On the afternoon of Wednesday, February 23, 2011, 24 hours before the Shuttle Discovery was set to launch, low clouds hid the top of the 525-foot-high Vehicle Assembly Building at the Kennedy Space Center (KSC) in Florida. For regular space shuttle watchers, the clouds were a sober reminder that even relatively mild weather can delay the launch of a shuttle. Airliners and other aircraft taking off or landing routinely pass through clouds as low as those at the KSC that afternoon, and even storms do not deter some airliner landings. But such weather can result in the cancellation of a space shuttle launch, which must adhere to rules far stricter than those that would have kept a cautious 1920s open-cockpit, biplane pilot on the ground.
Caption: The Shuttle Discovery 18 seconds after launch as seen from the media area at the Kennedy Spacecraft Center. It is now 4,093 feet high traveling at approximately 521 mph.
For example, shuttle launch rules require that if the sky is more than 50 percent covered in clouds, the cover must be at least 5,000 feet above the ground, and the visibility must be at least four statute miles. While these rules might seem extreme, there is, in fact, solid reasoning behind the restrictions; in case of an emergency immediately after lift off, the shuttle needs to be able to land back at the KSC without being inhibited by cloud cover. Given the inherent risks involved in space flight, there are myriad rules like this that govern how the weather must be before a launch can take place, and so the process involved in making the “go” or “no-go” decision for a shuttle launch is intricately tied to what can only be described as “extreme forecasting.”
Caption: Discovery astronauts pose before boarding van to the launch pad.
Despite the clouds 24 hours before launch, the weather was looking good for launch day. At a press conference a few hours earlier, NASA's mission management team announced its unanimous approval for launching Discovery between 4:50 and 5:00 p.m. EST the next day. Kathy Winters, the launch weather officer, said the odds were still 80 percent that launch time weather would be favorable. The only “slight” issue would be a localized offshore shower in the late afternoon.
On launch day, only a few clouds dotted the blue sky as thousands of people crowded into the area around the KSC to view the launch. Two hours before launch time, Lt. Col. Patrick Barrett from the Air Force's 45th Space Wing Weather Squadron pointed to a weather radar display at the KSC media center showing a few showers—as predicted the day before—over the Atlantic east of the Space Center.
However, if one of those showers happened to move over the launch pad between 4:50 and 5:00 p.m., the launch would be canceled. Discovery was hauling supplies and equipment to the International Space Station, and in order to link up in orbit, it had to blast off during that 10-minute period. If it didn't, the launch would have to wait for a similar, exact 10-minute period with suitable weather the next day.
“The clouds coming ashore are dissipating. This is a blessing,” Barrett said. The showers were developing over the Gulf Stream's warm water offshore, but they were falling apart as they drifted over the water close to the coast and over land, both of which were still relatively cool. “In the summer the clouds coming ashore would balloon when they reached the warm land,” Barrett said.
The Air Force's Weather Squadron based at Patrick Air Force Base near the KSC and the National Weather Service's Spaceflight Meteorology Group (SMG) at the Johnson Space Center in Houston, Texas, each provide shuttle forecasts.
The Air Force handles forecasts for the shuttle when it's on the ground. This includes moving the shuttle from the Vehicle Assembly Building to the launch pad, beginning to load the external fuel tank, and weather at the launch pad. The Air Force also forecasts wind, weather, and waves for the two ships that recover the shuttle's solid rocket boosters from the Atlantic Ocean. The SMG in Houston handles all forecasts for shuttle landings, including a possible emergency landing back at KSC immediately after launch.
SMG's support starts several days before a launch with forecasts for the launch abort landing sites in the U.S. and Europe. SMG continues forecasting after the shuttle lifts off until it lands at the KSC, the preferred landing site, or one of the two backup sites at Edwards Air Force Base, California, or White Sands, New Mexico.
The Air Force directly supports the launch director and Launch Control Team at KSC, while the SMG directly supports the flight director and Flight Control Team at the Johnson Space Center in Houston. Shuttle command and control switches to the Mission Control Center in Houston when the shuttle clears the launch pad.
“Before either group issues a forecast product, our lead forecaster talks with the Launch Weather Officer,” says Frank Brody, the SMG's chief. “Coordination with the Launch Weather Officer is routine. SMG and the LWO talk on the phone at prescribed times to coordinate the portions of the weather outlook that overlap on launch day. Coordination can of course occur any time when the weather or situation dictates. We are able to talk with the LWO directly via phone and over the dedicated NASA voice circuits”
Caption: Lt. Col. Patrick Barrett from the Air Force's 45th Space Wing Weather Squadron, in the KSC media building a few hours before Discovery's launch.
Unlike most meteorologists, who never meet the people using their forecasts, shuttle forecasters work directly with NASA officials and engineers on various aspects of shuttle operations. One example is the NASA engineering team that calculates what is known as the I-Load update before launch. Doris Hood, an SMG upper-air forecaster, said that for Discovery's launch, the engineering team used weather balloon observations five hours before launch to “generate steering commands for the Space Shuttle” that keep it “within the structural constraint envelope and maximizes the payload lift capability.”
Upper-air wind forecasts, such as those Hood prepared for the launch, give the I-Load team “an indication of the wind issues of the day and also how much the winds will be changing during the hours before launch,” Hood said.
Shuttle forecasters have another weather data source: reports from an astronaut weather reconnaissance pilot. For Discovery's February 24 launch, this was Lee Archambault, who had made two flights prior to the shuttle lift-off time. His job was to give a pilot's-eye view of the current weather.
Caption: Astronaut Lee Archambault in the shuttle cockpit mock up at the Johnson Space Center in Houston.
Archambault took off in one of NASA's T-38 jets about three hours before the scheduled launch to check the winds and weather around the KSC. After landing the T-38, he took off an hour and 45 minutes before the launch in the Shuttle Training Aircraft to fly approaches to both ends of the runway in order to experience what Discovery's pilots would find if they had to make an emergency landing at KSC. The Shuttle Training Aircraft is a Gulfstream II business jet that's been modified to fly like the shuttle on an approach from approximately 35,000 feet down to a landing.
Archambault reported that conditions favored an emergency landing toward the southeast at the Shuttle Landing Facility, which is the 15,000-foot-long runway built for shuttle landings. Archambault said that whether it's an emergency landing or a normal end-of-mission landing, shuttle pilots need the required ceilings and visibility so “we can land even if our navigation system is a little bit off when we visually see the runway. We only have one shot at this, so we have to be certain” His “one shot to land” refers to the fact that the orbiter is a heavy and fast glider. If the shuttle orbiter isn't lined up exactly right for a landing, the pilots can't apply engine power and take off for a go-around to try another landing, as pilots of ordinary airplanes do.
On February 24, the clouds stayed offshore all afternoon, no mechanical or electrical problems cropped up, and a last-minute computer glitch was resolved in time for Discovery to lift off at 4:53 p.m. Only four minutes, 10 seconds later, it left the weather behind—until time to land—as it entered space at 328,100 feet (100 kilometers) above the Atlantic Ocean and 143 miles from the launch pad.
Tim Oram, a lead forecaster at the SMG, said, “We can consider 328,100 feet as the top of the atmosphere, because it includes the mesosphere and part of the thermosphere.” Also, this is the altitude that the Fédération Aéronautique Internationale says a person needs to attain to be recognized as having flown in space.
More Weather Restraints
In addition to the strict ceiling and visibility requirements for launch and landing, a shuttle needs to avoid rain because at the shuttle's launch speed, rain drops would be like bullets hitting the thermal tiles that keep the spacecraft from burning up when it re-enters the atmosphere at the end of a flight.
Many other criteria aim to avoid even a small chance that lightning will hit a shuttle. In addition to the obvious restrictions against flying into or near thunderstorms, shuttles are not allowed to fly through clouds that offer even a slight chance of the shuttle triggering a lightning stroke or when the measured electrical fields exceed certain values.
A shuttle, of course, is vulnerable to lightning on the ground as well as during its brief time flying through the atmosphere. Elaborate lightning rod systems protect a shuttle on the launch pad and in the Vehicle Assembly Building (VAB), where the orbiter is mated with the booster rockets and the external fuel tank.
Lightning could target the shuttle on its six-hour trip from the VAB to the launch pad, which is why this journey is usually made late at night or early in the morning when lightning is less likely, and on a night when the Air Force meteorologists forecast that lightning is unlikely.
Col. Barrett noted that one danger of lightning hitting a shuttle after launch would be the external fuel tank exploding. “The debris field would be wide; it could be a civilian disaster.”
Discovery was scheduled to land back at KSC the morning of Wednesday, March 9, but only if the weather forecast met the strict landing requirements. For a landing, the forecast peak head winds can't exceed 25 knots, crosswinds can't exceed 15 knots, and the tailwind can't exceed 15 knots. No thunderstorms, lightning, or precipitation can be within 30 nautical miles of the runway.
The required ceiling and visibility for an end-of-mission landing are 8,000 feet and five miles, respectively, instead of the 5,000 feet and four miles required for a landing immediately after launch. Archambault explained that these are needed because “after 12 or 14 or 15 days in orbit, the biggest problem is the brain and body being able to be reconditioned to gravity. Our reflexes are a little slower, and the approach is very dynamic at steep approach angles.”
A shuttle landing begins in orbit with the pilots turning the orbiter around, so that the rear-facing orbital maneuvering system engines are facing in the direction the orbiter is traveling. These engines are fired in what's called a deorbit burn, slowing the orbiter at a precise time to land at either KSC, Edwards Air Force Base in California, or White Sands, New Mexico. This is done only after the SMG tells the flight director that weather meeting the landing flight rules is forecast for the landing time and place.
Once the engines are fired for the deorbit burn, there's no going back. Slowing the orbiter drops it into a lower orbit, and eventually it begins to enter the atmosphere. Here, air flowing over the shuttle's wings and other surfaces creates a lifting force that keeps the orbiter from dropping to the ground and enables the pilots to maneuver it. But, without any engines, they can't maneuver to land at Edwards if the KSC weather unexpectedly deteriorates.
Caption: Discovery on the launch pad the night before it launched.
Kurt Van Speybroeck was the SMG's lead forecaster for Discovery's landing on Wednesday, March 9, working with Tim Oram and Doris Hood. He and other SMG forecasters had begun closely tracking landing site weather on Sunday and keeping flight controllers and directors informed on how it was developing for the scheduled landing day. On Tuesday, Van Speybroeck briefed the flight director, who would make the landing decision.
On landing day SMG meteorologists began monitoring the weather and updating their forecasts around 3 a.m. EST.
By around 10:30, using the forecasts and weather observations, including those from Astronaut Rick Sturckow who was that day's weather reconnaissance pilot, the SMG told Flight Director Tony Ceccacci that the weather was still a “go” for landing at KSC. Clouds in the area were not a threat, and the wind was from the southeast was at 15 knots, peaking to 20 knots with a headwind straight down the runway.
At 10:42 a.m., Ceccacci radioed Shuttle Commander Steve Lindsey to make the deorbit burn at 10:52 a.m., which he did.
After the deorbit burn, “We started seeing winds briefly peak to 28 knots and settling down to 25 to 26 knots,” Van Speybroeck said. “At moments we were exceeding the head wind limits. But, after the burn you're going to come down and going to land; you have to make the best of it. The flight dynamics officers were plugging in new numbers, making sure not to land short. This caused a little bit of concern”
Caption: Shuttle Discovery is traveling at 224 mph as it touches down to end its last flight at the Kennedy Space Center on March 9, 2011.
At 11:58:04, Discovery's main landing gear touched down at KSC into a direct headwind of 25 knots, ending its last flight into space. The shuttle landed 12 days, 19 hours, 4 minutes, 50 seconds after launch. Except for the 4 minutes and 10 seconds immediately after launch and approximately 25 minutes before landing, Discovery was in space, far above the reach of weather.
If the go-ahead had not been given to land that morning, Discovery could have tried to land at KSC later that day. But, Brody said, “If we had gone to a second opportunity, there's a very strong chance the SMG forecast for KSC would have changed to no-go before the deorbit burn decision.” If that had happened, NASA probably would have decided to land at Edwards Air Force Base the next day. In that case, Discovery “would be sitting out at Edwards costing the taxpayers a million and a half dollars” to fly it back to KSC atop the special Boeing 747 shuttle transporter and then to be prepared for its final flight on the transporter to the Smithsonian Air and Space Museum at Dulles International Airport near Washington, DC. “Of course,” Brody noted “the first priority is the safe return of the shuttle”
“I told Kurt ‘you're never going to forget this one,’” Oram said. “To make that go/no-go decision…it's a team effort. The right decisions come out of that”
JACK WILLIAMS, the author of the AMS Weather Book and editor of the USA TODAY Weather page from 1982 until 2005, is a freelance science writer. He answers weather questions on his Web site, www.weatherjackwilliams.com