At 11 a.m. Eastern time Thursday, from the Aircraft Operations Center in Lakeland, Florida, the National Oceanic and Atmospheric Administration printed the number: 8 to 14 named storms, 3 to 6 hurricanes, 1 to 3 major hurricanes. A 55 percent chance of a below-normal season, 35 percent near-normal, 10 percent above-normal. The first time since 2015 that NOAA's May outlook has called a below-normal Atlantic year. [1]

The paper's Wednesday feature argued that households on the Gulf and Atlantic coasts already had the maps they would need: cone, ENSO, drought, HeatRisk, medicine-storage. The Thursday print confirms what those maps implied. The Climate Prediction Center now puts El Niño odds at 98 percent for August through October, with La Niña effectively at zero. [2] That is the climate signal that holds the headline down.

The number is lower than the paper expected to write about. Pre-print signal across forecasters pointed to a 13-to-19 range; NOAA came in beneath even Colorado State's 13 and beneath AccuWeather's 11-to-16 range. [3] [4] The paper said Tuesday that El Niño would arrive as a sea-level multiplier on the same coastlines the hurricane cone would test. Today the El Niño moved the cone instead.

The agency was careful with the qualification. "Even though we are expecting a below-average season in the Atlantic, it is very important to understand that it only takes one," said Dr. Neil Jacobs, the NOAA administrator. "We have had category fives make landfall in below-average seasons." [5] Ken Graham, director of the National Weather Service, added a sharper sentence: "Every category five that's made landfall in this country was a tropical storm or less three days out." [5]

That is the print's first usable line for a household. The forecast range is a planning instrument, not an all-clear. The 1992 season produced six named storms — and one of them was Hurricane Andrew. The 1983 season produced four — and one was Alicia. Below-normal is not a synonym for safe.

The El Niño driving this year's number also drives the contradiction the paper has been tracking on the sea-level brief. NOAA's William Sweet calls El Niño a high-tide-flooding multiplier on the East and Gulf coasts, particularly during September and October king tides. So while the cone is quieter, the inundation map is louder. A coastal Memorial Day weekend driver in Charleston or Tampa will see fewer named storms but more nuisance flooding days, and that gap is the year's structural service-journalism story. [6]

The forecaster reasoning matters. Sea-surface temperatures in the western tropical Atlantic are warmer than normal but cooler than 2023; the Atlantic Multidecadal Oscillation remains in its high-activity phase, but vertical wind shear is expected to be stronger because El Niño shifts the Pacific upper-level winds eastward. ACE — accumulated cyclone energy, the measure that captures both intensity and duration — is projected at 45 to 115 percent of the median. [2] That is a band centered just inside the below-normal cutoff, which is how NOAA hedges against the years the May print misses.

It has missed, and it has missed in one direction. Across the 23 seasons from 2003 to 2025, the actual named-storm count landed inside NOAA's May range 14 times. Of the 9 misses, 8 were busier than NOAA called for; 2005 was 12-to-15 forecast against 28 storms delivered. Only 2006 went the other way. [7] The structural lesson, the one a household should hold in its head while it lays out its medicine and its evacuation map, is that NOAA's May outlook misses by underforecasting, not by overforecasting. A below-normal print does not retire the cone.

The new hurricane-cone graphic this year does what NOAA's experimental phase argued for last summer: it shows inland watches and warnings inside the same shaded cone, instead of pretending the cone is only a coastal artifact. [8] The agency is also rolling out storm-surge watches and warnings for Hawaii for the first time, and an experimental ellipse-based cone that captures 90 percent of forecast tracks instead of the traditional 67 percent. [8] These are the household-facing changes the paper's lost-science thread has argued for since March: more usable maps, not more numbers.

What did not change today: the field-laboratory story 200 miles inland from any beach, and the institutional contradiction the Bundibugyo lead named on Tuesday. NOAA can run its hurricane outlook on time; the CDC's Title 42 order from Monday still contradicts the World Health Organization's no-border-closures instruction; the field lab cartridge that should have caught Bundibugyo in Ituri five weeks ago still has not been replaced. Federal science is decommissioning unevenly: weather modeling holds, public-health procurement does not.

For a Gulf Coast family laying out the Memorial Day weekend, the headline number does one job and one job only. It permits a slightly slower hurricane-prep timeline this year — water by mid-June, generator fuel by July, medicine refills before peak — without permitting the false comfort of skipping the steps. The El Niño that suppressed the headline number is also the climate state that will raise the king tide. The household map is bigger than the cone.

-- DARA OSEI, London