> But adjusted for the actual CO2 emissions, they come remarkably close. Take a look at the table I mentioned (line 326), in particular "ΔT / ΔF skill". That's

> But adjusted for the actual CO2 emissions, they come remarkably close. Take a look at the table I mentioned (line 326), in particular "ΔT / ΔF skill". That's the change in temperature over the change in forcing (ie, CO2 etc). Hausfather et al call it "implied TCR". In (overly) simple terms, it's the predicted increase in temperature per unit of CO2 emitted. Look at the numbers: 0.51, 0.41, 0.63, 0.42, 0.83, etc. The best is 0.87. The average was 0.69. The authors attempted to adjust for actual CO2 emissions and found that, after the correction, the models were predicting 40%-90% of the observations, with an average score around 70%. Is that "remarkably close"? Well, we might disagree about the meaning of that term. But there's another observation that we should agree on. Note what the authors say after that table: > The average of the median skill scores across all the model projections evaluated is 0.69 for the temperature vs time metric. > Using the implied TCR metric, the average projection skill of the models was also 0.69. In other words, the models are, on average, as accurate with or without the correction. Adjusting for actual greenhouse emissions doesn't make the models more accurate (nor less).