Pacific Climate Impacts Consortium

University House 1, PO Box 1700, STN CSC, University of Victoria, Victoria, British Columbia, Canada, V8W 3R4 Phone: 250-721-6236 | Fax: 250-721-7217 | pacificclimate.org 1. Here ‘likely’ is defined as having an assessed likelihood of 66-100%. 2. These figures are from the Technical Summary of Working Group One’s contribution to the Fifth Assessment Report of the IPCC. This Summary is available—along with the full report and a high-level overview document, the Summary for Policymakers—from the IPCC’s website: http://www.climatechange2013.org 3. For more on the Coupled Model Intercomparison Project, see: http://cmip-pcmdi.llnl.gov/ 4. That part of the Earth’s atmosphere that is closest to and directly affected by the ocean’s surface, generally about 1 km to 2 km thick. 5. Here, σ is a standard deviation, a measure of the spread from the average of the measurements. For the 2σ range, about 95.4% of observations will lie within the shading if observations have a bell-shaped “normal” probability distribution. Recent findings published in the journal Nature, by Sherwood, Bony and Dufresne (2014), indicate that a part of the variance in climate sensitivity among climate models can be traced to vertical mixing in the atmosphere. Using this result, combined with mixing values derived from observations, the authors suggest a lower bound of 3 °C on the warming that would result from a doubling of the atmospheric CO2 concentration. The magnitude of the impacts that are projected to occur as a result of anthropogenic climate change depend, in part, on the magnitude and rate of the changes to the climate system. These, in turn, depend largely on how sensitive the climate system is to changes in the concentration of atmospheric greenhouse gases. The equilibrium climate sensitivity of the Earth’s climate system, often just referred to as “climate sensitivity,” is defined as the amount of warming that would eventually be seen in the Earth’s average surface temperature if the atmospheric concentration of carbon dioxide were to be doubled and then held constant. The Intergovernmental Panel on Climate Change (IPCC) states that the likely1 range2 of climate sensitivity, from observations, studies of climate feedback mechanisms, paleoclimate data (such as ice cores and tree rings) and Global Climate Models (GCMs) is between 1.5 °C to 4.5 °C. The mean value for modern GCMs is 3.2 °C and the models that most closely simulate the modern climate have values in the range of 3 °C to 4 °C. Using satellite observations, reanalysis data and GCM output from models that participated in the third and fifth phases of the Coupled Model Intercomparison Project3 (CMIP3 and CMIP5), Sherwood, Bony and Dufresne (2014) examine how much of the variation in GCM climate sensitivity between GCMs can be explained by the differing ability of the models to represent vertical mixing of the atmosphere. The vertical mixing that they analyze in is made up of two shallow (~ 5 km), vertical, overturning air circulation patterns that dry the marine boundary layer4, which reduces the formation of low clouds that cool the Earth by reflecting sunlight and increases the formation of higher clouds which absorb infrared radiation that is trying to escape the planet and thus have a net warming effect. If this mixing is strengthened by a warming climate, then low PCIC SCIENCE BRIEF: SPREAD IN MODEL CLIMATE SENSITIVITY TRACED TO ATMOSPHERIC CONVECTIVE MIXING