I thought I would follow up yesterday’s post with one that highlights another paper that looks at CMIP6 climate sensitivity. It’s a paper by Femke Nijsse, and colleagues, and considers [a]n emergent constraint on Transient Climate Response from simulated historical warming in CMIP6 models. The basic idea is to use the warming since the 1970s to constrain the transient climate response (TCR).The figure on the right illustrates the basic constraint. The light grey vertical band is the observed warming since 1975. The darker grey horizontal band shows the range of TCR values from models that satisfy this constraint. This result suggests a likely TCR range of 1.5K to 2.2K, with a best estimate of 1.8K. The figure below shows the TCR plotted against the ECS, for each of the models considered. Models with TCR values between 1.5K and 2.2K, can have ECS values that range from ~2K, to ~5K. So, although this analysis might suggest a better constrained TCR, it doesn’t particularly constrain the ECS.
Something I did wonder (which maybe someone else can clarify) is if this implied anything with respect to carbon budget estimates. Carbon budgets are typically estimated using the Transient Climate Response to Cumulative Emissions (TCRE). According to this paper (H/T Femke Nijsse) the TCRE for the CMIP6 simulations is 1.8K ± 0.4K per 1000 GtC. The reason this can then be used to estimate the carbon budget is that there is essentially no warming committment; global temperatures are expected to stabilise when emissions cease.
However, I think (although I’m not sure) that this does depend somewhat on the TCR-to-ECS ratio, which is typically thought to be about 0.6. However, if it’s possible that the ECS is more than twice the TCR, would we still expect global temperatures to stabilise when emissions cease? If not, would this then imply that carbon budget estimates could be too low? I don’t know the answer, so will stop there.