Assessment of CMIP6 Climate Model Simulations of Recent Changes in Earth's Radiation Budget Using CERES Observations
Abstract
It is well established that clouds have a profound influence on Earth's energy budget and that cloud feedbacks are responsible for much of the uncertainty in climate model projections of global warming in response to radiative forcing by greenhouse gases. While many model evaluation efforts focus on model representation of the observed mean state—for example, the regional mean climatology of key climate variables—there is also tremendous value in testing how a model represents observed variations in the climate system. Here we use CERES top-of-atmosphere (TOA) radiation observations during and after the so-called global warming "hiatus" to evaluate AMIP simulations from seven CMIP6 climate models. The AMIP simulations were extended through 2017 from the CMIP6 official end date of 2014. The hiatus period is defined as 07/2000-06/2014 and the post-hiatus period is 07/2014-06/2017. The latter is characterized by dramatic warming over the eastern Pacific Ocean off North and South America and over the central Pacific along the equator. In response to the warming, the CERES data show a marked 0.83 Wm-2 reduction in global mean reflected SW TOA flux during the post-hiatus period that is primarily driven by a pronounced decrease in low cloud cover over the eastern Pacific Ocean. We examine how well the AMIP simulations reproduce the observed TOA flux changes and explore possible reasons for model-observed differences. Recognizing that early analyses have shown higher Equilibrium Climate Sensitivities in some CMIP6 models compared to their prior versions (e.g., in CMIP5), and given the critical role of low cloud feedbacks in explaining model spread in ECS and TOA flux interannual variability, we also test whether or not there is any relationship between a model's ECS and its ability to reproduce the CERES-observed changes.
The seven models considered are: CanESM5, EC-Earth3-Veg, ECHAM6.3, GFDL_AM4, HadGEM3-GC3.1, IPSL-CM6, and NCAR_CESM2. In the CMIP6 AMIP simulations, SSTs and sea-ice boundary conditions are constrained by observations and forcings are time-varying through 2014, and held fixed at 2014 levels thereafter.- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2019
- Bibcode:
- 2019AGUFM.A33O3011L
- Keywords:
-
- 1620 Climate dynamics;
- GLOBAL CHANGE;
- 1627 Coupled models of the climate system;
- GLOBAL CHANGE