For decades, NOAA's Climate Prediction Center classified El Niño and La Niña using the Oceanic Niño Index (ONI): a simple three-month running average of sea surface temperature anomalies in the Niño 3.4 region of the equatorial Pacific. If that anomaly sat at or above +0.5°C for five consecutive overlapping seasons, you had an El Niño. Simple, consistent, and the standard for almost 30 years.
The problem: the global ocean has been warming for decades because of climate change, independent of any ENSO cycle. ONI doesn't separate "this is El Niño" warming from "this is just the whole ocean running hotter than it used to." As baseline temperatures crept up, ONI started reading recent El Niño events as proportionally stronger than they actually were relative to the rest of the tropical Pacific, and La Niña events as weaker.
What RONI Actually Does
The Relative Oceanic Niño Index (RONI) fixes this by subtracting the tropical-mean sea surface temperature anomaly (averaged across 20°N–20°S) from the Niño 3.4 anomaly before classifying anything. In other words: instead of asking "is this region warmer than its 1991-2020 average," RONI asks "is this region warmer than the rest of the tropics are right now." That's a meaningfully different, and more diagnostic, question.
There is a third step that usually gets left out, and it matters. After the subtraction, the variance is rescaled to match the variance of the original Niño 3.4 index, and a three-month running mean is applied. Without that rescaling, subtracting the tropical mean would shrink the spread of the index and make every event look weaker on a scale that is no longer comparable to the historical record. So RONI is not simply "Niño 3.4 minus the tropics." It is that difference, restretched so the numbers still mean what they used to mean.
The dataset matters too. RONI is computed on ERSSTv5, NOAA's reconstructed historical SST analysis. The weekly Niño region values CPC publishes for real-time monitoring use a different dataset, OISSTv2.1. Weekly and seasonal values are therefore not directly comparable even before you account for the averaging period.
The practical effect, confirmed by NOAA's own analysis: RONI tends to dampen the apparent strength of El Niño events and amplify the apparent strength of La Niña events compared to the old ONI. A textbook example: using ONI, 2020–2023 was read as a choppy "triple-dip" La Niña with several neutral-looking gaps. Using RONI, that same period reads as one continuous, unbroken La Niña.
The Thresholds, Stated Precisely
A persistent myth is that RONI uses a higher bar than ONI did. It does not. CPC's operational definition is that El Niño is characterised by a positive RONI greater than or equal to +0.5°C, and La Niña by a RONI at or below -0.5°C. That is the same numerical threshold ONI used. What changed is the quantity being measured, not the line it has to cross.
That last row is worth internalising. Because of the filtering applied to ERSSTv5, the most recent RONI values are estimates and can move after first posting. A single season's reading is never the final word, and neither is a single month's.
Two Weekly Tables, Both Live, 0.8°C Apart
This is the part that trips up careful readers, including professional ones.
Since February 2026 the Niño region values quoted in CPC's ENSO products, the monthly Diagnostic Discussion and the accompanying slide decks, are relative values. But CPC also still publishes the conventional weekly table it has always published, at a separate URL, and never switched it. Both are current. Both are correct. They disagree, by a lot.
The tell that you are looking at an index difference rather than a data revision is that the offset is nearly uniform across all four Niño regions. A genuine revision nudges one region slightly. It does not move all four by the same 0.75°C. That uniform gap is the tropical mean being netted out, and in mid-2026 the tropical mean anomaly is large enough to be worth about three quarters of a degree.
So if you see a Niño 3.4 figure quoted somewhere that runs roughly 0.8°C above the one on this site, you have almost certainly found the conventional table rather than an error. This site tracks the relative series, which is what CPC's own ENSO products quote.
What Changes When You Reclassify History
Switching indices does not just change today's number. It reorders the record.
Measured on raw Niño 3.4, the strongest event in the modern record is 2015-16. Measured on RONI, which strips out the background warming that 2015-16 enjoyed and 1982-83 did not, the strongest event is 1982-83. Same ocean, same observations, different question, different answer. Any headline calling something "the strongest on record" without naming the index, the dataset and the baseline has skipped the step that decides the answer.
The current event illustrates the same gap in real time. For April to June 2026, NOAA's unrounded ONI value is +0.98°C, comfortably in El Niño territory on the old scale. The RONI value for the same season is +0.47°C, which rounds to the +0.5°C threshold but sits just below it at full precision. Half a degree of difference, one season, two defensible numbers.
Why It Matters for You
If you are tracking this site's RONI panel alongside other sources still quoting legacy ONI numbers, expect divergence, especially during a developing event like the current one. Both metrics track real phenomena. RONI is the more climate-change-aware version and it is now NOAA's official standard, which means it is the one that determines how this event enters the historical record.
Three practical habits follow. Check which index a number comes from before comparing it to anything. Check whether it is weekly or seasonal, because those use different datasets. And treat any recent value as provisional, because it is. For a worked example of two organisations publishing different figures for the same week, see One Ocean, Two Numbers.