MJO On the March, Late Season Atlantic Activity (or not)?: AI Models Discussion 11-25-2025

  • Andy Hazelton
    Andy Hazelton

Share post

MJO On the March, Late Season Atlantic Activity (or not)?: AI Models Discussion 11-25-2025

After Cyclone Fina made landfall in Australia, our attention shifts back to the Northern Hemisphere for now, as an MJO pulse moves through the Maritime Continent and West Pacific and helps spawn some potential TC activity.

Atlantic

There has been a little bit of intrigue for the possibility of a very late-season Atlantic TC, as the aforementioned MJO pulse moves through in a week or two. At this point, however, the GFS is the main global model showing development, and given its known biases, this seems unlikely. Google DeepMind currently only has one member that is showing development in the Caribbean in the next 10-12 days (Figure 1).

Figure 1: Google DeepMind forecasts of TC tracks (shaded by intensity) over the Caribbean, initialized at 06 UTC November 25, 2025, and valid out to 312h. Figure from www.weathernerds.org.

Upper-level winds do look unusually favorable for this late in the season (Figure 2), so it’s possible that something could sustain itself if a disturbance is able to get going and stay over water.

Figure 2: GraphCast forecast of 250-hPa wind initialized at 06 UTC November 25, 2025, valid at 06 UTC December 5, 2025.

With a large upper ridge (Figure 3) over the Gulf, however, it seems likely that any nascent disturbance will make its way into Central America before it can get going. We’ll see if anything changes over the next week or so.

Figure 3: AIFS forecast of 500-hPa height and anomalies initialized at 06 UTC November 25, 2025, valid at 06 UTC December 5, 2025.

Indian Ocean

The Bay of Bengal is quite convectively active with the MJO pulse moving just to the east, and there are two systems of note: TD 04B and Invest 96B. TD04B is near the coast of Sumatra, and will move inland soon. Models have generally backed off of any significant development with this system in the Indian Ocean (Figure 4).

Figure 4: Google DeepMind forecasts of TC tracks (shaded by intensity) over the Eastern Indian Ocean, initialized at 06 UTC November 25, 2025, and valid out to 312h. Figure from www.weathernerds.org.

Another system, 96B, is located near Sri Lanka. Most models show this system staying fairly weak, with southwesterly upper-level winds shearing it as it parallels the coast of India (Figure 5).

Figure 5: GraphCast forecast of 250-hPa wind initialized at 06 UTC November 25, 2025, valid at 06 UTC December 1, 2025.

West Pacific

In the West Pacific, TD 33W is moving into the South China Sea. Models indicate that it could intensify some as it moves to the West (Figure 6).

Figure 6: Google DeepMind forecasts of TC tracks (shaded by intensity) over the Western Pacific Ocean, initialized at 06 UTC November 25, 2025, and valid out to 312h. Figure from www.weathernerds.org.

It’s not clear if the system will make it all the way to Vietnam, as a weakness in the ridging to the north will probably cause the storm to slow down and eventually potentially turn to the North. This will also help to induce some southerly shear over the storm (Figure 7), which may cap the intensity before it gets too strong.

Figure 7: FourCastNest-v3 forecast of 250-hPa wind initialized at 06 UTC November 25, 2025, valid at 12 UTC November 28, 2025.

In the medium range, models indicate yet another system may form east of the Philippines, aided by a westerly wind burst associated with the MJO passage (Figure 8). Some of the Google DeepMind members show it becoming fairly intense, though it may be a fairly large and monsoonal circulation initially. This will be something to monitor as we head into next week.

Figure 8: GraphCast forecast of 10-meter wind and MSLP initialized at 06 UTC November 25, 2025, valid at 06 UTC December 2, 2025.