Scientists increasingly assess climate changes at specific levels of global warming rather than at specific times in the future. This has numerous advantages over other methods of diagnosing future climate changes and allows for a more direct connection with various climate mitigation targets like the 1.5° or 2°C Paris agreement goals. A key assumption inherent to global warming levels (GWLs) is that all climate models and all future experiments that cross a given global temperature change threshold can be averaged together – that the response of the climate at a given GWL is largely independent of the emissions scenario or the sensitivity to carbon dioxide in the models comprising the average. This assumption has not been systematically vetted in the literature. Here we mine the entire CMIP5 and CMIP6 archive to determine the extent to which the climate response of several fields at various GWLs is sensitive to the future emissions scenario and to the models’ climate sensitivities. We then quantify how much uncertainty in future California climate at a given GWL comes from internal variability, scenario uncertainty, and model uncertainty, and how much of the latter is tied directly to spread in climate sensitivity. This information will be valuable for establishing the extent to which constraints on climate sensitivity can have tangible benefits for regional climate resiliency planning.

This work was performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344. It is supported by the Regional and Global Model Analysis Program of the Office of Science at the DOE. IM Release # LLNL-ABS-838335.