InfoScan

Cracking the Microburst Code confronts a mysterious threat to flight safety.

BY RICK DARBY

BOOKS “From 1964 to 1985, a number of microburst-related plane crashes in the United From ‘Adverse Winds’ to Microbursts States killed hundreds of people at a time,” he Warnings: The True Story of How Science Tamed the Weather says. “Today, this type of fatal airline accident Smith, Mike. Austin, , U.S.: Greenleaf Book Group Press, 2010. has practically been eliminated. From 1986 to 304 pp. Photographs. 2008, the number of microburst fatalities in the ark Twain’s quip that everyone talks United States was 37, a decrease of 93 percent, about the weather but no one does in spite of a near doubling of airline flights Manything about it has been made obso- during this period.” lete by today’s meteorology, to aviation safety’s But the path to prediction and avoidance of great benefit. microbursts was anything but smooth. Mike Smith, himself a meteorologist, has Smith credits meteorologist Ted Fujita been involved for most of his career with with pioneering many of the methods used measuring and forecasting extreme weather to today in storm analysis and prediction. enhance safety. Despite minuscule financing “Without Fujita’s techniques, increasing compared with that for cancer research, heart our knowledge of and similar disease research and traffic safety innovations, small-size meteorological events during the meteorology has resulted in “a far more im- 1950s, 1960s, 1970s and 1980s would have pressive reduction of deaths,” Smith says. been nearly impossible,” Smith says. “He had The annual death rate from tornadoes in a creative perspective and a mind that viewed the United States has decreased from three the world in four dimensions, the north/ per million people in the 1920s to 0.068 per south dimension, the east/west dimension, million in 2006 through 2009, he says. Weather vertical dimension (altitude) and time. … To science’s influence on aviation safety is also perform his analysis, Fujita almost single- significant. handedly created the art of meteorological

52 | flight safety foundation | AEROSafetyWorld | March 2011 InfoScan

photogrammetry [measuring by photogra- reach speeds of 70 mph [113 kph] or more. A phy]. He triangulated the photographs from pilot flying through the sinking air, with its different locations in order to track the evolu- rapid change in wind speeds and directions, tion of the storm’s features.” would be severely challenged to keep control Smith cites, as an example of the prevailing of the plane.” attitude in aviation at the time, the accident in- Other meteorologists were unconvinced or volving Eastern Air Lines [EA] Flight 66 at John expressed outright disbelief. No one had seen F. Kennedy International Airport (JFK), New or recorded such a . Fujita perse- York, in June 1975: vered in his research and identified “a smaller, “Reports to the JFK control tower from more intense form of downburst he named a an aircraft awaiting takeoff that it was be- ‘microburst.’ Yet even as Fujita’s body of evi- ing buffeted by the storm’s high winds were dence grew, many in both the meteorological disregarded by the air traffic controllers and, and aviation communities remained deeply ‘Yet even as Fujita’s thus, not relayed to the aircraft on approach. skeptical.” Another airplane landing ahead of EA 66 Smith himself, along with a companion body of evidence barely avoided crashing. The flight crew of in “storm chasing,” helped provide additional grew, many in both the EA 66 knew there was bad weather ahead — evidence in the form of what he describes as it was visible on their on-board radar — but “the first microburst ever photographed,” near meteorological and pressed on anyway. One hundred and twelve Wichita, Kansas, U.S. The photo is included in people were killed.” the book. aviation communities The U.S. National Transportation Safety “ were further confirmed by remained deeply Board (NTSB) determined the probable cause Project NIMROD (Northern Meteoro- to be “the aircraft’s encounter with adverse logical Research on Downbursts), conducted in skeptical.’ winds associated with a very strong thunder- the area around O’Hare International storm located astride the ILS [instrument land- Airport,” Smith says. ing system] localizer course, which resulted The Flight 66 accident helped motivate the in a high descent rate into the non-frangible U.S. Federal Aviation Administration (FAA) approach light towers.” Smith comments, “To to begin installing the low level most people in aviation, the NTSB is the final alert system (LLWAS) at U.S. airports. word, and the NTSB believed the cause of the The issue of downbursts and microbursts accident to be ‘adverse winds.’ And it was — received renewed attention 10 years later with sort of.” the fatal accident involving Delta Air Lines But the probable cause was not the sort of Flight 191 at Dallas/Fort Worth International adverse winds the NTSB had in mind, Smith Airport in August 1985. says. Fujita, called in to investigate the Flight Flight 191, a Lockheed L-1011, was on 66 crash, “conducted a detailed study of the final approach. The cockpit voice recorder 11 aircraft that landed safely ahead of EA 66. recorded the first officer, the pilot flying, He studied the weather, the radar and flight saying, “Stuff is moving in.” The captain, the paths, and he talked with the surviving crews.” monitoring pilot, radioed to the tower, “Delta That led to a new and unorthodox theory, One-Ninety-One heavy, out here in the rain. described in a 1977 paper by Fujita and Hor- Feels good.” ace Byers, describing a previously unknown Smith says, “Now under the base, the weather phenomenon they called a downburst crew could see what appeared to be a light rain — “a rapidly sinking column of air that origi- shower between them and the runway. Other nated in a and then spread out, planes were flying through the shower and and accelerated when it reached the ground,” landing normally. But once Delta 191 entered Smith says. “As the air spread out, it could the rain shower, all hell broke loose.” Soon the www.flightsafety.org | AEROSafetyWorld | March 2011 | 53 InfoScan

ground-proximity warning system was generat- of the microburst, the wind direction shifted ing its “whoop whoop pull up” automated voice from out of the south (a headwind) to the north message. The captain called for a go-around, but (a tailwind), causing an instant drop in airspeed it was too late to avert the accident. The air- and even more sink.” craft touched down a mile short of the runway, The book devotes a chapter about the bounced, touched down three more times, skid- controversy — including a lawsuit brought by ded, and struck a large water storage tank. The the captain’s widow against the FAA and the jet fuel ignited. (NWS) — to why “The microburst, a phenomenon that many Flight 191 was flying an approach in a thunder- meteorologists said did not exist, had claimed storm at all when “just about everyone on the another commercial jetliner and 137 lives,” east side of the airport — traffic controllers, says Smith. He notes that airport weather pilots preparing to take off, the crew of Delta instruments measured winds gusting to 100 191 and the airport weather observer — had all mph (161 kph) at the eastern runways, while seen the storm.” the west side runways were dry, with partial The NWS was legally charged with the sunshine. responsibility for providing weather information Once again, Fujita was invited to investigate to the FAA, which in turn passed it on to con- the accident. Smith says, “Fujita himself flew trollers and pilots. Smith says that the hand-off over the airport, surveyed the on-site instru- worked better in theory than in practice at the mentation from a cherry picker [boom lift], time. photographed the exact location of the an- National weather radar charts were sent emometer [wind speed sensor] and wind vane, from Kansas City, Missouri, but were not re- and collected every scrap of data he could. With ceived by airports or air traffic controllers until Fujita, one never knew which type of data might nearly an hour after the radar observations had turn out to be crucial. been made, a delay that can be like a cen- ‘Almost instantly, “In addition to the data collected by the tury for aviation purposes. Local NWS radar instrumentation at the airport, Fujita collected facilities were often located not only outside the plane went from weather satellite imagery, radar data, eyewit- airports, but outside the cities they served, to ness reports, and data from the cockpit voice provide better advance warning of approaching being too high to recorder and flight data recorder. Once he storms and reduce clutter on the radar screens. nose-diving toward had all of the data, he began to weave it into a The NWS radar for the Dallas–Fort Worth area coherent picture.” was in Stephenville, about 50 mi (80 km) from the ground.’ As Smith describes it, the picture looked like the airport. this: “The data from that radar [were] fed to two “As the L-1011 neared the north end of the NWS facilities in Fort Worth: the Fort Worth runway, it gradually slowed and descended forecast office, located in the federal build- along the glideslope. When the plane initially ing downtown, and the NWS Center Weather encountered the microburst, the leading edge Service Unit (CWSU) inside the FAA’s air route of the wind struck the aircraft. … Delta 191 traffic control center near the Dallas/Fort encountered high winds and rising air. This Worth airport,” Smith says. increased the speed of the aircraft and lifted it Among other problems, “the NWS radar above the descending trajectory it was supposed technician in Stephenville was at dinner, away to follow. from the radar console, when the microburst- “This is where the insidious nature of the producing thunderstorm developed just north microburst presents itself. Almost instantly, the of Dallas/Fort Worth International [Airport]. plane went from being too high to nose-diving Right after he finished eating, he helped toward the ground. As it reached the south half launch the evening weather balloon, leaving

54 | flight safety foundation | AEROSafetyWorld | March 2011 InfoScan

the radar still unmanned. He did not return the hundreds of millions of dollars of aircraft to the radar until 6:00 p.m. At 6:04, two losses prevented.” minutes before the crash, he telephoned the Fort Worth downtown office to inform it of REPORTS the storm near the airport. The 6:04 p.m. call came too late to allow a warning to have been Look Away issued because … it took the NWS office six Laser Hazards in Navigable Airspace to 10 minutes to prepare an aviation weather U.S. Federal Aviation Administration (FAA) Civil Aerospace Medical warning for the Dallas/Fort Worth control Institute (CAMI). AM-400-10/3. Available from CAMI, Shipping t ow e r.” Clerk, AAM-400, P.O. Box 25082, Oklahoma City, OK 73125 U.S.A. Also available via the Internet at . 2010. to instantly convey a threat directly to aircraft,” Smith says. “This faulty system is one that con- he FAA has recorded more than 3,000 tinues, to some extent, even today.” reports of aircraft targeted by laser beams. But progress has been made in mitigating TWhile aeromedical researchers work to the danger from microbursts and wind shear. understand the physiological aspects of laser “The NWS, recognizing that the split radar/ beams on vision (ASW, 12/10–1/11, p. 50), warning responsibility contributed to some CAMI has issued this brochure to explain the of the worst failures in its history, changed its nature of the threat to pilots and recommend entire forecast structure in the 1990s so that actions to minimize the effects of a laser beam radar data and warning responsibility were striking the cockpit. co-located 100 percent of the time,” Smith says. Among the recommendations are these: “Much follow-on work based around Fujita’s • “When operating in a known or suspected research began in order to train pilots how to laser environment, the non-flying pilot avoid microbursts, and if they were to inadver- should be prepared to take control of the tently fly into microburst wind shear, how to aircraft”; escape it (if possible).” The FAA contracted with an industry • “Check aircraft configuration and (if avail- advisory group to create the Windshear Train- able) consider engaging the autopilot to ing Aid and established wind shear training maintain the established flight path”; requirements for U.S. Federal Aviation Regula- tions Part 121 and Part 135 operators. • “Use the fuselage of the aircraft to block The last crash attributed to a microburst the laser beam by climbing or turning was U.S. Airways Flight 1016 at Charlotte, away”; North Carolina, U.S., in July 1994. Smith • “Inform Air Traffic Control of the situ- says, “Given the ever-increasing number of ation. Include location/direction of the people and planes in the air, the number of beam, your present location, altitude, etc.”; lives saved due to Fujita’s pioneering research and, that eventually led to implementation of microburst avoidance procedures in the Unit- • “Turn up the cockpit lights to minimize ed States is well over 2,000, not to mention any further illumination effects.” 

www.flightsafety.org | AEROSafetyWorld | March 2011 | 55