A Systematic Relationship between Intraseasonal Variability and Mean State Bias in AGCM Simulations

Abstract

Abstract Systematic relationships between aspects of intraseasonal variability (ISV) and mean state bias are shown in a number of atmospheric general circulation model (AGCM) simulations. When AGCMs are categorized as either strong ISV or weak ISV models, it is shown that seasonal mean precipitation patterns are similar among models in the same group but are significantly different from those of the other group. Strong ISV models simulate excessive rainfall over the South Asian summer monsoon and the northwestern Pacific monsoon regions during boreal summer. Larger ISV amplitude also corresponds closely to a larger ratio of eastward-to-westward-propagating variance, but no model matches the observations in both quantities simultaneously; a realistic eastward-to-westward ratio is simulated only when variance exceeds that observed. Three sets of paired simulations, in which only one parameter in the convection scheme is changed to enhance the moisture sensitivity of convection, are used to explore the common differences between the two groups in greater detail. In strong ISV models, the mean and the standard deviation of surface latent heat flux is greater, convective rain fraction is smaller, and tropical tropospheric temperatures are lower compared to weak ISV models. The instantaneous joint relationships between daily gridpoint relative humidity and precipitation differ in some respects when strong and weak ISV models are compared, but these differences are not systematic enough to explain the differences in ISV amplitude. Conversely, there are systematic differences in the frequency with which specific values of humidity and precipitation occur. In strong ISV models, columns with a higher saturation fraction and rain rate occur more frequently and make a greater contribution to total precipitation.