Summer ENSO gave no clear forecast gain for the California South Coast to North Coast winter precipitation ratio in 1981-2026

We tested whether July-September Nino 3.4 improves forecasts of the natural log of South Coast to North Coast November-March precipitation in California. We used fixed NOAA ERSSTv5 and nClimDiv sources. Each fit used earlier winters only. The test has 46 winters, 1981-2026, after 30 initial training winters. Mean squared error skill was -1.97% against the earlier mean log ratio. A conditional 95% five-year block interval was -27.58% to +17.78%. Skill was +17.80% in the first 23 test winters and -20.05% in the last 23. Planned model and target checks also left gains uncertain. The full-record summer ENSO correlation with the log ratio was positive, at 0.227, but this association did not establish forecast gain. These results do not show that ENSO has no effect. The ratio compares mean precipitation depths, not water volumes. Retrospective products, two selected regions, a short test, simple models, and intervals conditional on fixed forecasts limit inference. The package gives retained source rows, hashes, annual forecasts, all block draws, code, a figure, an independent arithmetic audit, and an exact offline repeat.

ENSO · California · precipitation ratio · chronological forecast · uncertainty · NOAA · spatial contrast