Validating Interannual Variability in an Ensemble of AGCM Simulations

Abstract

This paper is an effort to explore general statistical procedures for validating the interannual variability in ensembles of atmospheric general circulation model simulations. The authors propose an estimation method for the correlation coefficient between the SST-forced components of simulated and observed seasonal means. The proposed methodology has been applied to study the interannual variability of New Zealand surface temperature as simulated by the Australian Bureau of Meteorology Research Centre nine-level atmospheric general circulation model. Compared with observations, it is found that the model simulates this variability well in most cases, except for summer maximum temperatures. It is suggested that deficiencies in the simulated storm tracks during summer may account for some apparent deficiency in the model. The failure in simulating the summer maximum temperatures also indicates the need for model validation.