“WHAT DOES CLIFF SAY?” tens of thousands of us obsessively asked each other last December. We were in the middle of one of the longest runs of snow in Seattle history, and the pronouncements of local meteorologist Cliff Mass had the urgency of dispatches from a war zone. A week before Christmas, four and a half million people with a zillion errands to run, roads already impassable from the previous storms, garbage uncollected, planes grounded, holiday parties canceled, the few Metro buses that were inching along treacherous emergency routes crammed to capacity—in short, the city essentially paralyzed—and what Cliff was saying on the blog that he had taken to updating two, sometimes three times a day was that “a really threatening event is possible for Saturday night and Sunday. Not a wimpy convergence zone…but a major Pacific weather feature.”
Already famous locally for his popular Atmospheric Science 101 class at the University of Washington and his weekly weather prognostications on KUOW, Mass had been riding higher than ever all that autumn on the strength of his newly published best seller The Weather of the Pacific Northwest. And then the snow started falling and didn’t quit and suddenly Cliff Mass was the indispensable seer to a desperate population. On that ominous morning of December 18, some 30,000 local weather nuts were hunched over computer screens, reading his blog, 25,000 of us clicking back again (and again) to revel in the words major Pacific weather feature.
What our meteorological prophet did not reveal was that while frozen Armageddon was bearing down on us, Mass himself was 3,000 miles away caring for his ailing father in New York. “I could see everything on my laptop,” Mass, trim and fit in his mid-50s from years of biking to work, told me six months later in his office in the UW’s atmospheric sciences building. “These days, with so much information available, there’s no need to be on the scene.”
Mass paused before delivering one of his signature provocative jabs. “The fact is,” he murmured in a Long Island accent undented by three decades in the Northwest, “meteorologists are the last ones who should be allowed to forecast the weather. The weather service guys over at Sand Point tend to overforecast snow because they like it so much. Inevitably judgment is impaired by excitement.”
Mass’s relish in mixing it up with the National Weather Service and his absolute conviction that he is “the preeminent authority on Northwest weather” (to quote the jacket of his book) have made him something of a local meteorological rock star (one television forecaster calls him the Eminem of weather). But the pokes and boasts go beyond bad-boy posturing. As everyone with a professional stake in Pacific Northwest weather will tell you, Cliff Mass and Brad Colman, the meteorologist in charge of the National Weather Service forecast office at Sand Point, have been tangling for years over how best to probe, analyze, predict, and communicate the mysteries of the atmosphere in this damp corner of the world.
Some of it is just a pissing contest—who nails the numbers better, Mass and his academic colleagues and grad students in the UW’s atmospheric sciences department or the 18 government forecasters who work with Colman at the NWS? But at a deeper level, the Mass-Colman clash reflects a profound schism that runs through this region’s weather community.
“The weather service lags behind. They’re not nimble, they lack resolution…once they make a forecast, getting them to change when conditions warrant it is like turning a big ship.” —Cliff Mass
You may be wondering why a place where nine months of drizzle cycles predictably into three months of serene sunshine even needs a weather community. Seattle doesn’t have hurricanes or severe thunderstorms. Tornadoes are exceedingly rare, extremes of heat and cold uncommon, blizzards unheard of (even last winter’s snowstorms were puny compared with the storms that routinely clobber New York, Chicago, and Minneapolis). Nonetheless, despite its often humdrum weather, our region has emerged as one of the nation’s leading centers of atmospheric research—and contention. Important progress is being made here in both the daily business of forecasting and the larger sphere of climate change (another area in which Mass has weighed in controversially—most notably in his recent rebuke of colleagues at the UW who claimed that Cascade snowpack has dwindled by 50 percent in the last half century).
On your typical day nobody really cares if Mass says there’s a 30 percent chance of rain while Colman’s people insist on 40 percent. But when the weather does turn hazardous—when black ice slicks our roads or Pineapple Express rains overflow our rivers or low-pressure systems spin vicious winds off the Pacific or snow accumulates in our streets for half a month—the way those numbers get crunched and communicated becomes a matter of life and death. This is a pissing contest with consequences.
No one died as a result of the snowstorms that pummeled the region last December—but it was truly a month for the record books. By the time the weather wound down, snow had fallen on 11 of the 14 days between December 13 and December 26, with a record cold streak keeping the white stuff frozen and in place for 357 hours. Few roads aside from the freeways and the runways at Sea-Tac were ever plowed, and countless streets were closed for the duration. Numerous flight cancellations stranded thousands of passengers at Sea-Tac. Metro buses provided severely limited—or nonexistent—service for days. Amtrak and Greyhound shut down simultaneously. Garbage collection ceased; schools and businesses closed down; and the city, according to the Preliminary Damage Assessment, ended up $3.4 million poorer.
Cars spun out, flipped over, careened between cavernous ruts, and on December 19 a pair of charter buses circumnavigating closed downtown routes turned onto Thomas Street, lost control, and slid down a hill. One of the buses crashed through a freeway guardrail and hung perilously over I-5.
Brad Colman is definitely one of the snow lovers at the Sand Point forecast office—in fact, he may be Snow Lover in Chief. Tall, balding, reserved, and soft-spoken, Colman combines the mild manners of the Northwest native (Tyee High School, class of 1973) with the high-beam seriousness of the career scientist (MIT Doctor of Science, 1984). But get him talking about snow, and Colman kindles. In his mid-50s, Colman is an avid skier who lives in the snow-prone Cascade foothills south of Issaquah and owns a winter cabin at Snoqualmie Pass. He harbors a fervent wish that global warming will hold off for a few more decades of deep Cascade snowpack.
So naturally Colman was ecstatic last December when snow started falling on the Puget Sound lowlands and didn’t quit. He was also incredibly busy. “I practically lived here during the series of storms,” he told me when I met with him in his big sunny corner office at the NOAA complex at Sand Point. “We did so many briefings with emergency managers and media I almost didn’t get to enjoy it. The work was endless.”
The fact that Colman was glued to his desk, his monitor, his phone, and his picture windows overlooking Lake Washington while Mass was sitting in New York tapping away at a laptop says a lot about the fundamental differences between them. It would be too simplistic to boil it down to a conflict over the relative importance of human versus machine—but they’re playing out a subtle weather—geeky variation on that theme.
To understand the fine points it helps to have a little background on the recent history of weather prediction. Back in the stone ages—pre-1950s—weather forecasting was largely empirical and intuitive: The forecaster plotted observations on weather maps, analyzed major features like fronts and pressure systems, and foretold the future by comparing those features to similar conditions in the past. Old-timers with long memories were prized for the dependability of their hunches.
“We look like jerks when Cliff insists on undercutting us publicly…he comes off as the hero—we are the goat. It’s hurting the weather service.” —Brad Colman
Everything changed in April 1950, when computers entered the picture. An early task assigned to the ENIAC (Electronic Numerical Integrator and Computer) was to run a series of calculations that could be used to forecast weather. ENIAC -assimilated data from 768 weather stations to -construct a grid that simulated atmospheric conditions—and then applied the laws of physics to calculate how conditions at each grid point (the imaginary spots in the three-dimensional chess board) would evolve in the near future. It took ENIAC nearly 50 minutes to crunch the data, and the resulting model was so crude that only large, sprawling events were captured. The meteorological impact of our mountain ranges and irregular channels of salt water did not even show up on the early models. But, as the speed and capacity of computers grew exponentially over the past half century, the models became correspondingly more refined and realistic.
So realistic, crows Mass, that the high—resolution MM5 model that he began to apply to Pacific Northwest weather a decade ago, and still runs out of the UW, can almost perfectly simulate a Puget Sound convergence zone. Plug in the numbers, hit enter—and voilà—showers over Everett, heavy snow at Stevens Pass, and a big blue hole from the King-Snohomish county line down to SeaTac. Even the national models that are the daily workhorses for National Weather Service forecasters remain too crude to capture a convergence zone in all its subtle, evanescent glory.
Not so fast, counters Colman. For years now, Colman has been challenging Mass’s insistence that more detail and higher resolution are better. “Cliff’s fine-grid models are sexy and seductive physically, but what’s their utility if he ends up with more and more detail that is wrong in terms of the actual timing and placement of the event?”
Mass responds with some pointed pokes of his own: “The Weather Service lags behind. They’re not nimble, they lack resolution. And they insist too much on continuity. Once they make a forecast, getting them to change when conditions warrant it is like turning a big ship.”
“We look like jerks when Cliff insists on undercutting us publicly,” Colman fires back. “Some of his remarks are really hard to swallow. He can say things and get away with them because he’s a rock star. He comes off as the hero—we are the goat. It’s hurting the Weather Service.”
So, yes, some of it is schoolyard stuff about who is bigger, faster, better, more accurate. But what’s also going on is a power struggle between two institutions with fundamentally different missions and constraints. Nick Bond, a meteorological rock star in his own right who has flown into epic rainstorms off the coast of Alaska and towering thunderheads in the tropics, sheds some light. Bond is a research meteorologist with a joint appointment at the UW and at the Pacific Marine Environmental Lab (a climate and ocean research facility run by NOAA out of Sand Point), so he has a foot in both camps. “Some of the tension between the department and the Weather Service is natural because of their different agendas and different demands,” says Bond. “The Weather Service has the mandate to forecast every day. The researchers at the UW have the luxury of cherry-picking particular storms to analyze. If Cliff is busy or his cat throws up, he doesn’t do the blog. But at the Weather Service the clock is always ticking for the next set of zone forecasts. There’s always that harsh deadline.”
Two charter buses lost contorl and slid down a hill. One crashed through a freeway guardrail and hung perilously over I-5.
There’s no question that the institutional tensions are exacerbated by the size and prickliness of the egos involved. “Cliff has sharp elbows—and never shies away from a fight,” one local meteorologist commented off the record. “Cliff is a lightning rod,” another colleague comments, also insisting on anonymity. “He sometimes takes credit for stuff that others have been involved in. There are those who roll their eyes whenever he talks. He definitely has his enemies.”
Not even Mass’s allies would claim that diplomacy is one of his strong suits. Take the one-sentence diss of television meteorologists in his recent book: “Local TV weathercasters generally do not make their own forecasts and rarely stray far from NWS predictions.” Try telling that to seasoned, highly trained professionals like Jeff Renner, chief meteorologist at KING 5, or M. J. McDermott, who does the morning slot at Q13 Fox (and was a student of Mass’s some years back).
In 2006 McDermott won the spring forecast contest that the UW’s atmospheric sciences department has been holding for 40-odd years—the only woman and the only television weathercaster ever to have come in first. “They hold the contest in the April-to-June period because the weather is especially tricky to predict then,” McDermott notes. “I got a lot more respect after I beat out the guys.”
“That is a pretty informal forecast game,” Mass sniffed when I ask how his horse placed that year. But he also cracked up when pushed to talk about how it feels to be a local superstar. “When I have lunch with Jeff Renner no one comes up to ask for my autograph.” Not that he’d mind if they did.
Strip away the pettiness and the politics, and what fundamentally divides Mass from Colman and his crew at the NWS is a profound disagreement over the role of the individual forecaster. “It is often said that weather prediction is as much art as science,” remarks Mass, “but this is no longer true.” Mass firmly believes that with enough data, the right equations, and sufficient speed and memory, our computers can conquer the world, or at least the atmosphere.
Mass is putting this idea into practice with a fully automated experimental local forecast system called Probcast that he and colleagues at the department of atmospheric sciences devised in collaboration with the departments of psychology and statistics, and the Applied Physics Laboratory. Instead of assigning a single value for tomorrow’s high and low temperatures and chance of rain, Probcast, implementing probabilistic forecasting techniques, includes values for the upper and lower ranges in each category and the likelihood that each value will occur. In essence, it’s a way of sharing the inherent uncertainty of weather prediction with the user—a tipping of the forecaster’s hand. But what’s truly unique about Probcast is that no snow- or sun-addled human being messes with it: All values are generated automatically through equations that assimilate weather data, numerical models, statistics, and the laws of physics. “The beauty of Probcast is that it runs by itself,” says Mass.
For Colman the problem with automated forecasting is that it rubs up against his basic belief in the importance of having a person in the middle of the forecast process. “Early in my career one of my mentors advised me to spend time on the front line as a weather forecaster,” he told me, “which is not a typical move for someone with a doctorate in meteorology. I spent two years in Juneau as a line forecaster—and I now see that doing it every day, in both boring and exciting weather, is the best way to learn to visualize the atmosphere and bring the science to life.”
Colman has no illusion that human skill will ever rival that of the numerical models—but he strongly disagrees with the idea that human input does nothing but introduce bias. “In many ways our job is becoming more important,” he insists. “As society becomes more vulnerable to the impacts of weather, the demand for reliable weather information has grown—at the same time that the skill of the forecaster has grown. We have really come to recognize that there’s more to weather prediction than getting the forecast correct.”
The person at Sand Point who does the finest job of supplying that “more” is, hands down, Ted Buehner. A 53-year Northwest native with a passion for umpiring baseball games, Buehner is the office’s warning coordination meteorologist—which means that when dangerous weather threatens, Buehner is the one on the phone briefing emergency managers, elected officials, local businesses, and media.
Soon-to-be-former Mayor Greg Nickels dug himself into a deep political pit last winter when he gave Seattle’s snow removal efforts a “B” at a time when just about the only snow-free surfaces in the city were a couple of interstate lanes and the airport runways. (Nickels’s handling of the situation is widely cited as the top reason he lost the primary election this past August.) Buehner wisely refrains from grading the Weather Service’s forecast record during last December’s snowstorms—but he will say that the response to his online briefings and numerous media interviews during the onslaught was “overwhelmingly positive.” “Somebody needs to interpret and spell things out,” he notes. “This human input is critical for making wise decisions. Let’s face it, we have 4.5 million people living in our forecast area, and not all of them read Cliff’s blog.”
When the next big storm is bearing down, who are you going to trust—an automated website with cute icons and a disclaimer about “uncertainty information” or a guy who knows Ichiro’s career batting average, the best route to drive to Leschi on icy roads, and how to make meteorological threats riveting to high school students and community groups?
Everybody in the local weather community has a war story about the long-standing debate between Mass and Colman: jabs traded at the annual Northwest Weather Workshop, testy exchanges on the weather chat room where area meteorologists hash out their postmortems of major events—verbal pokes and head butts. But those in the know also believe that we all stand to gain from what is, in essence, an intellectual clash between two brilliant scientists.
“Yes, there’s tension,” says the UW’s Nick Bond, “but by no means is there animosity. Ultimately, this is how progress in science is made. If we’re all singing ‘Kumbaya,’ no one will ever get out of their comfort zone.” Colman stresses how respectful and productive the relationship has been: “The bottom line is that we’re delivering stronger results and better weather information because of what Cliff and I have learned from each other. The community as a whole has benefitted.”
One thing the two of them agree on is that, with a mild to moderate El Niño event building in the tropical Pacific, this coming winter will more likely than not be warmer, drier, and more tranquil than normal. As for lowland snow, you can keep your shovel and sled safely tucked away in the garage.