Talk:Gol Transportes Aéreos Flight 1907/Archive 1

Archive 1

Archive 2

Archive 3

→

Archive 5

"recieved"

What means the word "recieved" in the text ? Is it an alternate spelling of "received" ? —Preceding unsigned comment added by 201.90.132.168 (talk • contribs) 20:30, September 30, 2006 (UTC)

Where do you see it? What section? Crum375 20:39, 30 September 2006 (UTC)

Called a typo, an easy one to make --SGGH 20:45, 30 September 2006 (UTC)

Distance from Peixoto de Azevedo

This CNN article says that the wreckage was found only 20 miles (30km) from the city of Peixoto de Azevedo, while this article says 200 km (124 miles). Is there a source for the 200 km fact or should it be changed to what CNN is reporting? -albrozdude 20:42, 30 September 2006 (UTC)

The source is in the article. [1] antiuser 20:45, 30 September 2006 (UTC)

The best possible source currently is [2], which does say it's 200km. --cesarb 21:05, 30 September 2006 (UTC)

Looking at the crash site coords on GoogleEarth I am pretty sure CNN is wrong here. But there are closer towns than 200km away, so let's wait to get a better fix. Crum375 21:14, 30 September 2006 (UTC)

Per this source[3]: "Segundo a FAB, a localização exata é de 10º 29' sul e 53º 15' oeste.", i.e. 10º29'S, 53º15'W. Crum375 21:17, 30 September 2006 (UTC)

The closest GoogleEarth town I see to the crash site is Sao Jose do Xingu, from which it is about 41 miles nw. But we can't use it (WP:OR) until we get a WP:RS using it (or something else closer or bigger). Crum375 21:29, 30 September 2006 (UTC)

It is likely that the authorities are referring to the closest city with hospitals and/or rescue equipment, instead of just referring to a small village with no infra-structure. This is complete speculation from my part (which is why it's in the Talk page). - antiuser 21:34, 30 September 2006 (UTC)

Makes sense. The accident report will usually refer to the nearest populated landmark, if available, but we'll see. Crum375 21:40, 30 September 2006 (UTC)

It seems our Brazillian colleagues have this entry: São José do Xingu. Crum375 21:47, 30 September 2006 (UTC)

Hm, just found out that the (recently added to the article) sources [4] and [5] say it's 30 km... And, being the airline's own press releases, it should be more reliable than CNN (it's harder, however, to compare it with the DAC). If only the FAB site had the information... --cesarb 21:52, 30 September 2006 (UTC)

The FAB says its crash coords are 'exact'. That spot is about 192 east-south-east of Peixoto de Azevedo, so it is consistent. But there are closer (smaller) towns, such as São José do Xingu, only 41 miles away. Crum375 21:57, 30 September 2006 (UTC)

Globo TV Internacional reported yesterday evening that the search-and-rescue base of operations was established at Jarinã cattle ranch, at a location some 40 km from the spotted wreckage. Helicopters used this spot to go back and forth to the crash site.

There is also another report about preparations at the cattle ranch: http://g1.globo.com/Noticias/Politica/0,,AA1292257-5601,00.html

I guess there is a confusion because in some South American contries "Municipio" means city, but in Brazil it refers to an administrative division analog to a county: a much larger area, including rural areas. Some reports may refer to 30 km from the county border.

--Aldo L 20:22, 1 October 2006 (UTC)

There is a complementary report about goverment operations at the cattle ranch:

(in portuguese) http://g1.globo.com/Noticias/Brasil/0,,AA1294156-5598,00.html

Headline: Jarina ranch turns into support base for searches at the crash site.

End of second paragraph: "There are more than 200 people (from the Ministry of Aeronautics) here", affirms Ademir Ribeiro, who manages the ranch.

Aldo L 06:22, 2 October 2006 (UTC)

Legacy section

I added this section in the assumption that as events unfold and the accident is investigated, more information will be avalaible and added here about the Legacy. It is common in collision articles to have at least one section for each of the aircraft. Crum375 22:51, 30 September 2006 (UTC)

Language

Especially concerning the external links and notes, please notice that this is the English-language Wikipedia. Even if you are linking to a document in Portuguese (and please, only do that if there's absolutely no other reliable source in English), it does the reader of this Encyclopedia little service to have the link read "Nota à Imprensa". The regular English-speaking user cannot understand that. Have it say "Press release" or something to that effect. We must keep in mind that this encyclopedia is written in English for the English-speaking audience. Use of the Portuguese language in the article must be kept to a minimal, only the essential. Thanks, Redux 23:09, 30 September 2006 (UTC)

I'm just copying the page's title (or, where it has no useful information, the name used on the link to the page from the site's main page). Feel free to change it to something better; these titles are highly noninformative ("Nota à Imprensa(3)", for instance, says nothing useful beyond it being the third press release). --cesarb 23:29, 30 September 2006 (UTC)

I'd also like to point out that, in this case, most of the international press is getting their information from the Brazilian press, so Brazilian sites are more likely to have the most up-to-date info, while the international ones have to first get the info translated and then cross-checked. As long as we stick to reputable sources (Globo, Terra, JB, Estado and Folha) the information will be accurate. When an English language reference for new info cannot be found, posting Portuguese ones is ok. I guess translating the title is a good idea.antiuser 00:01, 1 October 2006 (UTC)

If you do add a Portuguese link, please consider it to be temporary only, and when you find the same information in english, replace the link. The ideal is to have no foreign language links!213.112.249.100 20:50, 2 October 2006 (UTC)

"Excel Airlines?"

It seems unlikely that the embraer aircraft involved in the accident is unlikely to be owned by Excel Airways, which is a holiday charter company based in the UK, as the link suggests? It is more likely to have been being operated by Excel Air (www.excel-air.com), which operates Private Business Charter Aircraft.

There is also ExcelAire based in Long Island, NY. Obviously better references are needed. Crum375 13:48, 1 October 2006 (UTC)

I recall hearing it was "Excel Air" yesterday on TV Globo, but didn't add because wasn't sure of the spelling. --cesarb 14:15, 1 October 2006 (UTC)

There is also a company in KY, unrelated to the others, with the same "Excel Air" name, also a 'possible' if you google them. Crum375 14:21, 1 October 2006 (UTC)

• This site says it was an Excel Airways jet, they usually should know what they are writing about, maybe it was a jet for internall use? According to List of air carriers banned in the EU another airline with a similar name the Excel Air Services, Inc. (EXI) is supervised by Liberian (!) authorieties and is blacklited in the EU. Mieciu K 14:42, 1 October 2006 (UTC)

Well MAO is Manaus, and at this time all sources seem to say that the Legacy was going to the USA (perhaps via fuel stop, but MAO is not far enough for that). Also, the collision point does not make sense (unless the Legacy made a big deviation) if the destination was Manaus. So, if the MAO is wrong, then the spelling of the company could be wrong too. And I have found that site to be fairly reliable too, but maybe it (like us here) has some 'stabilization time' for accuracy. Crum375 14:51, 1 October 2006 (UTC)

I just found this article that says there was a scheduled fuel stop for the Legacy in Manaus. So I take back what I said above about the www.airliners.net source - they seem to be on the ball. Crum375 15:43, 1 October 2006 (UTC)

The aircraft belongs to ExcelAire out of Ronkonkoma, NY. Here's the FAA info matching the registration number on the Legacy (N600XL):

Type Reservation  	 	 Fee Paid
Mode S Code		51744527
Reserved Date		06/15/2006
Renewal Date		None
Purge Date		07/15/2007
Pending Number Change		None
Date Change Authorized		None
Reserving Party Name		EXCELAIRE LLC
Street		200 HERING DR
		LI MACARTHUR AIRPORT
City		RONKONKOMA
State		NEW YORK
Zip Code		11779
County		SUFFOLK
Country		UNITED STATES

antiuser 17:57, 1 October 2006 (UTC)

Good job antiuser! I guess it is in a 'reserve' status hence not on the usual place, probably pending delivery and US arrival (which may take a while longer now). Crum375 18:17, 1 October 2006 (UTC)

Registration of aircraft is confirmed by pictures released by Brazil Air Force - I have uploaded the image on commons but don't know how to add it to the article... It's here. antiuser 18:25, 1 October 2006 (UTC)

Image of the Legacy aircraft

Can someone who knows how to do this replace the current one with the image at wikimedia commons? I uploaded it to wikipedia by mistake and wikipedia doesn't have the proper licensing template for Agência Brasil. I've marked the image on wikipedia for speedy deletion but don't know how to include images from commons. Help, anyone? - antiuser 20:24, 1 October 2006 (UTC)

Hmmm... not sure, but, as soon as the one on Wiki is speedied, the version from Commons will automatically take its place in the article. Fvasconcellos 20:30, 1 October 2006 (UTC)

Ah, ok. Thanks! — antiuser 20:40, 1 October 2006 (UTC)

Looking closer at the wingtip and tail damage on the Legacy, it is amazing how that tiny amount of damage on the smaller aircraft can be enough to take down the bigger one, even if we assume it hit the 737 in the tail area. One more mystery to solve, besides the obvious ones of how the collision could occur with so many safety systems in place. Crum375 20:59, 1 October 2006 (UTC)

Warning: serious speculation ahead! The rumours I've heard floating around the pilot networks so far say that the Legacy pilot wanted to climb to a higher altitude to be able to fly faster, turned off his transponder (which would make it impossible for ATC or the 737's TCAS to see him) and climbed to FL390. Somewhere around the climb, it hit the bottom/rear of the 737, hitting the horizontal stabilizer. This could've caused damage to the rear pressure bulkhead AND the flight controls, which would be consistent with the nose-dive situation. I find it extremely hard to believe, however, that a pilot (especially if the pilot was indeed one of Embraer's) would turn off his transponder, especially when it's known they were flying over a military-controlled area. We'll have to wait for the CVR and flight data recorders from the 737 to be found to know the real answer. - antiuser 21:03, 1 October 2006 (UTC)

I am also disturbed by my (primitive) graphic over the map image. If the Legacy was really going to Manaus, as was reported by ANAC and others, then he seemed to head too far east to meet up with the 737 which seemed to be on course to Brasilia. Also he would have been getting close to Manaus - why climb illegally out of his assigned 360 to 370 or higher at that point? He would need to descend soon. Also, turning off the transponder, if he knew he was landing in Manaus, would not make sense as he knew that he could be inspected for it in Manaus (including the CVR/FDR). Having a bunch of people from Embraer and Excel Air watching would also make it unlikely. Also, if he needed higher altitude, why not just ask for it? I can't believe he wouldn't have gotten had he just asked for it. OTOH, why all the independent safety mechanisms failed is very mysterious. Crum375 21:23, 1 October 2006 (UTC)

There's more - the newly released pictures of the crash site [6] show the 737 had its landing gear lowered and locked, as if configured for landing or to create extra drag to slow the aircraft down. The gear bay is in one piece, which contradicts a nose-dive trajectory, but the fact that the gear is down and locked implies the crew was conscious when loss of control occurred, lowering the gear to slow the plane down for an emergency landing attempt - however, an emergency landing in a wooded area would require the gear to be UP, to make it harder for the plane to get tangled onto a tree. This gets weirder and weirder as more info is released... User:Antiuser/sig

(outdent) Lowering the gear when the airplane is out of control, or nearly so, is standard procedure (regardless of speed) as the lowered gear can help in regaining control and the gear lowering has saved a few flights in the past from disaster. This would be independent of any imminent intent to land. In addition, the minor damage that occured during the collision to the Legacy means the energy transfer to the 737 would have been similar, so odds are the 737 crew were able to at least try to retain control, and the gear lowering would confirm it. Crum375 21:56, 1 October 2006 (UTC)

Brazilian news sites are using images of both left and right sides of the Legacy to put the damage in perspective, specially for people who are not familiar with this type. Wouldn't that be a good idea for Wikipedia too?

Aldo L 06:36, 2 October 2006 (UTC)

If you can get identical (i.e. perfectly symmetrical) views of the left and right, I would agree. If they are not perfectly symmetrical then many people may still have a problem comparing. I think the current shot we have is pretty good as most people can clearly see the (surprisingly) minor damage to the left winglet and horizontal tail tips. Crum375 12:01, 2 October 2006 (UTC)

The picture of the undamaged side released by Agência Brasil is almost identical to the one of the broken winglet, so I've added it for comparison. User:Antiuser/sig

Just heard on TV

I just heard on TV:

• The Legacy black boxes show it did collide with the Boeing
• It looks like it was the Legacy's pilot's fault; he turned off both transponders, to be able to fly higher and faster (and with them turned off, it disappeared from the radar screens, and the system to warn about colisions couldn't work)

Can someone find online sources for these? I can't be sure I didn't mishear something. (Forgot to add: the TV channel was TV Bandeirantes.) --cesarb 00:09, 2 October 2006 (UTC)

As I said above, it would not make sense at all for them to do it. I'd like to see some confirmation from a reliable source. Crum375 00:15, 2 October 2006 (UTC)

This plane was not built with "stealth" technology, so it should not disappear from radar screens when the transponder was turned off. It should still be a dot on the screen, but perhaps without an identifier or altitude. Do planes of the size of this and the 737 have radar for collision avoidance? Is it completely dependent on the transponder being on? Does the flight data recorder show transponder status?Edison 17:48, 4 October 2006 (UTC)

Both planes were equipped with collision avoidance systems, but those systems need transponder data to work. The Legacy would've shown up on radar but with a weak signal, meaning inaccurate positioning and no altitude/airspeed readings. ||| antiuser (talk) (contribs) 18:08, 4 October 2006 (UTC)

For the record, I believe some military ground radar installations may be able to get an approximate altitude readout on primary (non-transponder) radar returns, whether the transponder is off, on and squawking mode A (non-altitude), or C (altitude). But this capability deteriorates rapidly with increasing target distance from the ground radar antenna (as the target angle with the ground approaches 0), and we have no idea (and may never know precisely) what capability the military has along the affected flight routes, and I have no idea what coordination there would be, if any, between military and civilian capabilities for normal operations. Regarding FDR showing transponder status (off, and mode A or C if on) - it's an important question that needs some research (or someone who knows it off the cuff). Crum375 18:32, 4 October 2006 (UTC)

The Honeywell avionics the Legacy uses does have an integrated FDR which does record Transponder Mode C information (It's the same as the ones on the ERJ-145 and some biz jets like the newer Learjets use a scaled-down version). You can switch off Mode C/S information on this transponder as well as enable or disable TCAS). All this will come out with the FDR analysis. While the SIVAM system installed radar coverage in the north of Brazil, apparently the incident occured in the "handoff" zone between Manaus Center and Brasilia Center. So it could've been that each center was oblivious to the other center's aircraft.--Dali-Llama 19:07, 4 October 2006 (UTC)

I already noted earlier, when the 'pilots turned off the transponder or mode C' theory first surfaced here, that it really wouldn't make sense for them to do it, precisely for this reason. Assume the flight had been uneventful (e.g. the Gol missed the Legacy by an inch and nobody noticed), or even that the Gol flight hadn't been there at all. Having a functional transponder in today's environment is critical. There were reportedly 2 Embraer reps on the flight, including an Embraer test pilot. As the flight landed in Manaus, had there been reports of Transponder or Mode C 'outage' on the flight, which is effectively a 'shake down' flight prior to U.S. delivery, there is non-trivial odds that the FDR would have been inspected in Manaus. If the pilots were so stupid as to turn it off in positive control airspace, at 37,000', they could have ended up in all kinds of hot water once the FDR were inspected. I see no motive at all for them to do something so senseless and take such a risk with their license and careers. It simply doesn't make sense. But of course if in fact the mode C will show up, as you say, on the FDR, then soon we should know more. Crum375 20:38, 4 October 2006 (UTC)

You're only playing with your career if you're caught. The odds of a mid-air collision are so tiny it can just be disregarded. (Really!) Especially over Brazil at 37000 ft. It's plausible deniability: That your transponder signal vanished at a certain point in a flight is very unlikely to be reported or even noted by ATC. If they do report it you say "how strange" and write a line on the maintenance log, if you feel like it. Not particularly honourable behaviour but that type of thing is almost forced on pilots from time to time. {scurrilous speculation}: Salesman says "can this thing really go to 48000 ft?" Pilot says sure. Go on then. Pilot then contemplates asking ATC for permission. Already ATC could have not been particularly compliant, or there could be a newbie at the desk (you can tell) and you don't want to wait 5 minutes while he/she consults and then denies the request. Or you're just a bloody lazy pilot. You flick the transponder off and climb. {/scurrilous speculation} Paul Beardsell 11:20, 6 October 2006 (UTC)

the collision

Is it only me, or is the page actually not describing the actual details of the collision at all?!? I for one don't understand how this accident came to be. Writing "no details of how the collision occurred is currently known" would help visitors understand this, so they don't have to hunt the page - in vain, I might add.213.112.249.100 20:53, 2 October 2006 (UTC)

pre-collision altitudes

The current altitudes (relative to the flight directions) don't sound right, although they could be temporary ones. More verifiation is needed. Crum375 21:35, 2 October 2006 (UTC)

Swapping pictures

Does anybody have a problem if we swap the "crash site" and the "Boeing wreckage" photos? IOW, use the wreckage photo in the crash box, and the "crash site" photo in the recovery section. If we get a better (clearer) wreckage photo it could replace the existing one. Any thoughts? Crum375 22:16, 2 October 2006 (UTC)

I say go for it, you can't see much from the "crash site" picture anyways, and the map does a much better job at showing where it actually happened. Browsing through the Agência Brasil site (which by the way is seriously bogged down following the general elections yesterday), that one was the best one I could find. The wreckage was spread over a 20km radius and this is the largest chunk of wreckage found so far. — antiuser 22:23, 2 October 2006 (UTC)

A 20km radius would seem to rule out a final vertical dive, wouldn't it? Maybe that should be revised. Crum375 22:27, 2 October 2006 (UTC)

Indeed, the new theory is that the aircraft broke up while still in the air. I'm trying to find the reference to that so I can update the collision section with the new data. ||| antiuser (talk) (contribs) 22:37, 2 October 2006 (UTC)

Joe Sharkey's report

Sharkey's report that he posted on his blog has been taking a lot of flak lately and might be interpreted wrong. I see he's also published an article about the accident on the New York Times [7], which is much better-worded and more detailed. If there's no objection, I will change the part about his blog report with the NY Times one. ||| antiuser (talk) (contribs) 05:42, 3 October 2006 (UTC)

I agree that Sharkey's NYT report is excellent and adds a lot of details (e.g. passenger names, crew ages, etc.). It is definitely better than his blog entry. Crum375 11:53, 3 October 2006 (UTC)

Conflicting reports

There have been a lot of conflicting reports lately in the press. This abcnews report does a reasonable summary, but leaves a lot of speculation open. There is also a lot of info in Portuguese on Globo and Folha, rife with speculations. I think we should stay away from trying to insert the conflicting speculations into the article until the dust settles.

A point I'd like to clarify, since it's not in the articles I've read, is that an aircraft under IFR, as both these flights were (above 18,000 feet all flights are by default), may only fly a cleared altitude and routing. The cleared routing is normally the latest one assigned via radio by ATC, but defaults to an older clearance or the original 'flight planned' one (submitted and approved prior to flight) in the case of a radio communications failure. Since in this case, at this point anyway, we don't know what the exact state of the ATC communications were at each moment for the 2 individual flights, we really don't know what the 'correct' altitudes were for each flight at each point. Speculating at this time, or listing a lot of controversial theories, would not be encyclopedic - all we need to do is wait for more solid evidence to come in (e.g. ATC communication transcripts, CVR transcripts, FDR data, etc.). I hope we are in agreement here. Comments? Crum375 15:02, 4 October 2006 (UTC)

Yeah, there are a lot of conflicting reports going on right now. The current hypothesis is the following: The Gol flight was flying at FL370. The Legacy, after taking off from São José dos Campos was flying at FL370 as well, which despite being against the Flight Level rules was permissible since it was a one-way airway from São Paulo to Brasília. After Brasilia the Gol flight should've changed altitude to either FL360 of FL380, since he was now on a two-way airway. We know he did not, and thus the collision occured. Questions: 1)Did Brasilia Center authorize him to continue to fly on FL370 or did he tell him to change altitude? If the latter is true, did he disobey an ATC instruction? 2)The crash near the boundary between Brasilia Center and Manaus Center. Was there a mix-up in the hand-off? 3)And the question on my mind: Why did the TCAS not light up like a christmas tree and tell the pilots they were about the collide?

BUT, we'll know more by Friday when they open the FDR and CVR of either plane. We can provide lots of speculation until then or wait until we have factual data. --Dali-Llama 15:18, 4 October 2006 (UTC)

It was not the Gol flight, but the Legacy that should have either climbed or descended to an even-numbered flight level. The latest reports-cum-speculations by the media imply that the crew of the Legacy did not reply to radio calls and their transponder off or malfunctioning (which would be weird since they squawked 7700, signaling an emergency after the collision and had full mode C established) and Globo states the Air Force has discarded ATC error as a cause for the accident. They don't mention, however, if the controllers tried to warn the 737 that an aircraft could possibly be flying in the same altitude and airway as them, in the opposite direction. ||| antiuser (talk) (contribs) 07:52, 5 October 2006 (UTC)

You're right, I misspoke.--Dali-Llama 17:26, 5 October 2006 (UTC)

I agree with antiuser, and would add the following observation which I have not seen published: If in fact, ATC wanted the Legacy at FL360 when northwest-bound from Brasilia to Manaus on UZ6, regardless of the original flight plan filed, Brasilia ATC would have instructed the Legacy to descend to FL360. Presumably the Legacy, arriving at FL370 over Brasilia, would have been in good radio contact with Brasilia ATC for quiet a while, both before and after arriving overhead. There is no mention of any radio outage, nor is there any known 'black hole', over Brasilia. If the Legacy mode C was not reporting properly, then Brasilia ATC would have simply asked the Legacy to verify mode C was on, and in any case to 'verify level at FL360'. All this would have been done in the Brasilia area, as the Legacy started its northwest-bound leg on UZ6 toward Manaus. Assuming even worst case (that we are not directly told) that a radio outage with the Legacy started over Brasilia, and Brasilia ATC could not confirm, either by radio or mode C, the altitude change, then assuming normal ATC practices they would have 'blocked' at least both altitudes (FL360-FL370) and sent an urgent message to Manaus that those altitudes are blocked on the UZ6 airway. Bottom line: there is more here than we know at this point, and this is why I think we should refrain from reporting speculations in the article that (at least to me) don't make sense, until we get some hard evidence. Thanks, Crum375 12:12, 5 October 2006 (UTC)

I believe it should be made clear in the article that the main hypotesis is that the Jet was flying, northbound, in an unnappropriate FL (in this case, 37 thousand feet)since it should be using either 36 or 38 thousand feet, and that the Gol flight was cruising normally at the cleared altitude of 37,000 feet. I feel the article as it is today is partial, leading the reader to believe that there´s an equal chance of failure of either party.201.19.195.23 12:19, 5 October 2006 (UTC)

Please read what I wrote above and earlier. The 'correct' flight altitudes to fly are determined by a sequence of priorities, starting from the highest: number 1 is what you are directly cleared to, number 2 is what you were told to expect, number 3 is what you filed (and was cleared for). You only get to numbers 2 or 3 if you are out of radio contact and cannot raise ATC after several attempts, on multiple frequencies. The west-east altitude rules are not in that equation at all - they have to do with the original clearance and flight plan filed, not with any pilot discretion while flying IFR, as these flights were. Bottom line: we cannot speculate at this point without more evidence. After each aviation accident the media, many ill-informed about aviation, are rife with misconceptions and speculations. We as WP need to try our best to strive for good quality, i.e. reliable and verifiable information. Crum375 12:37, 5 October 2006 (UTC)

Reviewing the regulations some more, it actually gets even more complicated. First, of course the Brazilian rules would apply, not ICAO or U.S. FAA, and I have no online access to Brazilian IFR regulations so far. Second, with the FAA and presumably ICAO and others, there is a concept of 'uncontrolled IFR' which is rarely if ever used at high (>18,000') altitudes - I am fairly certain that airway UZ6 at FL360+ is not 'uncontrolled', but someone who knows the actual facts in Brazil should opine here. Also, the normal altitude clearance rules in case of 'communication loss' is to use the highest of the possible applicable altitudes: cleared, expected or minimum, which could easily be interpreted to mean to stay at the last cleared altitude of FL370 upon communication loss. Bottom line again is that this requires a lot of solid factual input before jumping to any conclusions in article space. Crum375 14:16, 5 October 2006 (UTC)

The article says the airliner had been cleared by ATC to climb from FL350 to FL390 so discussion of the flight level of the bizjet above seems almost moot. ATC must have had a part to play in the collision if both aircraft were flying IFR in controlled airspace as that means that ATC were responsible for aircraft separation. If the pilots were following instructions - there's nothing to the contrary so far. Paul Beardsell 16:25, 5 October 2006 (UTC)

There was some back-pedalling in the Brazilian press about the initial reports that the Gol requested a climb from FL350 to FL390. Now some reports just say that it climbed to FL370 and stayed there, despite prior claims to the contrary. So, due to a lot of conflicting information in that regard, and without ATC or CVR transcripts, I think it makes sense to just leave that part of the article as is for the time being, pending more solid evidence (which should be forthcoming). Crum375 16:35, 5 October 2006 (UTC)

I agree: The joining the dots deduction and speculation here is not for the article. Nevertheless: IFR in controlled airspace points at ATC. Paul Beardsell 16:39, 5 October 2006 (UTC)

I agree as well. I originally put the climb story, and admit it has now been proven false, but the stories are changing every day. Let's let the dust settle a bit.--Dali-Llama 17:28, 5 October 2006 (UTC)

Gainsaying something said far above: If you are IFR outside controlled airspace (and this is routinely done) then no cleared altitude and routing will be followed. However, then the pre-allocated altitude levels depending upon direction must be! Paul Beardsell 16:37, 5 October 2006 (UTC)

Correct, hence one important validated input we need to establish is whether, as expected, both flights, operating along UZ6 at FL370, were in fact under IFR in 'positive control' and in 'controlled airspace'. Crum375 16:50, 5 October 2006 (UTC)

One had just received a clearance to climb (OK, this is unclear), the other was at the cleared altitude. If "cleared" then you must be in controlled airspace. If one was the other was otherwise there would be no article. Controlled means, inter alia, the default rules as to flight level depending upon heading does not apply. The pilot under IFR in controlled airspace does what he is told. Heading and flight level. Paul Beardsell 16:55, 5 October 2006 (UTC)

Media reaction

There is one aspect that I consider is now appearing in the brazilian media and is not common for air crashes: there seems to be some kind of frenzy to quickly point a culprit, rather than a cause. And in this case, most reports from Brazil are blaming the Legacy pilots.

Risking to enter in conspiracy-theory land, let's not forget that the civil aeronautics in Brazil is run by the military, and we are reading reports that top officials have already ruled out any wrong-doing from the ATC, which ironically is there exactly to prevent this kind of national disasters.

There are many interests in this case: the brazilian military investigating their own participation in the mishap, the brazilian military investigating a plane from Brazil's main aerospace firm (actually, Embraer's facilities are located right next to the military-run Aerospace Technical Center), the brazilian military investigating a downed brazilian airliner, which from its crash site they themselves are right now retrieving thousands of body parts, and the brazilian military investigatig two foreing pilots.

This report is very interesting: (in portuguese)

http://www1.folha.uol.com.br/folha/cotidiano/ult95u126723.shtml

Headline:Pilots deny having turned off Legacy equipment.

Eighth paragraph: American company ExcelAire, which bought the Legacy, is fearing to turn into a "scapegoat" and decided to go out of a defensive position and started an offensive, hiring lawyer José Carlos Dias' bureau and the communications agency Burson-Marsteller, represented in Brazil by Francisco de Carvalho.

Some aviation forums are already talking about similarities with other multicultural air crash investigations, such as Egypt Air 990.

Aldo L 19:35, 5 October 2006 (UTC)

While the media is publishing a lot of speculative stories, it is important to note that neither the Air Force nor the DAC/ANAC have published any accusatory material on their own websites. Furthermore, Brazil and the US have very close diplomatic ties, the NTSB and Boeing have personnel assisting with the investigation — the last thing they want is a diplomatic incident. ||| antiuser (talk) (contribs) 19:26, 5 October 2006 (UTC)

Aviation regs I have seen (not Brazilian ones) say that no one involved or who has any priveleged info can speak to the media. And aviation regs have the force of law (can't speak for Brazil). Might explain the lack of official comment. Paul Beardsell 11:02, 6 October 2006 (UTC)

I think this would be a good test for WP to show that we can continue to present a well founded and properly balanced picture as events unfold and new information comes in. Crum375 19:36, 5 October 2006 (UTC)

Details about the accident from the ATC's point of view

LEGACY FICOU FORA DO RADAR POR 15 MINUTOS E RECEBEU 5 ALERTAS. I will merge that information to the article later (replacing part of the speculation) if nobody beats me to it. --cesarb 04:45, 6 October 2006 (UTC)

Thanks. So, with the help of Babelfish, the transponder signal (if any!) from the Legacy was not being received. And the article speculates the pilots deliberately turned it off so they could experiment. The journalist on board certainly didn't report aerobatics but surely he would have noticed anything more than a light turn or very gradual climbing or descending - doesn't sound like experimenting to me. (On the other hand, that pilots sometimes deliberately turn off the transponder is well known.) The article seems well reasoned and not inconsistent with that which is known. Am I right in understanding the article to say that transponder/radio transmissions are not reliable in the area? Also, when you squawk 7700 it would be usual to press the "ident" button which makes your radar blip flash (or does it? I'm unsure), and that would draw ATC's attention to the blip, if they had missed it. Also, why, if they were so intent in contacting the Legacy, why didn't they inform the 737 of the conflicting traffic? Several times flying in the UK I have heard ATC advise of "potentially conflicting traffic altitude unknown" (i.e. no transponder). Paul Beardsell 10:57, 6 October 2006 (UTC)

As mentioned above, we have to be extremely careful in building and updating this article, sticking to well supported facts and staying clear of speculations and misinformations. There are 2 types of relevant facts in this case: those that pertain specifically to this accident, and those that relate to aviation in general. Unfortunately, many of the statements in the press include speculations by both journalists as well as some officials, that are conflicting. What I suggest again is that we apply very strict standards of quality, and try our very best to stick to raw data and hard facts. I read the Globo article cited above, as well as previous Globo, Folha and other articles. Many have factual technical errors as well as sheer speculations. Since this article obviously involves living persons, both named (the Legacy crew) and as yet unnamed (the Brasilia, Manaus and other ATC personnel involved), we must adhere to WP's living persons sourcing policies. I see no problem to discuss and report here in the Talk page about speculations and allegations, but I don't think anything but hard, well supported (e.g. by raw data from CVR/FDR or ATC transcripts, ANAC, CENIPA or NTSB official statements) facts should go into the article. I hope we are in agreement. Crum375 12:59, 6 October 2006 (UTC)

To clarify, the above strict sourcing applies to anything related to potential liability; other facts, e.g. related to recovery efforts, would rely on WP normal reliable sourcing rules. Crum375 13:06, 6 October 2006 (UTC)

I (and others) have already specifically agreed with this. I'm not sure why the reminder is necessary. But, in the same spirit, and hopefully unnecessarily, let me add I hope we will apply what Wikipedia policy actually says, not what we think it says or ought to say. Paul Beardsell 13:56, 6 October 2006 (UTC)

I concur with Crum and Paul.--Dali-Llama 14:39, 6 October 2006 (UTC)

I added the part I think can be considered reliable (it was both radio and secondary radar, it's not "unclear" anymore, or at least it's less unclear from a similarly reliable source). While the article says the informations come from the ATC records and CVR records, I think it's better to wait before adding the rest (such as how many times they were contacted, that the primary radar was showing 36000 feet because it's less accurate, that they were going to ask other airplanes to help with the communication, etc); there have been way too many conflicting reports lately. --cesarb 15:17, 6 October 2006 (UTC)

I agree with CesarB's comments and edit (which I modified just slightly). As CesarB says, there are a lot of things we don't know, and most relevant facts will be easy to establish once we have all the ATC communication transcripts (including landline) and the Legacy CVR transcripts (both should be easy to obtain in theory). The Gol CVR would likely (as it seems at this time) not add much, but it will complete the picture. My own personal prediction (which of course will stay here and here only) is that, barring surprises, once we have those transcripts, the reasons for the crash should be fairly clear. But let's wait and see and follow the hard data as it comes in. Crum375 18:38, 6 October 2006 (UTC)

In fact, I've been wondering lately if we shouldn't remove some of the earlier speculation that's still in the article (from before it became clear we would be having conflicting reports all the time, and at a time information was scarce). For instance, the initial analysis of the ATC recordings; most media lately seems to be considering the Gol as flying level at FL370, with no climb at all. I wonder if said "initial analysis" wasn't some offhand remark of someone related to the investigation, instead of an official statement. In fact, the reference given uses the preterite future tense, which is used "quando o locutor não quer responsabilizar-se pela informação do enunciado" (when the speaker doesn't want to have responsability about the information on the sentence) [8], making a stronger case for it being speculation at the time. --cesarb 21:13, 6 October 2006 (UTC)

I tend to agree with you. Maybe we should just say that "Reports indicated the Gol flight was assigned FL370." The reason I haven't touched it so far is that I have seen so much conflicting and technically incorrect information published in Globo/Folha (and BTW I don't blame them - the reporters are not technical experts - you often see the same in the US and other countries). The problem even with my above proposal is that we really have no hard evidence for it, and in fact we have seen some statements that the Gol first climbed to FL350 and requested FL390. So given all this uncertainty, and given that we did have those early reports (which are duly sourced), I would leave it for now pending more concrete data. All we need are the ATC transcripts - they should be available almost immediately to the 'right people' - the problem is for them to propagate to the public, which may take a long time. Crum375 21:29, 6 October 2006 (UTC)

We could just nuke it. If we don't know which FL the Gol was at, we don't say which FL the Gol was at. What matters is only that the "apparent collision" happened at FL370. For instance:

Initial reports suggested [...] for the accident. ^[14][32] The Gol flight was southeast bound, while the Legacy was cruising in the opposite direction northwest bound, level at its previously assigned altitude of FL370 (flight level 370, approximately 37,000 feet above mean sea level), when the apparent collision occurred.^[citation] There were reports ATC lost contact (both radio and secondary radar) with the Legacy shortly before the collision.^[34]

What do you think? --cesarb 21:40, 6 October 2006 (UTC)

Yes, that sounds good to me. I may add the airway, as that seems non-contentious, but it's not critical for now. Crum375 21:52, 6 October 2006 (UTC)

I ended up removing the whole sentence. The flight level information is already at the next paragraph, and I didn't find a good source for the northwest/southeast part (which, by the way, wasn't in the cited source). I might add it back after I find a good source among the hundreds of news reports; but somehow I think the way it is right now flows better for some reason. --cesarb 15:48, 7 October 2006 (UTC)

You edit looks good to me - it's simple and tells us what we know with fairly high certainty for now. I did have sources for the nw/se, but I prefer your simpler version for now. Thanks, Crum375 16:13, 7 October 2006 (UTC)

Addition of BLP Template

I added this template to the top of the Talk page to remind us all that even though this is not a biography article per se, it does deal with a high profile case with living persons and their reputations, careers and legal liabilities involved. Hence we should continue to carefully adhere to the ultra strict sourcing and validation policies of WP:BLP. Crum375 18:49, 6 October 2006 (UTC)

Legacy transponder AD

These excerpts may or may not be related to this accident. Please be careful.

From: http://www.airweb.faa.gov/Regulatory_and_Guidance_Library/rgAD.nsf/0/1bddc176ac629452862571e70059100d!OpenDocument Airworthiness Directive Federal Register Information Header Information DEPARTMENT OF TRANSPORTATION Federal Aviation Administration 14 CFR Part 39 [Docket No. FAA-2006-24639; Directorate Identifier 2005-NM-171-AD; Amendment 39-14761; AD 2006-19-04] RIN 2120-AA64 Airworthiness Directives; Honeywell RCZ-833J/K, -851J/K, and - 854J Communication (COM) Units, Equipped with XS-852E/F Mode S Transponders; and Honeywell XS-856A/B and -857A Mode S Transponders; Installed on But Not Limited to Certain Transport Category Airplanes PDF Copy (If Available): 2006-19-04.pdf Preamble Information AGENCY: Federal Aviation Administration (FAA), Department of Transportation (DOT). ACTION: Final rule. SUMMARY: The FAA is adopting a new airworthiness directive (AD) for certain Honeywell COM units and transponders, installed on but not limited to certain transport category airplanes. This AD requires a revision to the Normal Procedures section of the airplane flight manual to advise the flightcrew to check the status of the transponder after changing the air traffic control (ATC) code. This AD also requires replacing certain identification plate(s) with new plate(s), testing certain COM units or transponders as applicable, and corrective action if necessary. For certain airplanes, this AD requires replacing the transponders of certain COM units with new or modified transponders. For certain other airplanes, this AD requires installing a modification into certain transponders. This AD results from the transponder erroneously going into standby mode if the flightcrew takes longer than five seconds when using the rotary knob of the radio management unit to change the ATC code. We are issuing this AD to prevent the transponder of the COM unit from going into standby mode, which could increase the workload on the flightcrew and result in improper functioning of the traffic alert and collision avoidance system. DATES: This AD becomes effective October 17, 2006. The Director of the Federal Register approved the incorporation by reference of certain publications listed in the AD as of October 17, 2006. Another excerpt from further down in the web page: Request To Revise Compliance Time Empresa Brasileira de Aeronautica S.A. (EMBRAER) requests that we extend the compliance time for the AFM revision from 5 to 30 days. EMBRAER asserts that the loss of the transponder does not pose so great of a hazard to justify such an urgent compliance time. As justification for extending the compliance time, EMBRAER states that some of the affected airplanes might be on international trips, where it may not be possible to return an airplane to a convenient location and accomplish the AFM revision within 5 days after the effective date of this AD. We agree that the compliance times can be extended somewhat. We have determined that extending the compliance time to 14 days will not adversely affect safety. Therefore, we have revised paragraph (f) of this AD accordingly. And yet another excerpt from further down in the web page: Applicability (c) This AD applies to the Honeywell parts identified in paragraphs (c)(1) and (c)(2) of this AD, approved under Technical Standard Order TSO-C112, installed on but not limited to Bombardier Model BD-700-1A10 and BD-700-1A11 airplanes; Cessna Model 550 and 560 airplanes; Cessna Model 650 airplanes; Dassault Model Falcon 900EX airplanes, serial number (S/N) 97 and S/Ns 120 and subsequent; Dassault Model Falcon 2000EX airplanes, S/N 6 and S/Ns 28 and subsequent; EMBRAER Model EMB-135BJ, -135ER, -135KE, -135KL, and - 135LR airplanes; EMBRAER Model EMB-145, -145ER, -145MR, -145LR, - 145XR, -145MP, and -145EP airplanes; Learjet Model 45 airplanes; Lockheed Model 282-44A-05 (C-130B) airplanes; Lockheed Model 382G series airplanes; Raytheon Model Hawker 800 (including variant U- 125A), 800XP, and 1000 airplanes; certificated in any category. (1) Communication (COM) unit RCZ-833J part numbers (P/Ns) 7510700-763 and -863; RCZ-833K P/Ns 7510700-765 and -875; RCZ-851J P/N 7510700-813; RCZ-851K P/N 7510700-815; and RCZ-854J P/Ns 7510700-725 and -825. (2) Mode S transponder XS-856A P/Ns 7517400-865 and -885; XS- 856B P/Ns 7517400-866 and -886; and XS-857A P/Ns 7517400-876 and - 896. Unsafe Condition (d) This AD results from the transponder erroneously going into standby mode if the flightcrew takes longer than five seconds when using the rotary knob of the radio management unit to change the air traffic control code. We are issuing this AD to prevent the transponder of the COM unit from going into standby mode, which could increase the workload on the flightcrew and result in improper functioning of the traffic alert and collision avoidance system. Compliance (e) You are responsible for having the actions required by this AD performed within the compliance times specified, unless the actions have already been done. Airplane Flight Manual (AFM) Revision (f) For all airplanes: Within 14 days after the effective date of this AD, revise the Normal Procedures section of the applicable AFM to include the following statement: "After completion of any 4096 ATC Code change (also referred to as Mode A Code), check the status of the transponder. If the transponder indicates that it is in standby mode, re-select the desired mode (i.e., the transponder should be in the active mode)." This may be done by inserting a copy of this AD in the AFM. Accomplishing the actions specified in paragraph (h) or (j), as applicable, of this AD terminates the requirement of this paragraph. For the complete report please refer to the aforementioned Internet address. Aldo L 19:02, 6 October 2006 (UTC) [Editing note: Moved down from other section, squeezed out spaces.Crum375 20:25, 6 October 2006 (UTC)]

I think this AD could be very relevant, potentially, as it could explain the transponder 'outage' on the Legacy. It still would not explain why ATC would not deal with this situation as it deals with many transponder outages routinely, e.g. by blocking off altitudes. But it definitely could be an important piece of the puzzle, especially if it can be shown that the Legacy was asked to change its transponder squawk code just prior to the 'outage'. OTOH, if the outage happened unrelated to a squawk code change, then this would not be applicable. Crum375 19:14, 6 October 2006 (UTC)

This is relevant!!! "This AD results from the transponder erroneously going into standby mode if the flightcrew takes longer than five seconds when using the rotary knob of the radio management unit to change the ATC code." A transponder going into STBY mode when you take too long to switch the ATC frequency? That's nuts. Now, it might be likely that either a) the problem has extended itself to the COM frequency as well, or b) because of operator error, in trying to reach Manaus/Brasilia Center, the pilot started fidgeting with the ATC code instead of the COM frequency and took more than 5 seconds doing it, turning off the transponder.--Dali-Llama 22:05, 6 October 2006 (UTC)

Yes, it's possible in principle for the pilot to change the squawk code without being told. Not knowing the Legacy panel layout I wouldn't know, though I suspect it's probably unlikely. Also, taking 5 seconds to change the squawk code is fairly long, long enough obviously that the glitch wasn't detected earlier. However, all these are sheer speculations (though they could be absolutely correct). At this time, I would let the hard facts flow to us: ATC tape transcripts - where are you? Crum375 22:19, 6 October 2006 (UTC)

I'm quite familiar with the Honeywell Primus system they use. The transponder button is directly below the radio/com button, and use a single knob to change the frequency. This is (admittedly) pure speculation. The FDR of the Legacy should clear everything up...--Dali-Llama 00:35, 7 October 2006 (UTC)

I think I heard on the news yesterday that the FDR unfortunately doesn't have the transponder data (something about it being optional). And I just thought of something: if the pilots lost radio contact, couldn't they have changed the transponder code to 7600? --cesarb 15:34, 7 October 2006 (UTC)

That's very well possible, but I think they would've stated if they did that. In any case, this has broken on the news: [9].--Dali-Llama 15:45, 7 October 2006 (UTC)

Yes, that's a good point - it's possible that the pilots decided at some point to declare 'radio communications loss'. If they did, you are right, they would punch in '7600' and if they did it really slowly, over 5 seconds to set that code in, it could possibly put the Transponder into Standby per that AD. But, I would think that prior to doing that, they would try many other options. First, they would try other frequencies. Then, they would try other radios (at least 2 different ones). Then, they would also try to raise other flights on the frequency. They would also try the emergency frequency (on 2 radios). Bottom line, it's possible that after all that was tried and failed they set the 7600 code and knocked out (inadvertently) the transponder. But as you say, it's sheer speculation at this point. Crum375 15:55, 7 October 2006 (UTC)

Flight International published this more than one year ago:

DATE:02/08/05

SOURCE:Flight International

Transponder drop-outs spark AD

Regulators urged to act quickly as fault on Honeywell device causes flight details to disappear from radar screens

With aircraft suddenly disappearing from radar screens, air traffic controllers are calling on European safety regulators to speed up approval of a solution for a recurring fault afflicting Honeywell’s Primus II transponder.

European controllers lose identification and flight details for an aircraft around five times a month, says the International Federation of Air Traffic Controllers (IFATCA). The data loss also disables air traffic control safety-net systems – short-term conflict alert and the airborne collision avoidance system (ACAS) of the aircraft concerned.

IFATCA has written to the European Aviation Safety Agency (EASA), Eurocontrol and the Joint Aviation Authorities to try to accelerate the process of approving a permanent modification to the transponder to supersede the current operational “fix” proposed by Honeywell. IFATCA says EASA is shortly to publish an airworthiness directive (AD), but under present rules it would give carriers a year to comply, which it says is too long. Meanwhile, Honeywell says it has had a software fix available “for a while”, and airlines can obtain it by returning the transponder. It also has a rotable pool of units for loan.

The Honeywell Primus II Mode S (elementary) transponder has been fitted in some aircraft since September 2004. The problem has usually involved Embraer ERJ-135/145 regional jets and occurs when pilots are asked to change their four-digit “squawk” code – the flight identification code allocated by the ATC unit; if they take more than 5s to input the new code the transponder automatically switches to “standby”, making it unresponsive to interrogation by ground radar or by the ACAS transponders of other aircraft. In areas where there is only secondary radar with no primary radar back-up, the contact disappears from the controller’s display. IFATCA cites a case this year over France where an ERJ-145 flew for more than 30min in busy airspace without radar contact.

Honeywell has told pilots to select the transponder to standby before inputting the new code, then back to active once the code has been set; or they can check the annunciators on the unit or on the ACAS display after a reset, and if they indicate standby they should select the transponder to active.

The latter advice was accepted in January as “a mitigation procedure” by Eurocontrol’s Mode S regulatory group. “In themselves [these measures] are not a permanent solution,” says Eurocontrol, adding that EASA’s proposed AD is “expected to be released shortly”.

DAVID LEARMOUNT / LONDON

Aldo L 02:21, 8 October 2006 (UTC)

Yes, that sounds like very pertinent info, obviously related to the AD. I wonder if the Legacy crew knew or were told the 'mitigating procedure' (called 'workaround' in the software business) of recycling to standby on each code change by Embraer when they picked up the plane, or if Embraer had already installed the software fix to make it a non-issue. We definitely need more facts here. Thanks for the good reference. Crum375 03:20, 8 October 2006 (UTC)

New York Times reported:

"Device on Jet in Amazon Crash Had Not Been Recalled as Faulty (...) Some transponders built by Honeywell, an American company, have been ordered repaired by the Federal Aviation Administration because they have stopped broadcasting unexpectedly in some cases. But a review of the part numbers on the Legacy shows that the model on the jet was not recalled, Bill Reavis, a Honeywell spokesman, said Sunday."

Full text at http://www.nytimes.com/2006/10/09/world/americas/09crash.html

Aldo L 06:32, 10 October 2006 (UTC)

Apparently though, the transponder in the Legacy is not included in the recall, according to FAA spokeswoman Laura Brown [10]. ||| antiuser (talk) (contribs) 09:00, 10 October 2006 (UTC)

Secret airbase

I reverted the "airbase was not secret" and speculation about Sharkey's emotions because it was not sourced. We do have Sharkey on record saying it was secret (probably because someone told him so at the time), we don't have a reliable source saying it isn't. I did see a forum somewhere that mentioned it was originally a secret nuclear airbase during the 1980's; interesting if we can get more data on it. The US NTSB just calls it "Cachimbo air base" in the preliminary accident report. I'm sure we'll find out more in the future. If anyone has a reliable source about the base, feel free to add. Crum375 01:02, 7 October 2006 (UTC)

I believe it used to be a top-secret facility in the 70s and 80s but has since been converted to an experimental airfield (the full name is Campo de Provas Brigadeiro Velloso). They have a website here: [11] ||| antiuser (talk) (contribs) 01:08, 7 October 2006 (UTC)

Very interesting. Apparently they are also involved in accident prevention, I see this page on their site (under SPAA):

---

Seção de Prevenção de Acidentes Aeronáuticos ________________________________________________________________________________________

DEFINIÇÃO

Prevenção de Acidentes Aeronáuticos, de Incidentes Aeronáuticos e Ocorrência de Solo é o conjunto de atividades destinadas a impedir essas ocorrências, evitando assim custos adicionais desnecessários à operação através da preservação dos recursos humanos e materiais.

Os elementos que constituem a base e o objeto de toda a atividade de Prevenção de Acidentes Aeronáuticos são definidos no trinômio “homem-meio-máquina”.

PRINCÍPIOS BÁSICOS DO SIPAER

a) “Todo acidente aeronáutico pode e deve ser evitado”;

• All aviation accidents can be prevented

b) “Todo acidente aeronáutico resulta de uma seqüência de eventos, e nunca de uma causa isolada”;

• All aviation accidents result from a chain of events; never from a single isolated cause

c) “Todo acidente aeronáutico tem um precedente”;

• All aviation accidents have a precedent

d) “Prevenção de acidentes aeronáuticos é uma tarefa que requer mobilização geral”;

e) “O propósito da Prevenção de Acidentes Aeronáuticos não é restringir a atividade aérea, mas estimular o seu desenvolvimento com segurança”;

f) “Os Comandantes, Diretores e Chefes são os principais responsáveis pelas medidas de segurança”;

g) “Em prevenção de acidentes aeronáuticos não há segredos nem bandeiras”;

• There are no secrets nor borders (flags) in the prevention of aviation accidents

h) “Acusações e punições agem contra os interesses da Prevenção de Acidentes Aeronáuticos”.

• Accusations and punishments do not serve the interests of preventing aviation accidents

---

It seems so eerily apt. (feel free to add to and fix my loose translation) Crum375 01:52, 7 October 2006 (UTC)

I also realize now that what Sharkey actually said was "formerly secret air base", which would be perfectly correct, if in fact it used to be secret in the past. Crum375 02:43, 7 October 2006 (UTC)

That's why I left it alone, even though I disagree with the tone on his blog post. And your translation is just fine. ||| antiuser (talk) (contribs) 02:58, 7 October 2006 (UTC)

Lead phrasing update

I revised the lead to better express the current known status, as I felt the old version was out of date. My only concern was about using the word 'collided' instead of 'probably colliding', but that would sound awkward and I think by now there is very little doubt left, and we certainly have plenty of sources. If anyone feels otherwise, please comment. Crum375 14:50, 7 October 2006 (UTC)

ATC communications transcripts

I think that these easily obtained pieces of evidence, that should be available hours if not minutes after an incident or accident, hold the key to this specific accident. If you had all the time-stamped transcripts, of all relevant frequencies, plus the landline contacts between the relevant ATC sectors, I think this case would probably be largely solved. Despite this, I see no reference anywhere in the Brazilian (or other) press to these crucial tapes or transcripts. They do not require a major effort to obtain or even analyze (at least superficially). Obviously with the Brazilian ATC being quasi-military (I am assuming) it may take longer than other countries where ATC is totally civilian. Can anyone shed any light on this issue? Crum375 16:29, 7 October 2006 (UTC)

One has to remember that despite the ATC in Brazil being a hybrid of civilian and military for ATC, the investigation is being conducted by a civilian agency (ANAC). The current rationale for the transcripts and the black boxes is that a)they're not all in (part of the Gol CVR is still missing) and b) their analysis is not complete (the Gol's FDR had to be sent to Boeing in WA, for example).--Dali-Llama 16:58, 7 October 2006 (UTC)

I understand that the 'black boxes' take longer to analyze, both voice and data. But my point is specific to the ATC tapes, which are easily available, easy to transcribe, and in a case where procedure appears to be in focus, could easily contain the keys to the solution. Crum375 18:25, 7 October 2006 (UTC)

"Legality" of FL370

As I mentioned elsewhere on this page, we need excellent sourcing on anything that casts a doubt on the proper conduct of any living person. In this case, it is premature to speculate on whether FL370 was 'correct' or 'legal' for the Legacy crew or not. There are multiple issues and rules to consider, for example:

1. If you are cleared to FL370 by ATC, you are required to be at FL370, regardless of your direction of flight
2. If you lose radio contact with ATC, there is a specific procedure to follow for routing and altitudes under IFR
3. The 'lost comm' procedure may require the pilot to continue on the last cleared altitude, or to fall back on previous clearances, including the 'expect' or the original clearance
4. The actual altitude to fly for lost comm, if there is a conflict among several possibilities, could be the highest one

These rules may vary between ICAO, FAA and Brazilian ANAC; obviously the latter would be most relevant. If someone can get the online version of the ANAC lost comm IFR procedures, that would be most helpful. But in the meanwhile, we can't just insert editorial comments or even quote from newspapers where we know that these issues are complex, unclear and conflicting. Thanks, Crum375 18:42, 7 October 2006 (UTC)

For reference, here are the applicable U.S. FAA IFR 'lost comm' procedures: FAR 91.185 operations: Two-way radio communications failure. Now we need the Brazilian ANAC equivalent. Crum375 21:36, 7 October 2006 (UTC)

As I understand, they were cleared from São José dos Campos to Brasilia at FL370, and their flight plan after Brasilia was filed for FL360 followed by FL380 about 200mi before where the accident happened. They were told to expect the flight levels as filed in their flight plan. In the loss of radio contact, you are expected to follow a previously cleared instruction (in this case FL380). This is explained in this article which, unfortunately as many of the articles in this case, is in Portuguese. What I'm honestly concerned about is at what point did they lose contact with the tower--If they were supposed to descend to FL360 right after Brasilia, they couldn't possibly lose contact with Brasilia Center, as (short of a technical malfunction) they were flying directly over it!--Dali-Llama 22:09, 7 October 2006 (UTC)

I agree about the FL370 to FL360 part that you mention, but the other question is about the supposed FL370 to FL380 climb requirement at some fix northwest of Brasilia. My understanding of lost comm procedures is that you are supposed to go the highest of your last cleared altitude and your last expected altitude (i.e. "expect FL380 at fix X" would count as 'expected clearance'). OTOH, my understanding of ATC procedures is that ATC would block all relevant altitudes for you until they have positively verified your change of altitude. We need to see the ATC transcripts to see what the exact clearances and fixes were, because the exact wording of the enroute clearances (and their verification/acknowledgment by the flight crew) are the determining factors, not the original flight plan. Crum375 23:00, 7 October 2006 (UTC)

I had check the flight level rules, under all but metric (which would definitely not be the case) the Legacy should have not been at FL370. I think it's relevant and should be mentioned, albeit until we have the transcript like mentioned above we may want to mention the possibility ATC messed up and the pilot followed incorrect orders anyways.

Here are the Regras do Ar e Serviços de Tráfego Aéreo (Brazilian Air Rules). I believe it's current.

http://www.icea.gov.br/ead/ima100-12intraer/sumario.htm

Aldo L 06:16, 10 October 2006 (UTC)

Equpiment use

Both these plans were brand spanking NEW. You know when you pick up a car at the rental, you have to figure where the lights/wipers/ect are?? Well, maybe, just maybe these people didn't read the NEW owners manuels. Maybe the "another plan is about to hit us" alarm wasn't properly set. These plans were HOURS old...thats my take, we'll have to see....--68.228.148.52 22:41, 7 October 2006 (UTC)

Not really. First of all, aircraft don't just roll out of the factory and on to the customer - there are stringent flight test goals defined by international authorities as well as the FAA that must be met. A new plane is always flown first by the manufacturer's test flight crews, and they do it at least a dozen times and all instruments are tested even before the aircraft leaves the hangar. Living in Seattle, I saw PR-GTD a few times sitting at the tarmac at Boeing Field, where 737s go for painting and final trim fitting. Embraer does the same thing with their jets. Also, the pilots were type-rated to fly the aircraft and had more than just a few hundred flight hours. One of the pilots on the Gol 737 was even rated as an instructor on the 737. Getting a commercial pilot licence and flying a jet aircraft is a much more complex task than getting a driving licence and driving a new car.

There is, however, the possibility of TCAS failure, as noted by the Airworthiness Directive issued for the same model transponder that the Legacy is equipped with, and I'm sure the investigating authorities will test the equipment to check if that could've been the cause for the accident. ||| antiuser (talk) (contribs) 23:01, 7 October 2006 (UTC)

Some more points to consider:

1. New planes or cars may often have more glitches (teething problems) than older ones
2. The ATC system is designed to handle hardware breakdowns or communication problems without major consequences
3. In this case, we don't know at this stage whether the planes or pilots in the air were any more responsible for the accident than the support people on the ground, and it may well be some combination

Bottom line: let's not jump to any premature conclusions about any hardware or people until the fat lady sings (i.e. the final report is in). Crum375 23:22, 7 October 2006 (UTC)

TO add a little more, thoughthe plane may be new the same model or a close variant of both had already been around for a while. Not sure on ExcelAir, but GOL only flies 737's thus both pilots were experienced. I doubt they could tell the difference between the -800 and the -800SFP variant. Also, the transponder in qustion in the Legacy is used on many planes.

"Situation of pilots involved" section

I removed this section because I think that it is mostly WP:OR. If the contributor wants something like this to go into the article, every statement and factoid need to be meticulously sourced, per WP:V, WP:RS and WP:BLP. Crum375 23:53, 7 October 2006 (UTC)

Reversion of Jggouvea edits

I reverted these edits as I believe them to be a combination of WP:OR and improperly sourced statements. Since this article must conform to WP:BLP rules, it must have excellent quality sources for each allegation related to living people. Please discuss each intended change here first if at all possible. As example in this case (but only one example), we have no sources showing that Joe Sharkey said "the accident was caused by the poor flight control". Again, at WP we don't accept any facts that are not extremely well sourced when it relates to living people. Please note that per WP:BLP, anyone can remove unsourced or poorly sourced material in such cases, with no need for discussion or WP:3RR limits. Thanks, Crum375 00:09, 8 October 2006 (UTC)

I'd also like to suggest that we be extremely careful with statements that might be observed as POV. While the authorities involved in the investigation are being praised for their neutral handling of the case, the media has been jumping to conclusions in certain points. ||| antiuser (talk) (contribs) 00:30, 8 October 2006 (UTC)

As a heads-up to everyone collaborating to the article, Brazilian media is already reporting that the Air Force's preliminary findings are that the Legacy pilot is basically at fault. I recommend we hold any edits in that direction (assigning guilt) until the actual report is released, early next week.--Dali-Llama 00:54, 8 October 2006 (UTC)

Agreed. ||| antiuser (talk) (contribs) 01:06, 8 October 2006 (UTC)

Also agreed. Just the fact that some sensational allegation is made is no reason for WP to repeat it - we are not a tabloid. We need to double check all facts for reliable sourcing. In this case, as I noted above, I don't think anything but hard, well supported (e.g. by raw data from CVR/FDR or ATC transcripts, ANAC, CENIPA or NTSB official statements) facts should go into the article. Crum375 01:39, 8 October 2006 (UTC)

Updating Needed?

I think some things need to be added and some need to be cut.

At this point, I think the metion of the farmer spotting the plane low can be snipped out. Also, the discussion on whether there was a collision or not at first should be shortened.

We also need to add more. Supposedly the FAA was considering a recall of the transponder in use on the Legacy. Also, Joe Sharkey seems to be changing his story a bit, it's gotten more dramatic over the last few days. I would reconsider using him as a reliable source, though it could just be the "glad to be alive" wearing off. Also, to add more to the GOL Flight and the search for it, some relatives that were criticising the long wait were flown over the site by the Brazilian AF to see how challenging it was, and they thanked them for the hard work afterwards. Might be worth a mention.

And for human interest when the whole thing blows over, there's a photo of the Air Force throwing 154 roses over where the plane fell that imght go well in closing.

Just floating some ideas...

Actually, the transponder in the Legacy would not be included in the recall, according to Globo, so at least for now I don't think it should be included in the article. ||| antiuser (talk) (contribs) 07:55, 8 October 2006 (UTC)

One of my own guiding principles in writing an 'encyclopedic' article for this type of a dynamically unfolding event is to try to shoot for 'permanence', i.e. ideally we should be gradually building, the 'final version'. Of course there is never anything really final on WP, but there is a 'stable version' which is close. My own crash articles, many of which were very sensational as they occurred, rarely get touched nowadays. The point here is to try to keep to a minimum both wrong info that will have to be reverted as we uncover the truth over time (although in principle WP allows us to include 'well sourced falsehoods') as well as anything that is clearly transient and will not belong in the 'final' or 'stable' version. The way to achieve that, IMO, is to keep insisting on 'excellent' sourcing for anything that we include, which we are required to anyway by WP:BLP rules. BTW, even if everyone in a crash dies, unlike here, you still need to practice care because first there are always live people potentially liable: mechanics, dispatchers, controllers, etc. and second, even casting premature (i.e. pre-final-report) blame on any deceased party carries legal liabilities.

Specifically in your suggestion list, I think the 154 dropped flowers, can certainly be added. The farmer seeing the plane flying low - I would leave it as the only eyewitness for now - it can be dropped down the road if it doesn't make the final report. The transponder issue I would leave until we know (from ultra reliable sources) if the Legacy actually had the model and serial number affected by the AD, and whether Embraer claims it had been complied with (and how - workaround, software fix). As to Joe Sharkey - I would stay away for now from subsequent bloggish statements made by him but keep the early ones, made as a direct and 'fresh' witness. Regarding the collision speculations, I would wait for some evidence of paint transfer - that would be fairly definitive. BTW, I read that one theory that was unlikely but could not yet be excluded is that the Gol broke up in midair for another reason, and one of the debris pieces hit the Embraer. Of course the FDR's from both planes could exclude that one based on precise timing.

All of the above is of course just my own opinion and is open to challenge and debate. Thanks, Crum375 11:54, 8 October 2006 (UTC)

External links

With the overall theme of keeping this article tightly within WP's policies and guidelines, please consider the WP:EL guideline before adding any external links to the article. The one I just reverted does not necessarily contain technically incorrect data, but that's not a criterion for EL. In this case it is a clear case of WP:OR in a non-notable web site. One cannot evade WP:NOR by just setting up a web site and linking to it. In any case I think we are in general agreement that we should only allow hard evidence (as defined elsewhere) into the article and no speculations. Thanks, Crum375 16:59, 8 October 2006 (UTC)

I have reverted another EL by the same anon IP, 82.225.218.187, this time linking to an accident web site that proclaims that the Legacy crew was 'flying at the wrong altitude'. Since that hasn't been established, it implies misconduct on the part of living persons, and the site overall does not meet WP:EL (e.g. it does not add new acceptable facts to what we already have), I removed it per WP:BLP, WP:EL, WP:RS, etc. Please, if you consider editing this article, be extra mindful that this is a WP:BLP related article, and only excellent well validated sources may be introduced. Thanks, Crum375 18:44, 8 October 2006 (UTC)

Flight plans section

I reverted this new section due to improper sourcing which fails WP:V, WP:OR, WP:NPOV, WP:RS, etc. Right now we have the Brazilian ANAC and CENIPA officially investigating this accident. They have the actual flight plans that were filed by the two flights, the initial IFR clearances that were issued by ATC (which supersede any filed plan) as well as the enroute clearances (which supersede the initial clearance). Clearances include both routes and altitudes. The ANAC also have the transcripts of the readback for each clearance, which are crucial for any potential miscommunication. They also know the specific variations of the Brazilian flight rules from the ICAO rules, if any. The newspapers at this point are printing a lot of conflicting reports and speculations. It is important for us to keep our eyes on the ball and ensure that all our sources are of the best possible quality, as this article relates to living people and to potential liablities. I believe that high quality sources in this case are the ANAC, CENIPA, the U.S. NTSB, ATC tapes/transcripts, CVR/FDR transcripts, etc. I think the other editors on this page are in agreement, as noted above. Crum375 03:00, 9 October 2006 (UTC)

Concur.--Dali-Llama 05:00, 9 October 2006 (UTC)

I don't agree on this particular issue. At best the above reasoning is flawed. At worst we are suppressing information which is known to be true. A flight plan is filed by the pilots or the operator before a flight and is indicative only - it need not be followed by the pilots and clearances are still required for all stages of the flight through controlled airspace. Clearances are obtained piecemeal immediately before and during a flight. Clearances requested by the pilots and/or provided by ATC need not and often do not agree with a flight plan. (Hence the use of the word "plan".) So, let's be 100% clear: The flight plans and the clearances are only very tenuously related. So you cannot argue that the flight plan info is not well enough known until the clearances are known. If the flight plans are a matter of record then they should be in the article. If they are not yet a matter of record then as soon as they are they should be part of the article. The only relevant question to ask is: Are the flight plans now known? If so, include the info. Not to do so is to verge from caution to suppression of information. To delete relevant information which is correct from a WP article violates WP policy. To hide behind the living biography on facts is not appropriate. Speculation? Sure, let's not include that. Paul Beardsell 10:04, 9 October 2006 (UTC)

Yes, I agree that any properly sourced fact should be included. In this particular case, we have no official body telling us what the original flight plans were, let alone the subsequent clearances and readbacks. We also, to date, do not have any reliable official source for the airways involved. As an example, I can tell you that to my knowledge, in most countries of the world, there are no outwardly marked 'two-way' or 'one-way' airways. Yes, there are lots of airways that are internally designated as such by ATC, but that designation may fluctuate over time, and in any case is not depicted on navigation charts. Similarly, altitudes are not directly depicted on charts. Now, this could be different in Brazil - someone in the know with Brazilian (low/high altitude) IFR charts may want to chime in here - but at this point we have no reliable data either way. Newspapers have so far published many contradictory claims. They are not official or reliable sources of information in this type of sensitive WP:BLP related case. There is no need to jump the gun - we will have ample hard information as the ATC/CVR transcripts and FDR data come in, in addition to official statements from the responsible government agencies: ANAC, NTSB, CENIPA. etc. If someone does have access to official Brazilian government aviation publications, such as airway charts, regulations, etc. - that would be most helpful too, especially if they can be accessed online. Crum375 13:04, 9 October 2006 (UTC)

I'm not sure that any knowledge about what airways are commonly in use in Brazil is helpful at all. As I think you know, much marked on an IFR chart is not binding on either the pilot nor ATC. What is important is: (i) What was planned, (ii) what was cleared, and (iii) what was flown. (i) is fairly uncontroversial and not disputed and relatively unimportant in comparison with (ii) and (iii) which are still controversial and so should be left out of the article until some real knowledge is forthcoming. But to say "It is widely reported [1][2][3] that the flight plans filed were as follows..." is very interesting and lets readers of the article come to appreciate what some of the issues are before the issues are resolved. And, it seems, would satisfy your criterion (not a WP criterion) that the info should not need changing much in the future. Paul Beardsell 15:31, 9 October 2006 (UTC)

My main issue with using the flight plans is that as it stands even the pilots disagree with the ATC as to what the filed Flight Level was, with the pilots stating that the flight plan was filed for FL370 all the way to Manaus, and the ATC saying that it's technically impossible as the filing computer would declare a conflict with flight rules of the airway. And, as Crum mentioned, we don't know what clearances and changes were made afterwards. I would be okay with a "he-said, she-said" edit containing the flight plan data, with both the going ATC version and the pilots' versions, but we can't be stating it as fact until the preliminary report is out in the next couple of days.--Dali-Llama 15:52, 9 October 2006 (UTC)

I didn't see the report you refer to although it doesn't ring entirely true. There are several ways of filing a flight plan and not all of them involve a computer at the time of filing! Also, the default flight level is the same for all segments of the route in question so whoever is claiming to speak on behalf of ATC is inconsistent. If the computer would have barfed on FL370 for the 2nd part of the route it should have bletched on the 1st part too. If there is a conflict over what was filed then that is interesting in itself and will doubtless be in the "final" version of the article. So now I am going to have to go off and google for the info myself as some are not allowing it to be here. Perhaps the fault is WP attempting to cope with news at all. But I don't see how we could leave it out. Paul Beardsell 17:04, 9 October 2006 (UTC)

The claim in some of the media that I read is that the first airway segment, from the Sao Paulo area to Brasilia was supposedly 'one way' and hence any altitude would work, odd or even, while the second segment from Brasilia to Manaus was 'two way' and hence only even values for westbound would be accepted. But as I noted elsewhere, this is all conjecture and speculation. The original flight plans were filed with ANAC, whether by computer, phone or paper. Then they were possibly modified by ATC as the official IFR clearance for the trip, and read back by the crews. All of this is normally recorded, and I am sure the actual ATC (and possibly CVR) transcripts are being carefully analyzed by the investigators. At this point all we really have in the press is hearsay and speculation about this process and data. WP is not a play-by-play blog; we try to get it right the first time, especially for WP:BLP related cases, by relying on the best possible sources, which would be ANAC, ATC/CVR tapes, etc. Any other less reliable data source would degrade our quality. Crum375 17:34, 9 October 2006 (UTC)

Globo has published a part of a facsimile of the actual flight plan for the Legacy. The charts for the area where the aircraft was say UZ6 is a two-way airway (as indicated by the blue colour), so RVSM would apply. I'm hoping they'll publish the whole image some time soon. ||| antiuser (talk) (contribs) 00:13, 10 October 2006 (UTC)

Also, note how that excerpt of flightplan says the aircraft must climb to FL380 at the TERES waypoint. The collision occurred about 400km northwest of TERES. ||| antiuser (talk) (contribs) 00:15, 10 October 2006 (UTC)

Where is the descent to FL360 past Brasilia? Is it chopped off the fax? Crum375 00:41, 10 October 2006 (UTC)

I see it now - they just put up a new version. Crum375 00:46, 10 October 2006 (UTC)

Well, I'm glad this part's over--phew!--Dali-Llama 03:04, 10 October 2006 (UTC)

ICAO altitude rules and RVSM

I removed a statement about ICAO altitude rules for two reasons: first, it is unsourced (and per WP:BLP must be removed from this article on sight, as it could be seen as casting blame); second, there is no indication that ICAO rules were applicable or relevant for this particular flight. Let me mention another point while we are at it. This point was also raised earlier by Antiuser, that if the Legacy flight was expected to use FL360 or FL380, this could only be accomplished via RVSM. Up to a few years ago, in most of the world, only the odd altitudes were available and legal from FL290 to FL410. The even altitudes of FL300 to FL400 were simply unheard of (to my knowledge) and never used by normal civilian flights, leaving a 2,000' separation at the higher cruising altitudes. One of the main reasons for the original larger separation (vs. 1,000' at the lower altitudes) was that altimeter error of +/-1% at 35,000' is +/-350', enough to bring two flights separated by only 1000' unacceptably close to each other vertically. The RVSM is an attempt to squeeze more flights vertically into the same airspace by insisting on higher accuracy standards, more equipment and training, etc. and is still somewhat controversial, quoting from WP:

Critics of the change are concerned that by reducing the space between aircraft, DRVSM may increase the number of mid-air collisions and near-collisions. In the US this program was known as the Domestic Reduced Separation Minimum (DRVSM).

Anyway, I suspect that this issue will also come up during the investigation, but I don't see how ICAO can mandate even altitudes for the higher flight levels where clearly using RVSM requires special approval from the aviation authorities.[12]

Of course, Brazilian flight rules could very well be different. If anyone can shed more light on this issue, please chime in. Crum375 12:20, 12 October 2006 (UTC)

RVSM cannot be accurately described as controversial. Such operations are provided for in ICAO annexes and the regulations of various states and have been routinely employed for more than 5 years in suitable airspace with suitably equipped aircraft. RVSM was adopted throughout the CAR/SAM (Caribbean/South American) region from early 2005. In this incident, it is highly probable that both aircraft were suitably equipped and RVSM approved. While it is true to say that voices were raised against RVSM implementation in CAR/SAM, this was based on the expense or inability of older aircraft achieving RVSM approval and the fact that they would be excluded from the airspace, not on direct safety concerns. As for the relationship between ICAO and aviation authorities, ICAO produces Standards and Recommended Practices (SARPS) which the authorities should either adopt or declare a difference, so the default position is compliance. treesmill 13:12, 12 October 2006 (UTC)

Do you know if in Brazil the use of RVSM is automatic or do you need to get special operator/crew/aircraft approval there for it? And when you say 'highly probable that both aircraft were suitably equipped and RVSM approved', isn't RVSM also crew training and operator approval dependent? Quoting from the FAA RVSM document: "RVSM operators must receive authorization from the appropriate civil aviation authority." Do you know for a fact that the ExcelAire company received formal RVSM approval from the Brazilian authorities as an operator? And is it possible that a brand new aircraft just leaving the factory on a maiden delivery flight (ignoring local factory test flights) would not be immediately RVSM approved? Crum375 13:22, 12 October 2006 (UTC)

Also, AFAICT the FAA in the U.S. requires an operator to obtain a specific "Letter of Authorization" (LOA) as one of the rquirements to be approved for RVSM.[13] Does anyone know if Brazil has a similar rule? If so does ExcelAire have such an LOA from Brazil? Crum375 13:41, 12 October 2006 (UTC)

Whereas I don't disagree with the removal of the default altitude material it is for a different reason than Crum's: Quoting details of this particular rule at this time gives it far too much weight. It just does not matter what that particular rule says: What is important is twofold: Firstly, the altitude for which each aircraft was actually cleared by ATC and, secondly, if either of the aircraft deviated from their clearance. The actual details of RVSM seem completely irrelevant at this time and I bet will remain so. Paul Beardsell 17:56, 12 October 2006 (UTC)

A different point: The idea that quoting an aviation rule can defame anyone seems a little far fetched to me. I agree with WP:BPL - I just think that policy is being cited in this case entirely inappropriately. Certainly, if a rule was relevant I would want to see it cited in the article and now. Indeed, already in the article there is info present which allows one to see how blame might be apportioned depending upon facts as they are revealed. But no one is removing that info. Good! But why the inconsistency? Paul Beardsell 17:56, 12 October 2006 (UTC)

I had two problems with the so called ICAO altitude rules: it was not sourced, and it made it sound like the Legacy was at the wrong altitude, despite the fact, as you correctly say, that we don't know what altitude it was actually cleared to. I am still waiting for anyone here to show me the ICAO high altitude rules, and they absolutely do have to do with RVSM, since if RVSM is something you need special approval for, I don't see how even altitudes can be mandated above FL290. Again, this could be different in Brazil, and I'd like to see some reliable source of either the Brazilian or ICAO rules. And for me 'reliable' is not a newspaper reporter quoting someone, but the actual organization's site, as I have referenced above to the US FAA site for its applicable rules. Crum375 18:27, 12 October 2006 (UTC)

Yes. I agree with the 1st sentence. The rest gybes with the 1st sentence because it just does not matter what the default altitudes mandated by RVSM or any other regs as both aircraft were in controlled airspace and therefore MUST fly at the altitude for which they had a clearance. RVSM is just irrelevant. If ATC says "cleared to X at 32451 feet" the pilot says "say again?" to check the weird instruction and then flies at that altitude. Repeating: The only altitude issues are: What was the clearance altitude given to each aircraft? Were the aircraft at their respective altitudes? Paul Beardsell 22:37, 12 October 2006 (UTC)

While googling this issue on the web, I have seen some apparently conflicting information. AFAICT officially in the domestic US and much of the world, since 2005 RVSM is used/required between FL290-FL390.[14] However, I have seen some reports in online forums of pilots who claim that sometimes non-RVSM qualified flights are allowed in the RVSM airspace if they stick to the odd altitudes. I also see that RVSM makes official exemptions for various types of flights, including delivery flights of brand new airplanes to their customer's home base. So it seems the Legacy wouldn't have needed an LOA to go into RVSM airspace, as it was on a delivery flight from the factory. Again, any hard info on this subject is welcome. Crum375 21:14, 12 October 2006 (UTC)

But the rules are simply defaults outside controlled airspace. Inside controlled airspace ATC may tend to allocate altitudes which are in complicance with the defaults or they may not. The issue is did the pilots comply with clearances they were given. So we can just ignore the ICAO default altitudes. Paul Beardsell 22:37, 12 October 2006 (UTC)

You can ignore the defaults, to a point. But, once you determine you have a 'lost comm' situation in IFR, you have a procedure to follow, that relies very much on 'defaults'. Here are the FAA regulations that I cited above. I am not sure how they are affected, if any, when in RVSM class A airspace. I guess we'll find out when this case is over. Crum375 23:00, 12 October 2006 (UTC)

In my opinion the RVSM issue is irrelevant. While it is true that it is possible for non-RVSM compliant aircraft to be cleared to operate in RVSM airspace, they would be kept separated from RVSM traffic by ATC, just as RVSM traffic is. Whatever the planned levels, it is clear that the aircraft operating levels coincided for some reason. Whether this was due to an erroneous clearance should be established from ATC tapes. If not, the Legacy FDR should show at what level it was operating at the time of the collision and therefore whether it was following its clearance. The 737 FDR is apparently in poor condition, but the information from the Legacy should be sufficient to show whether the 737 was at its cleared level. What may be difficult to discover is why one of the aircraft was at the wrong level, particularly if the 737 CVR is unreadable. It is significant that neither aircraft seems to have responded to a TCAS warning, which may have been because no such warning was generated. At this stage the reasons for that are just speculative. treesmill 01:00, 13 October 2006 (UTC)

I'm a little unclear about this but that the Embraer pilots consciously experienced a comms failure or thought that they had a communications failure is not the case. So they should have been following the clearance they had. End of discussion, if I'm right. If I am wrong, and if the FAA regs were to apply, it is interesting that the discussion on selecting an altitude in the rule you quote does not apply. It was an IFR flight but in VFR conditions. Therefore paragraph (b) applies and the correct action upon the Embraer experiencing a communications failure is to land, not to select another or the same altitude. But, as I said, it is ATC alleging a comms failure. Either there wasn't one or the Embraer pilots were blissfully unaware of it. In which case they should have been at the cleared altitude (I repeat myself ad nauseum.) Did the Embraer pilots say they had a comms failure? Paul Beardsell 23:16, 12 October 2006 (UTC)

(outdent) I believe the Embraer crew testified they did have a comm failure, but having read so many different and conflicting versions, I am not sure exactly when it started. OTOH, a brief 'comm failure' is very commonplace, certainly to be expected over undeveloped and uninhabited areas, and when such an outage occurs, you don't instantly start lost comm procedure - mostly you try to raise ATC on some other frequency, and this may take a while. You may punch in the lost comm code. If there are terrain obstacles coming up then you may have to climb, but in this case at FL370 this was not an issue. So yes, there was an outage, of unclear duration, which along with the exact procedure followed by the Embarer crew remains to be revealed. Crum375 23:31, 12 October 2006 (UTC)

Checking here: No. So they had no reason to change altitude from that cleared. Which means all the discussion about default altitudes / flight levels falls away UNLESS they were not in controlled airspace. I think that is settled and they were? Paul Beardsell 23:23, 12 October 2006 (UTC)

If they were told by Brasilia ATC, for example, "expect FL380 at Teres", and they reached Teres while in a lost comm mode, one could argue they should have climbed, but I can also see arguments why not. I guess we need to find out what they actually heard, thought and did. Crum375 23:40, 12 October 2006 (UTC)

I cannot find any reference to the Embraer pilots thinking anything was untoward before the collision. When it comes to comms failures I think the ATC version of events, that they tried to raise the Embraer seven times, without trying to talk to the 737, is just a little odd. Paul Beardsell 23:42, 12 October 2006 (UTC)

The 737 was under CINDACTA-4, while the Legacy was under CINDACTA-1 when it lost communication. According to Globo's Air Force sources, the Legacy lost radio contact upon crossing the BRS VOR, the point when they were supposed to descend to FL360. The question though is did ATC tell them to mantain FL370 before comms were lost? If so, that would override the filed flight plan (unless they said something like mantain FL370, expect FL360 after BRS). ||| antiuser (talk) (contribs) 00:53, 13 October 2006 (UTC)

If BRS is Brasilia, then I find it hard to imagine that radio contact would be lost right over Brasilia. I am sure there are a lot of frequencies one can try in the area of a big city, not to mention other planes on the frequency. The Embraer had at least 2 comm radios, if not more, each of which could be tried. And obviously at least one did work since they used it at Cachimbo. I read a version that said the comm was OK over Brasilia, and was lost subsequently. If 'subsequently' is far enough into the wilderness, then I can see how lost comm could happen there. But by then they would have had to pass BRS VOR, where presumably ATC would have wanted them at FL360. Bottom line is that I am still very confused. If we only had those ATC transcripts, all of this would be very clear. Crum375 01:13, 13 October 2006 (UTC)

I think we are confusing loss of secondary radar (transponder) contact with loss of radio contact. And, it seems, the Embraer pilots were unaware of either. Is that correct? Paul Beardsell 02:08, 13 October 2006 (UTC)

I saw confusing claims in the media of various combinations all 4: radio, primary radar, secondary radar Mode A, and Mode C at different points. But I believe both Antiuser and I are talking about radio loss in this case. Crum375 02:20, 13 October 2006 (UTC)