Diagnosing and Evaluating Meteorological Mechanisms Associated with Extreme Temperatures over North America in Global and Regional Climate Models

A Department of Atmospheric and Oceanic Sciences seminar presented by Paul Loikith, Caltech Postdoctoral Scholar, NASA Jet Propulsion Laboratory

Wednesday, February 12, 2014
3:30 PM - 4:30 PM
MSB 7124


Warming due to anthropogenic climate change is often posed as a shift in the mean of the temperature distribution; however, changes in extremes (i.e. distribution tails) are anticipated to have the most severe climate impacts. It is therefore crucial to evaluate the fidelity of current-generation climate models in simulating extreme temperature events in order to properly assess risk of future changes to society. To this end, several novel methodologies designed to systematically evaluate the ability of climate models to simulate the key physical mechanisms and meteorological processes associated with temperature extremes are presented for the North American region. Composite analysis is employed to describe and diagnose synoptic-scale atmospheric circulation patterns associated with extreme temperature days in observations and global climate models archived as part of the Fifth Phase of the Coupled Model Intercomparison Project (CMIP5). While the fidelity of individual ensemble members varies, in most cases the multi-model ensemble mean captures the key features of these patterns, suggesting that within-ensemble bias is not systematic. Evaluation of the shape of daily temperature probability distributions, focusing on the tails of the distribution, in regional climate models participating in the North American Regional Climate Change Assessment Program (NARCCAP) shows higher model fidelity in winter than summer. A dynamical component to error in distribution shape is apparent in analysis of the meteorological patterns associated with days in the tails of the distribution. These methodologies lend themselves to regional, process-focused model evaluation across the globe and provide a holistic analysis of the interaction between weather, climate, and extreme events in climate models.