Hurricanes: Structure, the Cone & Model Tracks

In short: a hurricane is a heat engine fed by warm ocean water, with its strongest winds in the eyewall ring around a calm eye. The National Hurricane Center’s cone shows where the center will probably go, not how big the storm is or where the damage stops. Model tracks show the range of possibilities the forecasters weigh. Read them for the spread, not for one line.

The parts of a hurricane

• The eye is the calm center. NOAA’s JetStream notes an eye usually forms once sustained winds pass about 74 mph. People in an eye can think the storm has ended, when the other side of the eyewall, with winds from the opposite direction, is minutes away.
• The eyewall is the ring of tall thunderstorms around the eye, where the strongest winds and the heaviest rain are. When a new eyewall forms outside the old one, a process called an eyewall replacement cycle, the storm’s peak winds often dip and then the wind field grows wider.
• Rainbands spiral in toward the center. They bring bursts of heavy rain and wind, and tornadoes in landfalling storms, most often in the bands on the right front side.

The right side matters. In the Northern Hemisphere, the storm’s forward motion adds to the spinning winds on the right side and subtracts on the left, relative to its track. The right side also pushes water onshore, which is why the highest storm surge is usually just right of where the center comes ashore.

In MesoRadar: on the Satellite tab, Infrared (free on GOES-East, short loop) shows cloud-top temperature: a ring of very cold tops around a clearing, warming eye is a storm getting organized, and an eye that fills in or a ring that warms is often a pause or weakening. Full disk · Infrared (Pro) reaches storms far out in the Atlantic or Pacific. Local radar shows the eyewall and bands once a storm is within range of a NEXRAD radar; far from the radar the beam passes over the lower part of the storm. Visible, Infrared & Water Vapor.

What the category tells you, and what it leaves out

The Saffir-Simpson Hurricane Wind Scale rates a hurricane by its peak 1-minute sustained wind:

Category 3 and up are major hurricanes. Below 74 mph the storm is a tropical storm (39 to 73 mph) or a tropical depression.

The scale is about wind and only wind. It does not include:

• Storm surge, the rise of seawater pushed ashore. Over 1963 to 2012 it caused about half of direct U.S. tropical cyclone deaths.
• Rain flooding, which can be extreme from slow or weak storms, far inland. In the decade 2013 to 2022 it caused more than half of direct U.S. tropical cyclone deaths, more than surge; Hurricane Harvey’s Texas flooding alone accounts for about a quarter of those.
• Size. A large Category 1 can spread damaging wind and surge over a far wider area than a small Category 4.

The forecast cone

The cone is the most-seen and most-misread graphic in weather. NHC’s definition, in plain words:

• It shows the likely path of the center. Not the storm’s size, not the area of impacts.
• It is built from past errors. Along the forecast track, NHC draws circles at each forecast time, sized so that two thirds of its official track errors from the previous five years fell inside them. For 2026, the Atlantic circle grows from 25 nautical miles at 12 hours to 200 at five days. The cone is the area those circles sweep.
• So the center leaves the cone about a third of the time. That is how the cone is designed.
• Impacts reach well outside it. Hurricane-force wind, surge and flooding rain often extend far beyond the cone’s edge, especially for large storms.
• Since 2026, NHC’s cone graphic also shows inland watches and warnings. Tropical storm and hurricane warnings can reach well inland, and the graphic now draws them, which is one more reminder that the cone is a track product and the warnings are the impact product.

No name yet does not mean no threat. When a disturbance could bring tropical storm or hurricane conditions to land within about 72 hours, NHC starts advisories, a cone and warnings before it becomes a tropical cyclone, under the name Potential Tropical Cyclone. Treat its warnings exactly like a named storm’s.

For impacts, read NHC’s watches and warnings, its storm surge products and its wind probabilities, not the cone.

Watches and warnings. A Hurricane Watch means hurricane conditions are possible, issued about 48 hours before tropical-storm-force winds are expected to begin; a Hurricane Warning means they are expected, about 36 hours ahead, because preparations get hard once the wind starts. Storm Surge Watches and Warnings follow the same timing for life-threatening flooding from the sea, and they can cover places outside the hurricane warning. Near landfall, local NWS offices issue Extreme Wind Warnings for eyewall winds of 115 mph or more, and Tornado Warnings for tornadoes in the rainbands. Evacuation orders come from local officials: follow them.

Reading the wind chances. Each percentage is the chance that winds of at least that strength reach that spot at some time in the next five days, allowing for errors in track, strength and size. Far from landfall the numbers look small: 10 to 20% for hurricane-force winds at one town days ahead is a strong signal, not a reason to relax. Watch whether they rise advisory by advisory.

In MesoRadar: the Hurricane tab draws each active storm with NHC's official cone and track, the coastal watch and warning line, and, when NHC issues it for a U.S. threat, its storm surge map in feet of water above ground (all free). For your location or a saved place, a card says whether it is in the cone, the storm's closest pass and whether damaging winds are possible there. Chance of winds, under On the map, shows NHC's probability of winds of at least 39, 58 or 74 mph over the next five days, one at a time (Pro).

Model tracks, the spaghetti plot

A spaghetti plot draws the forecast track of many computer models at once. NHC forecasters weigh all of them, and more, when they make the official forecast.

How to read it:

• Read the spread, not a line. Tight bunching means the models agree and the track is more certain. A wide fan means real uncertainty. The cone is the same size whatever the spread on a given day, so the spaghetti adds information the cone does not.
• Consensus usually wins, and the official forecast usually matches it. NHC’s verification shows that an average of several good models beats almost any single model over a season, and that the official forecast does about as well as the best consensus while changing less from one advisory to the next. A single model, even a good one, flips more.
• Not every line is equal. Some public spaghetti maps mix in simple statistical and trajectory aids that are poor guides past a day or two. The models that matter are the global models (GFS, ECMWF, UK Met) and NOAA’s hurricane models (HAFS-A and HAFS-B).
• Mind the run times. Some tracks are a model’s newest run; others, such as the European on many maps, are the previous run shifted to start at the storm’s current position, because the newest one was not ready in time. Two lines can disagree partly because they come from different runs.
• Watch the trend. Lines that shift run after run in one direction are telling you more than any single run.
• One wild line is not a forecast. A single model taking the storm somewhere new is a reason to watch, not a reason to act.

In MesoRadar: Model tracks, under On the map in the Hurricane tab, labeled "Consensus and the models" (Pro). The map notes that model guidance is not an official forecast. Tap a track to see its name and the run it came from, with Only this and Hide; press and hold a line to hide it straight away.

Intensity is the harder forecast

Track forecasts have improved steadily for decades. Intensity, how strong the storm gets, remains harder. The case forecasters worry about most is rapid intensification: NHC defines it as an increase in maximum sustained winds of at least 30 knots (about 35 mph) in 24 hours. Warm, deep ocean water, moist air and weak wind shear allow it, and it can turn a tropical storm into a major hurricane in a day or two, sometimes right before landfall.

The practical lesson: prepare for a category or so stronger than the current forecast, and never base a decision on the current category alone.