Understanding the Intermodel Diversity of Simulating Surface Temperature under Global Warming in Terms of Internal Variability

Internal variability (IV) is characterized by nonlinear and unforced processes in a climate system. This plays a role as a huge uncertainty factor to predict our climate system. Previous studies used large ensembles simulations with same external forcing and very small different initial conditions to examine the role of IV in a climate system. On the other hand, some recent studies suggested that IV of surface temperature (IVST) can be parameterized by statistical method in a single member of climate model. Using this statistical methodology, we calculated the IVST using a single run of CESM pre-industrial simulation and then we compare with the IVST obtained from the CESM large ensemble (CESM-LE) simulation. We found that the results are similar, supporting that the statistical methodology based on a single run of climate model is useful. We apply this methodology to the Coupled Model Intercomparison Project Phase 5(CMIP5) multi-model datasets. In this study, we examined how the CMIP5 multi-models differently simulate the IVST using their pre-industrial runs and then how the regional pattern of IVST would change under global warming. This would help to understand the inter-model diversity of simulating the surface temperature under global warming.