A Dark, Stormy Night over the Atlantic


Photo of F-GZCP - the airliner that crashed - from JetPhotos.Net

So what happened to Air France Flight 447? It is early and speculation at this juncture is often wildly wrong. And remember, there are usually several factors that conspire to bring an airliner down. But here is what we do know for sure. Keep this in mind as you process the often inaccurate reporting on aviation that is so prevalent in the mainstream media.

The Timeline – The flight, carrying 216 passengers and 12 crewmembers, left Rio de Janeiro at 2203 GMT (7:03 PM local time). It flew beyond radar coverage 3 hours and 33 minutes later (at 0133 GMT). A half hour later (0200 GMT) – now four hours into the flight – the plane encountered heavy turbulence. Fifteen minutes later (0215 GMT), now a long way out to sea, it transmitted automated signals indicating the plane was in serious trouble.

“A succession of a dozen technical messages (showed that) several electrical systems had broken down,” according to Air France CEO Pierre-Henry Gourgeon. He described the failures, which included (most ominously) the pressurization system as “totally unprecedented situation in the plane.”


Weather over Atlantic during crash - From Naval Research Lab

It was a dark and stormy night – in a place that is home to the world’s worst thunderstorms. Just as it disappeared, the Airbus A330-203 was flying into a thick band of convective activity that rose to 41,000 feet. This equatorial region is known as the Intertropical Convergence Zone – it is where Northeast and Southeast Trade Winds meet – forcing a lot of warm, moist air upward – which condenses – an efficient thunderstorm producing machine.

The crew had “Sully-esque” seasoning – The Captain had 11,000 hours total time (1700 in the Airbus A330/A340). One Copilot had 3,000 hours total time (800 in the Airbus A330/340) and the other Copilot had 6,600 hours total time (2,600 in the Airbus A330/340).

The Airbus A330 has a good record – and this was the first crash of a twin-engine A330 in revenue service in its history. In 1994, seven employees of Airbus died when a 330 went down during a test flight. The accident report says it was a case of pilot error. The airplane that crashed last night – tail number F-GZCP – had no accidents or incidents in its history. It went into service on April 18, 2005 and had logged 18,870 hours. In 2006, it’s wing collided with the tail of an Airbus A321 on the ground at Charles de Gaulle Airport – the damage was classified as “minor”. It was last in the hangar on April 16, 2009l for routine maintenance. No serious squawks reported.

No reason to believe terrorism – While you cannot take the possibility of a bomb off the list just yet, no groups have claimed any responsibility for downing the plane. What good is a terrorist attack if the perpetrators don’t, well, terrorize us?

So consider this as a possible scenario: The crew is flying toward a line of storms in the dark, out of range of land-based radar. They are equipped with on board weather radar however – and can use it to thread their way through the bad cells if need be.

It is quite likely the airplane was struck by lightning – or it could have triggered lightning by the mere act of flying at Mach .8 through storm clouds. It is not impossible that could have sparked a fuel fire – but that is highly unlikely. In fact, it has been four decades  since lightning alone caused an airliner crash in the US. A lot of time and effort is spent protecting airplanes from this clear and present danger (interesting piece here). And airliners get hit by lightning all the time – you don’t hear about it because nothing bad happens. Remember, it is seldom just one thing that brings a modern airliner down.

Many of those airliners that get hit by lightning are so called fly-by-wire aircraft (meaning the controls in the cockpit are linked to the movable surfaces on the airplane by electrical wires and computers). Airbus pioneered FBW control systems in commercial airliners and the engineers in Toulouse have gone out of their way to demonstrate their products are safe in stormy weather. There are four fully redundant electrical systems on an Airbus – and if the worst happens, a manual flight control system that allows the crew to fly the plane (barely) using the rudder, differential thrust on the engines and horizontal stabilizer trim. [You may recall that is how the crew of United flight 232 managed to get a DC-10 on the ground in Sioux City, Iowa in 1989 after a complete hydraulics failure]

Ironically, one of the systems most vulnerable to lightning strikes is the on-board weather radar located in the nose cone. It cannot do its job if it is shielded from lightning like the rest of the airplane is – and so it is more likely to go down when bolt strikes (which is, of course, when you need it most). So it is possible this plane was hit by lightning, knocking out the radar.

You can imagine the crew was suddenly preoccupied with multiple electric failures that left them in the dark, over the ocean and without weather radar as they hurtled toward some epic cumulus nimbus thunderheads. This would have been a serious emergency that should prompt a pilot to do a 180 and head for the nearest suitable size slab of concrete.

The fact that the airplane sent automatic warnings that it had an electrical problem means, by definition, that it was not a total, instant failure. But did things cascade from there? They might have found themselves inside a huge storm only able to control the airplane manually – which means minimally – with the rudder primarily.

And then there is the Airbus rudder. You may recall the crash of American Airlines flight 587 on November 12, 2001 as it departed New York’s JFK airport. The plane encountered some wake turbulence and the copilot apparently stepped too hard on the rudder pedals – breaking off the graphite vertical stabilizer and rudder (the tail).

As long as we are talking about pilot inputs leading to broken airplanes, consider this important point: when the Airbus FBW system is up and running as it should, there are all kinds limits placed on the pilot’s ability to move the control surfaces of the airplane. It’s sort of like a governor on a car engine. If you move the controls too far, too fast in any direction, the computer, in essence, ignores the human being’s commands and keeps the plane inside the flight envelope. This is designed to stop a plane from stalling, spinning, gaining too much speed or pulling too many “G’s” because a pilot is over-correcting (which of course, is not correct at all).

But as the electrical systems start failing, the machines lose their authority to trump the humans fairly quickly. Depending on how many multiple failures of redundant systems there are, the so called flight control laws change to “Alternate”, “Abnormal Alternate” and finally “Direct Law”. At each level, the pilots get more authority to move the control surfaces without the machines intervening. So a combination of loosened fly-by-wire reins, cruise speed and extreme turbulence would increase the potential for an in-flight breakup.


AA 587 crash site in Queens - from NOAAEven today’s advanced - seemingly invincible - airliners are no match for Mother Nature on a bad night. If a big airplane ends up in the teeth of a powerful thunderstorm, it could be torn to pieces in an instant.

We do know whatever happened on that airplane in its last few minutes was nothing short of horrifying. It is hard to imagine the kind of turbulence that would break up an airliner. My heart goes out to the passengers and crew.

Will we ever know what happened? This one will be hard. The wreckage will be likely strewn over a wide area – and locating the Flight Data and Cockpit Voice Recorders won’t be easy since they are likely at the bottom of the sea – possibly 24,ooo feet below the surface. Even if they are transmitting their homing signals, you would need a lot of luck and a pretty stout submersible to retrieve them. But that may be moot – as simply knowing where to search will be difficult.

One thing which may help: those automatic messages indicating system failures – which are designed primarily to give mechanics a heads up about problems so they can turn a plane around on the ground faster – no doubt contained much more information than is now in the public realm.

Which brings me to this wild idea: why not send steady streams of telemetry from airliners to the ground all the time – ala the space shuttle? This effectively places the “black boxes”, safe and sound – on the ground. Imagine how invaluable that much data would be right now – given the the distinct possibility this could remain an unsolved mystery.

We all need to know what happened to Air France 447. Is there something that makes the A-330 fleet unsafe in certain conditions? In the absence of real facts, will conspiracy theorists spin a tale of terrorism and government cover ups? Did the flight crew make crucial errors in judgment? Or was this an unavoidable scenario – bad luck with odds so long that nothing or no one is really to blame?


Tags: , , , , , , , , , , , , , ,

98 Responses to “A Dark, Stormy Night over the Atlantic”

  1. windsorsean Says:

    Has there been any talk in the industry about using real-time satellite tracking with airliners? After all many private pilots use the SPOT service for just that reason, to help track down missing aircraft.

    • Miles O'Brien Says:

      The airlines have been slow to embrace GPS technology. This is a big reason we need to make sure the FAA has the funding it needs to build its Next Gen air traffic control system – it makes GPA the primary means of locating and separating aircraft. Today, the system we use should be in museums.

  2. rockyinlaw Says:

    Gosh, Miles O’Brien, you sure know how to grab a reader — plus get my curiosity up about further research. Thank you! You’ve got one more fan.

  3. patimc Says:

    Very well done. If it were not for your Tweets on Twitter early this morning I would not have received any information, let alone facts without speculation.

    You have managed to pull this off at an incredible pace, too. No one else was even on it. Even friends in France and other Eurpoean nations were pointed to you for information.

    Once again, Miles O’Brien is an invaluable source on anything aviation and space related. You are a gem, sir, and I for one just wanted to say “thank you”.

  4. becca00 Says:

    I’ve heard this evening that Air France is reporting the aircraft carried on (after the warning messages) for another 3 hours. This puts the aircraft near the African coast–hence the involvement of the Senegalese.

    Which leaves one to wonder where the debris field will be located (if there is one to be found).

    Has this been reported in the U.S.?

  5. ronstx Says:

    Spot on, Miles. Your analysis at this level of information is excellent. Your mention of the Airbus crash in NYC (due to the vertical stabilizer coming off the airplane) and the possibility of this being a cause is the first mention of it I have heard from any “expert analysis” comments on TV today. Speaking of TV, I have been missing your commentaries on CNN.

  6. johnsmith Says:

    How did you figure that @0200GMT the the plane encountered heavy turbulence?
    At reuters you can read some words from pilots of two Lufthansa aircrafts, who didn’t encounter “anything special” flying nearly the same route 30minutes before and 2 hours after AF flight.

  7. kevinjdunn Says:

    This may be redundant – I thought I already posted it but I don’t see it up now.

    Miles, great to see you’re still in the game with informed, intelligent reporting. Always liked that about you, and miss it on our air.

    Kevin J. Dunn
    CNN International

  8. timashton Says:

    Miles avoids the great unspoken.
    When we learnt to fly we wre all told (sic)”never fly into big black clouds with lightning coming out of them!”
    No matter how much experience the crew has there is the elephant in the room that is drives the crew forward when maybe prudence would dictate a diversion or a return to the departure point.
    That elephant is management and the passengers with their committment to completing the flight on time and delivering the passengers on time.
    I have no doubt that this sort of pressure has a tremendous affect on the judgement of even the most experienced crews.
    IMO Pilots who turn around should be awarded medals, especially if I am on their flight!

    • Miles O'Brien Says:

      I guess that is the great unspoken. Get-there-itis is another way pilots describe it – and there are a lot of pressures that induce that dangerous desire to press on and get home.

    • orangeanchor Says:

      Even an ‘elephant’ is not going to disregard the prime motivation of self preservation. As a long time aviator, I have been pushed by various factors but risking self and passengers to get to point B is irrational.

      Yes, there have been incidents/accidents where weather was a factor. But you will NOT hear on the voice recorders, “We’ll probably get killed doing this but let’s try it anyway.” Aviators are people also and to assert that they will knowingly choose to act in a manner that can result in a fatality (including their own) is an argument that can not be supported.

  9. conorbod Says:

    Mr O’ Brien,

    You ought to be ashamed of yourself for this article. You start the article by criticising the inaccurate reporting on aviation eventsw and the dnger in speculating as to what happened. The end if your article is nothing short of sensationalism.

    Please explain this paradox!

  10. andyandyandy Says:

    Miles wrote: “And remember, there are usually several factors that conspire to bring an airliner down.”

    But remember, those factors don’t really “conspire”, in most of these tragedies. Human error, mechanical failure, and weather are the most common causal factors in aircraft crashes. But to use the term “conspire” in this context is entirely inappropriate. But I think I know what you meant 😉

  11. that1guy Says:

    Mr. O’Brien, thank for this analysis; it’s by far the most salient one I’ve seen so far on this tragedy.

    Re real-time telemetry from aircraft: I think it’s impressive that we have what we have from this plane; it’s certainly more than would have been possible less then ten years ago. However, I don’t know how much value would be added by a continuously broadcasted & recorded data stream from all transoceanic flights. The statistical likelihood of crashes is, after all, pretty low, and the amount of support infrastructure needed to operate & sustain such a system would be relatively high.

    A better proposal might be integrating the FDRs with onboard satcom systems (which I assume are fairly widespread by now in commercial aviation). In this scenario, a potentially catastrophic event such as loss of a hydraulic or electrical system, multiple engine shutdown, etc., would trigger an automatic data dump of the last 30 min of flight followed by real-time data until the aircraft was contacted & the crew commanded the uplink off.

  12. heather Says:

    As a squeamish but frequent flyer, this line terrifies me:

    “If a big airplane ends up in the teeth of a powerful thunderstorm, it could be torn to pieces in an instant.”

    Is that true of “strong” thunderstorms anywhere? I was under the impression that the types of thunderstorms in the Intertropical Convergence Zone are dangerous, but our typical U.S. midwestern thunderstorms are not. Is that true, or would you tornado-producing midwest thunderstorm be capable of bringing down a plane? And if so, why don’t we have more frequent crashes due to thunderstorms?

    I’m definitely going to need more Xanax before my next flight after reading this article.

    • Miles O'Brien Says:

      because we do not fly into them.

      • heather Says:

        But are thunderstorms over land as powerful as those in the intertropical convergence zone? I’ve flown through thunderstorms before – maybe they weren’t the heart of the storm but I’ve been on plenty of flights with rain, lightning in the distance and severe turbulence, nearly all in the midwest U.S. during the summer thunderstorm season. I keep hearing these storms near the equator are especially high. Can pilots fly over thunderstorms on land? Is that skill a risk?

    • dna550 Says:

      I used to travel by air frequently in Florida which has a mini, convergence zone, afternoon Thunderstorm Mountain Range [caused by the sea breezes on the Atlantic and the West Gulf Coasts colliding in the middle of the peninsula]. Towering thunderheads are present up and down the Florida peninsula most afternoons in the summer.

      I mention this because the description of this flight reminds me of an afternoon flight I had in July a few years ago from Tallahassee to Orlando.

      We took off at 1:30 in the afternoon on what would normally be a short flight to Orlando. Along the way we encountered the most frightening roller coaster ride I have ever experienced in an airplane. I was prepared for the end.

      After ten minutes of this the pilot found some better skies to fly in. About twenty minute later, when I was thinking we should have already landed, the Captain came on to announce that we had a new flight plan that would take us very close to Cuban air space and that we had a new destination: Fort Lauderdale.

      Needless to say, that afternoon there was no getting over or through the mountain of thunder heads in Central Florida – which is, btw, the lightening capitol of the U.S.

      After spending some time in Fort Lauderdale, in I finally caught an evening flight into Orlando – arriving around 10:00.

      It had been a long event-full day of flying, but I was greateful that the pilot, the airline, and the Air Traffic Controllers were not concerned enough about an on-time arrival in Orlando, to force the issue.

      Looking at the radar this afternoon in the Atlantic in the ITCZ, I wished that the Air France Captain had made a mid-course decision to fly northwest towards Venezuela, thus avoiding the huge thunderstorms in the Atlantic on his direct route to Paris

  13. rcp727 Says:

    Hello Miles,

    Thank you for the concise article with glimpses of what else may have transpired on AF 447. As the former owner of an Airbus A319 (N320NP) this tragic event has me questioning what open issues we’re either missing or ignoring.

    I’ve been a loyal fan of yours, particularly from the NASA reporting.

    I’m looking forward to attending The Paris Air Show in two weeks, although the mood will be more somber after todays loss.

    Best regards,
    Richard Page
    US (1) 612-889-6640

  14. quarkdoll Says:

    You made a factual error in stating that the aircraft “had no accidents or incidents in its history” — the plane was involved in an incident in 2006.
    Data as follows:
    DATE: 17.08.2006 LOCAL TIME: – LOCATION: Paris-CDG Intl AP (LFPG) COUNTRY: France
    AIRLINE1: Air France TYPE: Airbus A321-211 REGISTRATION: F-GTAM C/N: 1859 AGE: 3 y + 9 m
    AIRLINE2: Air France TYPE: Airbus A330-203 REGISTRATION: F-GZCP C/N: 660 AGE: 1 y + 5 m
    OPERATION2: ISP FLIGHT No.: – FROM: Paris-CDG TO: Ouagadougou VIA: –
    Both aircraft suffered damage in a ground collision at Charles de Gaulle Airport. The tail of the A321 was substantially damaged when it was hit by the taxiing A330. Damage to the latter was considered as minor.

  15. astrodeb Says:

    Excellent article, Miles! My husband and I (astronomers by trade, aviation/space enthusiasts by choice) talked about a similar scenario earlier today. I do rather doubt that the Captain stayed asleep long with the heavy turbulence upon entry to the ITCZ. My own experience in airliners encountering thunderstorms suggests that there is a tendency to aim for local minima in the weather radar returns and hope for the best rather than turn tail when confronted by a wall of storms. Most of the time this results only in jangled nerves and more airframe fatigue, but losing the radar would be deadly. Normalization of deviance again? I also thought about those nearly radar invisible convective tops common in Africa. Real-time telemetry for airliners is a great idea although quite a data challenge. Thanks as well for the excellent article on the Hubble repair mission.

    Best wishes!

  16. bsalvin Says:

    Miles-You were the best source of information for me on this topic today. I trust your reporting so the information you tweeted and posted had superior credibility to the temporary Airbus experts that filled the airwaves today.

    Very, very well done.

    • Miles O'Brien Says:

      Many thanks for the kind words from all. Gratifying to help people understand something so sad, complex and scary.

      • arriba Says:

        Thank you so much for your clear and concise information on this subject. It sure wasn’t out there on the MSM from any of them. None of whom appeared to have anyone who had a clue as to what he or she was talking about. I was appalled when CNN had Richard Quest do the report on this. Or the longest report on it that I saw on that channel. Mr. Quest’s knowledge of aviation and science appears to be limited to being able to call a flunkie to book his airline tickets, get his employer to pay for said tickets and a moderate ability to read a teleprompter. Sorry coverage IMO. I’m one of your former employer’s watchers who misses your presence there greatly. But I guess that that was a corporate decision based on the fact that nobody else in the cable news world was doing any better and most were doing worse so they “didn’t need no stinking science”. Glad to have found you online, hope things are going reasonably well for you.

        Diane – AKA Arriba

      • Miles O'Brien Says:

        Thanks Diane – There are plenty of ways I can share my passion – post CNN!

  17. derekpaton Says:

    My understanding is that the aircraft has not been located yet. Until such a time that it has been located, stewn over the Atlantic, should there not be more hope of survivors? Is it possible that the pilots successully ditched without any power (including communications)? A Sully-esque ditch would be extremely difficult at night and in a rocky ocean, however it is possible, right?
    Sydney, Aust.

    • Miles O'Brien Says:

      A stormy night…a turbulent sea state…and airplane with multiple systems and maybe structural) failures – as opposed to a windless winter afternoon…a fairly open river as flat as glass – and a plane that was all good – except for the lack of thrust. Two very different scenarios. Sad, but true.

  18. wdfarmer Says:

    There’s a Google map in Portuguese noting the location of events in the flight: http://tinyurl.com/noy5pu . In that map, it’s indicated that ACARS information was being transmitted from the plane. Wikipedia has quite a bit about ACARS: http://tinyurl.com/n3dh6e . I’d guess that ACARS was the system that transmitted the messages of electrical failure and depressurization that are currently being reported in the news.

  19. sactom Says:

    Miles, A very good article on the AF 447 disappearance. Air France was quick to point out that the captain had over 11K hours of flight time. More important is how much time he had on that or similar routes. If he accumulated the vast majority of his time flying in Europe or on trans Atlantic flights, he would lack experience in dealing with those dangerous equatorial cumulus buildups. Trying to pick your way thru a wall of thunder heads at night near the equator takes skill which only comes from experience. If lightning knocked out his wx radar it would increase the difficulty 10 fold. My heart goes out to the passengers & crew.


  20. ronaldmaustin Says:

    It starts “It is early and speculation at this juncture is often wildly wrong.” So when it got to “So consider this as a possible scenario…” I said, okay, here we go with the speculation.

    But by the end of the article, I felt this was one of the best written stories on an air disaster that I’d ever read. As a non-pilot I came away with knowledge that I’d never received from the hundreds of stories on aircraft crashes over the years. I signed up here just to say thank you for excellent writing.

  21. lordfalmouth Says:

    Mr O’Brien,
    I discovered you today via Reuters. Thank you for such detail and colour. I think you have got closer to the plane and the pilots than anyone. I am bowing from Japan.
    Do you remember the BOAC flight from Tokyo to Hong Kong [way back in the 60s] that was battered over Mt Fuji. Similar?
    Keep writing.

  22. overeasy Says:

    Mr. O’Brien –

    I need to try to understand something. If there is a telemetry being sent back to Air France from this plane, which indicated the failure of critical systems, why is it not monitored? Is this just some sort of data dump which is used to review problems after that fact? If so, that seems crazy. I’m not saying that it would have changed anything about this particular scenario, but perhaps, in another situation, it would have allowed search and rescue teams to get a jump on things.


  23. alanj Says:

    I ran across this link which claims that the aircraft that was lost out of Brazil experienced an electrical problem in March 2009. The story cites a “technical flaw in the avionics bay”.

    Here’s the link: http://www.bangaloreaviation.com/2009/03/air-france-airbus-a330-grounded-at.html

  24. tuebor Says:

    The pressurization problem makes me wonder if they suffered a dual engine failure. With no engines running there would be no bleed air to pressurize the cabin, and the cabin altitude (normally around 8,500′) would leak upward above 10,000′ triggering the warning. The crew would be very busy trying to restart the engines and as the aircraft slowly loses altitude, it could have descended into storm cells where it broke up or the crew lost control. The weather radar would be unavailable at least until they got the APU running but even at the single-engine altitude it was capable of (probably low 20,000′ range) the storms may have been unavoidable. An emergency call on HF radio, difficult in normal operations near storms, might have been a low priority, especially with an oxygen mask on and when neither your native tongue nor the contoller’s is English.

    What could have caused a dual-engine failure? Severe turbulence from the storms? Icing from the anvil tops or blowoff from the cells? Heavy rain ala Southern Airlines 242 in Rome, Georgia?

  25. vmorrison60 Says:

    Miles: Please, for my own edification, what is the everyday occurence of this “intertropical covergence zone”? Do all flights deal with this on a regular basis? I mean, I have read, where pilots have said this is a very “difficult” area to go through. What sort of continual update is monitored by the crew? What is at their disposal for reference! I would say, that as a passenger, would I have been made aware of what the flying conditions would be before the plane left?

    • Miles O'Brien Says:

      Pilots crossing the Atlantic get a big picture look at the weather based on satellite information primarily. It is not as accurate as weather forecasting over land where ground stations greatly enhance the situational awareness.

  26. tuebor Says:

    The reported pressurization problem leads me to believe the aircraft suffered a dual-engine failure. Without the engines running there would be no bleed air to pressurize the cabin (normally kept around 8,500′) and the cabin altitude would leak upward above 10,000′ triggering the warnig. The aircraft would lose altitude and drift down to a single-engine cruise altitude in the low 20,000′ range. As it descended it probably couldn’t avoid the storm cells and either broke up or encountered turbulence so severe the crew lost control. The weather radar wouldn’t be available at least until the APU was started, but the crew would be so busy in the dark attempting to get the engines restarted that picking their way through the storms may have been a secondary consideration. Making a distress call on HF radio, difficult in normal operations in the vicinity of storms due to the static, would have been the least of their worries, especially when wearing an oxygen mask and when neither your nor the contoller’s native tongue is English.

    What could have caused a dual-engine flameout? Turbulence? Icing from the anvil storm tops or blowoff from the cells? Heavy rain ingestion ala Southern Airlines 242 which crashed in Rome, Georgia?

    • Miles O'Brien Says:

      So maybe the engines ingested a lot of hail or too much precipitation. This would not be unprecedented.

      • mrhelio Says:

        If the Flight 447 ingested a rather heavy dose of hail, rain, coupled with wind shear there still would have been time to activate a distress via ATC, or automated system. The time it took for the airliner to impact water was extremely short in duration.

    • jonjosserand Says:

      Interesting possibility (both engine loss).

  27. Coates Bateman Says:

    Looks like some debris has been found: http://www.nytimes.com/2009/06/03/world/europe/03plane.html?hp

  28. mrhelio Says:

    I have 6000 hours and not to much turbine time, however living in the midwest has taught me to stay clear of any severe weather. The Airbus A330 is an advanced medium seat jetliner with the latest in glass cockpit technology and SAT NAV (satellite navigation)and EVS. When you encounter towering supercell activity at night over water, loss of horizon is emminent. What ever took place did so with vengence. CAT can occur at least 150 miles from thunderstorm activity at high altitudes. Massive decompression has to be looked at here as well.

  29. phillipbrant Says:

    Dear Miles:

    I have an idea that must surely have been thought of before. Why not have the cockpit voice recorders broadcast all cockpit conversations in real time to ground-based receivers located at the controlling airports along the route – in addition to the onboard recordings that are made? I suppose that the privacy of crews would be an initial issue, but given how much is at stake, could we not regard that need as subordinate to the vastly greater good to be obtained? The ground-received recordings would fill up a lot of disk space around the world, but there could be procedures for erasure according to a schedule.

  30. woodall Says:

    Hi Miles…GREAT job with great factual info. It is wonderful to see you back in the news stream! I had been searching the major networks but no info at all on this story. CNN and FOX just have the info babes with one sentence every hour about this lost plane.

  31. jonjosserand Says:

    Thank God I found your blog.

    Your ability to describe nuances of flight/space systems intelligently to a very broad audience is rare, and has certainly not been replaced on CNN.
    Now my question, if your sources have answers…. I understand that information transmitted via ACARS protocol/system varies widely by Airline & plane type & location. I am very interested in learning, if available, more information about the ACARS “fault” transmissions reportedly involved in this instance. (electrical issue, cabin depressurization) including how triggered and significance of actual transmissions. Were these two faults part of a single communication? Did it require 1 sec or 30 secs to negotiate and transmit. What other fault systems on this type craft existed that did not trip/transmit for this instance. Are you Airbus experts talking?

    • Miles O'Brien Says:

      I will try to get an answer.

      • jonjosserand Says:

        Miles, I found this info on web site of Aviation Herald. Perhaps we can start by identifying meaning of acronyms (other than ACARS):

        New information provided by sources within Air France suggests, that the ACARS messages of system failures started to arrive at 02:10Z indicating, that the autopilot had disengaged and the fly by wire system had changed to alternate law. Between 02:11Z and 02:13Z a flurry of messages regarding ADIRU and ISIS faults arrived, at 02:13Z PRIM 1 and SEC 1 faults were indicated, at 02:14Z the last message received was an advisory regarding cabin vertical speed. That sequence of messages could not be independently verified.

      • jonjosserand Says:

        Things evidently went sour quickly… perhaps in as short of a 3-4 minute timespan.

  32. sactom Says:

    Miles, Now that we know where the jet went down & that it lost cabin pressure, a picture begins to take shape. The loss of cabin pressure, whether by natural causes or by a bomb, would force the pilot to descend to 10K feet & return to Brazil – the nearest airports are there. The wreckage was found south of the planned course, so he probably turned to the right. He had enough fuel to make it to Brazil even at the lower altitude. The flight plan fuel load was for a 10+ hour flight to Paris & the trouble occurred about 4 hours into the flight. Obviously the damage to the aircraft & its systems was so severe the aircraft could not make it back to a landing site. It had to be a terrible time onboard for the crew & passengers.


  33. filippomaria Says:

    I live in Asia. I often fly through the nastiest weather conditions especially during typhoon season. Just recently I flew , at night, over the bay of Bengal and there was a powerful cyclone. It was scary, lightning all around, and nasty turbulencs.
    The plane was a boeing 737-800.
    Airplanes do not fall off the sky. Why are we immediately speculating that this accident is weather related and not, as it might be, simply a pilot error or a problem with the A330? Because the airline is called Air France?

    • Miles O'Brien Says:

      Not sure what you mean about you last rhetorical question. But if you read my post carefully, I delve into the issues of pilot judgment and the soundness of the airplane as well.

  34. hlunde Says:

    It would seem prudent to begin equipping Flight Data/Voice Recorders with an acoustic transponder in addition to a pinger. The transponder would conserve power by remaining in “receive” mode until it was interrogated by an defined acoustic signal from a ship. Doing this would extend localization opportunities from weeks to perhaps a year. Transponder technology has been available since 1970’s or earlier. The interregator of course has not power limitation whatsoever.

  35. petrysl Says:

    good article. the thing that is bugging me is that even though there is no radar coverage, there is limited radio capability out on the open sea. yet there was no apparent distress call. puts me more in mind of a Challenger-type disaster. total structural failure? explosive decompression? unlikely given that the debris field was relatively contained, but just can’t see this thing just falling into the ocean without warning. as you mentioned both captain and FOs were veteran pilots.

  36. bsmcconnell Says:


    I wrote an article in 1996 (never published) that described an inflight Internet that was designed to support a range of aviation services including forensics (consumer communication was just one usage, as it was intended primarily to be first and foremost an aviation service and support weather, notams, automatic telemetry, science payloads, etc). The government would have built and operated the ground based infrastructure, and set a standard for transceivers, network protocols, etc.

    The system was not envisioned as a satellite based system, to keep costs reasonable, although that could have been an option. Over land, aircraft would communicate mostly in a direct air-to-ground mode to transceivers colocated at navigation beacons. Over the ocean, they would operate in an air-to-air mode where aircraft would act as repeaters to capture and forward data to over the horizon ground stations.

    Had such a system been in place, it is likely that nearby aircraft would have picked up telemetry from the Air France plane, and would have repeated it to a ground station within a few hours at most. At that altitude, line of sight range for a signal is several hundred miles between aircraft. This type of system could also have been implemented in general aviation, where the risks and accident rates are substantially higher.

    Another way to avoid this outcome is to design the black box so that it is embedded in a fragile structure, such as a small tab on the underside of a wing, that is designed to break free and float while transmitting an emergency locator signal.

    Brian McConnell

  37. aviateur Says:


    I have been told that because of the way the A330’s pressurization system is designed, loss of cabin pressure can result from certain electrical failures. Thus, the rapid depressurization message sent by the plane’s ACARS may not have been a structural issue. It may have been part of the various electrical faults that were happening — likely brought on by a serious lightning hit.

    The crew, meanwhile, is dealing with the loss of the plane’s most critical flight instruments and controls. There would be a lot of confusion at this point, with warnings going off, messages flashing, and some very urgent troubleshooting. If, as part of the mix, the cabin suddenly decompresses, they may not notice this right away.

    And without use of supplemental oxygen, they pilots would have about 30 seconds of useful consciousness.

    Could they have passed out? The autopilot was off at this point, and in stormy weather the plane would have gone out of control very quickly.

    Patrick Smith

  38. geoffs Says:

    Thanks for the excellent coverage of this disaster. It’s incomparably better than nearly everything else I’ve read (though Robert Wall at Aviation Week has some good coverage).

    One thing, though — I think this is a remarkable statement:

    “No reason to believe terrorism – While you cannot take the possibility of a bomb off the list just yet, no groups have claimed any responsibility for downing the plane. What good is a terrorist attack if the perpetrators don’t, well, terrorize us?”

    I don’t have statistics at hand, but Al-Qaeda and the loose federation of groups at Al-Qaeda’s periphery infrequently claim responsibility. I believe there were never _credible_ claims of responsibility for 9/11, the 2004 Madrid bombings or the 2005 London bombings.

    To be clear, I would be somewhat surprised to find out it was a bomb — Air France and Brazil seem to be unlikely targets, and the bad weather seems more than coincidental — but the evidence I’ve read about fits nicely with a bomb or other intentional device:

    1. The crew had no time to signal distress

    2. The ACARS CMU transmitted data over a three-minute period, which represented only about a minute of information.[1]

    3. One of the messages reportedly signaled loss of cabin pressure and the final one was reportedly an advisory on ‘cabin vertical speed’.

    4. The aircraft was lost far at sea.

    I suspect that if Richard Reid had been successful in his effort to bring down AA flight 63 in 2001, the early evidence would have looked very similar to what has been reported here.

    Again, I believe terrorism is unlikely, but the absence of a claim does little to rule it out, and other evidence correlates well, so to say “no reason to believe terrorism” seems surprisingly dismissive.

    [1] http://www.aviationweek.com/aw/generic/story.jsp?id=news/RECORD060309.xml&headline=Concerns%20Over%20Recovering%20AF447%20Recorders&channel=comm

  39. etaglio Says:

    Hi Miles:

    I saw a National Geographic TV program recently about a helicopter downed by a lightning strike, which ended up being caused by composite resins used in the tail rotor.

    As more newer planes are moving away from traditional aluminum materials into the realm of newer-age composites, we may discover they are much more susceptible to lightning strikes. My understanding is that composites do not have the energy-dispersing characteristics of aluminum-based shells and that this Airbus contained composite materials in the wings

    Here are a couple links related to composites and lightning strikes:


  40. lars Says:

    Aircraft just don,t fall out of the sky. This must be a terrorist attack.
    The motive: read Wall Street Journal
    Europe May 26 2009

    France with their president the same week inaugurated a
    NEW permanent air force base in the UAE about 160 km from the border of IRAN.
    The base houses about 600 French military personell + French military aircraft

    Understand this is the first time in modern time France is beeing engaged
    with permanent military bases in the
    Arab world and certainly on the Arab peninsula.

    Miles – give me your e-mail address and will mail you the article from the WSJ

    There are numerous of muslim organizations that don,t accept western bases (infidelities) on Arab soil and publicly have advocated that they will do whatever they can to stop or prevent it.


  41. thyngsd Says:

    Anyone have easy answers to these admittedly superficial questions:

    1) Why aren’t at least one of the black-box recorders designed to FLOAT ??!? Ideally, there might be 2 identical units:
    one that would be somehow ejected from the aircraft in the slightest indication of a crash and then float to be easily
    located, and the other would stay with the crashed aircraft like the current one so it could located the crash site itself.
    Then the current situation would be vastly simplified — no desperate search for the single pinger in the deep ocean,,,
    and possibly batteries dying after 30 days, etc.
    And isn’t it silly now that there be separate units (one for data, one for cockpit sound recordings — like in the old days
    because they were different) when both functions are now just streaming data that could simply be recorded on a
    single hard-drive ?? With such drives so small and cheap, there should be multiple ones on the aircraft,,, perhaps
    one in the front, one in extreme rear, maybe one in the middle, etc, then at least one that floats. And this outdated
    concept of only recording the last 30 min of voice — with compact high-volume data drives, why not record everything
    starting with pushback to landing ??

    2) Since some airlines already have cheap online satellite-capable Internet connections available at every seat for modest cost,
    why isn’t all the above critical safety data just transmitted online real-time also,,, and do away with the archaic boxes all together ??
    If there is a crash, they’d just replay the data already on a land-based server and determine the cause immediately !

  42. everglades Says:

    I am amazed at the similarities of this crash to one that happened 46 yrs ago in the Everglades (NWA 705 involving B-720B N724US0). That crash, studied intensely by Northwest Airlines, Boeing and the CAB (now NTSB) was caused by a combination of circumstances and the term “jet upset” was eventually used to label the then unique crash.

    You have elucidated the events surrounding the Air France crash unlike any writer I’ve read in a long while – knowledgable and to the point. Keep up the good work.

    Let me know if you’d like some good reading on the everglades crash and a similar accident that happened six months later that the jet pilot recovered from (the everglades crash occurred c. 17K ft.: the recoverable event started c. 35K ft.) I’ll be happy to forward it to you.

  43. minette Says:

    Miles — I am glad to have found you, as others have mentioned. Tragic incidents like this one make many of us wonder — Where is Miles O’Brien to provide his valuable, knowledgeable perspective!? Excellent articles on the challenges of this horrible tragedy.

    Anyway, wondering as others have asked — I didn’t realize that air traffic control systems are now “antique” relative to today’s technology. Granted, I know next to nothing on this topic so I feel like a five-year-old asking this question… but… how is that acceptable given the rather high amount of air traffic around the world AND the available technology? Is it all about money? And what do commercial pilots and crews think/feel about this? I’m having a major disconnect on this having only just learned of it.

    Thanks in advance for your input and again, glad to see you!

  44. cookeee57 Says:

    THanks for this info, Miles! My hubby is a pilot also with a major airline. I enjoyed your take on the idea of having the data relayed constantly from planes as they are in transit, so I asked my other half. He said sure that would be great but the airlines have no money and the govt would not pay, so there you go.
    Also regarding a video I saw you in and you mentioned that this is not training the pilots receive as they stress taking off and landing. My other half says that his airline constantly pushes these tests when they are training/certifying in the simulators. So maybe some airlines don’t do this, but at least one in U.S. does.

    Thanks again!

  45. mchova01 Says:

    Miles, I miss your expertise on TV – I’m glad to see your passion continue in your blog – well done, Sir. As an aviation enthusiast, and a technology expert I came to the same conclusion. It doesn’t seem out of the realm of possibility to have the plane’s data points and voice recorder streamed and do away with the “black boxes”. This way we could have a precise record of fact and just focus on recovery efforts vs. recovery and locating black boxes. It seems like a satellite-based, On-Star-like replacement would be in order and well within technology limits. I was surprised that the plane did phone home with updates of a catastrophic event, yet it was several hours later before anyone noticed. At the very least, loss of cabin pressure should trigger a workflow that wakes up the CEO of ay airliner. My sympathy goes out to all affected persons in this tragedy.

    • Miles O'Brien Says:

      It seems silly to me when people say there is not enough bandwidth. It is not a lot of data in the grand scheme.

      • mzen Says:


        It is actually a huge amount of data. You referenced the space shuttle, and rightly so. But that is one vehicle and it does not, can not, transmit all the data the flight recorder compiles. When you consider a fleet of several thousand planes, all transmitting, the technology is not ready yet and the manufacturers and operators would not foot the cost that it would require. What you are speaking of would require the ability to transmit possibly hundreds of GB of data per flight, wireless, while moving at 500 mph. When you place a plane over water with receiving stations possible thousands of miles away it just isn’t feasible.



      • Miles O'Brien Says:

        Why not use Inmarsat or Globalstar? And perhaps just limit it to over the pond flights.

      • mzen Says:

        Miles –
        Yes, that is a possibility. You would have to limit the diagnostic transmissions to only the super critical systems: propulsion (in which there are at least 10 sensors per engine recording data at the rate of milliseconds), avionics, controls, ect. However, I think that this accident was unavoidable at the time it became evident that a problem had occurred.

  46. aura Says:

    Hi Miles,

    I found you through an ABC vid on Yahoo! I was surprised to hear you are no longer with CNN. I enjoyed your reporting on CNN, but haven’t had time to watch TV lately with the depression. The suits at CNN must be crazy to let you go. I have been linking to your articles on the crash and it is burning up my blog. Links to your blog posts are #1, 2 and 3 on my wordpress blog. The horrific story of what happened to doomed AF flight 447 seems to be coming together slowly based in part on your excellent research, assessment and reporting, along with the condition of the wreckage found, information surfacing about the pitot malfunction problems and the weather. Taking it all into consideration, I have cobbled together my own amateurish scenario of what transpired in the final minutes of that tragically doomed flight. The pitot sensors were telling the plane’s onboard computer system that the plane was traveling slower than it actually was, thereby allowing the plane’s rudder to get into a position it shouldn’t have been in because in actuality the plane was going too fast for the rudder to be in that position. Since the rudder was in the wrong position for such excessive air speeds, the immensely important tail of the air craft was sheered off by the elements, kicking off a mortal catastrophe the likes of which Air France has never seen. With fuselage breached and the rudder ripped away, the situation was beyond dire. It was sheer doom. There was no action the flight crew could have taken to salvage the aircraft and safeguard the passengers. It was curtains for all aboard. With the loss of the all important rudder, which in turn breached the fuselage, the plane rapidly depressurized, sucking out some of the terrified passengers and crew. The panic stricken fliers, hearts pounding uncontrollably with adrenalin, were forced to face the unfathomable reality of their situation – the plane was coming apart around them. The passengers and crew not initially sucked out of the fuselage during the initial rapid depressurization (as well as the ones who were) were pummeled mercilessly by the full force the violent elements enveloping them, as they rode their stricken airliner to its doom in a death plunge so terrifying its enough to make one never want to fly again. What was left of flight 447 with remaining passengers aboard was torn to shreds by the brutal winds as it fell to earth until alas, its tragic journey ended in the dark depths of the Atlantic, where passengers and crew were scattered in a sea of death. How utterly heartbreaking. The terror experienced by those fliers must’ve been off the richter scale. This scenario is every fliers worst nightmare. We all know the odds are against us if we are unlucky enough to end up on a failed flight. I dare say this tragedy is spooking the flying public like no other. How could God in all His mercy allow those people to suffer such a cruel fate? How could God be so compassionate, yet so cruel? I hope those passengers asked Him that when they got to heaven. May their continued journey through paradise be a peaceful one. God bless them and you Miles for enlightening us to the facts behind this painful and terrifying tragedy.

  47. mzen Says:

    Hi Miles,

    Thanks for the follow up story. Couple of thoughts. Excessive hail in the engine is a possibility, but not completely probably. When those engines are in development they go through the “Frozen Turkey” test. They literally throw a frozen turkey into the compressor intake. But as we saw with the geese on the Hudson River miracle, a system can only handle so much.

    Secondly, you spoke of real time system diagnostics being sent to ground stations, a kind of black box on the ground. This type of technology is being heavily researched for the JSF and other systems, but it is just not mature enough, or practical at this point for commercial craft (due to the volume of commercial air traffic and the cost associated. Military applications will use it but the cost is significantly increased and sometimes not a deterrent). It may be probably for commercial transports traveling over ground, but the power needed to send that huge amount of data a thousand to several thousand miles is just not practical.

    Anyhow, thanks again.


  48. arlen Says:


    Thanks for your info on this crash. There is a lot of incorrect info being given out by the media outlets which only adds to the level of confusion and upset of the general public.

    I am a instrument rated private pilot and an Electronics Engineer.

    I agree that aircraft need some other means other than the CVR and FDR “Black Boxes” for sharing the information that may lead to the prevention of future aircraft accidents.

    As far as bandwidth concerns, I think that is a red herring, the more probable concern is privacy for the flight crew, and its operating airline.

    In a perfect world, this would not be the case, but that is not the world we live in.

    So, why can’t we strike a balance between broadcasting potentially sensitive information and having the needle in the haystack search for the FDR and CVR that often occurs after a mishap?

    My idea is to equip the aircraft with a small device in the tail that can contain copies of the FDR and CVR info and be ejected from the aircraft prior to the aircraft impacting terrain or water. It could be triggered by rate of decent through 10,000 feet to automatically eject, as well as be commanded to eject by the flight crew if they know they are about to loose control of the aircraft or ditch the aircraft.

    The device would have a parachute and ELT transmitter becon to aid its survival during its decent and its location. It would transmit both its current location and the last known position of the aircraft before it ejected.

    Your thoughts?

  49. True/Slant Anniversary: Some of the best analysis and perspective published in our first year - Michael Roston - Newsbroke - True/Slant Says:

    […] his knowledge of aviation to write a pair of brilliant, in-depth blog posts – the first on June 1, the second a week later – that debunked a lot of the myths and explained a lot of the facts […]

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: