If ENSO is the slow, months-to-years climate signal, the Madden-Julian Oscillation (MJO) is its faster, more restless cousin: a pulse of enhanced and suppressed tropical rainfall and wind that circles the globe roughly every 30 to 60 days. It was first described by Roland Madden and Paul Julian in the early 1970s, and it remains one of the most important sources of weather predictability inside the 2-6 week range.
How It's Measured
Meteorologists track the MJO using the Real-time Multivariate MJO (RMM) index, which boils the pattern down to two numbers, RMM1 and RMM2, that together define a "phase" (1 through 8) and an "amplitude." Phase tells you roughly where the active convective pulse currently sits as it travels eastward around the globe; amplitude tells you how strong and well-organized that pulse is. An amplitude below 1.0 generally means the signal is too weak to matter much for forecasting; above 1.0, it's considered active.
The two numbers are best understood as coordinates rather than as separate readings. Plotted together they place the MJO somewhere on a circle, with distance from the centre being amplitude and angular position being phase. That geometry has a consequence worth knowing: phase 8 and phase 1 are adjacent, not opposites, because the pulse travels in a loop. Treating the phase number as a linear scale is a common misreading.
The El Niño Connection
Here's where it ties back to this site's main subject: when the MJO's active convective phase sits over the Maritime Continent and western Pacific (roughly phases 5 through 7) with a strong amplitude, it can trigger westerly wind bursts, short pulses of west-to-east wind along the equator that push warm surface water eastward and help generate the downwelling Kelvin waves that carry heat toward South America. In other words, an active, well-placed MJO can act as an accelerant for El Niño development, especially during the critical spring-to-summer transition period.
A Modulator, Not a Driver
It is easy to overread this. The MJO does not cause El Niño and cannot prevent one. What it does is change the timing and pace of development by supplying or withholding westerly wind bursts at moments when the ocean is already primed. A well-placed active MJO during a developing event is an accelerant. A quiet MJO removes an accelerant. Neither settles the outcome.
The useful frame is that ENSO sets the odds over months to years and the MJO adds or subtracts a few weeks of push inside that. When coverage attributes an El Niño's strength to the MJO, or treats a quiet MJO as evidence that an event is failing, it has confused a modulator with a cause.
A Worked Example From This Event
July 2026 gave a clean demonstration. MJO amplitude fell from 1.31 on 9 July to 0.33 on 24 July, dropping well below the 1.0 activity threshold and staying there. CPC had forecast exactly that decay on 13 July, which is a reminder that MJO evolution is genuinely predictable at two to three weeks in a way that ENSO amplitude is not.
The correct interpretation is that El Niño lost a tailwind during a window when it might have used one, and that is all. The event continued strengthening on other mechanisms. We worked through that in detail in The MJO Goes Quiet.
RMM and ROMI Are Not the Same Index
One index caveat, because this site cares about them. The standard RMM index and the ROMI (OLR-based MJO index) are related but constructed differently, and they can disagree on phase and amplitude at the margins, particularly when the signal is weak. This site's MJO Assist panel tracks phase and amplitude in real time with a plain-language verdict on whether the current state is helping or hindering development, sourced from NOAA PSL's ROMI dataset. Where a number here differs slightly from an RMM-based figure quoted elsewhere, that is usually why rather than an error in either.