The Devil at 37,000 Feet

William Langewiesche writes about the collision between a Gol Airlines 737 and an Embraer Legacy 600 jet:

http://www.vanityfair.com/magazine/2009/01/air_crash200901?printable=true&currentPage=all

Quite a chain of unfortunate events... :frown3:

It always is... and while it's nice to think of how weak each link is, and how easily the chain can be broken, it's scary to consider how many of these chains form routinely, then are broken before a disaster can occur. I mean, most of us have caught some small error (a heading, an altitude, a switch position) in time, and didn't think much of it at the time.
 
What a great write up. I honestly now understand some of the Brazil government's position too. I do not agree with it, but I can see how they arrived at this point.

I found the following particularly compelling and well written.

I felt the descent in my stomach as I read this account.

The Legacy came streaking at the Boeing about 30 feet to the left of the fuselage and 2 feet lower. The displacement was infinitesimal on the scale of the sky, and a measure of impressive navigational precision. The Legacy’s winglet acted like a vertically held knife, slicing through the Boeing’s left wing about halfway out and severing the wing’s internal spar. The outboard section of the wing whipped upward, stripping skin as it went, then separated entirely, spiraling over the fuselage and demolishing much of the Boeing’s tail. In the Boeing’s cockpit the sequence sounded like a car crash. Instantly the Boeing twisted out of control, corkscrewing violently to the left and pitching straight down into a rotating vertical dive. The cockpit filled with alarms—an urgent klaxon and a robotic voice insistently warning, Bank angle! Bank angle! Bank angle!, as if the crew might need the advice. Back in the cabin the passengers screamed and shouted. The pilots reacted as one might expect, fighting desperately to regain control. They probably did not know what had gone wrong. They certainly never mentioned it. What is unusual is that they also did not swear. Ten seconds into the dive, one of them did cry “Aye!,” but the other urged him to stay calm. “Calma!” he said, and seconds later he said it again. If pilots must die in an airplane, all would choose to finish so well. Of course these two knew they were gone, but they did what they could, even extending the landing gear to slow the dive. The gesture was hopeless. Twenty-two seconds into the plunge the airplane’s over-speed warning came on with a rattle that continued to the end. Forces inside the airplane rapidly grew until, 30 seconds into the dive, they exceeded four Gs—the gravity-load threshold beyond which some passengers must have begun to black out as the forces drained blood from their minds. Maybe they were the lucky ones or maybe it didn’t matter. In the cockpit the pilots kept trying to fly, struggling with the controls and exchanging a few words which are impossible to discern over the bedlam of alarms. Forty-five seconds into the dive came another “Aye!” Seven seconds later, at 7,000 feet, the Boeing broke into three parts, which plummeted in formation into the forest below.
 
Any relationship between William Langewiesche and Wolfgang Langewiesche, author of Stick and Rudder?

Edit:
Answered my own question:

http://www.amazon.com/America-Air-Aviators-American-Classics/dp/0801878195/ref=sr_1_8?ie=UTF8&s=books&qid=1228768515&sr=1-8 said:
In a new foreword [to "America from the Air: An Aviator's Story"], Langewiesche's son, writer William Langewiesche, describes his father's love of the view from above.
 
Last edited:
Any relationship between William Langewiesche and Wolfgang Langewiesche, author of Stick and Rudder?

Edit:
Answered my own question:
Jr is quite a prolific writer. I enjoyed reading his book, Inside The Sky, and would recommend it to any pilot.
 
I teach to look for potential issues. If something is off or at the very least doesn't seem right, don't accept it. If you see a flashing light or any indication of a change, address it or disable it. Don't ignore and certainly don't accept it.

If you can't find a safe and viable solution that will not affect the safety of flight, put the plane back on the ground. To do otherwise is an invitation to further links and worse.

If my transponder appears to have issues, I'll get a call from center indicating such. BUT, the transponder is also (and SHOULD be) part of my scan. As PIC, I should be looking to see what altitude is being indicated on the transponder. I should be looking to confirm some occurrence hasn't changed the squawk code. I should be looking for the reply light to an interrogation by radar.

If Center isn't calling me, I should be calling them soon enough to inquire of a correct radar return.
 
If my transponder appears to have issues, I'll get a call from center indicating such. BUT, the transponder is also (and SHOULD be) part of my scan. As PIC, I should be looking to see what altitude is being indicated on the transponder. I should be looking to confirm some occurrence hasn't changed the squawk code. I should be looking for the reply light to an interrogation by radar.

FWIW, I've never flown any turboprop or jet that had a transponder with a reply light.
 
Did you listen to the MP3's of the cockpit audio? "Just fly the airplane". 600XL right after they got hit. "Just fly the airplane".

Good advice. They were remarkably collected, despite the F-bombs.
 
I'm annoyed by the moralistic tone describing the Legacy - the pandering to the environmentalists and the class-warfare types to whom a bizjet is a manifestation of the horned one.

There's also speculation in the article about the transponder status - that may or may not be backed up by evidence. If the evidence shows that the Legacy transponder went to standby, then that's one thing, regardless of whether it was a malfunction of the airplane or the crew.

The one thing I am taking away from this is the need to know transponder status, and probably there should be an alert when a system like TCAS becomes unavailable (whatever the reason) in flight.

I'm unpleasantly surprised by the Legacy's crew having problems with the FMS and the automation of the airplane. I'm aware that the most common thing heard in jet training is "What's it doing now?" in regards to the autoflight systems - the same thing applies in our G1000/GFC700 cockpits. But none of the instructors at the places I use would sign off a pilot as PIC in a G1000 airplane with the level of confusion that these guys showed. Looking up the airport configuration, field length, and frequencies, or checking the flight plan to figure out time remaining is all basic stuff.

The crew dynamics were interesting too. I am still muddling through my feelings about them - more later.
 
Last edited:
I'm annoyed by the moralistic tone describing the Legacy - the pandering to the environmentalists and the class-warfare types to whom a bizjet is a manifestation of the horned one.

There's also speculation in the article about the transponder status - that may or may not be backed up by evidence. If the evidence shows that the Legacy transponder went to standby, then that's one thing, regardless of whether it was a malfunction of the airplane or the crew.

The one thing I am taking away from this is the need to know transponder status, and probably there should be an alert when a system like TCAS becomes unavailable (whatever the reason) in flight.

...

The crew dynamics were interesting too. I am still muddling through my feelings about them - more later.

I'm with ya there Tim.
 
As an aside, many glass airplanes (including the latest G1000 software) allow you to program an "along-route offset" that will fly the airplane a certain distance left or right of your flight plan route. This ensures you won't smack into someone else tracking the same path with the +/- 20 feet precision we get from GPS.

In my old airplane, we'd set the autopilot (an analog King KFC150 IIRC) so that it would fly about 3/4 dot right of center in NAV mode, which had the same effect. There was a separate adjustment for the LOC/APPR mode, so it would track a localizer spot-on.
 
William Langewiesche writes about the collision between a Gol Airlines 737 and an Embraer Legacy 600 jet:

http://www.vanityfair.com/magazine/2009/01/air_crash200901?printable=true&currentPage=all

Quite a chain of unfortunate events... :frown3:
That was interesting. With the accuracy of navigation systems and autopilots these days I can see how the big sky theory doesn't apply any more.

The author seems to take a dim view of technology as he compares it unfavorably with the low-tech natives. It's true that technology can distract you and lead you astray pretty easily. That's why they teach you to get rid of the technology and go back to the green needles (VORs) if you are ever confused about where you are going. Speaking of confusion, I can relate to the confusion with the box. I fly an airplane that has very similar avionics made by Honeywell. It isn't very intuitive. These pilots were taking delivery of a new airplane which they had probably only flown a couple times with the factory pilot. If it was a conformity flight before delivery like I did with the Sovereign, the factory pilot probably managed the box. They don't go into the avionics that deeply in school, especially regarding what you would use on a cross-country flight, because there is not enough time. They basically teach you enough to be safe and pass the checkride. One of the pilots had flown a similar airplane for an airline so he was probably the one doing the explaining to the other pilot. As far as the transponder goes, there is no blinking reply light. As is says in the narrative, the only indication that the transponder was off would be that on the radio page the word STBY would be bold instead of another mode (usually TA/RA). The "TCAS OFF" message is in very small letters on the MFD and not in the place where you usually receive messages. I don't know if this accident will prompt them to change the software so that the transponder being off will produce a message in the CAS (crew alerting system) box where the informational, caution and warning messages are normally posted.
 
Well, now that I've listened to the Legacy voice recording... I'm ****ed. Neither one of the two pilots appeared to know how to operate the avionics in the airplane. Stick-and-Rudder skills are important and they did well flying, but if you're going to put your charts in the back of the cockpit and go paperless, you damn well better know how to operate the electronic substitutes as well as you know how to fly the airplane. They didn't know if the airport was suitable until practically on final! If the training for the type rating (or whatever subsequent checkout for insurance or PIC purposes) doesn't cover and test the avionics, that's a real problem with the standards.

The crew managed the emergency, but the captain, in my opinion, didn't retain command. The FO did a good job of stepping up, but there wasn't much crew coordination afterward. I keep thinking that the captain's job is to manage the situation and have the FO do the flying - here the FO was flying AND managing and the captain seemed sort of subdued, but this MAY have been his way of avoiding conflict in an emergency situation - which is a good thing.

I'm also pretty sure I LOVE US ATC after hearing the alternatives.

I'm now sure that there needs to be better alerting when TCAS is not on.
 
Last edited:
I'm also pretty sure I LOVE US ATC after hearing the alternatives.
Yeah. I think I'll stick to flying over the United States where I can understand what is being said....
 
Quite an article. Thanks for posting.

I, too, noticed the digs at the biz jet and biz jet culture. Earlier in the article it seemed there was an effort to stay balanced. I wonder how much of that can be assigned to the author vs the VF editor? Hard to say. . .

Once again, thanks for posting.

Terrible tragedy.
 
That was an amazingly good article. Riveting. Informative. Balanced. Wow.

My first thought was, we have already gone through this before. Why bring it back up again. My mistake. Amazing writeup.
 
The one thing I am taking away from this is the need to know transponder status, and probably there should be an alert when a system like TCAS becomes unavailable (whatever the reason) in flight.

Good point. On the G1000, you're told when traffic is unavailable. Does TCAS similarly alert?

I'm unpleasantly surprised by the Legacy's crew having problems with the FMS and the automation of the airplane. <snip> Looking up the airport configuration, field length, and frequencies, or checking the flight plan to figure out time remaining is all basic stuff.

I'm of the impression, perhaps incorrect, that the field length and frequencies were not in the FMS database, as it was a military field. However, the identifier and lat/long were in the database, so they were able to "nearest field" direct-to it. You can hear them on tape having to pull out paper charts to try to find data on the field, and after going through all the charts, they weren't able to find it. Had to get the info over the air from the tower.

The crew dynamics were interesting too. I am still muddling through my feelings about them - more later.

Agreed, and I'll be interested in hearing your thoughts. Sounds like the captain didn't exercise PIC; the SIC took over--even apologizing later for doing so, with the captain saying, effectively, that he was glad the other guy did. The SIC certainly appears to have known more off the top of his head about airplane configuration and speeds.
 
I'm of the impression, perhaps incorrect, that the field length and frequencies were not in the FMS database, as it was a military field. However, the identifier and lat/long were in the database, so they were able to "nearest field" direct-to it. You can hear them on tape having to pull out paper charts to try to find data on the field, and after going through all the charts, they weren't able to find it. Had to get the info over the air from the tower.

You may be correct, but I'm of the impression that they couldn't work the system to get the airport information, not that the information wasn't in the system. It's been my experience that if the ICAO identifier is in the database, the basic information like runways and runway lengths are in the database too. But this may be an erroneous assumption on my part.

In my trainers for the SMITHS and UNS FMS, SBCC does exist.

If the case is that the Identifier is in the database, but there's no information associated with it, that's another thing that should be addressed - perhaps the system shouldn't identify airports with the "nearest" function if it has no useful information about them.
 
If the case is that the Identifier is in the database, but there's no information associated with it, that's another thing that should be addressed - perhaps the system shouldn't identify airports with the "nearest" function if it has no useful information about them.

If that were the case--they would have never found the airport...which wouldn't have helped matters.
 
Barry Schiff wrote an article in AOPA recently about the trend towards more and more accurate navigation as time has gone on.

said that his original instrument instructor, while teaching him to fly an A-N range, said that there is no way in hell two individuals could both simultaneously fly an A-N range accurately enough to present a collision risk over the station. In the few instrument students ive had, ive seen several have a really hard time realizing that some of the more basic navigation (ADF/VOR) does not give pinpoint accuracy, and that is OK.
 
If that were the case--they would have never found the airport...which wouldn't have helped matters.
No - they'd have been given another airport that the system DID have information for. The airplane wasn't falling out of the sky, or leaking fuel, so landing right away was a choice they made, rather than had forced on them... Perhaps they could have found another field almost as close.

Having played with my FMS trainers now, I was able to pull up information for SBCC - so I'm back to my original thought that they didn't know how to get the info out of the system. I find runway 12/30 at 8537 feet.
 
No - they'd have been given another airport that the system DID have information for. The airplane wasn't falling out of the sky, or leaking fuel, so landing right away was a choice they made, rather than had forced on them... Perhaps they could have found another field almost as close.

I have a hard time faulting the crew's decision given this:

The left winglet had torn away, leaving a jagged stump, the left wing had bent upward, and along its upper surface some of the skin was separating from the internal structure. The passengers could see the heavily deflected aileron necessary to maintain control.

When he tried to slow below 230 miles an hour, he had a harder time maintaining wings-level flight, and had to accelerate again.



Trapper John
 
Good point.

I will stick by my original theory that the crew didn't know how to get the information they needed from the system.

Note that I'm not blaming the crew for the accident (at least not solely - there's plenty of contributing factors). I'm just saying that when the ca-ca hit the fan they weren't up to the standards of proficiency with the systems that I'd expect from a professional crew flying passengers. Caveat: This is solely the impression I get from the CVR. If I'd been in the cockpit with them I might have a different impression based on body language and other non-audible cues.
 
Last edited:
Good point.

I will stick by my original theory that the crew didn't know how to get the information they needed from the system.

And that was keeping their heads down at a very critical portion of the flight.

Overall, it would seem the most important info was "transponder standby" and "TCAS Off."
 
And that was keeping their heads down at a very critical portion of the flight.

Even if they were heads-up and eyes-peeled, would they have seen the oncoming 737? Let's say they were looking. At a 1,000 mph closure rate, if they would have seen the 737 at 3 miles away, they would have had only about 11 seconds to identify, process the situation and evade.


Trapper John
 
Troubling to me is the omission of the fact that shortly after this accident Embraer discovered a wire-routing flaw that meant that, if you stepped on or kicked the boot at the base of the control column just so, the transponder would put itself into standby mode.
 
Troubling to me is the omission of the fact that shortly after this accident Embraer discovered a wire-routing flaw that meant that, if you stepped on or kicked the boot at the base of the control column just so, the transponder would put itself into standby mode.


Can you cite a source for that statement?
 
Troubling to me is the omission of the fact that shortly after this accident Embraer discovered a wire-routing flaw that meant that, if you stepped on or kicked the boot at the base of the control column just so, the transponder would put itself into standby mode.

Can you cite a source for that statement?
Ditto, that would really make this case more interesting.

This is sure not as cut and dried as it appeared from the initial stories!
 
As for the cause, as I see it:

The primary system - ATC - failed to keep two airplanes heading in opposite directions on the same route vertically separated.
The secondary system - Transponder TCAS - was unknowingly disabled on the Legacy.

The causes for these two failure can (and will) be debated for years to come.
 
Even if they were heads-up and eyes-peeled, would they have seen the oncoming 737? Let's say they were looking. At a 1,000 mph closure rate, if they would have seen the 737 at 3 miles away, they would have had only about 11 seconds to identify, process the situation and evade.

Agreed. A 737 at FL370 cruise doesn't have all his lights on--just position lights. If the GOL flight had landing lights on, this incident might not have occurred, but they just don't run with all those lights on at cruise.

Rich: I wouldn't call "cruising level at FL370" a "critical phase of flight" -- am I misinterpreting your statement?

One more thing: I've wondered, since this first happened, how a collision that just ripped the winglet off an Embraer business jet could have caused enough damage to the 737 to bring it down.

Listening to the 737 audio, there is a loud collision, sounds like a lot of wind (pressurization breached?), the BBetty starts calling out "bank angle" warnings immediately, and repeatedly. Sounds like an overspeed warning horn, too.

Has any info been officially released as to the impact angle and locations? Were there any identifiable witness marks on the 737? Did the 737 lose part of its wing? Ingest the winglet? What happened to the 737?
 
The article describes what happened to the 737. I'm assuming that's based on crash evidence and not artistic license, but who knows?
 
The article describes what happened to the 737. I'm assuming that's based on crash evidence and not artistic license, but who knows?

OK, I have to go back and read it now. :-) I only had time to skim through it, listen to the post-impact audio on the two feeds, and bookmark it for later.
 
In an eerie coincidence, I nearly had a mid-air today on VATSIM, where I was at FL350 and traffic at the opposite direction was at FL350. He climbed to FL360 (which was appropriate for his direction of flight) and we both got TCAS advisories.
 
Interesting article. Thanks, Tim, for pointing out that it discussed the impact location... the severing of a portion of the 737's left wing correlates with the audio that I heard.

We've discussed, in other threads ad nauseum, how our "improved" navigation and autopilot systems have tightened tolerances to the point where accidents like this are possible. There is no "slop" to accomodate human error or inattention.

With this level of accuracy, and with GPS and NAV systems that can fly displaced courses to the left or right of a courseline, why do we not implement a "fly right of centerline", like we do on the roads? (would you have to fly left of course in Great Britain, India, and Australia?) Yes, you could still run into the back of someone, but the closure rate is greatly reduced, giving you more time to react. Heck, on IFR clearances, ATC could even assign same-direction same-altitude traffic specific offsets... me at .25 off course, Spike at .50 off course, Dave at .75 off course. We'd all still be WELL within the course protected area. If you don't have an autopilot that can fly an offset course, even being told to fly "two dots" to the right of the course centerline would provide the clearance needed.
 
Rich: I wouldn't call "cruising level at FL370" a "critical phase of flight" -- am I misinterpreting your statement?

By critical, I meant in the moments before impact, they seemed preoccupied with something inside the cockpit.

Paladino said, “Did we hit somebody? Did you see that? Did you see something?”Lepore was hesitant. “I thought I saw … I looked up … ”
I really don't know if see and avoid could've worked in the situation. I do like your idea of "right of centerline" though.
 
With this level of accuracy, and with GPS and NAV systems that can fly displaced courses to the left or right of a courseline, why do we not implement a "fly right of centerline", like we do on the roads? (would you have to fly left of course in Great Britain, India, and Australia?) Yes, you could still run into the back of someone, but the closure rate is greatly reduced, giving you more time to react. Heck, on IFR clearances, ATC could even assign same-direction same-altitude traffic specific offsets... me at .25 off course, Spike at .50 off course, Dave at .75 off course. We'd all still be WELL within the course protected area. If you don't have an autopilot that can fly an offset course, even being told to fly "two dots" to the right of the course centerline would provide the clearance needed.


We do - see my earlier post on offsets. They're common in North Atlantic airspace
 
Back
Top