Simulation of different types of ENSO impacts on South Asian Monsoon in CCSM4

Abstract

It has been found in observation that there are different types of influences of El Nino Southern Oscillation (ENSO) on the South Asian Monsoon (SAM). A correct description and representation of these teleconnections is critical for climate models to simulate and predict SAM. In this study, we examine these teleconnections in NCAR CAM4 and CCSM4 models, including the strength and weakness of these models in preserving different types of ENSO–SAM relationships. By using observational and simulation dataset, the composite analysis, based on specific selection criteria, is performed for both SAM rainfall and the eastern equatorial Pacific sea surface temperature (SST) anomalies. Anomalous SAM rainfall is characterized in three different types i.e. the indirect influence of the SST anomalies of preceding winter (DJF-only), direct influence of the SST anomalies of concurrent summer (JJAS-only) and the combined influence of both preceding winter and concurrent summer (DJF&JJAS). The analysis reveals that CAM4 uncoupled simulation can reasonably well reproduce the anomalous SAM rainfall in DJF-only and DJF&JJAS types whereas the model fails to simulate the anomalous rainfall in the JJAS-only type. The better performance of CAM4, particularly in DJF&JJAS type, comes from its realistic simulation of moisture content and thermal contrast. Its failure to preserve the ENSO–SAM relationship of JJAS-only type is due to the absence of ENSO induced warming in Northern Indian Ocean via atmospheric circulation which is indirectly linked to the lack of air–sea coupling. The role of Indian Ocean in controlling the ENSO–SAM teleconnections of the DJF&JJAS type is further investigated using CAM4 sensitivity experiments. It is found that in absence of Indian Ocean SST, the anomalous SAM summer rainfall suppresses in the DJF&JJAS type, suggesting the important modulation by Indian Ocean SST probably through the preceding winter equatorial Pacific SST forcing and the atmospheric circulations. On the other hand, CCSM4 shows large systematical errors in DJF-only and DJF&JJAS types and reproduce weak anomalous SAM rainfall. The failure of CCSM4 in simulating DJF-only and DJF&JJAS types is found mainly due to the errors in its SST simulation. The JJAS-only type is better reproduced in the CCSM4 simulation as compared to CAM4 and observation composites. Strong convergence over the SAM region which intensifies the anomalous SAM is seen to be responsible for its better simulation in this type. It is found that the atmospheric circulations in CCSM4 contribute more than the thermal contrast in modulating the intensity of anomalous rainfall in JJAS-only type. This study suggests that, although air–sea coupling is important for better SAM simulation and its relationship with ENSO, the SST bias in coupled model can significantly degrade ENSO–SAM relationship.