Observations suggest a slowdown in global surface temperature trends since 1998, whereas most climate models simulate continued warming.
Some have speculated that this difference occurs because climate models are running hot and overestimating climate change. Our paper published by the Max Planck Institute for Meteorology, Germany (MPI-M) and the University of Leeds gives a clear answer: There is no evidence for systematic model error and the observed discrepancy is mostly due to chance.
Writing in Nature, Jochem Marotzke, from MPI-M, and his colleague, Piers Forster from the University of Leeds, compared simulated and observed temperature variations over 1900 to 2012. They found that simulated 15-year trends show no systematic deviation from the observations. The differences between simulated and observed trends are dominated by spontaneous climate variability.
In particular, the climate sensitivity plays no role in the difference between simulations and observations. This latter fact occurs although sensitivity varies across models by a factor of three. The claim that climate models systematically overestimate the warming caused by increasing greenhouse-gases is unfounded.