Check out the brand-new hurricane ‘cone of uncertainty’ graphics arriving this season » Yale Climate Connections

by Bob Henson, Yale Climate Connections
May 29, 2026

It might have seemed exotic when it first appeared, but the forecast “cone of uncertainty” used by the NOAA/NWS National Hurricane Center (NHC) is now a familiar part of tropical cyclone readiness in U.S. states and territories. For 2026, NHC has made a couple of key tweaks to its standard cone product. It’s also testing an expanded version of the cone – one made feasible by a new way of understanding how and where forecast errors arise.

Since its debut in 2002, the cone has become what a University of Miami writer called “arguably [the center’s] most iconic graphic,” a mainstay of TV coverage and weather apps. Prior to the cone, hurricane maps simply showed a line depicting the official multi-day forecast for the storm center, as issued every six hours by NHC. Experts urged the public not to “focus on the skinny line,” keeping in mind that a hurricane’s path can easily deviate from the forecast track and that impacts will typically extend far beyond that center.

When you see a cone graphic, that ‘skinny line’ may or may not appear (NHC provides both versions), but the cone itself has gone a long way to fix the skinny-line problem.

However, just as a hurricane’s impacts do not just lie along a narrow line, a hurricane’s damage doesn’t stop when it comes ashore. Some of the worst U.S. hurricane disasters in recent years have occurred well inland, including billions of dollars in wind-driven destruction across Georgia in 2018’s Michael, and the catastrophic, deadly flooding from 2024’s Helene, which killed more than 100 people in and around western North Carolina.

Up through last year, NHC’s cone graphics only showed watches and warnings along the coastline. Starting this year, the full extent of inland watches and warnings will be portrayed. In the example shown in Fig. 1 below, the revised graphics make it crystal clear that the hurricane warning for 2024’s Milton extended almost completely across the entire Florida Peninsula, including the Orlando area.

Another improvement shown in Fig. 1 is the addition of a crosshatched area to denote locations that are under both a hurricane watch and a tropical storm warning. It’s an important way to show that being in a tropical storm warning doesn’t mean you are necessarily off the hook for potential hurricane-level impacts.

When bad stuff happens outside the cone

Maybe because it’s so visually intuitive, the cone can deceive. The most common way to misinterpret the cone is to assume that it includes all possible hurricane tracks and that all serious hurricane impacts will fall inside the cone. It’s a problem that experts across disciplines have dubbed the “containment effect.”

The misunderstanding has led to some painful lessons. One of the most dramatic was in 2022, when Hurricane Ian veered toward the right-hand edge of the cone. Ian made a high-end Category 4 landfall near Fort Myers less than 36 hours after the official skinny-line track forecast had projected a strike near Tampa Bay. Because Ian was such a large and potent hurricane, its storm surge extended well to the right of the cone, delivering major flooding as far south as Naples. Ian took at least 161 lives and inflicted $112 billion in damage (USD 2022).

Multiple lines of social science research confirm that many laypeople make the mistake of assuming hurricanes simply don’t stray outside the cone. One survey of more than 2,800 Floridians led by Scotney Evans (University of Miami) and published in 2022 by the Bulletin of the American Meteorological Society found that nearly half of respondents assumed that the cone showed all of the potential tracks for a hurricane. 

“Our analysis suggests that many residents have difficulty interpreting several aspects, suggesting a rethink on how to graphically communicate aspects such as uncertainty; the size of the storm; areas of likely damage; watches and warnings; and wind intensity categories,” Evans and colleagues wrote. In some cases, better-educated respondents were actually more likely to misinterpret certain aspects of the cone. 

READ: Building a better hurricane cone of uncertainty

The issue is especially acute because of the cone’s sheer popularity. “The cone is one of the most, if not the most, commonly shared hurricane visuals,” said Robert Prestley (NSF National Center for Atmospheric Research).

A 2020 overview of hurricane risk communication in the journal Weather, Climate, and Society, led by Barbara Millett (University of Miami), noted that during the five days as Hurricane Irma approached Miami in 2017, the cone map accounted for more than 70% of independent pageviews at the NHC website. In a 2023 study published in the same journal, Prestley and NCAR’s Rebecca Morss found that cone graphics were retweeted more often than watch/warning graphics on Twitter.

With all this in the mix, “it was taking people by surprise when the hurricane would move outside the cone,” said Robbie Berg, warning coordination meteorologist at NHC.

In fact, the cone’s width is calculated for each storm based on the previous five years of track locations in the official NHC forecasts, rather than on how well or poorly behaved a particular storm might be. Based on average track errors from those preceding five years, the cone width is calibrated to include about two-thirds (67 percent) of all potential storm positions. This means that by design, one would expect the center of a hurricane to stray outside the cone margins about one-third of the time.

Making the cone substantially wider might seem like an obvious fix, but this approach carries its own hazards. Evacuations are based largely on storm surge risk rather than the cone itself, and storm surge warnings can extend well beyond the cone. However, a greatly expanded cone could mean a larger number of people finding themselves in a cone year after year, perhaps without significant impacts each time.

“Research shows the public perceives the cone as an area of concern – an indication to continue monitoring the forecast,” said Gina Eosco, director of NOAA’s Weather Program Office and a pioneering researcher on how people interpret the cone and other forecast products.

Since 2007, forecasters at NHC have used the two-thirds index for the cone width as a working compromise between overly narrow and overly broad. But a new way of analyzing errors from past years has paved the way to an experimental cone that would alert more people without including all that much more territory.

The key, according to Berg, was to decompose the total track error. A track forecast can make mistakes that are either “cross-track” (erring in the direction of motion) or “along-track”  (moving the system too quickly or too slowly). Standard practice is to draw the cone’s edges along each side of a series of circles straddling the forecast track, with the radius of each circle set to include 67 percent of potential positions and the circles growing larger with each forecast day.

As it turns out, timing mistakes (along-track) tend to produce bigger errors than do directional mistakes (cross-track), as shown in Fig. 2 below. Using circles to pool all of these errors obscures the difference between the two types, thus making the cone wider and less elongated than it ought to be.

When NHC examined the two types of error, they discovered that only a minor widening and lengthening of the cone could enclose 90 percent of possible positions, as opposed to the current 67 percent. This 90-percent cone is being used in experimental mode for the first time this season (see Fig. 3 below), alongside the traditional 67-percent version. The center will solicit comments and feedback before any move to finalize and adopt the experimental version. It’s been well received at conferences, according to Berg.

“Especially as we get out toward day 4 or 5, most of the error is in the along-track part of the storm,” said Berg. “Going to 90% doesn’t increase the width of the cone much. It’s more that you’re increasing the length.”

Another benefit of the 90% cone: some other NHC products already use the same threshold. For example, peak storm surge forecasts depict the maximum inundation one would expect from a given tropical cyclone approaching a given stretch of coast. These forecasts are calibrated so that a value higher than the maximum shown would be expected only 10% of the time. “So we’re trending in this direction: reasonable worst case, trying to capture as much of the risk as possible,” Berg said.

“The changes to the cone show remarkable scientific advancement,” said Eosco.

Meanwhile, the traditional version of the forecast cone will slim down a bit this year. Because of reduced error in the forecasts for 2021–2025 compared to 2020–2024, the two-thirds probability circles for 2026 will be 4 to 8 percent smaller on average in the Atlantic and 3 to 8 percent smaller in the Northeast Pacific. Such incremental improvements over the past couple of decades have led to striking reductions in cone size (see embedded post from 2025 below).

A timely year for new storm surge products in Hawaii

With El Niño boosting the odds that tropical cyclones will affect Hawaii this season, it’s fortuitous that NHC is now launching the same type of storm surge products for the main Hawaiian Islands that are regularly issued for the U.S. East and Gulf Coasts, Puerto Rico, and the U.S. Virgin Islands. These will include the peak storm surge forecasts noted above.

Behind the storm surge forecasts are exhaustive calculations carried out across more than 20 years of work using the P-Surge (probabilistic storm surge) model. The resulting datasets show the potential inundations at coastal points separated by 2.5 kilometers (about 1.6 miles) based on winds and atmospheric pressures from as many as 1,000 simulated tropical cyclones. As this work continues, NHC is looking to expand storm surge forecasts more broadly through the Caribbean in the coming years.

As stressed by Eosco: “Regardless of the cone’s shape or size, monitoring the forecast is a critical first step in assessing personal risk and empowering personal decision-making.”

Jeff Masters contributed to this post.

This <a target=”_blank” href=”https://yaleclimateconnections.org/2026/05/check-out-the-brand-new-hurricane-cone-of-uncertainty-graphics-arriving-this-season/”>article</a> first appeared on <a target=”_blank” href=”https://yaleclimateconnections.org”>Yale Climate Connections</a> and is republished here under a <a target=”_blank” href=”https://creativecommons.org/licenses/by-nc-nd/4.0/”>Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License</a>.<img src=”https://i0.wp.com/yaleclimateconnections.org/wp-content/uploads/2020/10/ycc-favicon.png?resize=100%2C100&amp;ssl=1″ style=”width:1em;height:1em;margin-left:10px;”>

<img id=”republication-tracker-tool-source” src=”https://yaleclimateconnections.org/?republication-pixel=true&post=139120&amp;ga4=1401ERFF5Q” style=”width:1px;height:1px;”><script> PARSELY = { autotrack: false, onload: function() { PARSELY.beacon.trackPageView({ url: “https://yaleclimateconnections.org/2026/05/check-out-the-brand-new-hurricane-cone-of-uncertainty-graphics-arriving-this-season/”, urlref: window.location.href }); } } </script> <script id=”parsely-cfg” src=”//cdn.parsely.com/keys/yaleclimateconnections.org/p.js”></script>