MJO amplitude has fallen from 1.31 on 9 July to 0.33 on 24 July, and it last cleared the active threshold ten days before that reading. CPC forecast this disorganization on 13 July and the index has since verified it, though its 27 July update expects the signal to rebuild. Here is what a weak MJO removes from El Niño's intensification window, the cyclone that was inflating the index before the collapse, and why none of this is a reversal.
El Niño coverage right now is almost entirely about a forecast peak. The event itself still has to get there, and for developing El Niño events late boreal summer and autumn are commonly an important intensification period ahead of the typical winter peak. One potential accelerant has weakened sharply going into it, which is worth noticing precisely because nobody is writing about it.
The Madden-Julian Oscillation is tracked through the Wheeler-Hendon RMM index, which reduces the pattern to two numbers. Amplitude is the distance from the origin on the phase diagram, and by long convention a value of 1.0 or above means a coherent, active MJO. Below 1.0 the signal is weak or incoherent and its phase stops carrying much practical meaning, though that describes the index rather than a complete absence of MJO-related forcing.
The shape is the story. Amplitude climbed through early July, peaked at 1.31 on 9 July, held above the threshold for a few more days, then fell away. By 14 July it was at 1.07, the last reading at or above 1.0. Ten days later it stood at 0.33, with RMM1 at 0.02 and RMM2 at 0.33, which is close to the middle of the phase diagram where the MJO effectively has no coherent position at all.
One precision note, since it can produce apparent contradictions. The amplitude series above is computed from the ROMI values cached on this site, adjusted to the RMM convention. CPC's operational product is the RMM index itself. The two are close relatives rather than the same number, so a phase reported here and a phase reported by CPC on a nearby date will not always agree, and CPC's Phase 8 call for late July is the operational one. Amplitude, which is what this post rests on, tells the same story in both.
This is not hindsight. In its MJO update dated 13 July, CPC wrote that dynamical models favoured an eastward propagating signal for about a week, "followed by a gradual disorganization of MJO activity later in July." It went further, warning that the weakening pattern in the RMM forecasts pointed to "the return of a period of destructive interference with the dominant El Niño background state as the enhanced MJO phase draws closer to the Indian Ocean."
Constructive and destructive interference are worth defining, because the terms do real work here. Both the MJO and El Niño move tropical convection around. When the MJO's enhanced phase sits over the part of the Pacific that El Niño is already favouring, the two reinforce each other. When the enhanced phase drifts back toward the Indian Ocean, it pulls convection away from where the El Niño background state wants it, and the two start fighting. CPC described the MJO as constructively interfering in mid-July and forecast that the interference would turn destructive. The observed amplitude collapse is consistent with the first half of that call, the weakening. The destructive phase itself is still forecast rather than observed.
Two weeks on, CPC's 27 July update sharpens the picture, and one detail is worth getting right. The MJO did not simply dissolve where it was. It "resumed its eastward propagation" after stalling over the western Pacific and is now in the Western Hemisphere, which CPC places at Phase 8. It is weakening as it moves away from constructive interference with El Niño, and the RMM index is "projected to further weaken as the MJO moves into the Indian Ocean, and enters a period of destructive interference." So the destructive phase is still ahead rather than already here. What has happened so far is the loss of the constructive assist.
CPC also frames the whole thing as routine machinery: the MJO "continues to circumnavigate the globe, in roughly a 30-40 day cycle, alternating between periods of constructive and destructive interference with the low frequency El Niño background state." Dynamical models show that cycle repeating, and both the GEFS and ECMWF "indicate a rapid increase in signal strength as the enhanced convective envelope moves away from interference over the Indian Ocean." A fortnight of low amplitude is therefore one trough in a repeating cycle, not a new regime. The right description is a lull with a recovery expected.
One detail from that same update deserves more attention than it has received. CPC noted that the RMM signal had stalled over the western Pacific while "remaining high in amplitude," and then explained why that reading might be misleading: the behaviour "may be tied to Tropical Cyclone Bavi projecting onto the index, with robust TC-related OLR and lower-level wind anomalies falling within 15S-15N to increase RMM1."
The RMM index is built from outgoing longwave radiation and zonal winds averaged across 15S to 15N. A large tropical cyclone inside that band contributes to those fields whether or not it has anything to do with a propagating intraseasonal wave. So part of what looked like a robust MJO in early July was a cyclone leaving its fingerprint on the measurement. That is a useful reminder that an index is an instrument, not the thing itself, and it is the same class of caution we applied to reading the Niño region gradient earlier this week.
Timing is what makes this more than trivia. The benchmark weekly Niño 3.4 anomaly stood at +1.4°C in the week centred 22 July. CPC's 9 July diagnostic discussion places an 81 percent probability on a very strong event during October to December. Getting from one to the other requires substantial intensification, and because ENSO is seasonally phase-locked, late boreal summer and autumn are commonly an important period for that ahead of the typical winter peak. The window is opening now.
The MJO's contribution to that process is fairly specific. An active MJO in the right phases can trigger westerly wind bursts over the western equatorial Pacific, which push warm surface water eastward and can launch downwelling Kelvin waves that deepen the thermocline in the east. That is one of the main routes by which a developing El Niño gets an extra shove, and it is the mechanism behind the triple-cyclone wind burst that helped set this event up in April. With RMM amplitude at 0.33, that route is currently much less available. An amplitude below 1.0 means the RMM representation of the MJO is weak or incoherent, not that all MJO-related convection or wind-burst potential has ceased. For the underlying physics, see our deep dive on how the MJO steers El Niño or the shorter explainer.
A quiet MJO is not a stalling El Niño, and the distinction matters more than the headline. The MJO modulates ENSO development; it does not drive it. This event already has its own coupled feedback running, and the supporting evidence sits in the same CPC material: subsurface anomalies described as "remarkably warm" and exceeding 2.5°C in the eastern Pacific, and significantly negative traditional and equatorial Southern Oscillation indices. The 27 July update is blunter still on the winds: "highly anomalous low-level westerlies are observed across the Equatorial Central Pacific," which is a stronger statement than the stalled westerlies described two weeks earlier, which we covered when the atmospheric half of the signal deepened.
Two more details from the 13 July update line up with what the weekly SST data has been saying. CPC observed that "Niño 4 has gradually cooled during the past several weeks," and that weak negative subsurface anomalies were emerging between 150 and 160E. Both are consistent with the east-weighted structure visible in the Niño region readings. Separate instruments, same picture.
So the honest framing is narrow. The RMM-indexed MJO contribution has weakened sharply during an important development period, which removes a potential accelerant. It does not stop El Niño, it does not prove future intensification will underperform, and with models expecting the signal to rebuild it may not last long. Anyone reading it as a forecast of a weaker event is going well past the evidence.
Track the MJO alongside the other indices on our teleconnections dashboard, which carries the live RMM phase and amplitude.