ONI and RONI gave an uncertain difference in winter cold-state recurrence change across 1951-1987 and 1988-2024

We tested whether index choice changes the estimated change in cold-winter recurrence across two equal periods. We used fixed NOAA CPC ERSSTv5 ONI and RONI DJF values for 1950-2024. At a 0.5 C threshold, recurrence changed from 50.00% to 46.67% for ONI and from 25.00% to 52.94% for RONI. The RONI period change minus the ONI period change was 31.27 percentage points. Its conditional 95% five-year block interval was -2.92 to 65.20 points. Thus the main test does not establish a clear difference in period change. Six of 20,000 draws had undefined ratios and were excluded. A 0.6 C check had a positive interval; other planned checks included zero. This shows threshold sensitivity, not a robust physical change. Few cold winters, shared revised SST data, different reference rules, rounding and conditional uncertainty limit inference. These annual labels are not official ENSO events. The package includes public domain input facts, exact source hashes, original code, all period-change draws, a figure, an independent count audit and offline repeat instructions.

ENSO · RONI · ONI · La Nina · winter recurrence · index sensitivity · null result