Impacts of the Unforced Pattern Effect on the Cloud Feedback in CERES Observations and Climate Models

The equilibrium climate sensitivity estimated from different sources is inconsistent due to its dependence on the surface warming pattern. Cloud feedbacks have been identified as the major contributor to this so-called pattern effect. We find a large unforced pattern effect in CERES data, with cloud feedback estimated from two consecutive 125-month periods (March 2000-July 2010 and August 2010-December 2020) changing from -0.45 +/- 0.85 to +1.2 +/- 0.78 W/m(2)/K. When comparing to models, 27% of consecutive 10-year segments in CMIP6 control runs have differences similar to the observations. We also compare the spatial patterns in the CERES data to those in climate models and find they are similar, with the East Pacific playing a key role. This suggests that the impact of the unforced pattern effect can be significant and that models are capable of reproducing its global-average magnitude.

Automated summary

This study finds a large unforced pattern effect in CERES data, with significant changes in cloud feedback over different time periods. When compared to models, 27% of CMIP6 control runs show similarities to the observations. The study also reveals similarities between the spatial patterns in the CERES data and climate models.