North Swell in the Virgin Islands

Is there a north swell in the Virgin Islands today?

No north swell is in the forecast to Thu, Oct 15. A small swell from the NNE (0.5 m at 10 seconds) is running today; the waves at the islands this week are the trade-wind sea from the east-southeast, about 1.2 m today (Open-Meteo marine forecast, north of St. Thomas).

Open-Meteo Marine Weather API (the week's forecast), its best-match model, at 19.0° N 64.75° W (open water north of St. Thomas and Tortola), read again every hour. A forecast model, not a measurement; the National Weather Service in San Juan issues the High Surf Advisories for the U.S. Virgin Islands.

The north swells of the winter of 2025–26

NOAA's wave model put 18 north swells of 1.4 m or more at the islands between October 1, 2025 and April 30, 2026, each from between west-northwest and east-northeast with a period of 10 seconds or longer. The first two were Hurricane Imelda's: it passed over Bermuda on the night of October 1 and raced off to the north-east as a large post-tropical storm. The National Weather Service in San Juan issued 11 High Surf Advisories and 2 High Surf Warnings between October and April; the U.S. Virgin Islands were in 6 of the advisories.

The swell at 19.0° N 64.75° W in NOAA NCEP GFS-Wave (WAVEWATCH III), global 0.25°, analysis every 6 hours (NOAA Open Data Dissemination on AWS); the buoy’s largest significant wave height within 15 hours of the peak (NOAA National Data Buoy Center: buoys 41043 (NE Puerto Rico), 41052 (south of St. John), 41053 (San Juan)); the source traced as in the method below; the NWS products from the Iowa Environmental Mesonet, NWS VTEC archive: High Surf Advisories and Warnings, San Juan office.

The winter’s most powerful north swell: February 2–3, 2026

On January 31, 2026 a low off the Carolinas began to deepen. In the 24 hours to 2 PM AST on February 1 its central pressure fell 29 hPa, to 973 hPa, in NOAA's GFS analysis; meteorologists call a storm that deepens that fast a bomb. Its winds raised seas of 11 m in NOAA's wave model.

The swell that reached the islands was made on the storm's south-west side, where the wind blew from the northwest. Buoy 41043 measured it arriving with 14-second waves from 315°. Walked back along that line at the speed a 14-second swell carries its energy, about 930 km a day, it leads to 8 AM AST on February 1 at 28.0° N 72.8° W, north-east of the Bahamas and 760 km south of the storm's centre, where the model had northwest winds of 15 m/s (29 knots) blowing toward the islands over seas of 6.4 m. From there the swell needed 29 hours to reach the buoy and about 33 to reach the islands, 1,290 km away. By then the storm itself was off Newfoundland, at 965 hPa.

Buoy 41043, 306 km north of St. Thomas in 5,286 m of water, felt it first: 2.1 m at 8:40 AM AST on February 2, 4.1 m two and a half hours later and 4.6 m at 12:40 PM, with 14-second waves from the northwest. The National Weather Service issued a High Surf Warning, the second of the season, for Puerto Rico's north and west coasts, with breaking waves of 15 to 20 feet; the St. Thomas and St. John zone was under a High Surf Advisory that read 15 to 18 feet. Its post-event analysis put breaking waves at 17 to 22 feet on exposed coasts of the northern U.S. Virgin Islands.

South of St. John, buoy 41052 never measured more than 1.1 m during the swell. The islands are a wall.

The storm, every 6 hours (NOAA GFS analysis)

What the buoys measured

The winter’s biggest north-swell days at buoy 41043

Each day’s highest significant wave height at NOAA National Data Buoy Center: buoys 41043 (NE Puerto Rico), 41052 (south of St. John), 41053 (San Juan) buoy 41043 from between 285° and 75°, with the dominant period and the mean direction at that reading; energy is height squared times period, the energy a swell carries toward the shore. The highest reading of the winter, 4.67 m on October 10, 2025, came from the east-southeast, not the north.

How far a swell travels in a day

Swell sorts itself by period: the longer the waves, the faster their energy travels. In deep water the energy moves at the group speed, gT/4π, half the speed of the crests.

Deep-water linear theory (WMO-No. 702, Guide to Wave Analysis and Forecasting; Delpey, Ardhuin, Collard & Chapron (2010), JGR Oceans: group speed gT/4π and swell dispersion). “Off Newfoundland” is 45° N 50° W, 3,272 km from St. Thomas on a great circle.

Which shores get north swell

Over the Virgin Bank the water is 20 to 60 m deep, shallow enough for a long swell to feel the bottom. It slows, its crests bend toward the shallows and wrap around the islands. The north shores face it; the south shores sit in the lee. NOAA's Coast Pilot says Magens Bay "is open to the northwest and consequently exposed to rollers", that the ground swell accompanying northers "is especially heavy" off St. Croix's north shore near White Horse, and that Pillsbury Sound "is quite secure against rollers".

When a north swell runs

How this map is made

The sea is NOAA's GFS-Wave model (WAVEWATCH III, the wave component of the GFS), its analysis every 6 hours from October 1, 2025 to April 30, 2026, averaged to 1° for the film and shown every 12 hours (every 6 around the February swell). Between model times the wave height is carried along the dominant waves' direction at their group speed. GFS-Wave does not assimilate buoy measurements; the buoys are the check.

A north swell is a swell partition at 19.0° N, 64.75° W (open water north of St. Thomas and Tortola) from between 285° and 75° with a mean period of 10 s or more and a height of 1.4 m or more, its peaks at least 36 hours apart. Each one is traced back along the great circle it came in on: at each earlier time the swell was its group speed times the elapsed time away, and the 10 m wind there is resolved along its path; where the model's wind blew hardest toward the islands over a sea of about the swell's period is taken as where it was made, read over its arrival direction ±10° and period ±1 s. A swell whose candidates spread over more than 18 hours, or that traces beyond 4,000 km, is listed with its source as unclear.

The February swell is traced from what buoy 41043 measured instead: its waves at the peak (14 seconds from 315°, 12:40 PM AST on February 2) walked back the same way lead to the same model time as the model's own trace, 8 AM AST on February 1, about 200 km closer to the islands. The edge on the map is drawn from there at a 14-second swell's group speed, so it reaches the buoy when the buoy measured the peak.

The crests near the islands are a fast-marching solve of first-arrival time at the linear phase speed on the NOAA NCEI seabed at 0.004° (scikit-fmm): for the February swell a 13-second swell from 326°; through the season 11-second swells from every 15° between 315° and 75°, and a 7-second sea from every 15° between 60° and 120°. Every ninth or twentieth crest is drawn (the trade-wind sea's every 64th, faint) (the real crests are 100 to 300 m apart, too close to see), and a swell of another period is drawn on the nearest field with its own spacing.

Their brightness, and the surf line, is how open the water is to the waves' direction: the share of the swell's directional spread (a Gaussian of 15°, 12° for the February swell) that reaches each spot without crossing land or a reef shallower than 1.5 m; for the trade-wind sea, its regrowth over the open water upwind of each island (JONSWAP fetch-limited growth, fully grown at 85 km). It agrees with NOAA's Coast Pilot (Magens Bay open to the northwest, Pillsbury Sound secure against rollers) and puts the February swell at buoy 41052, south of St. John, at about a quarter of its open-sea height; the buoy, which also measures the local wind sea, never read more than 1.1 m. Shoaling, breaking and diffraction are not modelled: it is a picture of shelter, not a surf forecast.

Through the season the crests follow NOAA GFS-Wave at 19.0° N 64.75° W every 6 hours: the north swell's height sets their brightness, its direction picks the crest field and its period the spacing; the rest of the sea (everything but the north swell) is drawn as the trade-wind sea from the direction of the GFS 10 m wind when that is easterly. The film's clock slows through each north swell.

Hurricane Imelda is named from the National Hurricane Center's discussions: it passed Bermuda on the night of October 1, 2025 and was post-tropical at 33.2° N 59.5° W by 11 AM AST on October 2, racing east-north-east; the GFS lows the trace found on October 2 and 3 sit on that track.

1. NOAA NCEP GFS-Wave (WAVEWATCH III), global 0.25°, analysis every 6 hours (NOAA Open Data Dissemination on AWS)
2. NOAA NCEP WAVEWATCH III model description (grids, swell partitions)
3. NOAA National Data Buoy Center: buoys 41043 (NE Puerto Rico), 41052 (south of St. John), 41053 (San Juan)
4. Iowa Environmental Mesonet, NWS VTEC archive: High Surf Advisories and Warnings, San Juan office
5. NOAA NCEI Storm Events Database, 2026 (High Surf, 2–4 February 2026)
6. NOAA U.S. Coast Pilot 5, chapter 14 (Virgin Islands)
7. NWS San Juan, The Whispering Trades (2014): the wave seasons of Puerto Rico and the USVI
8. WMO-No. 702, Guide to Wave Analysis and Forecasting
9. Delpey, Ardhuin, Collard & Chapron (2010), JGR Oceans: group speed gT/4π and swell dispersion
10. Open-Meteo Marine Weather API (the week's forecast)
11. NOAA NCEI DEM Global Mosaic (the seabed under the islands)
12. NASA Blue Marble and the OSM land mask (NASA GIBS)
13. BVI Department of Disaster Management: High Surf Advisory (Antigua and Barbuda Meteorological Service)
14. Applesurf, Tortola: the north-swell surf spots