New paper with @piersforster now out in @NatureClimate - Energy budget constraints on historical radiative forcing nature.com/articles/s4155…
Provides a top down energy budget constraint on historical radiative forcing of 2.3 (1.7 to 3.0) [5-95%] Wm-2. Thread: [1/n]
Provides a top down energy budget constraint on historical radiative forcing of 2.3 (1.7 to 3.0) [5-95%] Wm-2. Thread: [1/n]
Radiative forcing is a fundamental quantity for understanding the drivers of climate change, yet significant uncertainty remains in our quantification & model representation of it. For example see any paper that includes @gunnarmy including this epic journals.ametsoc.org/doi/abs/10.117… [2/n]
What we do is combine measurements of historical temperature change dT & Earths heat uptake dN alongside model reconstructions of Earth’s radiative response Y. Constrains historical forcing F via energy budget eqn F=dN+YdT. Like Otto etal nature.com/articles/ngeo1… in reverse [3/n]
Provides a much tighter constraint on historical radiative forcing than that assessed by #IPCC AR5 (~40% reduction in the 5-95% range) and substantially reduce the probability of radiative forcings at the lower end (5th percentile increases from 1.0 to 1.7 Wm-2) [4/n]
How do climate models compare? We find as many as one-third of #CMIP5 radiative forcings (@piersforster et al agupubs.onlinelibrary.wiley.com/doi/10.1002/jg…) fall short of our constraint - their historical radiative forcings are too small [5/n]
Our method does not tightly constrain individual drivers, but we provide an aerosol forcing constraint of -0.8 (-1.6 to +0.1) Wm-2, in line with #IPCC AR5. Nearly all #CMIP5 models (@mzelinka etal agupubs.onlinelibrary.wiley.com/doi/10.1002/20…) fall in the lower half of this constraint [6/n]
Why are some model forcings low? Suggestive that aerosol forcing too large, but also need to consider weak GHG forcings (e.g. Etminan etal 2016) and other uncertainties e.g. strong land-use (e.g. @Tim_AndrewsUK @richardabetts @GarethSJones1 et al link.springer.com/article/10.100…) [7/n]
If historical forcing too weak then some models may struggle to accurately simulate historical temperature trends or have compensating inaccuracies in their historical heat uptake or sensitivities (not necessarily long-term #ECS see Andrews et al agupubs.onlinelibrary.wiley.com/doi/10.1029/20…) [8/n]
Didn’t make the paper, but accurate estimates of forcings in #CMIP6 models are now starting to appear. See @chrisroadmap et al in open review atmos-chem-phys-discuss.net/acp-2019-1212/ looks like a lot of #CMIP6 models very close to our lower limit [9/n]
Skimming Flynn and Mauritsen in open review atmos-chem-phys-discuss.net/acp-2019-1175/ they also argue that #CMIP6 models forcings maybe too low – based on their historical temperature evolution [10/n]
Hope our constraint will be used by modelling centres to improve their simulation of radiative forcing. Issues not just related to aerosols, and improved representation of RF will improve projections of 21st century climate change (see Soden etal science.sciencemag.org/content/361/64…) [11/n]
#Footnote Novel (#controversial?) step is to recognise that atmospheric climate models forced with observed patterns of historical temperature change give us the best estimate of the energy returned to space over the historical period (not same thing as long-term #ECS) [12/n]
#Disclaimer: all of the above subject to the usual caveats/limitations (e.g. unknown structural errors etc., see details in paper) [14/14]
I think this link provides an #OpenAccess version rdcu.be/b1tMH
