Missing eddy feedback may explain weak signal-to-noise ratios in climate predictions

The signal-to-noise paradox—in which climate models predict the real world better than their own ensemble members—highlights a serious and unresolved model error that undermines climate predictions and projections. Using multiple seasonal forecast systems, this study quantifies feedback between transient eddies and large-scale flow anomalies and shows that current systems underestimate positive eddy feedback. This deficiency is strongly linked to weak signal-to-noise ratios in ensemble mean predictions. Improved eddy feedback is also associated with stronger teleconnections between ENSO and the Arctic Oscillation and with larger predictable signals. A technique is presented to estimate potential skill gains from eliminating eddy feedback deficiency, indicating that skill could double in some extratropical regions, notably improving Arctic Oscillation predictions.