El Nino / Southern Oscillation Tracker

Alerts

Raised against thresholds and, for what changed, against the state a day before, from the local database; an alert fires once when its condition appears and clears itself when it lifts, and what was raised, raised a level or cleared is marked so for a day. Severity colour never travels alone: each carries an icon and a written level.

• critical
  Champi: major typhoon, Category 5-equivalent super typhoon

  140 kt sustained, 917 mb, at 14.3N 138.4E in the West Pacific. It is over open ocean.

  standing since 17 Oct 2026 07:26 UTC
• critical
  Record equatorial warm water volume

  Aug 2026 WWV anomaly is +3.46 × 10¹⁴ m³ - the highest of 560 months on record. Stored heat is the fuel available to the surface anomaly over the following two to three seasons.

  standing since 5 Oct 2026 13:23 UTC
• warning
  Projected to peak in very strong territory

  Ensemble peak +2.29 °C at NDJ 2026, 82% probability of exceeding +2.0 °C.

  standing since 5 Oct 2026 13:23 UTC
• warning
  RONI at or above +1.5 °C - Strong El Nino

  JAS 2026 RONI is +1.69 °C, which CPC prints as +1.7: Strong El Nino conditions this season. An event's category is set by its peak, which is still ahead of it while it is building.

  standing since 5 Oct 2026 13:23 UTC
• warning
  Heat content still rising during an active event

  WWV is +0.253 × 10¹⁴ m³ per month and still climbing while the surface anomaly is already large. Events normally begin discharging by this stage; continued recharge means the event has not yet spent its fuel.

  standing since 5 Oct 2026 13:23 UTC
• warning
  Third-warmest JAS or better

  Rank 2 of 77 for this calendar season on RONI.

  standing since 5 Oct 2026 13:23 UTC
• warning
  Ocean and atmosphere are coupled

  Composite atmospheric index +1.93σ, with 100% of indicators in agreement.

  standing since 5 Oct 2026 13:23 UTC
• watch
  El Nino conditions present, episode not yet formal

  El Nino conditions present for 4 consecutive seasons of RONI at or above +0.5 °C; 1 more needed before NOAA logs this as a formal episode.

  standing since 5 Oct 2026 13:23 UTC
• watch
  Still intensifying

  RONI rose +0.33 °C from the previous season (intensifying rapidly).

  standing since 5 Oct 2026 13:23 UTC
• watch
  South of Southern Mexico: 90% chance of a tropical cyclone within 7 days

  NHC's outlook of Mon Oct 19 05:10:14 2026: 40% within 48 hours, 90% within 7 days. Disorganized showers and thunderstorms have persisted with an area of disturbed weather located several hundred miles south of Guatemala. Environmental conditions appear conducive for development, and a tropical depression is expected to form around the middle of this week as the system moves westward to west-northwestward at around 10 mph, well to the south of the coast of southern Mexico. Formation chance through 48 hours...medium...40 percent. Formation chance through 7 days...high...90 percent.

  standing since 16 Oct 2026 23:37 UTC
• watch
  East and central Pacific: 2.6x normal activity for the date

  343 units of accumulated cyclone energy against 131 for a normal season by 19 October. 22 named storms, 12 hurricanes, 10 major.

  standing since 5 Oct 2026 13:23 UTC
• watch
  Trudy is 233 km from Cabo Corrientes now, moving away

  The 2026-10-19 06Z analysis has it as a tropical storm at 55 kt, 233 km from Cabo Corrientes, Mexico, and the official forecast takes it further away from here on. This is a centre-line distance; the wind field is wider than the line.

  standing since 13 Oct 2026 15:16 UTC
• watch
  Trudy is 289 km from Puerto Vallarta now, moving away

  The 2026-10-19 06Z analysis has it as a tropical storm at 55 kt, 289 km from Puerto Vallarta, Mexico, and the official forecast takes it further away from here on. This is a centre-line distance; the wind field is wider than the line.

  standing since 13 Oct 2026 15:16 UTC
• watch
  Repeated westerly wind forcing over the warm pool

  5 months in the last year averaged strongly westerly at 850 hPa in the western Pacific, which is how downwelling Kelvin waves get launched.

  standing since 5 Oct 2026 13:23 UTC
• info
  Legacy ONI and official RONI diverging

  ONI +2.16 against RONI +1.69, a gap of +0.47 °C. That much of the raw anomaly is the whole tropical ocean running warm rather than the Pacific, which is why CPC has classified on RONI since February 2026.

  standing since 5 Oct 2026 13:23 UTC
• info
  Atlantic: 0.17x normal activity for the date

  18 units against 109 normal by 19 October. A suppressed basin is the expected signature of El Nino in the Atlantic, and it is not a reason to relax: 1992 was a quiet season and it produced Andrew.

  standing since 5 Oct 2026 13:23 UTC
• info
  Westerly wind event under way: central region, 155E-180, 5S-5N

  70.5 days above +2 m/s since 2026-08-08 06Z in CCMP's near-real-time winds to 2026-10-17, peaking at +9.2 m/s on 2026-08-14; longer than any of the 351 events in Harrison and Vecchi's 1986-1995 list (at most 32.5 days): a sustained westerly anomaly more than a burst. Equatorial westerly anomalies force downwelling Kelvin waves east.

  standing since 15 Oct 2026 16:17 UTC
• info
  Westerly wind event under way: east region, 180-150W, 5S-5N

  16.5 days above +2 m/s since 2026-10-01 06Z in CCMP's near-real-time winds to 2026-10-17, peaking at +10.1 m/s on 2026-10-12. Equatorial westerly anomalies force downwelling Kelvin waves east.

  standing since 15 Oct 2026 16:17 UTC

Cleared in the last day: feeds_cached, tc_threat_ep212026_socorro_island, tc_threat_ep212026_manzanillo, tc_hurricane_ep212026, weekly_surge, tc_ww_ep212026_tropical_storm_warning, tc_hurricane_sh012027, tc_threat_ep212026_mazatlan, tc_ww_ep212026_hurricane_warning, tc_ww_ep212026_hurricane_watch

NOAA CPC official synopsis

Issued 8 October 2026 · next update 12 November 2026

Tropical cyclones — live tracks and forecast spread

1 active · decks as of 2026-10-19 09:00 · solid line is the best track, a dot every six hours; dashed is the official forecast, each point marked with its category; the faint lines in each storm's colour are its ensemble members, fainter with each day of lead

Also live, warned on by JTWC: Champi (West Pacific, Category 5-equivalent super typhoon); Wati (Southern Hemisphere, tropical storm). Drawn on the globe and the storm desk rather than on this map of the Atlantic and eastern Pacific.

Distances are to the forecast centre line. The wind field is wider than the line — hurricane-force wind reaches tens of kilometres either side of it and tropical-storm force roughly twice that — so a two-hundred-kilometre pass is not a miss. The faint lines behind each forecast are the honest width of it.

Storm Desk

3 live storms · 4 formation areas

Open the Storm Desk → Every live storm on the newest satellite frame, drawn at the time that frame was taken, with its forecast track, cone and wind field. Zoom to the eye, swipe between imagery layers, loop the last two hours, and move the forecast forward to see the places each wind threshold reaches. Each storm's Protect tab has its watches and warnings, wind speed probabilities, peak surge, and what to do for people and for animals.

Every storm is also on the El Niño map, over street maps and satellite.

• Champi · Category 5-equivalent super typhoon · advisories from JTWC
• Trudy · Tropical storm · advisories from NHC
• Wati · Tropical storm · advisories from JTWC
• Formation, NHC: South of Southern Mexico · 40% in 2 days, 90% in 7 days
• Formation, NHC: Off the East Coast of the United States · 0% in 2 days, 40% in 7 days
• Formation, JTWC: Invest 96A · low potential
• Formation, JTWC: Invest 93S · low potential

Season verification — what this El Nino did to the basins

Accumulated cyclone energy through 19 October · climatology 1991–2020 · El Nino composite from 22 seasons with ASO RONI printed at or above +0.5 °C

This is the one panel in the system where a forecast it made is scored against an outcome in the same season. The Atlantic suppression and the east Pacific enhancement are the two most reliable seasonal signals El Nino has, and they are being tested here in public rather than asserted. About 89 per cent of a normal season's energy has fallen by this date, so the comparison has that much of a season behind it.

On stopping a hurricane: it is not possible. Project STORMFURY seeded eyewalls with silver iodide from 1962 to 1983 and was abandoned because the changes it produced could not be separated from natural eyewall replacement, and because mature hurricanes carry far too little supercooled water for seeding to work at all. A Category 4 releases on the order of 10¹⁹ joules a day as latent heat, several hundred times world electricity generation. What can be delivered is lead time, and the tracks above are four days of it.

Project STORMFURY, recreated

The 1962-83 hurricane modification experiment, rebuilt as a decision procedure and run against the storms that are live right now. Three parts, kept separate because they failed differently: the mechanism, the trigger, and the measurement.

3 of 3 live storms fail STORMFURY's own criteria. Every one of them fails on supercooled water, which is the criterion no hurricane has ever passed and the reason the programme ended.

The mechanism, computed

Seeding was supposed to build a new eyewall further out. Air arriving at a larger radius with the same absolute angular momentum turns more slowly, so the wind falls - and the arithmetic (M = rv + fr²/2) genuinely predicts a drop about the size STORMFURY reported. That part was never wrong.

What real eyewalls actually do

Hurricanes relocate their own eyewalls, unaided, several times a season. That gives a live test of the prediction above: take every natural expansion in a hurricane this season, ask what conserving angular momentum says the wind should become, and compare it with the wind the storm actually had six hours later.

The prediction is consistently far too strong. A hurricane is not a spinning-down flywheel - it is fed high-momentum air continuously through the inflow, and the vortex restores itself while the eyewall moves. A closed-system momentum argument is an upper bound, not a forecast, and STORMFURY's expected effect came from exactly that calculation.

Why nobody could have told

The claimed result for Debbie was a 31% wind reduction. This is every 24-hour intensity change in every unseeded hurricane this season, on the same axis. An effect that has to be picked out of this distribution, from a sample of four storms, was never going to be separable from what hurricanes do on their own.

Would it even have been allowed to fly?

STORMFURY could not seed whatever it liked. Its own operating rules are checked here against each live storm - including the one rule no hurricane has ever passed.

The record

Four storms in twenty-one years, because the criteria above are almost never met at once. The seeding layer the hypothesis needed sits between 5.0 and 7.5 km, about 2.5 km deep, above a freezing level near 4.4 km - and the observation that closed the programme is that there is almost no supercooled water up there to freeze. Willoughby et al., Bull. Am. Meteorol. Soc., 1985

Formally ended in 1983. By then the natural eyewall replacement cycle had removed the evidence, the microphysics had removed the mechanism, and a move to the west Pacific in search of storms far from land had run into the problem that other nations' storms are not available to be experimented on.

What it did buy: Twenty years of research aircraft inside hurricanes. The eyewall replacement cycle, the modern understanding of hurricane structure, and the reconnaissance programme that still flies are all products of STORMFURY's observing arm. The measurement outlived the experiment it was built to serve. The deliverable from all of this is lead time, and the tracks and the forecast panels above are what lead time looks like.

The map

Open the El Niño Map → From the whole planet down to a single street, on Esri's street, satellite and terrain maps, with Google's own map, satellite view and Street View of any point on request. Over them go this tracker's El Niño composites for September to November, the season now: what the El Niño seasons since 1979 did to rainfall and temperature against the ordinary ones. Today's sea surface temperature anomaly and the last three days' floods come from NASA, and the documented impact regions are a tick away, solid where this event puts them in play.

• Tap anywhere for what the El Niño seasons since 1979 did there, season by season, and whether the record can tell it from an ordinary year.
• The documented effects whose regions take in the point, marked in play or on the watch list for this event.
• Today's ocean temperature anomaly and flood class at that spot, read from NASA's own tiles, while those layers are on.
• Google's map, satellite view and Street View of the point, and links that open it in Google Maps and Google Earth.

Where it lands

Every other tier here is about the Pacific. This one is about a point on the ground. The atlas holds what the El Nino events since 1979 actually did at each cell of the world - rainfall against an ordinary year, and surface temperature - and resolves a click to the nearest of 7,342 named places.

Open the interactive atlas → Drag to pan, scroll or use the buttons to zoom, click anywhere for the local record.

• Composited per season over the El Nino events since 1979 in which RONI reached +1.0, against the neutral years (RONI inside +/-0.5): DJF 9 events (23 months) against 18 neutral years, |t| 2.31; MAM 6 events (10 months) against 33 neutral years, |t| 2.57; JJA 4 events (9 months) against 33 neutral years, |t| 3.18; SON 10 events (26 months) against 20 neutral years, |t| 2.26.
• Each event counts once and each neutral year once, because the months inside one event are not independent. A cell is called wetter or drier only where Welch's t passes the 95% two-sided point of Student's t for its season's sample - the |t| quoted with each season. Everywhere else the record does not separate the event from an ordinary year, and the page says so rather than colouring it in.
• The El Nino under way when the grids were built (2026) is not in the sample: the composite is what past events did.
• The composite is what has happened in past events, not a forecast for this one. The forecast tier is the one that forecasts.

A verdict of 'no clear signal' is a measurement, not a gap: across that season's El Ninos and neutral years, the record does not separate the two at that place. A temperature in parentheses does not pass the same test as the rainfall beside it: the sign is the composite's, the size is inside the noise.

The season shown is the one the rainfall record is most sure about at that place, which is not always the one the event peaks in: Jakarta's strongest signal is the June-to-August dry season, not the December monsoon, which is why El Nino fire years there are counted in the middle of the year.

Fortaleza is the one to read twice: the Nordeste drought is a textbook El Nino teleconnection, and the composite has it at -41% of normal in December to February - yet across 9 events its t of -1.86 does not reach the 2.31 that season's sample needs. The sign is right and the sample is not big enough to say so. Calling that a drought signal is the most common way a composite like this gets misused.

Read the percentages against the millimetres beside them. Los Angeles composites at +63% of normal in September to November, which sounds large and is 0.3 mm a day on top of an ordinary 0.5: the percentage is large because the season is dry there, not because the anomaly is.

Where the anomaly actually is

Daily OISST v2.1 sea-surface temperature anomaly for 2026-10-17, NOAA’s preliminary analysis (the final replaces it about two weeks on), on a half-degree sample of the quarter-degree analysis, against its 1971–2000 climatology. The four Nino boxes are drawn in their real positions with their area-weighted means, so the shape of the anomaly can be compared against the index it produces. The eastern edge of the 28 °C warm pool on the equator sits at 94W; in a neutral year it is near 180.

The Nino boxes day by day

The four Nino boxes each day, on the Pacific map’s own half-degree cells: NOAA’s final analysis from 2026-07-05 to 2026-10-03, then NOAA’s preliminary analysis from 2026-10-04 to 2026-10-17. Nino-3.4 is +3.44 °C on 17 Oct 2026, +0.06 °C on the week. The anomalies are OISST’s own, against its 1971–2000 climatology. CPC’s traditional weekly Nino-3.4, the hollow dots, is against 1991–2020. On the 15 weeks both carry, the days’ Nino-3.4 averages 0.08 °C warmer than CPC’s. On the 15 days both analyses carry, the final runs 0.05 °C warmer than the preliminary in Nino-3.4 on average, and differs by up to 0.20 °C on a day (0.34 °C in Nino-1+2): about how far the shaded days may yet move.

The Indian Ocean Dipole day by day

The Dipole Mode Index each day, the SST anomaly of the west pole (50–70E, 10S–10N) less that of the east pole (90–110E, 10S–0), each area-weighted on half-degree cells: NOAA’s final analysis from 2026-07-05 to 2026-10-03, then NOAA’s preliminary analysis from 2026-10-04 to 2026-10-17. The DMI is +0.84 °C on 17 Oct 2026, -0.53 °C on the week. The Bureau of Meteorology counts a DMI above +0.4 °C as a positive IOD value and one below -0.4 °C as a negative one, and eight weeks of either as an event. The anomalies are OISST’s own, against its 1971–2000 climatology: a warming common to both poles cancels in their difference, but a difference in their warming does not, so the DMI here can run apart from the Bureau’s own, whose thresholds these are. IOD events typically develop from about May, peak from August to October and break down when the Australian monsoon arrives at the end of the southern spring, so a value past a threshold from December to April seldom marks one. On the 15 days both analyses carry, the final runs 0.03 °C lower than the preliminary in the DMI on average, and differs by up to 0.24 °C on a day: about how far the shaded days may yet move.

The thermocline, in three dimensions

Depth of the 20 °C isotherm across the TAO/TRITON array, as of 2026-09-27. Longitude runs west to east, latitude across, and the surface height is depth, so the sheet you are looking at is the top of the cold water. Colour is the anomaly against this mooring's own climatology for the calendar month. The east-west tilt along the equator is currently 25 m on this daily array; a neutral Pacific runs near 100 m, and a strong event flattens it. The cross-section below quotes the same tilt from the monthly moorings, over a wider span, so the two numbers are two measurements rather than one. Drag to rotate. Use the buttons for fixed views.

The thermocline in cross-section

Equatorial Pacific temperature against depth, 2026-09, from the TAO/TRITON moorings. This is the mechanism rather than the symptom: the surface warming that the indices measure is what happens after this slope relaxes. The 20 °C isotherm averages 151 m across the two westernmost moorings and 127 m across the two easternmost, a tilt of 24 m across the basin. A neutral Pacific runs near 100 m of tilt and a mature El Nino flattens it. The heavy line is the 20 °C isotherm, the conventional marker for the centre of the thermocline.

How the anomaly moved

Time runs downward, longitude across, so anything travelling east leans to the right as you read down. These are three independent instruments watching the same ocean: a feature that appears in all three is a wave, a feature that appears in one is an instrument. The dashed line is the dateline.