The World Climate Research Programme (WCRP) Coupled Model Intercomparison Project, now in its sixth phase, provides the basis for climate model intercomparison and the multimodel averages published in peer-reviewed literature and Intergovernmental Panel on Climate Change (IPCC) reports. There is evidence that, due to different model performances against observations and the lack of independence among models, it is not optimal to give equal weighting to each available model projection. A recent 'Perspective' paper in Nature Climate Change, 'Taking climate model evaluation to the next level', by Veronika Eyring and others discusses newly developed tools that improve the evaluation of model simulations compared to observations and that include process-based emergent constraints and advanced methods for model weighting. Advancement of these approaches is critical for providing the most credible information possible to stakeholders and policy-makers.

Eyring et al. 2019

Figure: Annual mean SST error from the CMIP5 multi-model ensemble. Figure reproduced from Wiley (see Eyring et al., 2019).