Turbulence/windshear/microburst

The Terminal Area Forecast (TAF) provided information to the crew that moderate turbulence was likely to be encountered during the flight. The meteorology information provided no warning of severe turbulence until after the flight had landed at the Gold Coast. The TAF that the crew had used indicated that the change in wind direction and strength at 1600 would signify the passage of the front over the Gold Coast.

The turbulence encountered during the Dash 8's initial approach and the visual observations reported by the crew of a roll cloud and water spouts, were consistent with the aircraft having encountered the leading edge of the frontal zone. This was about 2 hours earlier than forecast. The indications from the weather stations at Evans Head and Cape Byron confirmed that the front was moving to the north faster than expected. However, the severe turbulence associated with the front could not be determined from those weather stations and was therefore unexpected.

The drop in temperature of 7 degrees provided the pilot in command of the B717 with an indication that conditions at the Gold Coast Airport were changing earlier than forecast. This temperature drop was also recorded by the automatic weather station at the Gold Coast Airport. While the drop in temperature cannot by itself indicate the degree and/or severity of turbulence likely to be encountered, it did indicate the arrival of the front.

As the aviation special weather report issued at 1258 was not passed to the crew of TNW, they were not in a position to appreciate that the passage of the front over the Gold Coast Airport was earlier than expected. The crew were aware that the B717 had encountered turbulence in the area of the Gold Coast Airport, but were not fully aware of the severity of that turbulence.

During the missed approach, the flap 5 limit speed was exceeded for a short period of time. At the time, the crew were likely to have been preoccupied with the low oil pressure warning and maintaining control of the aircraft due to the turbulent conditions. The overspeed did not result in damage to the aircraft. The extreme roll rates encountered by the aircraft could have given the perception that the aircraft rolled to unacceptably high roll angles, however the data recovered from the flight data recorder revealed that the roll angles encountered were within normal operating parameters.

The changes recorded in vertical `g' readings were indicative of a severe turbulence encounter. The decision by the crew to conduct a missed approach due to the turbulence was in accordance with normal operating procedures.

The Australian Transport Safety Bureau did not conduct an on-scene investigation of this occurrence.

FACTUAL INFORMATION

At approximately 1330 Eastern Standard Time (EST) on 8 October 2004, a de Havilland Canada DHC-8-102 (Dash 8) aircraft, registered VH-TNW (TNW), with a crew of three and 18 passengers, encountered severe turbulence during approach to the Gold Coast Airport, Queensland. The aircraft was being operated on a scheduled passenger service from Brisbane to the Gold Coast.

The flight crew reported that conditions were quite rough with moderate turbulence during the flight. Approximately 25 NM from the Gold Coast Airport, at an altitude of 5,000 ft, the aircraft encountered turbulence that resulted in a wing drop, while operating with the autopilot engaged. The crew were then cleared by air traffic control (ATC) to descend to 4,000 ft and to reduce speed, as they were being radar vectored behind a Boeing Company 717 (B717) aircraft that had conducted a missed approach at the Gold Coast Airport due to encountering turbulence.

The crew of TNW reconfigured the aircraft for a flap 15 landing. They were then advised by ATC that they would be in front of the B717, were cleared for a visual approach and advised to contact the control tower. Passing through approximately 2,500 ft, the crew reported that they encountered a severe updraft that stopped the aircraft's descent. They then disconnected the autopilot and continued the descent. The crew reported that as the aircraft was passing through approximately 1,500 ft, they encountered severe turbulence, which required them to apply almost full control inputs to control the aircraft. The pilot in command then carried out a missed approach and the aircraft was reconfigured for a climb. During the missed approach, the number 1 engine `low oil pressure' warning light illuminated for a short time. The crew checked the cockpit indications and confirmed that the oil pressure was within limits.

They climbed the aircraft to 3,000 ft and manoeuvred for another approach. The pilot in command reported that during the downwind leg they observed that the water off the coast appeared to be `foaming'. He also reported a number of small waterspouts present, and that there appeared to be a roll cloud forming ahead of the aircraft.

After landing, the crew advised the ground engineers that they had encountered severe turbulence. Following discussions with the operating company's base, a decision was made to ground the aircraft to allow a thorough engineering inspection. That inspection was to include flap over-speed and severe turbulence encounter inspections.

A review of the recovered data from the aircraft's flight data recorder revealed that, at a recorded altitude of 1,460 ft, the aircraft encountered a turbulence event that recorded +2.26 vertical `g'. At that point, the aircraft was banked to the left to 21 degrees. Almost immediately following this, the aircraft banked to the right to 20 degrees. The data also revealed that during the missed approach the engine torque values exceeded 100% for a period of two seconds and that the flap 5 limit speed of 148 kts was also exceeded.

The aircraft was not damaged during the turbulence encounter and none of the occupants were injured.

The pilot in command of the B717 reported that, when they had started descent, they were advised that the automatic terminal information service (ATIS) had changed. The new information that they received indicated that the temperature had dropped 7 degrees from the temperature that they had recorded just prior to starting the descent. He also reported that the conditions became very turbulent as the aircraft descended below 10,000 ft. While on approach, the conditions became very turbulent, and he conducted a missed approach when the aircraft became unstable on the approach.

The general meteorological forecast for the morning of 8 October 2004 indicated that a trough line was moving across south-east Queensland. Coincident with the movement of this trough, a front was moving north along the New South Wales coast and was expected to move through south-east Queensland in the mid to late afternoon.

The Terminal Area Forecast (TAF) for the Gold Coast Airport, issued at 1104, indicated that the wind was from 300 degrees at 18 kts, with gusts to 30 kts. It further indicated that the wind was to change direction and strength at 1600. The TAF also indicated that moderate turbulence was expected below 5,000 ft from 1100 until 1700.

An amended TAF was issued at 1335, which indicated that the wind was from 160 degrees at 20 kts. This TAF also indicated that moderate turbulence was expected below 5,000 ft from 1300 until 1700.

Data from the automatic weather station (AWS) at the Gold Coast Airport indicated that the wind direction started to change from a north-westerly direction to an easterly and finally a south-easterly direction between 1230 and 1326. The recorded temperature also decreased from 34.6 degrees to 26.8 degrees in the same period. An aviation special weather (SPECI) report related to a change in wind direction and a drop of temperature of more than 5 degrees was issued by the Gold Coast Airport AWS at 1258. That information was not passed to the crew of the TNW.

A significant meteorological (SIGMET) warning of severe turbulence was issued at 1449 that covered the area surrounding the Gold Coast Airport. This turbulence was expected below 8,000 ft.

The passage of the front along the New South Wales coast was detected by automatic weather stations at Evans Head and Cape Byron. These stations did not have the capability to provide 1-minute updates to the data. Therefore, they could only provide aviation routine weather (METAR) and SPECI reports.

SPECI data from Evans Head and Cape Byron stations indicated the passage of the front through those locations at 1025 and 1131 respectively.

1. An intense thunderstorm developed in a short timeframe ahead of the main line of thunderstorms, producing heavy rain, hail and windshear, which the aircraft encountered shortly after take-off.
2. Air traffic control and Bureau of Meteorology staff did not mutually exchange information regarding the thunderstorm as it developed and approached Coolangatta aerodrome.
3. Coolangatta Air Traffic Services staff did not ensure that the crew were aware of the changed weather conditions and the new Automatic Terminal Information Service broadcast advising that thunderstorms were in the area.
4. The crew did not have a complete and timely picture of a hazardous and rapidly deteriorating meteorological situation from which to make an accurate assessment of that situation.

2.1 Introduction

The occurrence involving EAL involved a number of issues including the limitations of airborne weather radar, the mutual exchange of information between BoM and air traffic control, and provision of information to the B767 crew. Further, the occurrence involving EAL displayed a number of similarities with a Boeing 737 microburst encounter at Brisbane Airport on 18 January 2001.

2.2 The aircraft

2.2.1 Aircraft weather radar

The aircraft weather radar did not have the capability to provide predictive forward-looking windshear detection and avoidance information to the crew, nor was there a requirement for it to do so. That capability would have provided an early alert to the crew about the hazardous conditions that existed ahead of the aircraft and may have assisted them to avoid or minimise those hazards.

The presence of red or magenta on the aircraft weather radar display is a measure of rainfall intensity. The crew reported areas of red on the display, with no hooks, fingers, contours, scalloped edges or U-shaped returns that could have indicated the presence of hail or other adverse weather conditions. In addition, they reported that they assessed the areas of red on the weather radar display as heavy rain only. The crew could see heavy rain approaching the aerodrome, but were not aware of any associated adverse weather conditions, such as thunderstorm activity and hail. In the absence of any indications of adverse weather from either the aircraft weather radar or ATS, they would have been unaware of the presence of adverse weather conditions in the take-off flight path.

2.2.2 Aircraft flight path

The crew reported that they increased thrust in accordance with the operator's published windshear escape manoeuvre. However, recorded FDR data indicated that, throughout the windshear encounter, both EPR and thrust lever angle (TLA) remained in the take-off position until climb thrust was set when clear of the encounter. The investigation was unable to resolve the discrepancy.

2.2.3 Other aircraft movements

A number of other aircraft movements into and out of Coolangatta Airport occurred in the short time before EAL took off. None of those aircraft reported encountering adverse weather, including the Boeing 737 that took off from runway 32 about 15 minutes before EAL was issued clearance to take-off from the same runway. The absence of any reports of adverse weather encounters likely contributed to the crew of EAL assessing that the weather conditions that they could see visually, and which were displayed on the aircraft weather radar, consisted of heavy rain only.

2.3 Organisational issues

2.3.1 Bureau of Meteorology

The BoM reported that the thunderstorm encountered by EAL developed in a short timeframe ahead of the main line of thunderstorms. Forecasters became aware of the severity of the thunderstorm 9 minutes before EAL was cleared to take-off. Despite observing the storm cell move towards and pass over Coolangatta aerodrome, forecasters did not contact Coolangatta ATS to advise of the approaching hazardous weather.

2.3.2 Air Traffic Services

ATIS Echo was current until 1337:41. ATIS Foxtrot was issued at 1338:22. The crew requested taxi clearance, notifying receipt of ATIS Echo, 45 seconds after ATIS Foxtrot was issued. Accordingly, the crew had not been advised that thunderstorms were present within 5 NM of Coolangatta Airport.

The SMC controller did not inform the crew that the ATIS had changed and did not advise the crew of the change in weather conditions. Accordingly, an opportunity was missed to provide the crew with updated information of the prevailing meteorological conditions at the time that the aircraft was intending to take-off.

Coolangatta ATS controllers could see that the thunderstorm was approaching and that weather conditions were deteriorating, however, they did not contact BoM staff to ascertain the severity of the approaching weather. Consequently, neither BoM nor Coolangatta controllers had a complete picture of the deteriorating meteorological situation. In turn, the crew of EAL was not provided with a complete picture of the meteorological situation in the vicinity of Coolangatta Airport and their intended departure flight path.

Coolangatta ATS reported that they had access to the BoM PC-based METRAD/RAPIC display. Controllers were aware that information depicted on those displays could be up to 10 minutes behind actual time. Due to the rapid development of the thunderstorm cell encountered by EAL, the investigation was unable to confirm if the METRAD/RAPIC display would have been able to provide Coolangatta controllers with sufficient information to advise the crew of EAL of approaching hazardous weather.

2.4 Previous occurrence

The occurrence involving EAL displayed a number of similarities to a Boeing 737 microburst encounter that occurred at Brisbane Airport on 18 January 2001. The circumstances of the occurrence involving EAL indicate that the ATSB recommendations published in ATSB investigation report BO/200100213 remain valid.

1.1 History of the flight

1.1.1 Overview

On 26 October 2003, at about 1346 Eastern Standard Time, a Boeing 767-238 aircraft, registered VH-EAL, with two pilots, seven cabin crew and 207 passengers, took off from Coolangatta Airport, Queensland, on a scheduled regular public transport service to Sydney, NSW. The aircraft had arrived at Coolangatta earlier that day, having flown the first sector from Sydney to Coolangatta. Shortly after takeoff, passing through an altitude of about 800 ft, the aircraft encountered heavy rain, hail and windshear. The crew reported that they increased thrust in accordance with the operator's published windshear escape manoeuvre. During the windshear encounter, the aircraft descended about 130 ft and a ground proximity warning system (GPWS) Mode 3 aural alert 'DON'T SINK' sounded.

During the subsequent climb, the cabin crew reported to the flight crew that there was damage, in the form of dents, to the leading edges of the wings. After diverting out to sea around the weather, the flight continued to Sydney. The crew configured the aircraft early in the approach to Sydney, in the event of flap and leading-edge device extension difficulties due to the damage, however an uneventful landing was conducted.

1.1.2 Sequence of events

1.2 Injuries to persons

No injuries to persons were reported.

1.3 Damage to aircraft

A post-flight technical examination revealed substantial damage to the leading-edge slats, leading edge wedge panels, horizontal stabiliser, vertical stabiliser, radome, fuselage area above the pilots' windows, nose-cowls of both engines and fan blades on both engines (see Figures 1 and 2).

1.4 Other damage

Nil.

1.5 Personnel information

1.5.1 Pilot in command

Type of licence: Air Transport Pilot (Aeroplane) Licence
Medical certificate: Class 1
Flying experience (total hours): 11,126
Hours on the type: 5,156
Hours in the preceding 30 days: 47

1.5.2 Copilot

Type of licence: Air Transport Pilot (Aeroplane) Licence
Medical certificate: Class 1
Flying experience (total hours): 3,614
Hours on the type: 924
Hours in the preceding 30 days: 52

Both pilots last completed windshear training as part of the operator's recurrent training matrix during the period December 2002 to January 2003.

1.6 Aircraft information

Manufacturer: Boeing Commercial Airplane Group
Model: 767-238
Serial number: 23306
Registration: VH-EAL

1.6.1 Aircraft weather radar information

The aircraft was fitted with a Collins WXR-700X weather radar system, which did not include a windshear warning or a predictive windshear warning function. The weather radar display was superimposed on the aircraft navigation displays and indicated rainfall intensity in different colours, with green depicting light precipitation, yellow medium precipitation and red or magenta heavy precipitation. The radar manufacturer's documentation stated that red was equivalent to a rainfall rate of 12.7 to 50.8 mm per hour, indicating a storm category of 'strong to very strong'.

Specific guidance on the use of the weather radar was provided to crews in the operator's Flying Manual and a training CD-ROM. A document was also provided to crews on the operator's intranet site. Use of the weather radar was covered as part of the operator's recurrent training matrix.

The crew reported that the weather radar was set according to the operator's requirements for takeoff. During taxi, and the time taken to negotiate a departure heading with ATS, the crew scanned the weather ahead of the aircraft flight path. They reported areas of red on the display, with no hooks, fingers, contours, scalloped edges or U-shaped returns that could have indicated the presence of hail. They reported that they consequently assessed the areas of red on the weather radar display as heavy rain only.

1.7 Meteorological information

1.7.1 Prevailing weather conditions during the afternoon of the occurrence

An intense surface trough was moving across the south-east inland of Queensland, towards the south-east corner and was forecast to move off the south coast by about 1800 to 1900. The atmosphere was very unstable ahead of the trough and scattered to widespread showers and thunderstorms were forecast from about 1400. Computer model output and morning temperature and moisture profiles, obtained from weather balloon flights, indicated very favourable conditions for the development of thunderstorms, and the possibility of associated severe weather phenomena.

During the afternoon, the BoM issued a number of amended forecasts and warnings including area forecasts (ARFORs), TAFs and Airport Warnings for Coolangatta. Those forecasts included reference to thunderstorm activity expected to affect Coolangatta aerodrome after 1400. The Airport Warning for Coolangatta, issued at 1250, indicated that the thunderstorms may produce strong wind gusts and large hail. A SIGMET for the Brisbane Flight Information Region, valid from 1300 to 1900, was issued to warn of a line of active thunderstorms from Dalby to Stanthorpe moving east at 30 kts.

The aircraft operator provided the crew with a meteorological briefing package prior to departure from Sydney. In addition, the crew reported discussing the meteorological situation with the BoM meteorologist, positioned in the operator's flight dispatch organisation, prior to departure from Sydney. Further, they were aware of the line of storms to the west and south-west of Coolangatta, having negotiated past them during the first sector from Sydney to Coolangatta. During the turn-around at Coolangatta, the crew updated the Sydney TAF using the aircraft communications addressing and reporting system (ACARS) and obtained Coolangatta ATIS Echo.

The operator reported that the meteorologist was aware of the development of the thunderstorm in the vicinity of Coolangatta Airport, from the weather radar monitor located in the operator's flight dispatch area. That information was passed to the operator's port staff at Coolangatta, however, it was not passed to the crew of EAL, as the operator did not have a procedure for disseminating such information to crews once they had commenced taxiing.

1.7.2 Bureau of Meteorology - weather radar

The BoM received three-dimensional radar data for the Brisbane to Coolangatta area from weather radars located at Brisbane Airport and Marburg. The Marburg radar was situated on the Little Liverpool Range between Marburg and Rosewood about 50 km west of Brisbane.

The BoM reported that forecasters first became aware of the severe nature of the thunderstorm involved in the occurrence at about 1336. It continued to intensify and quickly moved to the Coolangatta area, passing over Coolangatta Airport between about 1340 and 1349. EAL was issued a takeoff clearance at about 1345. Images of the thunderstorm passing over Coolangatta at 1340 and 1350, including the aircraft flight path from takeoff to 10,000 ft, are depicted at Figures 3 and 4 respectively.

Two-dimensional images from BoM's weather radars were displayed at various air traffic control working positions by means of a PC-based system known within Airservices Australia as METRAD (Meteorological RADar) and within the military as RAPIC (RAdar PICture). The use of METRAD/RAPIC by Air Traffic Services (ATS) controllers is described at paragraph 1.17.1.

1.8 Aids to navigation

Not a factor in this occurrence.

1.9 Communications

All communications between ATS and the crew were recorded by ground based automatic voice recording equipment for the duration of the occurrence. The quality of the aircraft's recorded transmissions was good.

An ATIS broadcast provided advice of conditions pertaining to the operation of aircraft within 5 NM of the respective aerodrome reference point. Coolangatta ATIS Echo was current until 1337:41. ATIS Echo included runway 32 (active runway), wind from 350 degrees at 22 kts, crosswind maximum 15 kts, visibility greater than 10 km and three to four eighths of cloud at 2,500 ft.

At 1338:22 ATIS Foxtrot was issued. It included information to expect a VOR/DME approach, runway 32 wet, wind at 350 degrees at 20 kts, crosswind maximum 15 kts, visibility reducing to 5,000 m in rain and thunderstorms, five to seven eighths of cloud at 2,500 ft and one to two eighths of cloud at 1,500 ft.

At 1339:07, 45 seconds after ATIS Foxtrot was issued, the crew requested taxi clearance and advised having received ATIS Echo. The SMC controller did not advise the crew of the changed ATIS or provide them with the changed conditions.

1.10 Aerodrome information

Runway 32 at Coolangatta was sealed and level. It was 2,042 m long, 45 m wide and aligned on a magnetic heading of 319 degrees.

1.11 Flight recorders

The aircraft was equipped with a Honeywell solid state flight data recorder (FDR). The FDR data indicated that EAL encountered windshear at about 800 ft above ground level. The encounter lasted about 30 seconds and included an 11-second period where the aircraft descended, resulting in a total altitude loss of about 130 ft. The pitch attitude changed from about 20 degrees nose up to 6 degrees nose up during this period. A GPWS Mode 3 aural alert 'DON'T SINK' was recorded for 4 seconds. Take-off thrust was de-rated giving an engine pressure ratio (EPR) of 1.39 and the power setting did not vary during the event. The de-rate is equivalent to a thrust reduction of 9 per cent.

The recorded FDR data indicated that, throughout the windshear encounter, both EPR and thrust lever angle (TLA) remained in the take-off position until climb thrust was set when clear of the encounter.

Recorded FDR data, indicating aircraft pitch attitude and GPWS Mode 3 alert, is depicted at Appendix 1.

1.12 Wreckage information

Not a factor in this occurrence.

1.13 Medical information

The crew reported no physiological or medical condition that could have impaired their performance.

1.14 Fire

Nil.

1.15 Survival aspects

Not a factor in this occurrence.

1.16 Tests and research

Nil.

1.17 Organisational information

1.17.1 Air Traffic Services information

The manual of air traffic services (MATS) was a joint document of Airservices Australia and the Department of Defence. The manual was based on rules published by both those organisations and the Civil Aviation Safety Authority (CASA). Section 5.1.7 of MATS permitted ATS controllers to use the BoM PC-based METRAD/RAPIC two-dimensional weather radar displays, in conjunction with information obtained from other sources, to provide as much information that was available to crews on hazardous weather avoidance. Those displays, which were also available to the general public, were updated every 10 minutes and therefore, could have been up to 10 minutes behind the actual time. Coolangatta ATS reported that they had access to METRAD/RAPIC. ATS did not offer, nor did the crew request, weather information accessed from the METRAD/RAPIC display.

1.18 Additional information

1.18.1 Aircraft flight path information

After encountering the adverse weather, the crew reported that they stopped the cleared turn and attempted to divert around the conditions being experienced. ATS radar data indicated that the aircraft continued on easterly headings to about 70 NM east of the coast, before turning south towards Sydney. The crew was unable to regain the cleared flight-planned route.

1.18.2 Other aircraft movements

There were a number of other aircraft movements into and out of Coolangatta aerodrome preceding the occurrence flight. A Boeing 767 landed on runway 32 at 1328, a Boeing 737 took off from runway 32 at about 1330, a CT4 landed on runway 35 at about 1335 and a B200 Super King Air landed on runway 35 at about 1340. The landing aircraft had approached Coolangatta from the south. No reports of thunderstorm activity, heavy rain, hail or windshear were received from any of those aircraft, except to say that the line of storms was approaching from the west and south-west.

1.18.3 Previous similar occurrence

The occurrence involving EAL displayed a number of similarities to a Boeing 737 microburst encounter that occurred at Brisbane aerodrome on 18 January 2001 (see ATSB investigation report BO/200100213). As a result of that occurrence, the ATSB issued a number of safety recommendations to Airservices Australia, BoM and CASA. Those recommendations included, but were not limited to:

• a review of air traffic controller initial and recurrent training programs to ensure they adequately addressed the effect of convective weather on aircraft performance and the limitations of airborne weather radar
• expediting the introduction of an integrated weather radar/air traffic control radar video display system capable of providing multiple weather echo intensity discrimination without degradation of air traffic control radar information
• an increased emphasis in air traffic controller training programs to ensure that all appropriate sources of information, such as meteorological forecasts, controller observations, radar information, and pilot reports are provided to flight crews
• development of a standard scale of thunderstorm intensity for use within the aviation industry
• BoM meteorologists to act as focal points for liaison with air traffic control units.
• As a result of the occurrence and safety recommendations, the following safety actions were implemented:
• Airservices Australia produced a hazardous weather training program for its air traffic controllers
• with the assistance of BoM and the radar manufacturer, the operator produced a CD-ROM based weather radar training program for issue to its crews
• the operator included information regarding operation of the weather radar and flight in heavy rain in its Flying Manual
• BoM integrated a qualified meteorologist into the operator's flight dispatch organisation.

1.19 New investigation techniques

Not relevant in this occurrence.