On 4 May 2016 first information about the Etihad EY474 incident with turbulence encounter appeared in the press media. Ethad was cited that the Airbus A330-200, flying from the Emirati capital of Abu Dhabi, encountered the sudden turbulence 45 minutes prior to its arrival at Jakarta’s Soekarno Hatta international airport on Wednesday. A total of 31 passengers and crew were injured.
At that day the thunderstorm detection and nowcasting system “Cb-global”, which has been developed at the DLR Institute of Atmospheric Physics, was running in real time mode at WxFUSION , a spin-off company of DLR. Cb-global is based on satellite data, in this case from Himawari-8 available in real time with a space/time resolution of 2 km (IR channels)/10 minutes. Given the aircraft incident the question arose whether the encountered turbulence has been caused by convection and whether Cb-global was able to detect the hazard.
According to data from flightradar24.com the aircraft with flight number EY474 landed on 4 May 2016 at 07:24 UTC at Soekarno Hatta, Jakarta international airport. If the information from the press media “45 minutes prior to its arrival at Jakarta’s Soekarno Hatta” is correct, the incident must have occurred around 06:40 UTC at about the location as indicated in the following flightradar24 map, i.e. near the island Bangka east of Sumatra.
The next two figures show Cb-global thunderstorm detections in the respective area at 06:37 and 06:47 UTC. Taking a close look into the region where the turbulence encounter must have happened (marked by a white circle), we find that an orange cell, i.e. a convective cell of type “rapid development”, is popping up between the times 06:37 and 06:47 UTC at about the aircraft position indicated by flightradar24.
Cb-global thunderstorm situation at 06:37 UTC as overlay on the corresponding satellite image (VIS channel). Full contours mark analyzed clouds, stippled contours one hour nowcasts, yellow contours indicate convective initiation, orange ones rapid development, and red ones mature thunderstorms.
Cb-global thunderstorm situation as above, but at 06:47 UTC
Provided the time indications from the press media are correct, we can conclude that the aircraft has very likely come in contact with that rapidly developing cell and encountered convective turbulence. Indeed the fact that this cell does not appear at 06:37 UTC but 10 minutes later, indicates that this cell has developed just around the time when EY474 passed over that location. Furthermore, from the fact that this cell is labeled as orange and not red, which would indicate a mature thunderstorm, it is possible that the pilots did not see the rapidly developing cell on the onboard radar screen, because no strong radar returns might have been visible at cruise level.
Note that rapidly developing cell that was recognized by Cb-global, is not very large, nevertheless it takes about 2-3 minutes for the aircraft to pass through that area, enough time to experience several ups and downs as the aircraft travels through the turbulent layer. Another indication that the turbulence occurred in the position of the cloud is supported by the aircraft altitude changes which vary according to flightradar24 recordings as follows:
39.000 ft at 06:30 UTC
39.000 ft at 06:39 UTC
38.975 ft at 06:40 UTC → estimated time of turbulence encounter
39.250 ft at 06:41 UTC
39.000 ft at 06:42 UTC
This documents that altitude changes have occurred in the area of Bangka Island.
Under the constraint that the preliminary reports in the press media on this incident are correct, the Cb-global analyses allow the conclusion that flight EY474 encountered convective turbulence near the island of Bangka on 04 May 2016 at around 06:40 UTC. Between the satellite observation times 06:37 UTC and 06:47 UTC a thunderstorm cell rapidly developed at that location as indicated by an orange contour.
On the one hand, this analysis documents that it is possible to recognize hazardous turbulence areas connected to a thunderstorm by processing satellite data in near real time with the Cb-global algorithm. On the other hand, one must admit that in this case, due to the rapid development of the cloud between two consecutive satellite scans and at the position of the flight, it would have been too late to warn the pilots of EY474 of the suddenly developing threat. Nevertheless, if Cb-global data would be uplinked regularly to cockpit electronic flight bags (EFB) - this is a new approach in the airline industry to better inform pilots of weather hazards - the displayed information would raise the flight crew’s situational awareness, and allow them better to take proper mitigation actions.
Although orange cells in Cb-global are generally small and might develop and disappear within 10 minutes - as do convective clouds in reality - they are markers of rapidly developing thunderstorm clouds and can in many situations give pilots early indications of potential convective turbulence, advising them to scan the clouds carefully with the onboard radar and asking passengers to fasten seat belts.
Dr. Dennis Stich (firstname.lastname@example.org)
DLR Institut für Physik der Atmosphäre
Dr. Arnold Tafferner
WxFUSION GmbH, Gilching