Pacific Remains Very Busy, A Final Chance for the Atlantic? AI Models Discussion 9-30-2026

  • Andy Hazelton
    Andy Hazelton

Share

Pacific Remains Very Busy, A Final Chance for the Atlantic? AI Models Discussion 9-30-2026

We will take a tour of the global tropics, once again spending a lot of time on the Pacific, where things remain very active thanks to the ongoing El Niño event.

West Pacific

The big system of interest in the West Pacific right now is 26W, which is located just to the east of the Mariana Islands. The system is forecast to cross the islands and move parallel to or near them, intensifying as it does so (Figure 1).

Figure 1: Google DeepMind WNV3 64-member ensemble forecasts of TC tracks and intensity over the West Pacific initialized at 06 UTC September 30, 2026, valid out to 360h. Data from Google.

The reason for the northward turn is a weakness in the subtropical ridge with a broad trough to the north (Figure 2), which should allow the system to recurve to the southeast of Japan. This will definitely be a system to watch in the Marianas, though, since it could be very close to those islands as it intensifies.

Figure 2: AIFS forecast of 500-hPa height and MSLP initialized at 06 UTC September 30, 2026, and valid at 06 UTC October 4, 2026.

Beyond this system, Hurricane Nolo will also move into the West Pacific from the Central Pacific, and could intensify as it does so (Figure 1). There is also an indication for development of yet another system in the medium range.

East Pacific

The East Pacific basin is very active right now. Hurricane Polo has dissipated over Mexico, but Hurricane Rachel is moving to the South of Mexico, TD 19E is further west, and there is a signal for further development behind these systems (Figure 3).

Figure 3: Google DeepMind WNV3 64-member ensemble forecasts of TC tracks and intensity over the East Pacific initialized at 06 UTC September 30, 2026, valid out to 360h. Data from Google.

Unlike Polo, Rachel is forecast to move west rather than north, thanks to a narrow ridge to the north of the TC (Figure 4). This will keep it offshore from Baja California.

Figure 4: GraphCast forecast of 500-hPa height and anomalies initialized at 00 UTC September 30, 2026 and valid at 00 UTC October 5, 2026.

It looks like the next system will likely follow in the footsteps of Polo and Rachel, with Google DeepMind’s 1000-member ensemble showing a ~75% chance of development and many members significantly intensifying it in the medium range (Figure 5).

Figure 5: Google DeepMind 1000-member TC genesis timing, track, and intensity forecasts for the potential East Pacific system, initialized at 06 UTC September 30, 2026. Figure from www.tcatlas.org.

The environment continues to look favorable in this region, with light upper winds thanks to a large anticyclone aloft near Mexico (Figure 6). Upwelling of cooler water by Polo and Rachel could be an issue for the system based on the exact track.

Figure 6: FourCastNet-v3 forecast of 250-hPa wind initialized at 00 UTC September 30, 2026 and valid at 12 UTC October 7, 2026.

Atlantic

The Atlantic will head into the final 2 months of the hurricane season without a single hurricane-strength system this year. As the El Niño continues to strengthen, the opportunities for development are dwindling. One of the final chances could be from a system in the Bay of Campeche or Southern Gulf in the next 7-10 days (Figure 7).

Figure 7: Google DeepMind WNV3 64-member ensemble forecasts of TC tracks and intensity over the Atlantic initialized at 06 UTC September 30, 2026, valid out to 360h. Data from Google.

The source for this appears to be a stalled front, with possible contribution from the Central American Gyre. Upper level winds look generally favorable in the Southern Gulf, with stronger shear possible if the system lifts further north (Figure 8).

Figure 8: AIFS forecast of 250-hPa wind initialized at 00 UTC September 30, 2026 and valid at 00 UTC October 10, 2026.

This is still a long way out, so we will watch and see what comes of it!