If the day is a flying day, we call Flight Service for a briefing. Inside the airplane we get local weather information Romeo, Juliet, or Whiskey. If we have ForeFlight strapped to our leg, we may see the conditions at our destinations, which runway to use, the headwind and crosswind components—everything except whether the coffee is any good at the fuel stops.
If today is a flying day, we know a lot about the weather. Winds aloft. Cloud heights. Airmets. We know what it is now along the route and what it likely will become later in the day.
It seems easy, all the available information. Touch a screen. Click a mouse. But like so many things, the apparently simple has a complex beginning. The information has to be gathered. It has to be organized. It has to come from somewhere.
Grumbling about the weather on a day I could not go flying, I discovered the somewhere is 3 miles southeast of Kansas City International Airport, from a building you would never imagine holds something remarkable: the Aviation Weather Center (AWC). This is where pilot weather begins.
Bright and cheerful in the early morning, Jonathan Leffler, warning coordination meteorologist, meets me at the security desk. He leads me down two hallways and into a room that looks a bit like the command center on a movie set. Men and women sit at desks surrounded by computer monitors everywhere. No windows. A quiet gravitas. An illuminated sign on a far wall reads Aviation Weather Center.
There are eight operational desks inside the center: Ceiling and Visibility; Icing; Turbulence; Convective Sigmet; Traffic Flow Management Convective Forecast; Tropical; Global Graphics North; Global Graphics South. Those eight desks have an eye on the whole planet, tracking weather for pilots.
Soon we are talking about process, how the information gets to the center and how it gets out, and Leffler tries to explain a complicated system.
“You have ASOS and AWOS sensors all over the place,” he says. “That data is transmitted to what we call the gateway, in Washington, D.C. The gateway is the big communications node that intersects with all the other nodes around the globe. There’s a gateway in Brazil. There’s a gateway in Europe. There are several around the globe. And that’s how we exchange information. All of that gets sent to the servers we have here on the floor.
“We have airplane observations,” Leffler continues. “We have information that comes in from Doppler radar, sensing motion in the atmosphere. All that gets fed into models which help create the products we push out. Something like winds aloft might be a combination of model data and aircraft reports. We really try to leverage as much data as we possibly can.
“We might even look into road sensors,” he says. “Temperature sensors. Freezing rain versus rain versus snow. Meteorologically there are certain things that have to happen in the low levels of the atmosphere to change one into another, and that’s a big deal. Straight rain versus freezing rain could mean the world to somebody’s operation. How do you know what’s going on in the lowest 2,000 feet of the atmosphere?”
There are conversations, lots of conversations, with other meteorologists. There is a CWSU, Center Weather Service Unit, at each FAA air route traffic control center. National Weather Service meteorologists sit with personnel from the FAA, planning each day’s routes and reacting to changes.
“We interface with them,” Leffler says, “so what’s really good is that the message we’re creating here is consistent. Whether it’s at the national level with the NAMs—national aviation meteorologists—or at the CWSUs with a kind of regional focus, we’re all kind of dancing to the same tune.”
At the desks
In nonaviation weather, things such as tornadoes or severe thunderstorm watches come from the Storm Prediction Center in Norman, Oklahoma, while tornado or thunderstorm warnings come from a regional office closer to the storms. That is not the case in aviation. Any desk at the Aviation Weather Center can issue an airmet or sigmet for its specialty, and they are the only ones who can. While lots of people can offer specialty forecasts, the FAA uses the one from Kansas City.
At the turbulence desk, Senior Forecaster Mitch Sego scans monitors that display a product for the CWSUs showing METARs around the country; a satellite image; a convective probability chart; and a group chat for CWSUs where meteorologists can preview sigmets about to be released, correct errors, highlight urgent reports, and discuss tweaks to the outlooks.
“We issue six-hour forecasts for high-level turbulence,” he says. “Flight Level 180 and above. And a forecast for low-level turbulence. Low-level wind shear. Surface winds. Any sizeable area of sustained 30-knot winds at the surface.”
The data come from observations, numerical models, satellites, and radar. “Pireps are very important. But anything we can get our hands on,” Sego says. Model data is verified by experience—a lot of experience. “When you have dealt with a scenario before, you get a feel for what the telltale signs are. If you were to sit down with every person here, you would see different things up on their screens. There is no single template we use. We all know it’s the little details that matter.”
Lead Meteorologist Jesse Sparks works the Traffic Flow Manage-ment Convective Forecast Desk, or TCF. “The TCF is a four-, six-, and eight-hour forecast for convection across the whole country. It’s used by the FAA to plan all the day’s flight routes. You want to put flights where there’s no convection,” he explains. “And you want to plan that before you even get flights in the air.”
“If I forecast an area of moderate turbulence and I’m not getting any reports, is it because I over-forecasted that area and everyone has smooth rides and no one cares, or are they actually avoiding that area and the forecast is legit?”What does planning mean? If the weather in Chicago, for example, is forecast to develop strong storms to the south, arriving and departing traffic is routed to the north. Flights from the west coast heading toward Detroit or Pittsburgh might be shifted north or south, or held at a higher altitude longer to avoid the turbulence. Based on the work in Kansas City, the FAA may change what is called the rate—the number of airplanes that arrive or depart from an airport. Eighty or 90 or 100 airplanes an hour may show up at a major airport during a morning or afternoon push. Maybe the FAA will put in a ground delay program, or alter the distance between aircraft, which is called miles in trail. This helps the controller de-conflict airplanes and also get around the weather.
In other words, a decision made by the FAA in Virginia, about an airplane flying from Seattle to Houston, may begin in Kansas City.
“For example,” Sparks says, “we have a big system coming out of the west today.” He points at a monitor showing water vapor and high-level wind fields, showing a wave that goes from Idaho down through Colorado and into Arizona and New Mexico. “It’s a pretty deep trough axis for May,” he says. “You usually don’t see something this deep coming out in May. This is more like a January or February type of feature.” He points at another lobe that goes from Portland to Salt Lake City. “A lot of strong dynamics come out of these, and that’s why we’re expecting lots of severe weather in the Plains the next few days.”
At the international desk for the Southern Hemisphere, Aviation Meteor-ologist Amanda Martin is working on high-level significant weather charts, drawing out the speeds and directions and flight levels of the jet streams. She’s just finished her turbulence forecast and is about to begin a forecast for convection. She’s also monitoring three volcanoes. “That’s actually very, very low,” she says. “There are usually quite a few more.” In a short while she’ll participate in an online chat with her colleagues in Brazil, New Zealand, Australia, Taiwan, Canada, and England, just to make sure everyone is thinking the same way.
At nearly every desk, one monitor shows the airspace over the continental United States and the North Atlantic Ocean. Over the ocean, with a little flag for every pilot report, it looks like a college party, everyone talking at once. There are only a handful of pireps over the Pacific, and almost none over the lower 48.
“That’s how we get visibility on what’s going on,” Leffler says. “Which is why pireps are absolutely important to us. I can’t know for sure what’s going on over in North Dakota unless someone tells me. I can see it on the radar and satellite. I can use interpretive skills and things like that. But it’s really nice to verify. In the winter, let’s say I’m forecasting moderate icing and planes are shooting through there and reporting trace, trace, nil. I’m thinking something’s going on there. I need to check into that. The pireps are a validation of our forecast—in the same sense that no pireps are also important. If I forecast an area of moderate turbulence and I’m not getting any reports, is it because I over-forecasted that area and everyone has smooth rides and no one cares, or are they actually avoiding that area and the forecast is legit?”
It seems like most of the work in the center is for commercial and private jets, the pilots hanging out in Class A. If I’ve just left the Fargo Jet Center, I say, and am cruising around at 5,000 feet in a 172, do you want to hear from me?
“I come from Alaska,” Sego says. “I’m conditioned to think about impacts on GA. It’s easy to think we’re all about higher flight levels. But that’s not true. It’s a big deal if you’re getting thrashed around in a mountain wave, for example. The models don’t show the threat from terrain. Terrain influences are huge. That’s where reports come in. That’s where experience comes in.”
“Oh, absolutely,” Leffler adds. “Absolutely. The pilot is the eyes-forward for the meteorologist. Absolutely.”
What do you do, I ask, with the pilot whose pirep is exaggerated, when light to moderate turbulence is the worst air he or she has ever flown through?
“We understand that,” he says. “That’s part of reading pireps. And we back that up with surrounding reports. Sometimes, though, a pilot hits the jackpot and gets thrown around when no one else near him does. It can be very localized. The objectivity, based on the pilot’s experience, is just a part of the pireps.
“We’re kind of at a disadvantage in aviation,” he says. “In the same way that a forecast office can see the rotation of a radar-indicated tornado, I can’t normally get the same level of verification.” There are signatures, like wave clouds, but the meteorologists need pireps to validate some forecasts.
“But not all pireps are equal,” he says. “A pirep from a small, single-engine flier is going to be a lot different than an Airbus. How do you come to terms with multiple airframes that are flying out of the same airport but reporting different conditions? What is the weather for that airport? That’s something that we kind of constantly have to come to terms with. We only have one airmet now. Is that airmet for the larger aircraft, the smaller aircraft, all aircraft? Right now we treat it as all aircraft. But in reality, in winter, with gusty winds, you could have moderate to severe turbulence for the small aircraft.”
Leffler pauses. “As a forecaster, our job is to support those who fly. Flight safety is everything we do here. And how that message is consumed is really important.”
W. Scott Olsen is an aviation writer and private pilot living in Moorhead, Minnesota, who teaches at Concordia College.