In case you missed it: Many new CMIP6 climate models show very high future warming. But recalibrating these by past observed global temp trends makes the upper bound of the warming much less likely and more consistent with CMIP5 (Thread: 1/n)
A different approach (preprint!) to weight models by trends *AND* mean climatology, and accounting for model independence supports these conclusions: still a lot of warming for highest emission scenarios, but the high end of CMIP6 is unlikely. (2/n)
earth-syst-dynam-discuss.net/esd-2020-23/
earth-syst-dynam-discuss.net/esd-2020-23/
The constrained transient climate response TCR is 1.6-2.1°C (66% range), uncertainty reduced a lot relative to the raw model spread, and very consistent with other ongoing TCR/ECS assessments. (3/n)
As in previous CMIPs, we can identify similarities in model code/parameterizations based only on model simulated climate and draw a “family tree” of related models. The actual number of models is smaller than the database suggests, some are largely duplicates of others. (4/n)
Published paper by @kasia_tokarska et al.
Discussion paper led by @luki_brunner with contributions from @apuffycloud @ClimateFlavors @AnnaMerrifield1 and Ruth Lorenz (5/n)
Discussion paper led by @luki_brunner with contributions from @apuffycloud @ClimateFlavors @AnnaMerrifield1 and Ruth Lorenz (5/n)
