Internal Atmospheric Variability and Interannual-to-Decadal ENSO Variability in a CGCM

Abstract

Abstract The interactive ensemble coupling strategy has been developed specifically to determine how noise due to internal atmosphere dynamics impacts climate variability within the context of coupled general circulation models (CGCMs). In this study the authors investigate the impact of internal atmospheric variability on the ENSO variability using four CGCM simulations. In the control simulation, the interactive ensemble strategy is applied globally, thereby reducing the noise at the air–sea interface at each ocean grid point. In the second and third CGCM simulations, the interactive ensemble strategy is applied locally in the extratropics versus the tropics only, respectively. In addition, those results were compared with a standard CGCM. The results suggest that tropical internal atmospheric variability strengthens the interannual-to-decadal ENSO variability and leads to a broader spectral peak. However, the noise due to internal atmospheric dynamics plays different roles when the interannual and decadal ENSO variability is considered separately. There are noise-induced changes in the SST–zonal wind stress feedbacks from interannual to decadal time scales. The tropical atmospheric internal variability largely modifies the frequency as opposed to the amplitude of the ENSO variability on interannual time scales. In contrast, tropical internal atmospheric variability is effective in forcing decadal ENSO variability, resulting in a significant decrease of decadal ENSO amplitude in the central tropical Pacific in a CGCM when the noise is reduced. The authors argue that the decadal ENSO variability is directly affected by the low-frequency noise over the western part of the tropical Pacific in a linear sense. On the other hand, the impact of extratropical atmospheric noise on the ENSO variability is weaker than the noise in the tropics.