According to Mike Zolnikov, who tends a couple of acres of Pinot Noir and an acre of Chardonnay on a flat, slightly soggy patch of the central Willamette Valley, in Oregon, it had been a once-in-a-decade growing season. “Not too hot, not too wet,” he recalled, wistfully. “It would have been a really great year.” A few hundred miles south, in California’s Napa Valley, the winemaker Ashley Egelhoff, of Honig Vineyard and Winery, was feeling similarly about her Cabernet and Sauvignon Blanc. “That’s how 2020 was panning out: like Goldilocks, just right,” she told me.
For wine growers and makers, each season offers a series of fresh yet familiar opportunities for disaster. Drought shrivels the grapes; excessive heat deprives the juice of acidity; too much rain results in rampant mold. “But that’s the fun of it,” Egelhoff said. “Every harvest brings a surprise.” The gamble of spraying early or of picking the grapes late, the black magic of fermentation, the art of blending: it’s precisely the puzzle of chance and choice that keeps winemakers hooked. Plus, every now and then, as in 2020, you get perfect conditions. “Then everything went to hell,” Egelhoff said.
That August, the West Coast’s worst fire season in history began. More than eleven thousand bolts of lightning struck central and Northern California in the span of thirty-six hours, heralding the start of an orange-skied autumn in which flights were suspended, more than eight million acres burned across twelve states, and winemakers’ dreams of a perfect vintage went up in flames. “The lightning storm came over on the first day we were bringing Sauvignon Blanc in, and within a couple of hours there was smoke,” Egelhoff told me. “I was on the crush pad—we were unloading our first truck of fruit—and it was probably one of the most heartbreaking moments of my career.”
In the past few decades, as wildfires have become larger, faster, and more severe because of climate change, the focus has been on the considerable damage caused by the flames themselves; the smoke has been thought to be relatively harmless. Only recently have scientists realized that the opposite is true. In humans, smoke inhalation has been linked to heart and lung damage and to multiple forms of cancer; this year researchers in Europe concluded that they had underestimated death tolls from short-term wildfire-smoke exposure by ninety-three per cent. In the United States, smoke exposure is estimated to have caused tens of thousands of deaths every year between 2010 and 2020—an order of magnitude more than the number of lives lost to the actual fires.
Plants don’t have lungs, of course, but grapevines do breathe, absorbing oxygen and other atmospheric gases—including smoke—through small pores on the underside of their leaves, or by diffusion across the fruit’s thin, waxy skin. The result is smoke taint, a flaw in wine that has been described as tasting “like Las Vegas smells,” like “burnt salami served on an ashtray,” and, perhaps most evocatively, like the morning after a big night out, when “you’ve smoked a bunch of cigarettes and then you wake up, smell your hands, and regret your entire life.”
Although many wine drinkers have remained blissfully ignorant of this addition to wildfires’ already heavy toll, it has been disastrous for winemakers. One analyst concluded that the 2020 wildfires cost the California wine industry nearly four billion dollars, an amount that includes both direct fire damage and sales lost owing to smoke exposure. “We had brought in just twenty tons of Sauvignon Blanc, and we had to assume that everything else was ruined,” Egelhoff said. “It was a lost vintage.” The hundreds of thousands of tons of California grapes left unharvested that year were estimated to be worth more than six hundred million dollars alone. Oregon suffered similarly. “For a couple of days, it was a red sky, and then there was no sky,” Zolnikov said. “It was just solid smoke.” He painstakingly cleaned all the ash off his vines before harvest, but when winemakers shared the bottles they’d made with his grapes they still tasted acrid and smoky.
Clearly, the best way to prevent smoke taint would be to prevent wildfires in the first place. In the meantime, the wine industry is desperate to protect its grapes. As 2020 drew to a close, a trio of West Coast researchers—Tom Collins, at Washington State University; Elizabeth Tomasino, at Oregon State University; and Anita Oberholster, at the University of California, Davis—proposed an ambitious, “smoke to glass” effort aimed at finding an answer. “That year made it very clear we need to be better prepared,” Tomasino told me. The U.S.D.A., which normally has a puritanical reluctance to fund research that might be used by the beer, wine, and spirits industry, awarded the team $7.65 million in 2021. “As devastating as 2020 was, that’s the silver lining,” Egelhoff, who recalled sending the trio “a very angry e-mail” that year, complaining about a lack of help from researchers, said. “It really pushed them to get the solutions we need.”
In September, I joined Collins and a group of students on a trip to Washington State University’s experimental vineyards, in the Yakima Valley. It was early morning, and two sunrises lit the horizon. The false dawn, to the north, was a wildfire: overnight, a lightning strike had ignited the desiccated grasses of Rattlesnake Ridge, casting the hills around us into ominous relief. It was a stark reminder of the reason we’d woken up at this hour. Before the morning was over, we would simulate a rangeland fire of our own, to study the impact of smoke on wine grapes.
Collins runs the most impressive smoke-taint experiments in the country. Whereas Tomasino’s team, in Oregon, works with a handful of vines at a time, Collins smokes the equivalent of a quarter-acre vineyard in large hoop houses, allowing him to get closer to real-world conditions—and to make a decent amount of truly terrible wine. (Sadly, Oberholster died, from cancer, last year.) Each house encloses two hundred Merlot vines, and once we arrived we began pulling shade cloths over them, to the accompaniment of a portable speaker pumping out Fleetwood Mac. Three of the houses were to remain smoke-free, as an experimental control. In three others, we used zip ties to hang fat swags of vented plastic hosing along each row of vines, directly under the clusters of purple grapes.
The students and I wrangled tarps and zip ties while Collins, who volunteers with the Boy Scouts, issued instructions leavened with gentle ribbing and reminders to hydrate. He fussed with the hoses, hooking them up to three battered grills. The light turned salmon, then golden, as we worked. Collins told the students to gather a few clusters and leaves for a pre-smoke sampling but to avoid vines with pink or orange tags, as these had been treated with an experimental barrier spray. Finally, with the samples stashed safely in an ice chest, Collins opened the grills, blowtorched some pellets inside, and watched as the smoke got going. I poked my nose through a slit in one of the houses as it filled with a pungent haze: the pellets were handcrafted from more than a dozen local rangeland species, including sagebrush, cheatgrass, and tumble mustard, all painstakingly collected by summer interns.
Although fire has been mankind’s constant companion and wine likely predates most agriculture, smoke-tainted wine seems to be a relatively recent phenomenon. “People weren’t really aware of it, but it probably had been happening,” Mango Parker, a senior research scientist at the Australian Wine Research Institute, told me. She pointed me to a reference in an Italian enological textbook from 1892, which lists “smoky taste” as a potential flaw in wine—fortunately “found more rarely in Italian wines than in German.”
In modern times, Australians have taken the lead in the war against smoky taste. In January and February of 2003, at the start of harvest season, the state of Victoria experienced its biggest bushfire in decades: an area larger than Sydney burned, and the A.W.R.I. began receiving calls from winemakers whose vintage was exhibiting pronounced notes of ashtray. At the time, Collins was a researcher at Beringer, an Australian-owned winery in Napa Valley; after a relatively meat-and-potatoes upbringing, he’d fallen in love with wine in New York’s Finger Lakes region while studying Russian at Cornell. But it wasn’t until 2008, when a thick haze of wildfire smoke blanketed Northern California for much of the summer, that Collins became aware of the problem, and it wasn’t until 2017, when fires briefly left Napa with the worst air quality in the nation, that smoke taint became part of the vocabulary of West Coast winemakers. “Everyone was talking about smoke taint that year,” Esther Mobley, the senior wine critic for the San Francisco Chronicle, told me. “But, for the most part, it really wasn’t an issue because of the time of year that the wildfires happened.” Ninety per cent of grapes had already been picked, so the damage was limited.
“Everything changed in 2020,” Mobley continued—both on the West Coast and in Australia, where bushfires were catastrophic, generating their own weather and pushing entire species toward extinction. Andrew Spinaze, a winemaker in Australia’s Hunter Valley, told me that he received several calls from his Napa Valley counterparts asking what to do. “We were saying we don’t really know,” he admitted. Part of the problem was that harvesting is expensive, and no one could tell whether a batch of grapes would, in time, yield a ruined vintage. “It became a huge question: Were you going to make wine that year or not?” Mobley said. In California and Oregon, the laboratories that typically perform analytical chemistry for wineries were quickly overwhelmed. “The labs were backed up for months, so people were sending samples of grape juice to Australia and to Canada,” Mobley said. “By the time you got the information back, the moment to harvest had passed anyway.”
By then, Collins had been intentionally smoking grapes for years; the very first grant proposal he wrote when he joined the faculty at W.S.U., in 2015, was for the hoop-house program. As we watched the fumes rise in his makeshift wildfire, the morning’s chirps and caws gave way to coyote howls. “Our exposures are typically thirty-six hours, which we’ve found is enough time for us to definitely get smoke impact,” Collins said. He took the night shift, maintaining smoke levels while watching “Zoolander” and “Spaceballs” on his laptop. We reconvened the next evening, just before sunset, to tear down the hoop houses and stow the gear. Collins wheeled away the grills for winter storage; a freezer was stacked with ziplock bags of sample grapes and leaves, ready for analysis. The fruit was still a few weeks shy of harvest, but I popped a tiny grape in my mouth anyway. After thirty-six hours of smoke exposure, it was still a treat, the juice delightfully sweet as it trickled over my tongue.
How can a smoke-tainted grape taste delicious off the vine, only to become acrid and ashy in the bottle? This conundrum obsessed Tomasino, a cheerful scholar of flavor chemistry whose research, though focussed on wine, has also considered the intricacies of coffee roasting and the impact of terroir on cheese. After the 2020 debacle, one of the most frequent requests that Tomasino fielded from anxious winemakers was for a kind of palate training, to help them get to know the enemy. This is a common practice among sommeliers and vintners: to learn how to identify the musty, wet-cardboard notes of cork taint, for example, they’ll sip glasses spiked with TCA, the compound that causes it. Now they wanted a mixture that re-created smoke taint—which meant that Tomasino had to figure out how to mimic something whose chemical composition had yet to be determined.
This was not a simple task. The glass of red you reach for at the end of a long day, the champagne you raise in a toast, the funky orange you order to look sophisticated: each contains between two hundred and four hundred aromatic compounds. The earthy quality so prized in a Cabernet Sauvignon is due to a smidge of pyrazine, a volatile molecule more commonly found in peas and bell peppers; the gasoline note of an aged Riesling comes courtesy of a compound called TDN. Sommeliers are fond of saying that wine is the second most complex liquid in the world, after blood. Tomasino told me this is not necessarily a scientific claim; still, the sheer quantity of chemicals interacting in a glass gives the search for offenders a needle-in-a-haystack quality.
When Australian researchers started considering the problem, in 2003, they identified two chemicals—guaiacol and 4-methylguaiacol—as the primary components of smoke. But the presence of these compounds in grapes failed miserably as a diagnostic for smoke taint. In 2009, Australian growers and winemakers sent samples in for testing during another disastrous fire season; when the A.W.R.I. didn’t detect excessive amounts of the two chemicals, vintners went ahead and bottled as usual. “The wines were just terrible,” Parker recalled. “They became smokier and smokier over time, and the thing the winemakers really hated was an ashy aftertaste that developed even after they spat it out.”
Part of the issue involved the grapevines’ own defense mechanism. Guaiacol and other smoke-related compounds are poisonous to plant cells, but plants have developed an elaborate in-house detoxification system, in which dangerous chemicals are quickly bound to sugar molecules, making them less reactive, odorless, and easier to sequester. This process, which was already under way as Collins watched “Zoolander” outside his hoop houses, seems to be quite effective for grapes; unfortunately for the humans who want to turn them into wine, it’s a ticking time bomb. Fermentation, which uses yeast to break down the sugars in grape juice, transforming single-note sweetness into the complex, delicious alcohol enjoyed by so many, also unlocks those smoky compounds; so, too, can the enzymes in human saliva. Testing for guaiacol and its friends in grapes failed to account for the fact that the chemicals were still bound to the new sugar compounds, called glycosides.
After one of Parker’s colleagues pointed this out, in 2010, the A.W.R.I. promptly added a handful of glycosides to its list of least-wanted compounds. But the test was still far from perfect, Tomasino told me. In desperation, she assigned a graduate student, Jenna Fryer, to burn all sorts of things—marshmallows, paper, charcoal—and then dilute them, hoping to develop a recipe for replicating the smoke-taint sensation. After sampling at least fifteen of these tinctures without much success—“I still have nightmares about some of the things we tasted,” Tomasino said—it occurred to her that she’d heard the term “ashy” used in culinary circles, so she told Fryer to chat with some chefs. This led Fryer to incinerate leek tips, brew the cremains using boiling water and a coffee filter, and then mix the result into wine. During a meeting, Tomasino served Collins some of Fryer’s burnt-leek beverage. “Tom looked at me, and he’s, like, ‘That’s literally the ashy aftertaste.’ And I said, ‘Right?’ ” she recalled, triumphantly. “Then I said, ‘I’m giving you a vial—tell me what’s in here.’ ”
Collins’s lab revealed that the leek tea was indeed loaded with smoky chemicals, and that it also contained another class of compounds: thiols. Thiols are responsible for the distinctive aromas of skunk spray and ripe durian; they’re also added to natural gas to provide a detectable rotten-egg smell at even trace levels. When Tomasino spiked red wine with guaiacol and similar chemicals, the mixture tasted like Band-Aids; if she used only thiols it had a vegetal aspect, reminiscent of Brussels sprouts. It took adding both to create the particular licking-a-grill notes of smoke taint.
Since then, Collins, by testing grapes that have been flash-frozen every few hours during one of his hoop-house experiments, has shown that the production of both glycosides and thiols seems to ramp up significantly during a smoke event—and that, like glycosides, bound thiols break apart during fermentation. Case closed. “This is a once-in-a-lifetime thing for a career, finding this, and it was pure dumb luck,” Tomasino told me, with a laugh. Astonishingly, given their impact, the levels of thiols in smoke-tainted wine are so vanishingly tiny that they’re measured in nanograms per litre. The team would never have spotted them just by analyzing a defective bottle.
In a fluorescent-lit sensory-testing facility at Oregon State University, Tomasino led me to a cubicle with a single computer, told me to follow the instructions on the screen, and wished me luck. Camilla Sartori, the facility manager, had set up a tray holding six black wineglasses, a spittoon, and two clear glasses, one filled with water and the other with a slightly sugary solution designed to be a palate cleanser. During the next sixty minutes, I sipped, swished, and spat while answering a barrage of questions about how the wines tasted, my emotional response to them, and my willingness to purchase.
Discovering the chemicals responsible for smoke taint is, sadly, only half the battle. When it comes to wine, perception is reality, and humans turn out to be remarkably varied in our responses to smoke taint. Among Tomasino’s recent findings is that at least ten per cent of drinkers can’t detect smoke taint in wine and many more don’t mind it, depending on the grape variety, wine style, and degree of taint. In Australia, similar studies have found that between nineteen and forty per cent of consumers, as the winemaker Peter Leske put it, “don’t care as long as it’s wet and it’s got alcohol.” Even among people who can detect smoke taint, Tomasino estimates, sensitivity varies at least a hundredfold. “I’ve made it a career goal to never work on saliva,” Tomasino told me. “But people have a lot of enzymes in their mouths, and they’re very different.”
After I’d spent an hour in the cubicle pondering whether I was tasting red fruit or dark fruit, and whether it made me feel bored, worried, or excited, Sartori tabulated my results and explained what I’d been drinking: a control glass of classic Willamette Valley Pinot Noir that hadn’t been exposed to smoke, a glass of commercially available smoke-tainted 2020 Pinot from the same region, and a series of samples that blended the two in different ratios. We established that I am prone to overthinking, that I’m not a huge fan of the earthy and acidic notes that are considered the hallmark of the region’s flagship varietal, and that the only sample I’d willingly purchase was a blend of one-quarter smoke-tainted wine and three-quarters not. According to my checklist, it tasted like tobacco, black pepper, and dark fruit, with a slight herbaceous note and a bitter finish.
My own idiosyncrasies aside, running this sensory test has helped Tomasino discover that the effects of blending are nonlinear, meaning a blend that is just five per cent smoked wine is typically perceived as having much more of an ashy aftertaste than one that is fifty-fifty. Tomasino credits this finding to a phenomenon that researchers call “matrix interactions,” in which the threshold at which a compound can be detected depends on its context. Research has shown that fruity esters can mask smoke taint, and that the molecules responsible for the cut-grass and green-apple notes of Sauvignon Blanc enhance it. “It makes for really complicated math,” Tomasino said.
To add to the complexity, smoke taint is inflected by subtle distinctions between grape varietals, too. Pinot Noir is universally acknowledged to be the most vulnerable to taint, for reasons that probably have to do with its famously thin skin; Syrah, in which a savory bacon note is considered desirable, naturally contains significant quantities of guaiacol, no wildfire required. (In general, red wines are especially susceptible to taint, because many smoky compounds are concentrated in the skins, which are left on during fermentation.) Still more variables are introduced when you consider the ripeness of the grapes—a smoke event earlier in the season may give the vine more time to metabolize toxic compounds, as opposed to one nearer harvest. And all these considerations are preceded by the unpredictable, impossible-to-model vagaries of smoke itself.
Mike Kleeman, an environmental engineer at U.C. Davis who has been working with Collins and Tomasino, told me that gauging the harm of smoke exposure requires, at minimum, an understanding of how old the smoke is by the time it reaches the grapes, how reactive the atmosphere it travelled through was (sunshine chews up dangerous chemicals, making the smoke less harmful; chlorinated ocean air can have the same effect), what chemicals were in it in the first place (California’s coastal chaparral seems to have high concentrations of smoke-producing compounds), and even the topography of the vineyard.
With so many variables at play, the mental effort of evaluating smoke exposure in grapes can cause more of a headache than drinking the resulting bottle. Mobley, the wine critic, told me that, after tasting dozens of smoke-tainted wines over the years, she’s decided that it makes more sense to think of them as falling on a spectrum than as being defined by a single flaw. “Sometimes they can have that herbal, mezcaly smokiness, sometimes it’s more of a barbecue smokiness,” she said. “Sometimes you might not get any overtly smoky sensations but there’s kind of a shortness to the wine—it tastes fine, just a little lifeless compared with what it could have been.” Regardless, she said, industry professionals tend to prize a certain purity of terroir and varietal expression in their wines. For them, she said, “it’s just all bad.”
For grape growers and vintners concerned with making a living, the pursuit of smoke taint’s essence has thus far had relatively limited practical application. The industry wants inexpensive, field-based tests that can predict whether a wine will be tainted or not, effective techniques to remove smoke-taint compounds from wine, and—dare to dream!—technologies to protect grapes from smoke in the first place. Collins and Tomasino argue that understanding the nature of the problem is a precursor to achieving these goals, but they’re working on them nonetheless.
In his hoop-house experiments, for example, Collins has shown that spraying grapes with kaolin, a clay-based powder that is already used to shield grapes from sunburn, reduces post-exposure guaiacol levels by nearly half—a promising result. “The idea is that the kaolin absorbs things from the smoke and keeps them from getting to the fruit,” Collins said. “But we realized you have to wash the stuff off, because otherwise it just releases everything it has absorbed back into the grape.”
Indeed, wine made with unrinsed kaolin-sprayed grapes is the only thing I drank in the course of my reporting that was literally nauseating. “If you don’t rinse it off, you would have been better off doing nothing,” Collins said, as I sputtered and heaved. In his latest hoop-house experiment, he was testing just how quickly that rinse needs to occur, though any rinsing may be an obstacle for growers, since it requires more water, which is in short supply in many wine-growing regions, as well as more labor. Tomasino has recently been developing a different approach—a proprietary lipid spray, which doesn’t need to be washed off—but won’t be able to test it until the summer. “It works on the lab stuff, but we don’t know about in the field yet,” she said. Even if it succeeds, it may only reduce smoke-compound absorption—not prevent it altogether.
In other words, vintners still need a strategy for restoring the purity of their ferments. “An old winemaker friend told me about a technique where you use organic skim milk, and the proteins in the milk bind up some of the smoke compounds,” Andrew Jones, a winemaker at Field Recordings, in Paso Robles, told me. In desperation, he tried it; the proteins sank to the bottom and formed a sludge, allowing him to drain the wine off after twenty-four hours. “It did take away some smoke,” he said. “It also strips away some color.” Egelhoff and Leske experimented with mixing their smoky wines with activated carbon, which Leske referred to as “the tactical nuclear weapon of winemaking.” Carbon—like reverse osmosis, which requires costly equipment—can indiscriminately strip out smaller molecules, removing color and flavor along with any suspect chemicals. “There are a lot of treatments where you’re, like, ‘Well, there’s less smoke, but it’s also taken everything positive about the wine out,’ ” Egelhoff said. “So, is it really a better wine now?”
In search of a more targeted solution, Lik Rong Lim, one of Tomasino’s students, developed a clever way to extract thiols, inspired by his professor’s childhood dislike for the sulfuric smell of canned vegetables. Years ago, Tomasino puréed canned spinach and ran it through a rotary evaporator to remove any aromatics, then offered a sensory panel the result alongside the original. “I’ve never seen such a statistically significant result,” she told me. Removing sulfuric compounds made vegetables more palatable; could something similar save a smoked-out Cabernet? In the corner of Tomasino’s lab, Lim showed me his invention, which uses inert nitrogen to bubble the thiols out of wine and into gaseous form. They can then be piped into a separate flask full of a colorless liquid that has two useful properties: it exclusively traps thiols, and it turns yellow in the process. At Lim’s bench, we watched as the liquid turned the strawlike shade of dehydrated urine. Lim explained that the fluid, unlike carbon or other refining agents, is extremely selective and never comes into contact with the rest of the wine. “As a proof of concept, it totally works,” Tomasino said.
An actual, practicable solution will require more research—and more money. The team will be submitting a proposal for another four-year grant from the U.S.D.A., tweaking its language to suit the Administration’s focus on economic-loss prevention rather than climate change. “I won’t have something for you tomorrow, but we will be able to take care of this problem,” Tomasino said. “With all the research that’s going on, in three to five years a smoke event will happen and we’ll know exactly what to do.”
Part of that statement is definitely true: a smoke event will happen. Researchers have calculated that by 2050 the current rate of warming will lead to seventy thousand Americans dying annually because of smoke exposure. “Lives lost are a tragedy; smoke-tainted grapes are a challenge,” Leske reminded me. Indeed, smoke-tainted grapes are just one of many challenges facing the wine industry right now, as rising temperatures also increase pest and disease pressures; heighten the likelihood of drought, hail, and flooding; and threaten to ripen fruit so quickly that harvest becomes compressed to the point of logistical impossibility. Smoke is a mere facet of this new normal—it is, arguably, now part of a wine’s “aeroir.”
Within the industry, this perspective has been slow to gain acceptance. But, sometime in the fall of 2020, Cyler Varnum, a vintner who’d purchased grapes from Mike Zolnikov, in Oregon, had a breakthrough. When people visited Varnum’s tasting room, in the Willamette Valley, they often asked how that year’s vintage had fared, given the wildfires. Varnum decided to take them back to a barrel and pull a sample so they could see for themselves. Some made a face and spat it out; others could taste the smoke but found it curious rather than repulsive; still others loved it. “That was the realization: we don’t dictate people’s tastes,” he told me, as we sat in his tasting room. “I shouldn’t be trying to tell people that it’s a flaw. I’d rather be, like, ‘This was 2020: you might like it, you might not.’ ”
After all, he pointed out, that’s true of any vintage: hot summers make wine taste one way, cold snaps another; some people like their Zinfandel aged in bourbon barrels or their Pinot Grigio diluted with ice. In the past few years, natural-wine aficionados have embraced the funky barnyard notes imparted by Brettanomyces yeast, which have traditionally been regarded as a flaw. And a floral, almost roselike aroma, considered characteristic of wines from cooler regions, was recently discovered to be caused by the inclusion of dead leaves in the ferment. Scientists have dubbed it “frost taint.”
Varnum, in his tasting room, shared what little remained of his 2020 stock, starting with a traditional sparkling blanc de blanc he’d bottled under the moniker Toast, made entirely from Zolnikov’s Chardonnay grapes and fermented in neutral oak. “It’s interesting, because when you think about champagne, you want toasted-brioche, crème-brûlée notes—that’s actually a quality you’re looking for,” he explained. On first sniff, I was not optimistic: the nose, as Varnum delicately put it, was “more on the burnt side of toast.” But the taste was much more nuanced: light, clean, and bright, with a browned-piecrust quality that never built into the bitter charred note I’d learned to anticipate. Earlier that year, Varnum’s partner, Taralyn, told me, they’d had a bonfire and brought out the glasses. “I think I drank almost a whole bottle,” she said. “Around a campfire, it’s delicious.” ♦
