Updated IPCC-models, to be published in 2021, are showing greater sensitivity of climate to CO2. 31 models published thus far (of around 100 total to be published) predict an average of 3.5°C warming in a ‘middle or the road’-scenario (RCP 4.5-6*) from 2090 onwards.
These numbers represent average warming of the whole earth, including oceans. Land surface warming averages double these numbers. Resulting in an average land surface warming of 7°C, with upper limits of up to 10.5°C in Eurasia and up to 14°C in the Arctic.
Disturbingly enough, some state of the art models (models from France, the US Department of Energy, Britain’s Met Office and Canada) are showing an even greater sensitivity of temperature to CO2. They come up with around 9°C warming on average on land surfaces in the aforementioned ‘middle of the road’-scenario. The new and higher results are the product of more computing power and a better understanding of cloud dynamics.
If we consider the worst scenario of the IPCC (RCP 8.5, which some call ‘Business as Usual’, and others deem impossible without positive feedback loops, such as for example large quantities of methane released by melting permafrost) numbers really become mind boggling. Warming is projected to be 5.5°C on average (this equals 11°C on land). The 4 ‘state of the art’ models just mentioned come closer to 7°C on average. If the rule of thumb of double land warming still holds with these numbers, it would mean average land surface warming of 14°C, whereas parts of Eurasia and Arctic land masses could experience well over 20°C warming by the end of the century. No wonder scientists are looking for methodological glitches and inconsistencies that could indicate that the models are wrong.