Air France Flight 447: A detailed meteorological analysis | Hacker News Reader

Air France Flight 447: A detailed meteorological analysis

weathergraphics.com

100 pointspmikal26 comments

For those of us who don't follow news much, there's more factual detail here: http://en.wikipedia.org/wiki/Air_France_Flight_447

There were 228 people aboard the flight including three pilots and nine other cabin crew. The passengers were one infant, seven children, 82 women and 126 men

It is doubtlessly an unfortunate tragedy :-(

Nonetheless, everyday over 3000 people die in car accidents all over the world. Over 5000 people die everyday due to preventable water diseases. Over 3000 people commit suicide every day. One can also argue that these tragedies could be prevented. From this perspective, it's all but cold statistics...

Just the idea of a plane vanishing, gone, with well over 200 people on board, boggles the mind. No distress calls?

I can only imagine the sudden chaos that erupted.

Technology v. Nature.

perhaps there was rapid decompression caused by the turbulence caught the pilots without their masks on, and they fell unconscious, thus no distress call, and a crash. there could have been a structural failure and the airplane may have broken up on descent to a lower altitude. remember the American Airlines crash in New York in October, 2001, where the vertical fin broke off, it was an Airbus A300. The BOAC crash caused by turbulence near Mt.Fuji in 1966 was a massive structural failure as well. Hard to believe the pilots would not have seen this weather on the radar and steered around it. Then again, they depend on meteorological reports and guidance from the ground in these situations, and may have been waiting for instructions. I've never flown through severe turbulence (few have and lived to tell about it) but I've experienced light turbulence, the kind where your butt leaves your seat very briefly, and it's not fun. 4 hours of that and you have a back ache.

As a pilot, I'm tempted to speculate on various crashes.

I've learned that this is a bad habit.

Best case, you get the cause right, and you look like a jerk. Worst case, you get the cause wrong, and you look like a jerk.

I will say that in general, flameouts due to ice ingestion and the complete fly-by-wire system of the Airbus have been generating interesting conversations in the aviation community for some time. Whether this has anything to do with the current situation is anybody's guess. I feel very sorry for all of those involved.

For an interesting counterpoint on 'fly-by-wire-is-bad' read the Vanity Fair article about the Airbus that landed in the Hudson and how the fly-by-wire augmented the pilot's own ability.

http://www.vanityfair.com/style/features/2009/06/us_airways2...

Boy, there is some fabulous choice of words in that Vanity Fair article. For example in the part telling the story of an Airbus 330 that ran out of fuel over the Atlantic. Both engines are going out, the flight attendants have instructed passengers to don life vests, the main lights have failed, the PA system goes dead and then finally "as the cabin pressure leaked away, the oxygen masks automatically dropped, and this caused another round of fussing."

Yeah, I bet those passengers were "fussy" alright.

I don't think anybody doubts the benefits of fly-by-wire. It's the failure modes that concern some.

Not that this issue bears on this particular incident. I was just listing the topics of conversation that come up when talking about the airbus and turbines and T-storms.

I'm also about halfway through my private pilot license and while I don't know enough yet to even speculate on crashes, I really enjoy reading the NTSB reports or crash reports in different aviation publications. It probably strikes many people as morbid, but I find it fascinating to try and learn from the mistakes of others so that I can be a safer pilot myself.

Continuing visual flight rules into instrument conditions is the pilot error most likely to kill you. If the weather starts to look bad land - now. Even on a golf course or a field if you have to

http://www.humanfactors.uiuc.edu/Reports&PapersPDFs/isap...

The private ticket is one of the most exciting things I've ever done. Follow it up with some more training, like an instrument rating or some aerobatics (Emergency Maneuver Training)

Aviation Safety magazine, in my opinion, has the best crash reviews and safety material. I think by reading those reports they can keep you alive -- it's a very good habit.

Somebody once said that the first 1000 hours of flying are the time you're most likely to kill yourself -- especially in the 200-400 hour range, and especially if you let your skills lapse by a few months.

I really wish that technology would learn lessons like aviation does. If we had the same feedback and crash analysis as aviation, perhaps so many projects wouldn't waste so much time and money.

I believe you're talking about The Killing Zone, by Paul Craig. Highly recommended.

http://www.amazon.com/Killing-Zone-How-Why-Pilots/dp/0071362...

It is human nature to speculate. But at least the article was an attempt at a thorough analysis rather than "it was lightning! It was a bomb!". I have nothing against instructive speculation, even if it turns out to be wrong.

And anyway, there is very little opinionating really, all he really said is "turbulance was almost certainly a contributing [my emphasis] factor".

Does a gradient apply to fly by wire systems?

You and others seem to imply that Boeing/other aircrafts can be less(?) fly-by-wire. How does that work? Shouldn't there always be redundant systems?

Many things he discusses are well above my level of comprehension but turbulence is likely to be the main cause of the crash although other planes passed that same region without problems. It sounds likely to me that heavy turbulence just 'broke' inner electronic equipment/circuits/meters/connections?

This particular airplane (F-GZCP) was involved in a ground collision incident with another Airbus in 2006, where its wing hit the other Airbus' tail. The wing was repaired.

An improper fix would certainly lead to an in-flight breakup after many operating cycles (see Japan Airlines flight 123 - http://en.wikipedia.org/wiki/JAL_123; China Airlines flight 611 - http://en.wikipedia.org/wiki/China_Airlines_Flight_611 for example), especially in conditions of heavy turbulence that this airplane obviously experienced.

Either that or a bomb is my educated speculation.

However, considering the maintenance transmissions were spread over 4 minutes, I'd suspect the plane didn't entirely breakup within a shorter period of time.. although this doesn't rule out that a bomb or breakup was the straw that broke the camel's back - just that it probably wasn't the sole cause.

The only thing people really know right now is that the plane is missing.

Although Tim's report suggests structural failure from turbulence, it really could be one of many possible causes. If you enjoy the speculation, which I do, check out the Airliners discussion:

http://www.airliners.net/aviation-forums/general_aviation/re...

"Terrorism has been ruled out"

I hate it when they try to manipulate mass opinion when they haven't even found the damn plane.

The first thing that comes to my mind is a bomb in a suitcase blew the plane in pieces.

Occam's razor.

I think you're probably not being serious, but just in case.. you probably aren't going to be getting 4 minutes of maintenance transmissions from a plane that's been blown into pieces ;-)

Totally serious. Nowhere I've read about those 4mins you claim.

Here is what I've found:

The Washington Post: "among the key questions are how long the plane kept flying after its last automatic satellite transmission and why no mayday call was received from the pilots."

"the aircraft emitted a series of automatic messages via satellite indicating that its electrical system was not functioning and that it had suffered a loss of cabin pressure. Those were the final signals from the plane. "

As soon as it explodes, the transmitters start sending info about electrical failure and cabin pressure, of course, and they do so for as many minutes as they fall from 30K feet.

Why there was no mayday call?

They were already dead. Remember Columbia.

It's in the CNN Article (the four minutes). Here is the relevant bit:

"But about 4:15 a.m. Paris time, Flight 447's automatic system began a four-minute exchange of messages to the company's maintenance computers, indicating that "several pieces of aircraft equipment were at fault or had broken down," he said."

You can read the full article here:

http://www.cnn.com/2009/WORLD/americas/06/01/air.france.braz...

My instincts tell me that if power was lost, they wouldn't be able to send a mayday call. Of course, if total power was lost, how was the plane able to communicate with the server?

A bomb could have caused this explosion, but doesn't fall within the usual terrorist M.O. No group has stepped forward to claim the attack- half the "fun" of a terrorist organization is spreading fear through your message.

Additionally, it would not make sense for a terrorist to blow up the plane 4 hours into the flight over the Atlantic Ocean. It would make more sense to blow it up as soon as it takes off or lands, to maximize damage on the ground as well.

I am not a pilot, nor a security expert. This is all just my street-instincts talking.

> No group has stepped forward to claim the attack- half the "fun" of a terrorist organization is spreading fear through your message.

I don't have any reason to think this was a terrorist attack, but I have the impression that (credibly) claiming responsibility for attacks is less common than it used to be. The Mumbai attacks last year, for example, were just miscellaneous mystery terror.

I agree with you, but to use your Mumbai attack example, "everyone" (everyone I've read) basically believes that the Pakistan CIA-equivalent helped sponsor the attacks. An organization taking responsibility for the attack would lead back to this organization, which would involve a lot of political drama.

As long as Pakistan doesn't openly claim responsibility, there's no real loss in political capital to them- Indians hate them anyway, and any time an attack happens, India seems to rattle its sabers for a few months before backing down and they remember that Pakistan is also nuclear capable.

Then business returns as usual.

From my little understanding, a total loss of power usually means from the engines. There is backup generators that can be used to restart the engines during flight, however anything could have happened to stop the engines restarting, but allowed the electronics to remain active.

I personally highly doubt terrorism. It's a flight from Rio to Paris, it's not exactly high risk for terrorism, plus no group has claimed the downing (to my knowledge). This means it was a lone psycho, however the chance of getting explosives onto an airline is kind of low. Plus if it was an explosive detonation, the damage would have caused explosive decompression and damage to the air frame. I'm sure the information Air France received would have indicated an explosion, likely as there would have been a complete disconnect between cabin and electricity (engines or backup generator).

"if total power was lost, how was the plane able to communicate with the server?"

The same way black boxes keep beeping even twenty thousand leagues under the sea.

The lightning was the first red herring, but it got quickly dismissed by experts. Now they say it's gonna take years to know the cause, so we can forget about it.

Reply on news.ycombinator.com