Better WXR Training Needed?
Honeywell: Better wx training needed
Radar manufacturers should consider making equipment easier to use and displays easier to interpret, Honeywell safety specialist Dr. Ratan Khatwa told attendees at this year’s Flight Safety Foundation European Aviation Safety Seminar, held in Bucharest.
He added that better weather-radar training can improve pilots’ awareness and decision-making skills and help them avoid penetrating severe meteorological conditions.
NTSB studies show weather to be a factor in about 25 percent of all U.S. flying accidents between 1994 and 2003. Khatwa said initial and recurrent flight crew training should cover fundamental concepts in six areas of weather-radar operation: beam coverage, Earth-curvature effects, antenna stabilization, tilt and gain management, calibrated weather and range. System limitations, such as attenuation and the significance of green radar “echoes” at high altitude, also should be covered.
These recommendations arise from analysis of pilots’ difficulties and concerns uncovered while Honeywell was developing its RDR-4000 radar. Khatwa’s study included a human-factors evaluation of flight-crew radar use, a global survey to assess pilots’ fundamental understanding and perceptions of training, and analyses of weather radar-related incidents and accidents. He concluded pilots did not understand fundamental weather-radar concepts; typical equipment use precluded detection of severe weather; and dedicated training was not standard practice.
Khatwa pointed out that current radars are concerned primarily with weather analysis and avoidance, proper interpretation of which depends on pilots’ adequate understanding.
Honeywell’s RDR-4000 human- factors research showed that almost 70 percent of pilots were dissatisfied with weather-radar training. From the survey, Khatwa concluded most operators do not provide initial or recurrent weather-radar training; most available training takes place on the job; there is little incentive for operators to provide training since regulators do not require it; many pilots do not understand weather radar, including its limitations; fundamental concepts of weather radar are “poorly understood”; and pilots want recurrent training.
To understand pilots’ weather-radar use and any difficulties, Honeywell conducted a comparative evaluation of independent flight-crew groups using current equipment or new radar-display modes. The PC-based exercise observed the behavior of 13 pilots during several scenarios involving weather-radar use. Overall, “significant weather” events were detected on almost 82 percent of occasions, with pilots correctly deciding on action necessary to avoid penetration 70 percent of the time.
Incorrect pilot action involved improper management of weather-radar tilt, gain or range, continued flight toward significant weather and imprudent weather-avoidance decisions. All pilots failed to recognize the vertical position of each of two weather cells. Khatwa emphasized the need for pilots to be cautious about green radar echoes at high altitude, since these indicate “potentially hazardous” conditions.
Analysis of actual flight-crew radar operation and interpretation was ultimately restricted because the study could not include “many other critical factors, such as provision of timely weather information, accuracy of such data, role of ATC and regulator” and other considerations. This part of Khatwa’s study drew on data covering fatal and nonfatal accidents and incidents (fixed- and rotary-wing) that involved global single- and dual-pilot business, public transport and cargo aircraft operations reported by nine different worldwide safety agencies between 1987 and 2007.
Excluding occurrences involving training flights, sabotage, terrorism, military action or insufficient weather-radar information, Honeywell researchers found just 14 relevant events. A quarter of these instances were fatal, and half of the aircraft involved were substantially damaged or destroyed. Some 57 percent of the accidents or incidents took place in instrument meteorological conditions, 50 percent occurred during cruise, the same proportion was in daylight and another 35 percent occurred between top-of-descent and destination.
Switching off the radar, despite forecast weather and prevailing conditions, or pressing on in the face of adverse conditions were cited as examples of poor planning. Other examples of poor decision making included making landing decisions based on the experiences of preceding aircraft that successfully penetrated convective weather; flying through gaps between closely spaced storm cells, rather than around the thunderstorm; and flying close to squall lines.
In addition, the operation of weather radar or radar-display interpretation were “not necessarily optimal” in two thirds of occur-rences, said Khatwa, who cited five problem areas: improper tilt operation or management; improper use of gain control; misinterpretation of ground returns; weather radar “off,” despite known cumulo-nimbus cloud; and insufficient appreciation of radar limitations and their impact on displayed images.
The Honeywell study shows that crew weather-radar training had not been provided in half of the accidents/incidents. Pilots talked about “trial-and-error experience” and “information [obtained] from other pilots,” an approach that Khatwa concluded can “lead to improper radar operating procedures and techniques.”
Honeywell Aerospace Pilot Survey Findings
In conducting a survey about the RDR-4000 weather radar, Honeywell safety specialist Dr. Ratan Khatwa asked more than 50 ATP-rated pilots about their experience with weather radar. The average age of the respondents was 52 years; the average flight time was 12,500 hours. The answers these experienced pilots provided were illuminating.
• 62 percent of the pilots surveyed answered correctly that a straight radar beam is not aligned with an aircraft’s current flight level (because of Earth curvature)
• 15 percent mistakenly thought that antenna down-tilt was required to offset a nose-up pitch angle. (That is offset by antenna stabilization.)
• 63 percent did not appreciate the need for weather-radar antennas to be set to compensate for earth curvature, which blocks weather targets beyond, say, 150 nm ahead for nominal cruise altitudes. “Curvature [effects] become noticeable at ranges above 40 nm, and if ignored can lead to weather-image interpretation errors,” said Khatwa.
• 55 percent of pilots did not realize that a weather target falling inside the radar beam will not necessarily be shown in its true color on the display. “The color selected for display is a direct function of the power returned to the receiver. Where the beam is partially filled, the total power returned may not represent the calibrated value associated with the target cell,” he said.
• Five in every eight pilots incorrectly thought green (short-range) radar targets shown near to cruise levels above FL310 need not be avoided. “Typically, at these altitudes, targets are less reflective. At high altitudes, there is a possibility of unstable air and hail above the storm cell. It is therefore not advisable to penetrate the less-reflective part of the storm top,” Khatwa explained.
• 73 percent of flight crew understood that antenna tilt angle does not need to match a climb (or descent) angle to detect weather on their flight path. “The antenna should be pointed at the base of convective weather during climb. Generally, the lower 18,000 feet is the most reflective part of the storm.” Radar can be used to analyze weather characteristics (such as vertical extent of cells) and to avoid strong convective activity. “Returns along the flight-path angle may not provide full indication of storm intensity and turbulence levels [to be encountered within the cell].”
• Almost 90 percent of pilots did not know the range at which their current weather radar was no longer calibrated and did not show returns at their true levels. Radar beams broaden with distance, so a smaller proportion is filled with moisture. “At shorter ranges, returned power is more representative of the target cell, and it is more likely to be displayed at its true calibrated value. Typically, returns are calibrated within a range of 60 to 80 nm.”