Roy Warren Spencer

How temperature biases in both climate models and observations are adjusted in order to make comparisons of climate change signals has been seldom discussed, yet the choice of adjustment method has a large impact on the resulting conclusions. When the primary interest is how global warming evolves through time, how the models’ diagnosed equilibrium climate sensitivities (ECS) correlate with yearly temperatures is a logical test of agreement. Unlike other commonly used methods, it is shown that correlations are maximized when all of the time series are adjusted so their trend lines intersect at year zero. The issue is important to the interpretation of how climate models reveal the global warming signal over time, to how well models agree with observations, and to the policy impact and public debates regarding climate change.