Observational high-cloud feedback constraints indicate climate models underestimate global reductions in high-cloud amount with warming

Cloud feedback remains a leading source of uncertainty in climate model projections under increasing atmospheric carbon dioxide. Cloud-controlling factor (CCF) analysis is a method used to observationally constrain cloud feedback and, subsequently, the climate sensitivity. Although high clouds contribute significantly to this uncertainty, they have historically received comparatively little attention in CCF studies. Here, we apply CCF analysis to observationally constrain high-cloud feedback, focusing on feedback associated with changes in cloud amount due to its dominant contribution to uncertainty. Our observational constraints reveal larger decreases in high cloud amount with warming than climate models predict, yet the net high-cloud radiative feedback remains near-neutral due to compensating shortwave and longwave effects. We also show that including upper-tropospheric static stability as a predictor effectively captures the stability iris mechanism and associated changes in cloud amount. This work highlights the importance of using physically relevant CCFs for robust observational constraints on high-cloud feedback and improving mechanistic understanding of its underlying drivers.