The two most dominant modes of interannual climate variability in the tropics are El Niño−Southern Oscillation (ENSO) and the Indian Ocean Dipole (IOD). Realistic simulations of these modes by coupled GCMs (CGCMs) are a key indicator of the CGCMs’ overall performance. However, despite the substantial progress made during the recent past, the CGCMs still have difficulty in correctly simulating the spatial structures and amplitudes of the ENSO and IOD modes; simulating their mutual interactions realistically has also been a challenge. Considering their significant influence on global climate variability and change, it is important that efforts are continued to improve the fidelity of the simulated ENSO and IOD modes in state-of-the-art climate models, such as those participating in the CMIP phase 5. 

 Here, we investigate some key features of simulated ENSO and IOD variability, including the ENSO‒IOD interaction, using a subset of the CMIP5 experiments contributed by two versions of the Australian Community Climate and Earth System Simulator (ACCESS), ACCESS1.0 and ACCESS1.3. Specifically, we use the pre-industrial control simulations and the abrupt 4xCO2 experiments that have been extended beyond the CMIP5 requirement to allow for the derivation of stable statistics under a near equilibrium condition. We find that both ACCESS1.0 and ACCESS1.3 simulate the main features of ENSO and IOD reasonably well. In particular, the spatial structures are well simulated, although the simulated IOD amplitudes are somewhat larger than observed.  Under global warming, the ENSO (IOD) mode becomes stronger (weaker) and their frequency increases in both models. Also, the simulated ENSO‒IOD interaction becomes slightly stronger under the global warming. There are also significant changes in the related ocean-atmosphere feedbacks due to global warming. We investigate the mechanism(s) of these global-warming induced changes in terms of changes in the feedback processes, which will be discussed in the presentation.