In this project,
* We have demonstrated that the climate sensitivity varied over the last 100 years by a factor of two on multidecadal timescales both in reality and in climate model simulations, but the models are quite unrealistic in the timing of the variation. In the models, sensitivity reached its maximum in recent decades, whereas in same period it was at its minimum in the real world, due to an unusual pattern of temperature trends in the Pacific Ocean.
* We have found that this pattern is produced as a response to explosive volcanic eruptions in reality, but not in models. The reason for this difference remains to be found, but it suggests a systematic deficiency of models. (Explosive volcanic eruptions, such as that of Mount Pinatubo in 1991, inject small particles, called aerosol, into the stratosphere. These particles reflect sunshine and thus cause a global cooling for a couple of years, until they have dissipated.)
* We have examined how the climate sensitivity increases as time passes under constant elevated CO2 because of the way in which the pattern of surface temperature change evolves, and we have shown that it is greater for higher CO2 concentration and larger global-mean temperature change. We have produced evidence than the climate sensitivity to CO2 differs among models partly because they predict different patterns of surface temperature change.
* We have developed an improved description of the global energy balance, in which the climate sensitivity depends on the change in the vertical profile of atmospheric temperature. This in turn depends on the geographical pattern of surface temperature change. Greenhouse gases, anthropogenic aerosol produced by pollution, and volcanic stratospheric aerosol all cause different patterns of surface temperature change. Our new theory explains why climate sensitivity is different for these three and other kinds of forcing.
* We have achieved a new understanding of the pathways along which heat is absorbed by the ocean in models as the climate becomes warmer, recognising that different processes dominate at low and high latitude. We have discovered that heat is absorbed at high latitude more readily by the ocean in those models where the density of seawater increases more slowly with increasing depth below the surface. Moreover, in such models the Atlantic meridional overturning circulation (AMOC) is generally stronger. This explains a previously discovered correlation between ocean heat uptake efficiency and the AMOC strength. Using a refined method, we have made a new estimate of rate at which the ocean has warmed since the late 19th century, based on observations.
All these outcomes are consistent with the hypothesis that variations in climate sensitivity can be related to the geographical pattern of climate change, and show that the pattern interacts also with the vertical profile of atmospheric warming and the efficiency of ocean heat uptake. They provide answers to some existing questions, and raise some new ones. Following the conclusion of the project, we will continue to work on applying this new knowledge for refinement of climate projections.
