Bad news: Warmest climate models might also be most accurate

Enlarge / Be cautious out there. (credit: Bill Green)

Some people who reject the conclusions of climate science claim that the existence of any remaining uncertainty means few or no actions need be taken to reduce our greenhouse gas emissions. In reality, though, uncertainty is ever-present in science, and it's not necessarily our friend. A new study from Patrick Brown and Ken Caldeira of the Carnegie Institution for Science highlights the fact that uncertainty means climate change could just as easily be worse than our best current estimates predict.

The study sought to narrow the range of projected global warming presented in places like the Intergovernmental Panel on Climate Change (IPCC) reports. For each of several scenarios of future greenhouse gas emissions, these reports have simply taken simulations from every climate model available and combined the results—showing the average temperature trajectory and the range they span. For the highest-emissions scenario, for example, the last IPCC report projected about 4.3 degrees Celsius (7.7 degrees Fahrenheit) warming by the late 21^st century. But the range of model results stretched from about 3.2 degrees Celsius to 5.4 degrees Celsius.

One strategy for dealing with this variance has been to weight the results of the best-performing models more highly. The difficulty is in confidently assessing which models are the best-performing ones. A handful of studies have used some aspect of cloud behavior as the measuring stick. That work has found that the models best simulating current cloud behavior also happen to simulate more future warming.