A paradigm shift in hurricane risk communication is due. Ian has made that clear.
But this time, we don’t need another fancy scientific product. Nor do we need a new fluctuating number no one understands.
We need actionable words in simple colors that are created by a reliable source and delivered by a trusted face, with the freedom to adjust circumstantially.
Climate change is almost certain to magnify hurricane impacts such as rainfall and storm surge incrementally over the next century. But a more imminent threat to American lives may be our inability to collectively understand, communicate and adapt to the already well-known risks from naturally recurring events in the digital age.
The growing controversy in Florida over Lee County’s response to Hurricane Ian has exposed a widening gap in modern-day risk communication. Entrusting the public — much less officials — to understand probabilistic forecasts in a worst-case scenario is about as effective as “thoughts and prayers” are to preventing another mass shooting.
Major Hurricane Ian hit a vulnerable and densely populated coastline, seemingly with inadequate notice, where the belief that “worst case never happens” was validated by Hurricane Irma just five years prior.
In contrast, 30 years ago, one man single-handedly saved thousands of lives from an extreme event with even less lead time. Bryan Norcross was one of only a few reachable, trustworthy sources of information as Category 5 Hurricane Andrew plowed through South Florida on Aug. 26, 1992.
Andrew killed 15 people directly, which is far less than Ian. But the hurricane warning from the National Hurricane Center (NHC) was issued just 24 hours prior! And there was no such thing as a Storm Surge Warning then.
Hurricane Ian was well-advertised. It was trending on social media a week before landfall — before it was even a named storm, identified as 98L. The first official forecast path from the NHC was nearly perfect. The next 23 forecast cones did what they were designed to do: convey uncertainty on where the center of the storm may track.
Since Andrew, accuracy of the NHC hurricane track forecast has increased by 75 percent. Intensity forecasts have improved by 50 percent. The ever-shrinking cone of uncertainty has been around publicly since 2002, with improvements made in 2017 to focus more on impacts.
But the small black box on storm projection diagrams with the words “Hazardous conditions can occur outside the cone” isn’t cutting it anymore. When was the last time you saw those words in your favorite weather app?
It doesn’t matter how accurate the NHC line becomes, the average American has dozens of more lines to process: The ensembles, their run-to-run trends, the one or two that appease our denial of what could happen. But the stated purpose is to “show uncertainty” — not save lives.
Not all deaths from Ian may have been preventable. Social scientists have been researching evacuation response for years. Every story is unique. Every decision is personal.
But those decisions almost always start with a baseline of information: The cone. It gets updated every six hours. It shrinks every year based on a margin of error, and every mile closer the storm moves.
A monumental and successful effort has been made by the NHC on storm surge messaging in recent years, largely under the leadership of NHC Acting Director Jamie Rhome. Prior to Ian, storm surge related deaths had decreased dramatically since 2017, despite six major hurricanes hitting the Gulf Coast over that period.
But the antithesis to recent forecast advancements seems to be a perfect storm of their own success, the inconsistent methods in how they are communicated, and the irresponsible means by which they are shared.
“In 1992, we didn’t have access to the spaghetti plots at the TV station, so the National Hurricane Center forecast was the only projection we paid attention to,” Norcross wrote in his 30-year anniversary story of Andrew. “In a way, that made things less confusing than they are today, now that anybody can make a forecast or a comment just sitting at home,” he added.
The devastation from Hurricane Ian should not be a surprise. Especially to anyone who understands what the reasonable worst-case scenario was. But that rules out most Americans.
The facts will eventually shed light on what could have been done better, or sooner, to prepare for Hurricane Ian. Especially where impacts outside (or near the edge of) the cone weren’t believed to be significant.
But the reality is that the lesson from one storm may not apply to the next. Especially in today’s rapidly changing economic and political environment.
Craig Setzer, a broadcast meteorologist with 25 years of hurricane experience, recently made this point on Twitter: If we didn’t learn the lessons from Hurricane Charley in 2004 — with an eerily similar path — how can we be assured we will learn from Ian? And what if the opposite happens: A hurricane that is projected to hit Fort Myers wobbles to Tampa.
Replacing the cone outright won’t work because that would give more weight to the dozens of other lines (model paths) freely available. There simply cannot be a vacuum on precision. Nor can there be a delay on credible forecast adjustments.
Recent improvements made on storm surge messaging can certainly be a pillar to build upon. But even those maps can become laden with too much detail, difficult to adjust, or easily misunderstood.
The “key messages” social media concept recently adopted by NHC forecasters has proven valuable on the larger temporal scale, but it can also get easily shoved aside as local impacts become imminent.
Perhaps the most actionable tool few people know about is a dataset the National Weather Service (NWS) produces (with guidance from the NHC) called Hurricane Threats and Impacts (HTI). The HTI graphics include timing, location, magnitude and potential impact per hazard with uncertainty built in. However, little effort has been made to distribute them widely, for whatever reason.
A massive effort has been underway for more than a decade to improve accuracy and lead times of tropical threats. It’s called the Hurricane Forecast Improvement Program (HFIP).
The improvements in modeling and guidance through HFIP have been nothing short of impressive. My fear is that meaningful benefits to the public will bounce around in an echo chamber of bureaucracy, take years to implement and add unnecessary layers of complexity to a suite of already successful products.
In today’s world with machine learning, better modeling and greater detail of a storm’s potential, let us not forget what’s most important: the degree of human exposure along a storm’s path. But this can vary dramatically depending on population density, demographics and infrastructure.
What Hurricane Ian has made clear is that tropical cyclone risks are still being misinterpreted at a level that jeopardizes lives. And this is not likely to be improved by another probabilistic product or redesigned cone.
There may never be a time when one person or one product can be directly linked to saving lives. Yet, I believe deaths linked to a misunderstood forecast can be prevented with a simpler, more human approach to communicating risk.
Jeff Huffman is a digital meteorologist for Baron Weather. He also built and directed the Florida Public Radio Emergency Network from 2011 to 2021. He has more than 18 years of combined broadcast and social media experience as a weather and risk communicator.