO2 in the stratosphere,cloud height feedback is more intense.The larger surface forcing in the case of other trace gases can also influence atmospheric water vapor amount as well as the water vapor distribution,which will in turn have stronger feedback effects.All these indicate that it is incorrect to use "effective CO2" to replace other trace gases in the general circulation model.