You have seen the cone. Every September it turns up on the evening news and on the lock screen of your phone — a pale funnel opening westward across the Atlantic, with a thin dark line drawn down the middle of it.
Most readers take that shape to be a map ensemble spread behind the cone of where the hurricane will go. What it draws is something narrower and considerably stranger, and the distance between those two readings is how a careful person ends up making a poor decision with good information sitting in front of them.
The cone shows the probable track of the storm's center. It is drawn so the center stays inside it roughly two thirds of the time, which means one storm in three wanders out.
What The Cone Is Actually Drawing
The National Hurricane Center builds the graphic out of circles. At each forecast hour — 12, 24, 36, 48, 72, 96, and 120 hours out — it places a circle around the predicted position of the storm's center, then draws the smooth envelope that contains all of them.
That envelope is the cone. It widens with time for the obvious reason that a forecast four days out is less certain than one made for tomorrow morning, and the circles grow to match.
Every element of the image therefore refers to a single point: the geometric center of the circulation. Nothing in the drawing describes the storm's size, its wind field, its rainfall shield, or the water it pushes ahead of itself.
Where The Two-Thirds Number Comes From
The radius of each circle is not a guess. It is set so that, across the previous five seasons of the Hurricane Center's own official forecasts, the storm's actual position fell inside the circle about two thirds of the time.
Read that sentence twice, because the second implication is the one that gets lost. The cone is built with a one-in-three failure rate designed into it, and in a season producing fourteen named storms, several will step outside their cones at some point.
The National Hurricane Center sizes each cone circle from its own track errors over the previous five seasons. The cone shrinks as forecasting improves, and it measures forecast skill rather than storm size.
Because those radii are recalculated every year, the cone doubles as a moving record of how much better the science has become. Average 72-hour track error has fallen by roughly 75% since 1990, and today's five-day forecast is about as reliable as a three-day forecast was a generation ago.
This is the same distinction we draw elsewhere in the journal between being right and being sure — see how accurate a weather forecast really is and what forecast confidence actually communicates. A number can be perfectly honest about its own uncertainty and still be misread by nearly everyone who looks at it.
How Wide The Cone Is On Each Day
Abstract uncertainty is easy to nod along with and hard to act on, so here are the dimensions in miles. The figures below are approximate values from recent Atlantic seasons, and the Hurricane Center republishes the exact table each spring.
| Forecast lead time | Approximate radius | Approximate full width | What that covers on the ground |
|---|---|---|---|
| 24 hours | ~40 nautical miles | ~90 miles | About one metro area, edge to edge |
| 48 hours | ~70 nautical miles | ~160 miles | Several counties on either side of the line |
| 72 hours | ~100 nautical miles | ~230 miles | Tampa to Fort Myers, and then some |
| 96 hours | ~150 nautical miles | ~345 miles | Most of a state's coastline |
| 120 hours | ~200 nautical miles | ~460 miles | Two or three states at once |
All of these add up to one practical point. At day five the cone is wide enough to swallow an entire coastline, so being inside it tells you almost nothing actionable; at day one it is tight enough to plan a morning around.
At three days out the Atlantic cone runs about 100 nautical miles either side of the forecast track, roughly 230 miles wide. At five days it is closer to 460 miles across.
The Misreading That Puts People At Ease
Here is the failure mode, and it is remarkably consistent. A reader finds their town sitting just outside the white shape, exhales, and stops paying attention — because the eye reads an edge as a boundary, and a boundary as a wall.
The cone has an edge because the statistics had to stop somewhere. It is a percentile line, and the world just beyond it is only marginally less likely than the world just inside it.
Hurricane Ian is the case study most forecasters reach for. For several days in September 2022 the center line pointed toward Tampa Bay and a great deal of public attention followed it; the storm came ashore near Cayo Costa, roughly 100 miles to the south, in an area that had been inside the cone the entire time.
The forecast did not fail there. The graphic performed exactly as designed, and the public reading of it did not.
Standing outside the cone means the storm's center is unlikely to pass over you. Wind, surge, and rain reach far past the center, so the cone edge should never be treated as a safety line.
The identical misreading happens with percentages, which we took apart in what a chance of rain actually means. In both cases the number is doing honest work while the reader quietly translates it into a promise it never made.
What The Cone Leaves Out
It is worth listing the questions the graphic cannot answer, because nearly every one of them matters more to your week than the track does. The omissions include but are not limited to:
- Intensity. The cone looks identical for a 45 mph tropical storm and a Category 5, because it encodes position and nothing else. Strength lives in the advisory text and the wind speed probability products.
- Size. Tropical-storm-force winds can extend more than 200 miles from the center of a large system, and Sandy's wind field spanned roughly 1,000 miles at its widest. A compact, intense storm and a sprawling, weaker one can produce the same cone.
- Storm surge. Water is responsible for the large majority of deaths in US tropical cyclones, and surge depends on bathymetry, coastline shape, and the angle of approach — none of which appear anywhere in a track forecast.
- Rainfall. Inland flooding follows the moisture plume rather than the center; Ida came ashore in Louisiana in August 2021 and produced its deadliest hours two days later in New Jersey and New York.
- Timing and forward speed. Two storms with matching cones can deliver six hours of weather or thirty-six, and a stalled system is a categorically different disaster from a fast-moving one.
- Tornadoes. The right-front quadrant of a landfalling hurricane routinely spins up tornadoes well away from the center, often in outer rainbands hours before the core arrives.
- Anything past day five. The cone ends where the official forecast ends, and it vanishes entirely once a system is declared post-tropical, even when the rain carries on for another two days.
Taken together, that list covers nearly every decision a household actually makes during a hurricane week. Track is the one variable the cone handles well, and it is rarely the variable that determines what happens at your address.
The Products That Answer What The Cone Cannot
Fortunately the Hurricane Center publishes a graphic for each of those gaps, and every one of them is free and public. Here is what to open, and for which question:
- Potential Storm Surge Flooding Map. Shows reasonable worst-case water depth above ground, neighborhood by neighborhood. For anyone within a few miles of the coast it is the single most decision-relevant product published.
- Wind speed probabilities. Gives the percentage chance that your specific location sees tropical-storm-force or hurricane-force winds, which is the honest answer to "am I in this or not."
- Arrival time of tropical-storm-force winds. Tells you when outdoor work becomes impossible, which is the real deadline for boarding up or leaving — usually long before the eye is anywhere near you.
- Watches and warnings. A hurricane watch means conditions are possible within 48 hours; a warning means they are expected within 36. Since 2024 the cone graphic has also carried inland watches and warnings, which closes part of the misreading gap directly on the image itself.
- Key Messages and your local forecast office. The Key Messages graphic states the three or four things forecasters most want you to know in plain language, and the local National Weather Service office translates all of it to your county.
None of these require meteorological training to read, and all of them answer sharper questions than the cone does. If you open exactly one product this season, make it the surge map.
Use the storm surge flooding map for water, the wind speed probabilities for wind, and the watches and warnings for timing. The cone answers none of those three questions.
How To Read The Cone In Four Moves
The graphic is still worth your attention, provided it arrives fourth rather than first. Here is the order we use:
- Start with watches and warnings. They are the only part of the package carrying operational meaning, and they are already drawn around hazard extent rather than center position.
- Ask how far your hazard travels from the center. Surge answers in tens of miles, wind in hundreds, and rainfall sometimes in the better part of a thousand — the cone's width bears on none of the three.
- Ignore the center line entirely. Forecasters have asked for years that it be de-emphasized, and reading it as a prediction of landfall location is the specific error that sent attention to the wrong side of Ian.
- Re-read at every advisory. Full advisories come every six hours, with intermediate updates every three once watches or warnings are up, and a cone that shifts 60 miles overnight is ordinary rather than alarming.
Read the watches and warnings first, the hazard graphics second, and the cone last. The center line is the least useful mark on the image and should carry no weight in a decision.
Worth keeping for the season: the cone is a statement about the storm's center, and your plan is a statement about your address. Those are two different questions, and only one of them is answered by the picture on television.
Why This Reads Differently On September 1
The Atlantic season runs June 1 through November 30, and the climatological peak falls on roughly September 10. The stretch from mid-August to mid-October produces the large majority of the basin's activity, which is why this is the fortnight when cone literacy stops being an academic exercise.
Early September also favors exactly the storms that are hardest to read: warm water deep into the column, low shear, and a long Atlantic runway that generates five-day forecasts while the system is still thousands of miles from anyone. Those are precisely the forecasts where the cone is at its widest and the pull of the center line is at its strongest.
If you want the machinery underneath, the models behind the forecast covers how the guidance is produced, and what falling barometric pressure tells you explains the number people watch as a storm deepens.
How We Handle It In A Brief
Vesper's position on this is straightforward. A hurricane brief that reprints the cone and the category has not done any work, because the reader's question is about the day in front of them — what to do, and by when.
So we write the deadline. "Boards up by Tuesday noon, winds arrive Wednesday at dawn, power out into the weekend" is a brief; "Category 3 approaching Thursday" is a caption.
That discipline is the whole method described in how Vesper writes a daily brief, and it applies to a hurricane precisely as it applies to an ordinary Tuesday. If a storm is genuinely coming for you, the packing question follows immediately, and packing by the forecast rather than the calendar survives an evacuation order intact.
More of this sits in the Vesper journal, where the weather-literacy cluster works through the numbers people are shown every day and almost never taught to read.
Common Questions
No. The cone only describes where the storm's center is likely to travel, and hazards travel much further than the center does. Tropical-storm-force winds can reach more than 200 miles out from a large system, storm surge is governed by coastline shape rather than track, and inland rain flooding regularly kills people hundreds of miles from the drawn path.If my town is outside the cone, am I out of danger?
In recent Atlantic seasons the day-five circle runs roughly 200 nautical miles in radius, which makes the cone about 460 miles across at its far end. At three days it narrows to about 230 miles across, and at 24 hours to roughly 90. The Hurricane Center republishes the exact radii before each season because they are recalculated annually.How wide does the cone get at five days out?
It does not. The cone is a track product only, and the identical shape would be drawn for a 45 mph tropical storm and for a Category 5 with the same forecast path. For intensity, read the advisory text, the wind speed probability graphics, and the Key Messages product, all of which state expected strength and the confidence behind it.Does the cone tell me how strong the hurricane will be?
Because it is a measure of forecast skill rather than of the storm. The radii are recomputed every year from the Hurricane Center's own official track errors over the previous five seasons, so improving forecasts automatically produce a tighter cone. Average 72-hour track error has dropped by roughly 75% since 1990.Why has the cone been getting narrower over the years?
Start with the watches and warnings for your county, then the Potential Storm Surge Flooding Map if you are near the coast, then the arrival time of tropical-storm-force winds to set your working deadline. The center line is the least informative mark on the graphic and forecasters have asked for years that the public stop treating it as a landfall prediction.What should I look at instead of the thin center line?