Computer Climate Simulations Just Crashed

Ross McKitrick and John Christy have an important new paper out in Earth and Space Science.

This is the latest fusillade in the long battle over whether the climate simulations that lie behind demands for decarbonisation and other political action actually amount to nothing but a hill of beans (as they say on the other side of the pond).Computer climate simulations predict that manmade global warming will cause the troposphere over the tropics to warm much faster than the surface, and there have been a series of scientific papers arguing whether these predictions are being borne out in practice. In a blog post published yesterday, McKitrick relates some of the back story, including attempts by one mainstream scientist to withhold his data, and the subsequent revelation that he had truncated it in a way that fundamentally altered the conclusions that would be drawn. McKitrick also outlines a series of subsequent papers that have concluded that real-world warming in the troposphere is much less than predicted:

[W]hether we test the tropospheric trend magnitudes, or the ratio of tropospheric to surface trends, across all kinds of data sets, and across all major trend intervals, models have been shown to exaggerate the amplification rate and the warming rate, globally and in the tropics.

So it's not looking too good for the models. The next logical step is to consider what this means for the bigger picture, and this is where the new paper comes in. As McKitrick points out, if climate simulators get the rainfall in the Amazon wrong, it's perhaps not the end of the story - that part of the model might be adjusted. But he and Christy are suggesting that what the models indicate about the tropical troposphere is essentially a diagnostic of their structures - almost all climate models agree that it will warm rapidly and it should only be greenhouse gases that can cause such a warming.In other words, if the models get this wrong, something is fundamentally wrong.Which is why it's so important that the authors conclude their paper thus:

Comparing modeled to observed trends over the past 60 years...shows that all models warm more rapidly than observations and in the majority of individual cases the discrepancy is statistically significant. We argue that this provides informative evidence against the major hypothesis in most current climate models.