Wildfire Traffic Jams Are Deadly. Can Anything Be Done?

You did everything right.

You had your go-bag ready and you knew your evacuation route. You monitored the wildfire as it moved closer and closer to your home, and you kept the volume turned up on your phone so you could heed a “LEAVE NOW” notice if one came. When it finally does, jolting you awake in the middle of the night, you realize that you can smell the smoke inside. When did the fire get so close?

The power is out, so you make your way downstairs using your phone’s flashlight. You have to Google how to manually open the garage door since the electronic clicker doesn’t work (oh, so that’s what the red cord is for). Your heart is thumping, but you’ve made it, you’re in your car; you even remembered to keep it filled to half a tank in preparation. You pull out of your driveway and onto the dirt road that leads out of your rural neighborhood. The night sky ahead of you is a weird neon orange.

You have to hit your brakes when you reach the intersection at the main road. It’s completely backed up with other evacuees, their red taillights stretching ahead through the thickening smoke as far as your eye can see. Some of your neighbors are pulling their boats on trailers; there is an RV up ahead. And you can see the fire burning down the side of the hill now — toward you, toward the gridlocked traffic that isn’t moving.

Leaving your home is only the beginning of a wildfire evacuation. But the next step — the drive to a safe location — is usually given no more attention in preparedness guides than the reminder to “follow the directions of emergency officials.” In the best-case scenarios, where communication is clear and early and residents are prepared, that might be enough. But when communication breaks down, or fires move fast and unpredictably, traffic can reach a dangerous standstill and familiar roads can transform into death traps.

In 2015, some 20 vehicles were overcome by a fire while stuck in a traffic jam on Interstate 15 between Los Angeles and Las Vegas; on the same interstate in Utah five years later, a backup nearly became deadly as a fire burned up to the road’s shoulder and panicked travelers abandoned their cars. Fire evacuations in New South Wales, Australia, in 2020 resulted in a 10-hour backup, and Canada’s Highway 3 had bumper-to-bumper traffic earlier this month because it was the only road out of imperiled Yellowknife. In 2020, some 200 people had to be evacuated by helicopter from California’s Sierra National Forest after a fire cut off their only exit route.

And when people die in wildfires, they are often found in their vehicles. In Portugal, 47 of the 64 people killed during a 2017 forest fire were in their cars, trying to escape. At least 10 people were found dead in or near their cars after the 2018 Camp fire, the deadliest blaze in California’s history. And in Lahaina, Hawaii, this month, in what the Los Angeles Times has called “surely … the deadliest traffic jam in U.S. history,” the lack of advanced warning combined with inexplicably blocked roads led an untold number of people to perish in their cars while trying to evacuate, including a 7-year-old boy who was fleeing with his family; a man who used his last moments attempting to shield a beloved golden retriever in his hatchback; and a couple who were reportedly found in each other’s arms.

In a best-case scenario, emergency managers are able to phase evacuations in such a way that the roads don’t get backed up and residents have plenty of time to make it to safety. But wildfire is anything but predictable, and officials who call for an evacuation too soon can risk skeptical residents deciding to take a “wait and see” approach, where they only get in their car once things start to look dicey. In one 2017 study, only a quarter of people in wildfire-prone neighborhoods actually left as soon as they received an evacuation notice (other studies have found higher levels of compliance). This is the worst nightmare from an emergency management standpoint, since “evacuating at the last minute is probably the most dangerous thing you can do,” Sarah McCaffrey, one of the 2017 study’s authors, told The New Yorker.

Further complicating matters is the fact that many wildfire-prone areas are isolated or rural regions with a limited number of egresses to work with. One 2019 investigation found that in California alone, 350,000 people live in areas “that have both the highest wildfire risk designation, and either the same number or fewer exit routes per person as Paradise” — the site of the 2018 Camp fire, where backups on roads prevented many from escaping.

Evacuation traffic also doesn’t behave like the rush hour traffic we’re more familiar with. It’s “a peak of a peak,” with the congestion caused by “the sheer amount of people trying to leave and load onto the roadway at the same time in the same direction,” Stephen Wong, a wildfire evacuation researcher and an assistant professor of transportation engineering at the University of Alberta, told me. Burnovers and hazards like downed powerlines or trees can further reduce exit options, funneling all evacuees onto the same low-capacity roads. Worse, once that congestion starts to form, “you actually reduce the number of vehicles being able to go through that section,” Wong added. “So you go from 2,000 vehicles per hour [per lane], and it drops to, like, 500 vehicles per hour.”

Get one great climate story in your inbox every day:

Households will also frequently evacuate with multiple cars — rather than leave a valuable asset behind to burn — and tow trailers, boats, and RVs. As a result, the average vehicle length increases by 3% during wildfire evacuations, one recent study that looked at the 2019 Kincade fire in California found — leading, of course, to even worse congestion. (Agonizingly, Wong’s research further uncovered that over half of evacuating households “had at least two or more spare seats available”). The Kincade study also discovered that drivers significantly slow down during wildfire evacuations — contrary to the common misconception of careening, panicked escapees — likely due to a combination of factors such as lowered visibility and more cautious driving.

Because “most [evacuation] research focuses on hurricanes and then tornadoes,” Salman Ahmad, a traffic engineer at the civil engineering firm Fleis & VandenBrink, told me, “traffic simulations — how traffic moves during a wildfire — are still lacking.” When emergency planners use computer models to calculate minimum evacuation times for their jurisdictions, for example, their assumptions can be deadly. “If you plan for an allocation considering normal traffic as a benchmark, you’re basically not making the right assumption because you need to put in that extra safety margin” to account for “the fact that people slow down,” Enrico Ronchi, a fire researcher at Lund University in Sweden and the author of the Kincade study, told me.

Wong agreed, stressing that the number of variables fire managers need to juggle is dizzying. “Evacuations are really complex events that involve human behavior, risk perceptions, communication, emergency management, operations, the transportation system itself, psychology, the built environment, and biophysical fire,” Wong said. “So we have a long way to go for evidence-based and sufficient planning that can actually operationalize and prepare communities for these types of events.”

And that’s the scary thing: A person or a community might do everything right and still be at grave risk because of all the unknowns. Evacuation alerts might not get sent or arrive too late; exit routes might become unexpectedly blocked; fires might leapfrog, via flying embers, to create new spot fires that cut off egresses. Paradise, California, famously had a phased evacuation plan in place and had even run community wildfire drills, but even the best-laid plans can unravel.

Tom Cova, a geography professor at the University of Utah who has been studying wildfire evacuations for 30 years, told me that “too many communities may be planning for the roads to be open, the wireless emergency alert systems to work, there not to be tons of kids at home that day — you can just go down the list of things that [could go] wrong and think, What’s the backup plan?” The uncomfortable truth is that we need plans B, C, and D for when evacuations fail. Because they will fail.

Take Lahaina, where a closed bypass road concentrated outbound traffic onto a single, jam-packed street. When people started to panic and abandon their cars, it ultimately further obstructed the road for everyone behind them. “It’s like a chain reaction, where each car is seeing the [people in the] car in front of them run,” Cova said. “And then you look behind you, you can’t back up. If you look to the sides, you’re stuck. And then you say, ‘We’re going into the ocean, too.’”

That improvisation ultimately saved some lives. But “it’s hard for emergency managers to order this kind of thing because what if people drowned?” Cova went on. “So you’re trading one risk for another risk.”

But the need for creative improvisation is also a conclusion that’s been reached by the National Institute of Standards and Technology (NIST), the government agency tasked with issuing guidelines and regulations for engineers and emergency responders. In new guidance released last week, NIST used the Camp fire as its case study and found “evacuation is not a universal solution,” explaining there are times when “it may be better for residents to shelter in their community at a designated safety zone” rather than attempt to drive out of town.

This is a somewhat radical position for a U.S. agency since evacuations have long been the foundation of American wildfire preparations. But the thinking now appears to be turning toward asking “what shelters do we have?” if and when a worst-case scenario arises, as Cova further explained to me. “Temporary refuge areas, high schools, churches, large parking lots, large sports fields, golf courses, swimming pools — I wouldn’t recommend using any of these things, and I wouldn’t recommend people being told to use them,” he said, “but [people] have to know what to do when they can’t get out.”

In the case of Paradise, for example, NIST reports that there were 31 such “temporary refuge areas” that ultimately saved 1,200 lives during the fire, including 14 parking lots, seven roadways, six structures, and a handful of defensible natural areas, like a pre-established wildfire assembly area in a meadow that had already burned and ended up serving as a refuge for as many as 85 people. Once established, these concentrated refuge areas can be defended by firefighters, as was the case for 150 people who memorably hunkered down to wait out the blaze in a strip mall parking lot. It’s far from a best-case scenario, but that’s still 150 people who would’ve otherwise been stuck in potentially deadly traffic jams trying to get out of town.

Temporary refuges are unplanned areas of last resort, but establishing a larger safety zone network and preemptively hardening gathering places like schools and community centers could also potentially reduce exposure on roads by shortening the distance evacuees need to travel to get to lower-hazard areas. So-called WUI fire shelters — essentially, personal fire bunkers that NIST warns against because they aren’t standardized in the U.S. but are popular in Australia — could also be explored. “That’s the direction we’re heading in with wildfire communities,” Cova told me grimly, “because we don’t seem to be able to stop the development in these areas. That means we’re forcing people into a corner where shelter is their only backup plan.”

Maybe this is difficult for you to imagine: Your community is different; a wildfire couldn’t happen here. You’d evacuate as soon as you got the notice; there’s no way you’d get stuck. You’re a good driver; you could get out without help. But as Lahaina and other “unprecedented” fires show, it’s the limits of our lived experiences that we’re up against now.

“We should think about possible scenarios that we have not seen before in our communities,” Ronchi, the Swedish fire researcher, said. “I understand that it’s a bit of a challenge for everyone because often you have to invest money for something that you have not experienced directly. But we are [living] in scenarios now in which we cannot anchor ourselves on our past experiences only.”

Read more about wildfires:

Most Wildfires Aren't Forest Fires