Each core is a pair of columns: ONI and RONI. Scrub through time and watch whether they part ways.
Huascarán, Illimani and the Quelccaya firn core track DJF ENSO strongly since 1960 (p < 0.05 after correcting for autocorrelation).
Switching ONI → RONI barely matters. No core changes significantly; the δ¹⁸O signal follows the ENSO-specific SST pattern, not tropical-mean warmth.
Quelccaya's coupling broke down after strong 1870–1920 coupling, under both indices. Press ▶ Play to watch it happen.
δ18O, accumulation, dust and major ions from NOAA NCEI Andean ice cores compared with the Oceanic Niño Index (ONI) and NOAA’s Relative Oceanic Niño Index (RONI). All statistics are recomputed live in your browser.
Load any annual, monthly or sub-annual time series: an ice or firn core, a lake record, a tree-ring chronology, station precipitation or river discharge. It is correlated with ONI and RONI (1850–present, or 1–2000 CE from the LMRv2.1 reconstruction), tropical-mean SST, and every NOAA NCEI Andean ice-core series, using the same statistics as the Analysis tab. Your file is read in this browser tab and never uploaded.
NOAA’s Climate Prediction Center adopted the Relative Oceanic Niño Index (RONI) in February 2026. RONI is the Niño-3.4 SST anomaly minus the 20°S–20°N mean SST anomaly, rescaled to Niño-3.4 variance. It removes the background tropical warming that the classic ONI carries. Andean ice-core δ¹⁸O partly reflects tropical-mean tropospheric temperature and moisture transport. If so, its ENSO coupling should weaken under RONI. This site tests that idea with every annually resolved Andean core in the NOAA NCEI archive.
21-yr running correlations, 1861–2015 (ERSST.v6 DJF ONI/RONI recomputed with CPC’s algorithm). Column height = |r|, colour = r; the disc marks p = 0.05.
Figure 1 | Andean ice-core δ¹⁸O–ENSO coupling is largely insensitive to removing tropical-mean warming. a, NOAA NCEI Andean ice cores with annual resolution (circles; fill, Pearson r with DJF ONI, 1960–2019; heavy outline, p < 0.05 after effective-sample-size correction) and drill sites lacking annual archives (triangles); inset, elevation versus latitude. b, Huascarán Col δ¹⁸O (black, z-score) with DJF ONI (purple) and RONI (green); shading, El Niño (orange) and La Niña (blue) years. c, Correlation statistics. d, El Niño minus La Niña composites (95% CI). e, ΔR = r(RONI) − r(ONI) per core (Meng–Rosenthal–Rubin test) and reclassified events. f, 21-yr running correlation for Quelccaya δ¹⁸O using ERSST.v6-derived indices (dashed, 1851–2000) and 100 LMRv2.1 reconstruction members (shading, 5–95%). g, Standardized regression on ENSO-relative and tropical-mean SST. h, Correlation spread under 1% yr⁻¹ layer-counting error. RONI subtracts the 20°S–20°N mean SST anomaly from Niño-3.4, rescaled to Niño-3.4 variance.