Analyzing recent latitudinal and seasonal changes in simulated atmospheric temperatures from a global chemistry-climate model

Abstract

Recent work by Santer et al. (2018) in Science examined the usefulness of the latitudinal structure and seasonal behavior of warming for fingerprinting anthropogenic climate change using satellite data and the CMIP5 multi-model ensemble over 1979-2016. They identify the first seasonal fingerprint in the northern hemisphere annual cycle and structure of warming, but do not specify what forcing agent (e.g. ozone, soot, or greenhouse gases) is responsible for causing it. We further probe this phenomena using 3 ensembles-of-opportunity over 1955-1979 and 1995-2024 of the Whole Atmosphere Community Climate Model version 4 (WACCM4), one of the world's few best fully coupled interactive chemistry-climate models. While our ensembles' construction covers limited time periods, it has the advantage of avoiding the effects of El Chichón (1982) and Pinatubo (1991), which are difficult to capture in models and have different drivers (volcanic) than the ones of interest here. The key findings of this research are that added greenhouse gas forcings nearly fully determine the latitudinal structure of warming and change in the amplitude of the annual cycle, that WACCM4 does a much better job than the CMIP5 multi-model ensemble of predicting the magnitude and latitudinal structure of climate change, and that tropical expansion and a poleward shift of the jet may drive the key subtropical features Santer observed. Interactive chemistry is not found to be a defining factor in representing the rate and structure of warming in CMIP5, and is certainly much less important than other details of model construction.