Turbulence, the often violent and irregular motion of air currents, is an unavoidable flight hazard. But the day may soon arrive when turbulence no longer takes air crews by surprise. And this will help prevent potentially harmful jolts to flight attendants and passengers.
Avionics developers, researchers, airlines and government agencies hope to make flight safer by increasing warning time. However, this won’t be a smooth undertaking; turbulence is difficult to predict, and its causes are varied.
Where It Comes From
Turbulence often is associated with thunderstorms. Weather radar detects the intensity of rainfall. But estimates of turbulence can be difficult, and displays can be bedeviled by false returns from ground clutter.
Turbulence also arises from disturbances in air flow caused by mountains and the jet stream. When these disturbances arise in dry, clear air, causing clear air turbulence (CAT), radar is useless. But some industry experts think that as much as 65% of what is considered CAT today is associated with weather, according to Steve Paramore, radar marketing manager with Rockwell Collins’ Air Transport Systems. Radar often can’t see the upper third of a thunderstorm, but blundering into it can give planes a bad shake.
Weather/windshear radar is being enhanced to more accurately detect storm-related turbulence, including very low moisture activity. Avionics developers are advancing light-detection and ranging radar (lidar) and passive infrared (IR) approaches to true clear air turbulence. But these new systems will take time to mature and must prove themselves worth the cost.
Larry Cornman, a scientist with the National Center for Atmospheric Research (NCAR), a federally funded research and development (R&D) center in Boulder, Colo., defines turbulence as the "natural disordered state of the atmosphere." Although the causes are complex, a simple analogy is two garden hoses pointing streams of water end-to-end or "jets of air blowing by each other, interacting and creating a lot of changes in wind speed and direction." At 35,000 feet, cross-cutting wind currents can stretch over "tens and twenties of kilometers," breaking down into eddies 100 meters (328 feet) in size and affecting aircraft.
Wandering into regions of glaciated ice crystals found above the rainy portions of thunderstorms–and not often seen by today’s radars–risks encounters with severe turbulence.
In one encounter with severe turbulence over Greenland, "We lost several hundred feet," says Bob Massey, director of aviation weather programs for the Air Line Pilots Association (ALPA) and a pilot with a major U.S. carrier. "It was sudden. There was no warning. There were no Pireps [pilot reports] out and no forecasts for it." Severe turbulence generally means loss of control of the aircraft, maybe for a few seconds," he says. "Objects can go flying, if [they are] not secured."
Besides checking the weather radar–turbulence mode–pilots can look at temperature gauges. Changes in temperature indicate changes in air mass, "but it’s not foolproof," Massey says. En route radio contacts with air traffic control and airline dispatchers regarding "ride reports" from previous flights and weather conditions, respectively, also help.
There isn’t an easy answer to a problem brought about by various causes. Therefore, industry and government experts are mulling a multiprong approach. This would include more accurate forecasting, more sensitive sensors, improved cockpit/flight attendant communications, cabin furnishings (such as handholds), and training.
Delta Air Lines now provides pilots on North Atlantic routes with unique, color-coded charts predicting the location of clear air turbulence, says Capt. Jim Johnson, a Delta pilot and manager of applied meteorology. The charts, which have been used operationally for more than a year, are based on manual analysis of satellite weather data, water vapor data, pilot reports and computer model output. Eighty-two percent of the crews who have used the turbulence charts have found them accurate, according to an airline survey.
In a given 24-hour period, during the summer months, more than 1,000 aircraft cross the Atlantic, divided into two sets of five tracks, Johnson says. Although pilots can’t change tracks except in emergencies, the charts help Delta’s international dispatchers to plan for turbulence and at least predict where pilots may run into it. Beginning in 2002, Delta hopes to make similar charts available for crews and dispatchers on Pacific routes. The charts indicate predicted turbulence levels by color: green (light), yellow (moderate), red (severe) and black (extreme).
United Airlines, working closely with NCAR, has loaded NCAR turbulence measurement and reporting software algorithms into the on-board computers of 82 Boeing 737s and 757s, says Carl Knable, meteorology manager with the airline. The planes use existing sensors and avionics to collect data and report it at one-minute intervals. From the data is produced information known as the eddy dissipation rate (EDR), a measurement of turbulence intensity. EDR reports are data-linked to the ground and distributed to United, the National Weather Service (NWS), and NCAR for storage, archiving and analysis. Although the effort is still in the test mode, United plans to load another 118 aircraft with EDR software within the next six to 12 months, Knable says.
This turbulence-predicting capability, of course, will not benefit United exclusively. Delta may add this capability to as many as 200 aircraft in a year or two, Cornman adds. American, Northwest and other carriers also may join the group.
"The real push is to automate turbulence reporting," Knable says. If 1,000 aircraft can be equipped to report EDR data in the next three to four years, he says, researchers believe it’s possible to develop an approximately 80% accurate model for predicting clear air turbulence. That’s about twice as accurate as is available from the best automated models today, Knable estimates. NCAR is working on turbulence measurement algorithms for Boeing 767s, 777s and Airbus aircraft, as well, he says.
The hope is to use EDR, pilot reports, atmospheric modeling, and other data to improve turbulence forecast algorithms, so that weather reports can be transmitted to aircraft several hours away from the disturbance. A thousand aircraft reporting EDR measurements every minute would yield vast amounts of objective data and substantial coverage of U.S. airspace.
EDR reports also could be uplinked directly to pilots as they get closer to the trouble spot. A report’s location could trigger a retransmission of the data to other aircraft in the area, Knable says. The process could use the ARINC developed and operated ACARS (Aircraft Communications Addressing and Reporting System) communications infrastructure already employed to transmit wind and temperature data today. United already is using ACARS for EDR.
Current weather radar is limited in turbulence detection. It requires moisture to operate, making it impossible to see true clear air turbulence. Beyond 40 nautical miles (nm), radar provides little more than general weather information because its beam is so broad. And ground clutter can make the radar think it sees turbulence where there is none.
Refinements to current weather/windshear radar will help to improve storm detection and avoidance through automatic tilt control and advanced ground clutter rejection. Processing techniques also can help detect turbulence in lower-moisture conditions.
"Radar essentially sees only the bottom two-thirds of a thunderstorm," says Collins’ Paramore. The company is working on technology to better detect and avoid storm-related turbulence and to estimate a storm’s growth rate, direction and point of intersection with an airplane’s flight path.
Honeywell likewise is developing algorithms for enhanced turbulence detection. For CAT, Honeywell and Collins are pursuing lidar- and passive IR-based approaches, respectively.
Collins’ forthcoming MultiScan radar will automatically scan at three tilt settings, keeping the radar focused on weather, while electronically removing ground clutter. "A fairly low tilt setting skims the earth’s surface and looks out over the radar horizon," to get weather out to 320 nm, Paramore says. (Few pilots today use radar for weather beyond 160 nm, he adds.) A higher setting aids detection of short- and intermediate-range weather. And an "aggressive down tilt" looks for weather below the aircraft that is growing into the flight path.
Collins will offer turbulence detection improvement in two phases. The first will "compensate for vertical shear as a result of horizontal winds to yield a more accurate magenta turbulence [display] representation," operable to 40 nm, Paramore says. The next step, known as "enhanced turbulence," is a similar concept to windshear, he says. It will detect turbulence using very small amounts of moisture. Collins hopes to provide advisories at 40 nm and warnings at 20 nm.
Another feature of Collins radar is "OverFlight protection," the ability to better detect storm-activity occurring below the airplane and assess its growth rate. In the first phase of OverFlight protection, the radar will combine scans at two tilt angles, effectively broadening the beam to hold a storm within the beam longer than is possible today. The result will be better turbulence avoidance. The second phase will add a more downward-tilting, Doppler-pulsed beam, as well as storm-modeling algorithms to estimate the point where a storm will interfere with the flight path.
MultiScan will be available for new aircraft, and customers using Collins’ windshear radar can upgrade with no wiring changes. The company plans to offer initial capability to customers by early 2002 and complete capability by 2003. Customers would need to upgrade software and hardware in the receiver/transmitter and pedestal and add a new control panel.
Honeywell also is upgrading its weather/windshear radar to enhance turbulence detection and provide overflight protection. Enhancements will make it possible for an aircraft to detect about 60% of turbulence that could affect it, estimates Gordon Carter, Honeywell’s director of radar business development.
Honeywell already has certified its auto-tilt feature, which United Airlines uses today, says Steve Hammack, manager of technical marketing for Honeywell’s radar sector. The feature lessens the pilot’s workload and keeps the radar focused on weather, minimizing spurious ground returns, he says. Honeywell controls tilt settings based on inputs from the terrain database of the enhanced ground proximity warning system (EGPWS), as well as altitude, selected range and radar mode. Other than changing the control panel, this is a software upgrade, Hammack says.
For enhanced turbulence detection, Honeywell plans to apply its windshear radar processing power–17 digital signal processors–at altitudes where the windshear radar is not required. Enhanced turbulence detection likewise is essentially a software upgrade, Honeywell says. The company hopes to offer the feature by July 2002.
Enhanced turbulence detection may make it possible to indicate the level of turbulence–whether moderate or severe–Hammack says. Avionics designers also want to provide a turbulence icon on the screen with range and bearing information. Right now turbulence is alerted via magenta-colored dots on the display, which the pilot has to interpret.
Honeywell also plans a lidar-based system to detect clear air turbulence, measuring the motion of dust particles in the wind. This active, Doppler-pulsed technology could bring the overall detection rate to as high as 95% of turbulence incidents, Carter says. Laser radar technology is mature enough, but companies have to convince airline customers of its value and affordability, according to NCAR’s Cornman. Honeywell will work to decrease cost, size and weight
There is also a question as to the concentration of dust particles at cruise altitudes, which, if too sparse, could adversely affect performance. Although lidar measures turbulence only in the direction of the beam, rather than vertically, it does not require moisture.
Honeywell prefers lidar to passive infrared. "We feel that direct measurement is really best," Carter says. "You’re not inferring anything." The company hopes to have a system with a 10-nm range for airline customers in 2004 or 2005.
Collins is studying passive IR but won’t make a decision on its viability until the end of the summer, Paramore says. If it’s "a go," the company hopes to have a system by 2003. IR appeals because it promises a wider view and information in the vertical as well as horizontal plane, he says. Collins also has developed an IR ranging technique. Like lidar, passive IR will take up real estate and add cost. And passive IR infers turbulence from temperature data. Thus, there remains "the question [of] how to relate turbulence to temperature fields," Cornman says.
CAT Injuries and Costs
Turbulence is the leading cause of non-fatal flight injuries. Almost 60 passengers are injured each year, according to U.S. Federal Aviation Administration (FAA).
NASA’s Aviation Safety program estimates that airlines encounter severe turbulence nine times a month, resulting in an average 24 injuries per month, says Capt. Jim Johnson, manager of applied meteorology for Delta Air Lines. Most injuries are to flight attendants, he estimates. NASA estimates the costs to the airlines from turbulence run more than $100 million a year, Johnson says. After a bout of severe turbulence–often involving a pilot’s momentary loss of control–an aircraft must be taken out of service and inspected. Costs include aircraft downtime, medical bills, flight attendant time off, and passenger injuries and lawsuits.
The aviation industry and government regulators are tackling the problem as part of the Safer Skies initiative. A Joint Safety Implementation Team (JSIT) on turbulence is considering improvements in the cockpit, cabin and in turbulence prediction as part of the process.
"The JSIT is processing many recommendations," including preflight planning and forecasts, warning and response, communications and procedures, and training, says Rick Heuwinkel, manager of aviation weather, policy and external affairs with the FAA’s Liaison Division.
"Recommendations for intervention will be based on how many injuries would be reduced," he says, although the group does "back of the envelope" economic analysis.
Each intervention would have a different timeline, depending on its complexity. Some, such as changing guidance on training, "could be fairly rapid." Heuwinkel hopes to have them all in place by 2007.