Climate models are used to scare people. With only one exception -- the Russian IMN model -- they are not accurate. If their claims about CO2 are compared with actual observations since 1940, assuming ALL warming was caused by CO2 (very unlikely), their average prediction is a global warming rate double of reality. That model average represents the modern climate change consensus, which is obviously wrong.
Scary climate model predictions require two false assumptions:
(1) The relatively small effect of CO2 is tripled (2x to 4x) by an imaginary, unproven water vapor positive feedback.
That alleged feedback supposedly takes several centuries to develop (ECS) but the IPCC gives people the impression they are talking about 50 to 100 years in the future (TCS). The guess for TCS (average of +1.8 degrees C. per CO2 doubling) is considerably lower than the guess for ECS (average of +3.3 degrees C. per CO2 doubling).
(2) There are a range of five estimates for the future growth rate of CO2 emissions. The IPCC scaremongers prefer models to use the highest estimate, called RCP 8.5.
Climate models used to scare people use ECS and RCP 8.5, resulting in global warming predictions 100% faster than past worst case (blame all warming on CO2) measurements.
Climate models can be made to appear more accurate by using TCS, rather than ECS, and RCP 4.5, rather than RCP 8.5.
Doing that is the only way to create an appearance of accuracy. So that's what a Climate Howler scientist named Zeke H. did for his propaganda paper claiming climate models really don't over predict global warming by much. And using his propaganda, leftists claim their beloved climate models are accurate.
Reality Check:
Zeke H. is a Climate Howler, so you know in advance that he has to claim climate models are accurate. In fact, models can not be accurate -- accuracy requires a very good understanding of perhaps 10 climate change variables. which does not exist. Therefore, any "model accuracy" is merely the illusion of a lucky guess. Climate models are used to scare people, and that requires scary climate predictions. So predictions must be for global warming faster than past mild, harmless warming since 1975, and they are. So when Zeke H. claims climate models are accurate, he MUST be deceiving you, and I just explained how Zeke did it.
Here is another version of the explanation:
"Zeke Hausfather et al. (2019) (herein ZH19) examined a large set of climate model runs published since the 1970s and claimed they were consistent with observations, once errors in the emission projections are considered."
"the discussion in ZH19 is for a shorter period than the 70 years used for TCR computation."
"ZH19 ask whether climate models have overstated warming once we adjust for errors in ... faulty emission projections."
All QUOTES above are from the following complicated article,'which I tried to simplify: