A CMIP6-based multi-model downscaling ensemble to underpin climate change services in Australia

A multi-scenario, multi-model ensemble of simulations from regional climate models is outlined to provide the core data source for a set of climate projections and a climate change service. A subset of realisations from CMIP6 Global Climate Models (GCMs) are selected for downscaling by Regional Climate Models (RCMs) under a 'sparse matrix' framework using the CORDEX guidelines for Shared Socio-economic Pathways that feature low emissions (SSP1-2.6) and high emissions (SSP3-7.0). The subset excludes poor performing models, with performance assessed by the climatology over a large Indo-Pacific domain and an Australian-specific domain, the simulation of atmospheric circulation and teleconnections to major drivers, then incorporating other evaluation from the literature. The models are selected to be relatively independent by simply choosing one model from each 'family' where possible. The projected change in temperature and rainfall in climatic regions of Australia in the selected models are broadly representative of that from the whole CMIP6 ensemble, after deliberately treating models with very high climate sensitivity separately. A limited but carefully constructed ensemble will not represent statistically balanced estimates but can be used effectively under a 'storylines' style approach and can maximise representativeness within limits. The resulting ensemble can be used as a key data source for the future climate component of climate services in Australia. The ensemble will be used in conjunction with CMIP6 and large ensembles of GCM simulations as important context, and targeted 'convective permitting resolution' modelling, deep learning models and emulators for added insights to inform climate change planning in Australia.

Automated summary

A multi-scenario, multi-model ensemble of simulations from regional climate models is used to generate climate projections and a climate change service. The selected models are chosen based on their performance and representativeness, and they provide key data for future climate planning in Australia.