THE GOLDEN MILE of Canada’s Okanagan Valley is an imprint of the last ice age, its rocky bottom and steep sides carved by glaciers. Spring meltwater runs down its ravines, carrying with it loads of sand and gravel from the peaks. For the last 7,000 years, that alluvial earth has spilled out across the valley in a deep, fertile fan. The mineral soil and the balmy summers of the high desert make this swath of southeastern British Columbia ideal for growing grapes. Dozens of wineries dot the Mile, filling half of its 1,500 acres with vines. Fairview Cellars, one of the spot’s earliest vineyards, has grown grapes and bottled wine where the valley floor meets the mountain walls since 1997. In August of 2015, those peaks were burning.

Spring that year had come early, and growers predicted that the warmth would produce a memorable vintage—but that heat dried the earth, and set the stage for a particularly bad wildfire year. On the night of August 14, a lightning storm started a blaze at the southern edge of the valley. By morning, high winds were driving the flames north, through the Golden Mile’s dry sage and grass. 

On September 10, the valley was under an evacuation order, and Fairview’s Bill Eggert was only partly done with harvesting for the year. The vineyard owner, who retired earlier this year, stuck around. “There’s a golf course between myself and the fire,” he explains, and that highly watered turf wasn’t likely to feed the flames in a hurry. In the end, they stopped within a half mile. “It was just a forest fire,” he says when asked if he remembers anything from the day. “Lots of helicopters, lots of planes bombing, lots of bureaucrats running around justifying their existence. You just carry on, right?” 

When the fire was over, Eggert harvested the grapes, pressed them, and set them up in barrels. That winter, when he cracked open the containers for a first taste, there was an odor he didn’t recognize.

“I never really smelled it in my wines before,” he says. He thought a wayward colony of wild yeast might be to blame—a genus called Brettanomyces is known to create savory funks that range from smoke to rot. But he opened another and tasted the same thing. Same with the next. “I asked quite a few different winemakers, and none of them could identify what it was,” he says. Then he realized what it was: He was tasting the fire.

Grapes thrive in dry summers. So do fires. Over the last 30 years, persistent drought caused by climate change has turned the West Coast of North America into a tinderbox. Between 2000 and 2018, fires were twice as common and burned four times as much land in the Western US than in the previous two decades. Sometimes, the damage to grapes is direct: In 2017, fires around California’s Napa and Sonoma destroyed at least 25 wineries; 2020’s Glass Fire wrecked dozens more. 

Wineries that escape direct damage, like Eggert’s, instead have to reckon with the flavor of smoke—or, as the industry sometimes calls it, smoke taint. “Like licking an ashtray,” the common refrain goes, so many choose to avoid using the fruit entirely. In 2018, for example, a California winery backed out of $4 million in grape purchases, leaving growers near fires in Oregon’s Rogue Valley with 2,000 tons of fruit. And last year, close to 5.5 million acres burned across California, eastern Oregon, and British Columbia, which triggered the cancellation of vintages up and down the West Coast.

Plenty of winemakers are figuring out how to avoid smoke in their vintages, but Eggert is among a faction of vintners who think wine drinkers can learn to accept, and maybe even love, these tasting notes the same way whiskey lovers embrace a peaty scotch. “I consider it to be part of the terroir now,” he explains.

Adapting to a changing planet in this way presents a two-part problem. First, chemists and horticulturalists are analyzing grapes to understand how smoke in the air becomes a taste on the palate. Armed with that information, winemakers then must puzzle out if they can tame the flavor—or embrace it and hope the consumers catch on.

On a scale from zero to 100-percent-ashtray, Eggert places his 2015 harvest at about a 30—think Dewar’s. So he bottled and sold it as Fumé Franc. It was a hit. “It’s been sold out now for four years, but there’s still people that show up and hope I have a bottle or two kicking around somewhere,” he says.

Amanda Ringstad

A LITTLE BIT of smoke in a wine isn’t really a novelty. Many chardonnays are aged in lightly charred barrels, and winemakers talk about bacon, hoisin, and mesquite as tasting notes. Plenty of spirits go even further: Peaty scotches carry the boggy essence of the flames used to dry barley, and mezcal is born out of slow-roasted agave. Tom Collins, a professor of winemaking at Washington State University and the president of the American Society for Enology and Viticulture, has been digging into smoke’s effects on wine since 2008, when fire tainted grapes at the Napa Valley vineyard where he worked. He says two crucial things distinguish wildfire smoke: Drinkers don’t expect it, and winemakers can’t control it.

When a wildfire burns, it doesn’t fully vaporize its fuel. Instead, the heat energy shakes up raw plant material (sap, grass, wood, leaves) and turns it into thousands of chemicals called volatile phenols. These have a tendency to evaporate, making them easy to smell. They’re what give smoke—a cocktail of ash, carbon gasses, and assorted other organic detritus—its “smokiness.”

When the phenols settle over vines, some of them wind up soaking into the grapes. Phenols are toxic to plants, so the fruit begins to seal the excitable chemicals up with sugars, which disguises their smell. “Unless it’s so severely smoke-tainted that the grape is actually reeking of smoke when it’s fresh, you’re not going to taste smoke at harvest,” says Wesley Zandberg, a wine chemist at the University of British Columbia Okanagan.

It’s the fermentation that eventually releases the smoke back into the world. The main engine of that process is cultured yeast. Over the course of months, yeast eats sugar and dumps out ethanol; in fact, the microorganisms are so hungry that they’ll munch the sugar coating off of a phenol, releasing it naked into the wine. That’s the smell that Eggert noticed when he first opened his barrels in 2015. The longer a vintage sits, either in a barrel or bottle, the more phenols come unwrapped.

That time-release reaction helps explain why winemakers hesitate to work with smoked grapes. “This is a really, really big risk if you make a $60 bottle of wine that you expect people are going to buy and put in their cellar and pull out when they really want to impress their guests,” Zandberg says. “If it gets pulled out after five years and it’s just rotten, you’ve completely destroyed your brand permanently. And that is worst-case scenario.”

Some of those challenges would be solved by just being able to predict the effects of smoke. Horticulturalists in Australia have been working on this since 2003, when a drought-fueled bushfire season destroyed tens of millions of dollars’ worth of grapes. And agricultural schools up and down the dry interior valleys of California, Oregon, Washington, and British Columbia have been dialing in on the variables that might impact a vintage’s smokiness for more than a decade.

Variation, they’ve found, can start with the fire itself. The Australian Wine Research Institute has gathered up published findings into a series of fact sheets that many winemakers turn to when first trying to understand what’s happened to their grapes. That research has sketched the outlines of how smoke works: Late-season grapes that have already reddened appear more likely to pick up off flavors. Smoke from far away loses its phenol bouquet, even if it’s thick enough to turn the sky orange—and so soot that blows out over the ocean before settling in interior valleys is going to have less impact than smoke from the next valley over. “The age of the smoke, the distance it travels, whether it’s attached to particulate or water droplets—all of this stuff has a lot of value,” says Zandberg. 

“It’s actually been really rare for people to smell the wine and be like, ‘Oh, that is clearly tainted.’”

—Aron Weinkauf, Spottswoode Winery

The Aussie fact sheets have created a palate of specific chemicals (cresols, syringols, guaiacols) that commercial labs can test for, but they’re also full of gaps. They don’t pinpoint, for instance, how close a fire needs to be to impact grapes, what kinds of fruits are at risk, and whether anything can be done to treat specific wines—at least not with any degree of certainty. More questions became apparent as researchers studied North American fires: Differing forests, like North America’s conifer-filled ones, produce different flavors than Australian eucalyptus. “If you smoke fish or smoke meat, there are all kinds of ideas about what’s the best kind of wood to use,” says Collins. “Lots of people spend a lot of time online debating that.”

Collins started running experiments to connect discrete plants to distinct smoke effects in 2016. He set up huge tents across rows of grape vines and bathed them with the exhaust from a wood-fired smoker. So far, he can say that, yes, a grape hit by smoke from western red cedar ends up with a different chemical bouquet than sagebrush. 

He’s also trying to hone in on the way phenols blossom as they sit inside bottles. In one experiment, Collins separated the tiny chemicals with a microscopic filter—“stripped them out” with carbon—until he could no longer measure the free-floating smoke compounds in the wine. Then he and his lab rebottled 10 gallons of each. “We could watch month over month, and see the [phenol] levels coming back,” he says. In some cases, there was so much smoke stored in the wine’s sugar that, after a few months, the wine was back where it had started before filtering.

Yet, even if people like Zandberg and Collins can nail down the particular chemistry of a smoky vintage, there’s still no guarantee that everyone will taste the same thing. “We really, really, still have an early understanding of how exactly we can connect an objective chemical measurement with a subjective sensory one,” Zandberg says. Humans—even highly trained ones—can reliably recognize only about 10 flavor notes at a time, while there could be 25,000 chemicals in a glass of red wine. And those a person does detect can accentuate or mask one another. In its pure state, guaiacol might be smoky, or even medicinal, but in smaller concentrations it’s part of what makes a tomato skin spicy; in combination with pungent ethanol, bitter tannins, and fruity yeast esters, it could be transformed even more.

As wildfire becomes a part of the winemaking reality, knowing the molecular makeup of a particular bottle is only half the puzzle. What even the best flavor chemistry can’t can’t tell you is  who—if anyone—will enjoy it. 

Amanda Ringstad

ON A SUNDAY NIGHT this summer, I sat down with my partner and two friends to taste a few bottles of smoke-affected wine: a riesling grown in 2017 in the Columbia River Gorge, on the Washington-Oregon border, and a 2020 pinot noir and cabernet franc from Oregon’s Willamette Valley. The white, which had been affected by the massive Eagle Creek Fire, wore its smoky origin on its bottle. It was branded as a “rauchwein,” a nod to toasted German beers called rauchbiers. The two reds were made from less-smoky grapes that had still been grown during one of Willamette’s first brushes with megafires, including the 200,000-acre blaze known as Lionshead. 

None of us—a former cheese-seller, a food journalist, and two enthusiasts—can claim any expertise in wine, but naming every complexity of the bottles wasn’t the point. We wanted to know if we liked what we were drinking.

My three tasting partners agreed that the pinot was overly metallic, with lingering, sharp tannins. ”Rusty pipe,” one said. I found a mushroomy umami note enjoyable, and I didn’t taste even a hint of metal. Two tasters liked the way the riesling balanced green apple and a bitter finish, while I found it a little cloying. We agreed that the cabernet franc was the most surprising of the three, with a musty-sweet smell that my partner read as smoky and others thought was more similar to the funk of kombucha or apple cider vinegar.

What we didn’t taste was smoke, at least not in any way we would recognize in a drink like mezcal or scotch. I got a hint of rubber in the pinot. Was that it? I wondered? The clearest shared quality was an aftertaste that, depending on the sip, leaned bitter or acidic.

It turns out that people who work with wine for a living have had the same experience. “It’s actually been really rare for people to smell the wine and be like, ‘Oh, that is clearly tainted,’” explains Aron Weinkauf, manager and winemaker at Napa’s Spottswoode Winery, which was blanketed in smoke for nearly two months in 2020. “Oftentimes they’ll taste the wine and find that maybe the tannins are a little more dry, or there’s a little bit more of an acrid finish.” That could be a symptom of smoke—but it doesn’t have to be. Picking grapes a little early or late can produce the same effect.

And while the verdict on the wines was overall negative (I was the outlier), our tasting group agreed that with an enthusiastic guide telling us the story of each bottle, we probably could have been swayed. If anything, I want to try a smokier bottle.

Beyond our small taste test, almost everyone agrees that certain smoke notes taste terrible—specifically the ones that can leave a bottle bitter and acrid, more like sucking on coffee grounds than tasting bacon—and that winemakers may do well to avoid them. After the 2020 fires, Weinkauf sent his wines off for chemical analysis at nearly every available commercial testing lab. Armed with spreadsheets populated with raw chemical concentrations, his team tried to figure out whether any given bit of data corresponded to what they were tasting. They came up with a “line in the sand,” Weinkauf says, a rough rule of thumb that let them discard some grapes for being too chemically smoky and work with others. Even so, there were outliers: wines that were smoky on paper but tasted fine, and others well within normal ranges that were off. “We went back through and tested 10 years of historic wines that we knew were untainted. And these compounds exist in wine all the time,” he says. 

Some winemakers, like Doug Tunnell, of Brickhouse Wines in Oregon, have decided that those notes don’t have a place in his lineup. “With our customers, it’s not something that they’re going to embrace.” After a fire in late September 2020, he discarded some batches of heavily affected grapes  and blended the unaffected ones with leftover barrels from 2019. “We take a hit, six bucks a bottle, or so,” he says. “But our feeling was, ‘We need to make something out of this vintage.’”

Not everyone can afford to ditch some, or all, of a harvest, though. Fairview Cellars’ Eggert went ahead with Fumé Franc—as a pragmatic farmer rather than a romantic vintner. He makes just 5,000 cases a year, which puts Fairview at the small end of the “small winery” category—bigger than around 80 percent of US outfits, but tiny compared to the 71 biggies that  produce more than 500,000 cases a year. Larger wineries value predictability, so they might treat wine with filters or carbon to remove phenols. “A lot of places would rather strip it out and not talk about it,” says Rhys Pender, a master of wine and winemaker in British Columbia. But, he says, that often removes other sparks from the wine. “These small-production wineries don’t necessarily want to, and they don’t have to.”

There’s a parallel to the growing popularity of natural winemaking, a philosophy that avoids treating wine and highlights idiosyncrasies. “Natural wine is all about an open-mindedness among consumers,” Pender says. “You’ve got this new market that’s more interested in story, focused on where their food comes from, and what does that taste like?” In 2009, for instance, Eggert made a vintage from grapes that had been split open by a hailstorm and allowed to dry on the vine. He marketed it as The Wrath. 

The Mount Kobau fire which raged for weeks in 2015 immediately behind the vineyards on the Golden Mile Bench. The British Columbia Wildfire Service initiated a countermeasure called a controlled back-burn, seen here, that greatly reduced the blaze. Luke Whittall

Lots of things can go wrong in winemaking, and the resulting flavors are called faults. Some are climatological, others technical, and others bacterial. And plenty of flavors that could be faults are fine in moderation. “Oak itself isn’t technically defined as a flaw,” says Chad Stock, a winemaker at Oregon’s Limited Addition Wines. Too much, though, can be sickly and overpowering. “There are an infinite number of what I would call artistic examples of delicious wine that can defy and bend the rules,” he says.

Stock and his winemaking partner and wife, Bree, tasted Eggert’s Fumé Franc on a 2015 visit to British Columbia and encouraged him to work with the vintage. “The smoke profile actually felt a lot more like a barrel toast than it did acrid or ashy.”

According to Bree, a master of wine, many of the rules about what good wine tastes like were revised recently, by modern palates. Twenty years ago, red wines from Burgundy and Bordeaux were renowned for tasting like the floor of a stable—like the barnyard funk of a ripe goat cheese with a creamy core and a fluffy rind—because of Brettanomyces, the same yeast that Eggert first suspected of tainting his barrels. “I remember when I first had Burgundy in a restaurant,” she says. “My wine director gave me one, and I came back to her and I said, ’This tastes like horseshit.’ And she said, ‘Yeah, that’s burgundy.’”

In the past few decades, fermentation chemists developed tools to test for and eliminate the yeast—in part because American palates didn’t favor it. “Now it’s very difficult to find a wine from Burgundy with Brettanomyces,” Bree says. “It’s been that quick of a shift. Modern winemakers started showing sellers in these old traditional regions that, ‘Oh, yeah, this flavor is actually a fault.’”

Winemakers who see smoke as a fault have found different ways to adapt. Some release vintages that are specifically marketed to be drunk within a few months of bottling to head off gradually intensifying smoke. Some blend vintages to keep production running when they lose some berries. Others actively treat wine via reverse osmosis and carbon to remove volatile phenols. 

But some have come back to smoke after experimenting. JR Fletcher, one of the winemakers at the Hatch, near Fairview, has explored filtration in the years since the 2015 fires. “Honestly, sometimes it works, sometimes it doesn’t,” he says. They’d had success in just releasing their 2015 cabernet franc, which they called Smokeshow. “It was kind of funny, a little cheeky,” Jones says. When the tasting room heard about it, they wondered if anyone would like it. But the sellers dug in. “It was really a logical progression for us to point out the smoky and be like, ‘Hey, this is technically a wine fault. But why?’”

If a winemaker is working in pursuit of some ideal, Fletcher says, they’re going to miss surprises in front of them. And maybe they’re missing a way that regional flavors will change with the climate. The Okanagan Valley and the surrounding region have been covered in smoke every August for the past several years.

In fact, Fletcher and Jones want to keep making wine with smoke—on purpose. “We did set aside some cab franc this year for another Smokeshow,” he says. “It’s our biggest cult-following wine we’ve ever put out,” Jones adds. And, she jokes, they’re going to sell it for a lot more than they did last time.

Read more PopSci+ stories.