Talk:Gol Transportes Aéreos Flight 1907

From Wikipedia, the free encyclopedia

You have new messages (last change).
This article is part of WikiProject Brazil, a project to improve Wikipedia's articles related to Brazil. For guidelines see the project page and Wikipedia:Contributing FAQ.
This article is within the scope of WikiProject Disaster Management.


/Archive 1

Contents

[edit] Criticism of Joe Sharkey in Brazil

Government officials in Brazil have been among those being outspoken critics of comments made by jornalist Joe Sharkey. It became a major issue in Brazil and I would think it might deserve a session here.--Gkklein 18:21, 12 October 2006 (UTC)

I think it is more of a he said/she said type gossip at this point and is non encyclopedic. Since living people are involved, we would have to exercise extreme caution not to violate WP:BLP. Crum375 18:27, 12 October 2006 (UTC)

[edit] Leaked ATC transcripts

I have revised the wording somewhat, after reading the recent reports carefully. I would much prefer getting the 'real deal': actual transcripts, of the air-to-ground communication as well as landline, plus any additional black box voice transcripts. But at the moment all we have are the leaks, which are partial and shaky, but they do match the Legacy crew's depositions. I welcome any comments. Crum375 05:43, 4 November 2006 (UTC)

Regarding the "as filed" piece, I did read some speculation in the press about it, but don't consider it to be pertinent at this point. Yes, the use of "as filed", had there been one, would indicate that the originally filed flight (with its altitudes and route) would be resumed past a certain point. But as we don't really have the actual transcripts, only leaked snippets, I feel we should stick to what we know, and avoid at this point noting what was not said (e.g. maybe the clearance was amended downstream - not likely but we don't have an official version yet). I think just getting the 'leaked' initial clearance and showing it matches the crew's version is reasonably safe WP-wise at this point, and I suggest leaving speculations about what was not said and why for the investigators' final report. Recall that we are dealing with potential errors on the part of live human beings (ATC and/or the pilots), and per WP:BLP we really want hard facts as much as possible. Crum375 14:34, 4 November 2006 (UTC)

I disagree. We have an accurate enough picture from press reports about what happened with the altitude. We're not speculating: the air force confirmed that the ATC erred in its clearance (see [1]). What the "as filed" statement does is explain what the significance of the error was (And it's not even analysis or opinion--we both agree on what "as filed" means). Overall, I agree that the ideal situation is to obtain primary source documents to corroborate the article, but I think that when a news story is specifically confirmed by the air force, its validity is sufficient for the article. --Dali-Llama 18:46, 4 November 2006 (UTC)
I again re-read the Folha article. It mentions an interview with BAF, without giving us the actual transcript of the interview. It mentions ATC communications, without giving us their transcript. It mentions depositions of the Legacy crew, again without actual transcript. The Folha editor, I am fairly certain, is not an aviation safety expert nor otherwise IFR qualified, increasing the odds for misunderstanding. In addition, we have already clearly seen plenty of reporting and analysis mistakes in this case. The 'as filed' issue, IMO, does not add anything critical at this point. Yes, as I noted, leaving out the 'as filed' would further prove that the Legacy crew was correct in their understanding of their clearance, but we really don't know the whole story. There could have been other clearances, or nuances along the way. Since we are in effect casting blame on living individuals (ATC personnel in this case), let's wait for the hard evidence to come in. As I noted numerous times elsewhere on this Talk page, that would be actual transcripts (not leaked 3rd hand tidbits) of ATC comm tapes, CVR, etc. Let's leave the speculation to the media - our mandate is to get it right, to the best of our ability, at the possible cost of not including everything immediately. Crum375 19:56, 4 November 2006 (UTC)

Just because the info will improve over time (although it is not a certainty that all details will be made public) does not mean we should not reflect here at WP what is being widely reported. I think we have been and continue to be too cautious in this article in this respect. What is sometimes happening here is in danger of verging on censorship. There is far too much reluctance here to just report the facts as they emerge: A very interesting leak of a report has occurred must be reported as exactly that. The fact is that of the leak of the report criticising an aspect of the ATC and which corroborates the pilots' statements. I am not saying that WP must give greater credence to the leak than it deserves, just that the fact of this highly relevant leak must not be suppressed here at WP. Paul Beardsell 04:20, 5 November 2006 (UTC)

I agree.--Dali-Llama 04:25, 5 November 2006 (UTC)

The leak is not 'suppressed' -- it is in fact reported. What is suppressed is speculation about the wording or missing wording of the clearance, for which we have no exact copy, nor any subsequent amendments. An IFR flight consists of many communications with ATC, each of which could potentially contain an amendment superseding all previous ones. At this point, given the sensitive WP:BLP nature of this article, and its effect on human lives, I think the current version makes sense. Once we get hard data in the form of real transcripts as opposed to off-the-records leaks from unnamed sources, we can get more detail in. Up to now we have done a pretty good job of skirting the major gaffes of the media, like reporting on the AD that didn't apply, the filed flight plan that was superseded by the clearance, the altitude deviation by the Legacy that never was, etc. Let's keep up our good track record. We are not the media, eager to print to fill news space - as an encyclopedia we try to get the correct information from the beginning. Crum375 04:48, 5 November 2006 (UTC)

But the correct information is that some very well informed reasoning is happening as direct commentary on the leaks. This needs to be reported here. Paul Beardsell 22:47, 5 November 2006 (UTC)
Further, I think the speculation is not quite how you represent it. Of course it is possible that "as per flight plan" could have been tacked on to the end of the clearance. But it is unlikely - ATC phraseology is standardised and ATC is trained not to introduce potential ambiguity. [I have said to ATC "as per flight plan" (drawing their attention to the fact that I have filed one) but never have I heard ATC say anything like that back.] [And a flight plan is NOT NOT NOT a clearance!] Taking your argument to it's logical extreme you would have us consider the tiny (but non-zero) possibility that the two Embraer pilots said to each other "Let's see how close we can get to that 737". Your style of argument would require us to hold that UNTIL THE ENTIRE BLACK BOX TRANSCRIPT IS RELEASED WE CAN NOT BE CERTAIN THEY DID NOT SAY THIS. No. Obviously: The leak, if true, strongly supports the pilots' assertion they were cleared to be at that altitude and thus exonerates them and implicates ATC. And several are reported as saying exactly this. The article should report what is being said in this regard. this is not our speculation and it is not us connecting the dots. But refusing to report well informed and logical speculation which has great bearing is us being selective / editorialising / censoring. Paul Beardsell 22:47, 5 November 2006 (UTC)
For some time the inference to be drawn from the article is that the Embraer pilots were at fault because WP has been reporting what Brazilian ATC has been saying (one aspect of which is now shown almost certainly to be false) and what the media has been speculating. Why now, now the balance of commentary is swinging a little the other way, should WP be reluctant to report this? Paul Beardsell 22:53, 5 November 2006 (UTC)

WP does not have a stake in this, except to report whatever is reliable and neutral at any given point. So the 'balance' may swing arbitrarily depending on the available evidence. What we do want to avoid is a rush to judgment, which many news publications have done in this case. Let's take it one verifiable step at a time. Nameless allegations printed in papers, with a poor track record, are not my idea of WP:BLP-level reliability. And the Brazillian controllers deserve just as much presumption of innocence as the Legacy crew, until the facts are in. Crum375 23:57, 5 November 2006 (UTC)

[edit] alleged radio (voice) failure

I can't find a reference where the Embraer pilots say there was any radio failure. Anyone? Paul Beardsell 08:35, 6 November 2006 (UTC)

There are multiple references in the Brazilian press (which we include in the article) to the Legacy crew asserting in their depositions that they lost comm prior to the collision. In fact, they had to relay via a cargo flight to coordinate their emergency landing at Cachimbo. This is not a controversial issue, to the best of my knowledge. Crum375 13:14, 6 November 2006 (UTC)
Concur with Crum.--Dali-Llama 21:38, 6 November 2006 (UTC)
I used AltaVista Babel Fish to translate the text of the cited reference number 19. I quote one sentence from that translation: "The pilots had affirmed that they had not turned aside themselves from the flight route and at the moment of the shock with Boeing of the Goal they kept contact with the tower of Brasilia." Where are these "multiple references"? Paul Beardsell 07:05, 7 November 2006 (UTC)
I added another reference, that gives some more detail. We'll need to get the full deposition and/or the ATC tapes to get the complete picture, I suspect. Crum375 12:04, 7 November 2006 (UTC)

[edit] ANAC photo copyright

Does anyone know if Brazil has a similar law to the US where government photos are free of copyright? What about specifically the photo here, from the ANAC site, of the CVR memory module? It would great to have the image in the article, like that of the FDR, if we can get it. Crum375 00:44, 7 November 2006 (UTC)

Never mind - found it in the Agência Brasil site (somehow missed it before), and it was already on the Portuguese WP. Crum375 00:50, 7 November 2006 (UTC)

[edit] Heads-up

Just to let everyone know, the preliminary report on the crash is set to be released on Thursday. The report will point out the operational, technical and human factors that played a part in the accident (in the same way the FAA reports are usually structured). Press reports have already indicated the mix of factors is similar to what we've put in the article. It'll be nice put some affirmations on it, though. The final report is set to be released in the second half of 2007, but is only expected to elaborate on the findings of the preliminary reports (IE: don't expect any plot twists).--Dali-Llama 04:26, 15 November 2006 (UTC)


[edit] Confusing information

initial pre-takeoff IFR clearance given to the Legacy was "N600XL, Clear, 370, Manaus", clearing the Legacy to climb and maintain FL370 all the way to Manaus, and superseding the altitudes filed in the original flight plan

I am including a fact checking required reference there, because the current references only report laymen’s words, nothing authoritative to that effect. No one is questioning the clearance, but the air authorities in Brazil say that the original flight plan and the direction of the flight both should supersede the clearance, requiring the pilots to either question the clearance or stick to the plan.

We do include a reference for that clearance so adding 'fact' to it would be incorrect. OTOH, I agree that getting the clearance in the form of an official ATC communication transcript, e.g. with timestamps, vs. a news report, would be better, and I hope we get that version soon. As far as your point about "air authorities in Brazil say that the original flight plan and the direction of the flight both should supersede the clearance", I have never heard any 'aviation authority' say that. In Brazil, that would be ANAC, and I doubt they would state something that clearly contradicts the ICAO and international aviation regulations (including I am sure the Brazilian ones, we do have an online reference for those also). According to the international, ICAO, and US FAA regulations, it is the clearance or instruction at any given moment that supersedes any previous instruction, clearance or flight plan. A flight crew may optionally question the new clearance, but generally don't, in the assumption that some new situation requires new routing and/or altitudes. Crum375 18:56, 16 November 2006 (UTC)
The ICAO (international) and FAA regulations differ. It seems that the pilots were following FAA rules outside of the US, where the ICAO ones apply. The text I amended also assume FAA rules are universal. Leandro GFC Dutra 14:54, 18 November 2006 (UTC)
They do differ somewhat but they are a lot closer than they are different. What rule are you referring to specifically? Citation required. Paul Beardsell 17:23, 18 November 2006 (UTC)
Also to clarify, there is no rule that a flight crew must question a clearance, unless it is very clearly and plainly illegal or unsafe. The contention in the version I reverted, that a crew must question a routine ATC clearance is not sourced by any reliable aviation agency to my knowledge. Crum375 19:18, 16 November 2006 (UTC)
Sorry, these references you mean should be precisely cited in notes to the text, not just generally referred to in this discussion page. Also, if you take the time to read the references that are that, this is precisely what is being alleged by the Brazilian authorities. So I am reverting your reversal. Leandro GFC Dutra 19:29, 16 November 2006 (UTC)
At this point, the version I reverted to does not make any claims about Brazilian aviation regulations. It only states what the Legacy crew did, based on the clearance they received. If at some future point we discover a properly sourced reference that says all those things you state, we can include it. Note that this article deals with living people, hence it must follow the WP:BLP sourcing rules. We cannot use off-the-record statements by nameless persons alleging malfeasance by living people as a source in this situation. Crum375 19:43, 16 November 2006 (UTC)

[edit] Preliminary report released to press

Available here (PowerPoint warning), in Portuguese (emphasis on preliminary). I have also done some tense changes to some slightly outdated sections of the article. Fvasconcellos 11:00, 17 November 2006 (UTC)

Awesome! Thank you!--Dali-Llama 20:00, 17 November 2006 (UTC)
Welcome. Fvasconcellos 12:09, 18 November 2006 (UTC)

[edit] Commission structure

A question to Brazil and Portuguese experts: does anyone know what position Col. Rufino Antonio da Silva Ferreira holds, if any, within CENIPA? Looking in the CENIPA web site, I can't find him or the commission among the names in the CENIPA org chart. Is this a temporary assignment or position? What is the exact relationships between CENIPA and the commission? I realize it's all part of the Brazilian air force, as is much of the aviation system there as a whole, but this particular relationship is very unclear to me. Note that in the PPT report, on the bottom of the first slide, there is "Cel. Rufino/CENIPA" as author. Does 'slash' mean "member of", "working with", "attached to" or what? Crum375 22:22, 17 November 2006 (UTC)

Answering my own question, I found the info in the Agencia Brasil article, which states that Col. Rufino Antonio da Silva Ferreira is the head of the Inquiry Division of CENIPA. For some reason I couldn't find that in the CENIPA org chart. Crum375 23:45, 17 November 2006 (UTC)

[edit] Prelimary report details

I have added some more facts from the report. I think that just like CENIPA is doing, we should try to concentrate on the hard facts at this point. Speculations, especially when based on unreliable sources, should be minimized or eliminated, as this article involves living persons, and as such should conform to WP:BLP highest quality sourcing standards. Crum375 15:30, 18 November 2006 (UTC)

[edit] Question about the collision point

The preliminary report has the apparent coordinates of the collision point (CP), which it says is 10°44'S, 53°31'W, on top of a graphic slide of the collision sequence. Unfortunately, these coordinates don't make sense, as compared to the wreckage site (WS) coords we previously received as 10°29′S 53°15′W, and when compared to the two fixes (per the DOD chart) that bracketed the WS, Nabol (10°33.7′S 53°11.8′W) and Istar (10°20.2′S 53°26.5′W). The coords of the WS seem to lie roughly between Istar and Nabol, roughly 20km nw of Nabol. What doesn't make sense is the CP coords: it is clearly south of Nabol and west of Istar. Bottom line: someone somewhere is confused. It could easily be me, so please check me and find the bug. Normally government agencies are very meticulous with numbers and coordinates, so I am very perplexed. Crum375 01:49, 19 November 2006 (UTC)

37000 ft = 6 nautical miles. Glide angle 7:1 => 42 nautical miles. 1 minute of latitude = 1 nautical mile. Therefore difference between location of collision and location of wreck may differ by up to 42 minutes of lat and/or long. Paul Beardsell 09:32, 19 November 2006 (UTC)
Unfortunately you don't get to glide much with half your wing cut off. Also, the descent path of the semi-wingless Gol, which soon broke up into pieces in the air, is described in the report. In any case, the issue is not the wreckage site, which seems logical at the originally supplied coords, but the collision point (CP), which is well off (south of) the airway. One of the ironic reasons for 2 aircraft colliding over a vast and empty area could end up being the (unfortunate) accuracy of their GPS-driven navigation systems - hence one would expect the CP to be right on the airway's centerline. Crum375 12:45, 19 November 2006 (UTC)
Yes, but the difference is a lot less than 42 miles. Obviously: if you maintain some semblance of control for even a short period of time at 500mph you go a long way. Even a rock travelling at an initial horizontal velocity of 500mph at FL370 will land a long way away. Further: That a collision occurred does NOT imply that the aircraft were on track in the centre of the airway. Paul Beardsell 02:02, 25 November 2006 (UTC)
I don't have the numbers in front of me, but I believe the crash site was only a few miles from the collision point - even a rock flying horizonally at 500mph would come to an abrupt stop and fall down to Earth nearly vertically (essentially following wind currents) if it has a single wing attached to it. And the odds are very high, IMO, that the collision point was smack on the airway. I guess we need to wait to see if the location error is corrected. Crum375 02:10, 25 November 2006 (UTC)
Come off it! (1) 600mph = 10 miles per minute. Slowing from 600mph to 0 ("abrupt" is a nonsense) takes several miles. (2) Winds at altitude tend not to be zero (they were 40knots at 5500 in Northern Queensland yesterday - at FL370 much higher wind speeds are COMMON) so fluttering down with the wind creates extra horizontal displacement. (3) You beg the question by calling the difference between point of collision and point of impact an "error". (4) I didn't say that the collision did not happen "smack on" the airway - I simply pointed out a logical fallacy. But it's all a nonsense anyway: It all depends what you mean by smack on. 100 ft horizontal or 30 feet vertical is enough not to have a collision. Errors in GPS technology and in altimeters regularly(!) are of this order or greater. (And tracking VOR is worse). Additional errors are introduced by the autopilot in attempting to follow the track. (The one in the brand new Cessna 182 I've been flying this week has an annoying wobble.) And, on top of that, autopilots often have at least two algorithms and one of these gives a smoother ride than the other at the expense of accuracy but no autopilot algorithm creates sudden control movements to stay "smack on" track. If ATC tried(!) to cause a collision they would very rarely (1 in a 100) succeed. [That, by the way, is why there are so few collisions despite ATC often (much more often than you think) doing what they did in Brazil that day.] Paul Beardsell 07:27, 25 November 2006 (UTC)

(outdent) Let me try to respond to your points. First, the preliminary report supports my contention that a suddenly wingless plane comes to an abrupt halt and falls nearly straight down: it says the distance from the CP to the debris field is 3.48nm toward the SE, and that was the value I relied on in my description of the CP coordinate discrepancy.
Second, the error I am referring to is a contradiction between the given coordinates for the CP, the various slide graphics and the text. I suspect a typographical error in the CP coordinates; I think if you plot out the Nabil and Istar fixes, along with the given CP, plus the CP to debris field path, and the debris field's own coords, you'll reach the same conclusions. It's not a major issue, just strange and perhaps I am missing something. It would be nice if you, or someone else, could plot these points and reach your own conclusions to compare with mine. I think it's our duty as editors to vet information by performing sanity checks on it - this one fails my own sanity check (then again it could be my own sanity that's failing ;))
Now regarding the likelihood of the location of the CP on the center of the airway, I think that the navigation and autopilot equipment on these two modern RVSM approved aircraft would be somewhat superior to that found on typical light general aviation piston planes (none of which are RVSM approved, to my knowledge), such as the one which you encountered. Hence it would not be surprising at all that both aircraft were tracking the centerline of the airway (which BTW is designed specifically for GPS/RNAV type navigation, not VOR), to fairly high accuracy, possibly no less than +/-100 feet variation from true center (purely guessing here). Of course with that kind of precision, and wingspans of about 100 feet, the odds for a collision would be greatly enhanced. Crum375 15:41, 25 November 2006 (UTC)

As I have shown even the prevailing wind can account for much the difference in position between point of collision and point of impact. That an object starting off at 500mph at 6 miles altitude travels several miles before impact does not surprise anyone, surely? 3.48 miles is not abrupt. What is wrong here is the use of the word "abrupt" and the phrase "nearly straight down" without any attempt to define either. Paul Beardsell 06:48, 27 November 2006 (UTC)
No aircraft is kept as accurately on the centre of an airway as you claim. RVSM does not imply the accuracy you claim. GPS has inaccuracies but they're there. The air in which the aircraft fly is itself moving. But I forget: This is WP. The onus is on you to support your claims. By citation. Paul Beardsell 06:48, 27 November 2006 (UTC)

I am not at all surprised by the fact that the Gol travelled about 4 miles from CP to the ground, that sounds intuitively reasonable. My point was about the apparent mistake in the coords. I am still waiting for someone to plot it all out and give us an independent opinion to cross-check me. Regarding the 'abruptness', I believe the airspeed of the center of gravity of the Gol went from 600mph to near 0 in less than a minute, and fell vertically down to Earth in pieces. I am not sure about the exact acceleration forces experienced; that may come out later from the FDR. Regarding the centerline tracking accuracy, I only threw out guesses, which you are allowed to on the Talk page (when they are not potentially offsensive to someone). I don't know the actual tracking accuracy in this case, and I assume that will come out also from FDR GPS tracking data. In any case, the strict sourcing rule applies to the article, not the Talk page (although as I noted, unsourced negative comments about people are verboten anywhere in WP space). Crum375 13:01, 27 November 2006 (UTC)

An interesting recommendation to GPS navigation software programmers, as well as aviation regulators: promulgate a clear rule and implement in the navigation software a fixed lateral offset (e.g. 1 nautical mile) to the right while following an airway. This is something ICAO may want to look into. Crum375 13:03, 19 November 2006 (UTC)

[edit] Translation of preliminary report

I removed the link to the unofficial translation for several reasons:

  1. It does not meet the requirements of WP:EL, for example, a good quality translation would likely become included as a direct reference, so its inclusion now would fail Links normally to be avoided item #1.
  2. Being an unofficial and uncertified translation, it could be misleading, and we want the best possible reliablity and verifiability for a WP:BLP-related article - recall that we already have the original report in Portuguese included as an official reference.
  3. The copyright notice on each page is unclear - we really don't know how it was translated and by whom, and who owns the copyright to each step. It was clearly first translated into French and then into English (as leftovers remain), and each step could have its own copyright owner. Even the original version could well be copyrighted by BAF/CENIPA and there is no indication that there is permission for its redistribution by this site. Per WP policy we may not link into external sites that include possible copyright violations.

Having said all that, I don't see any major problem in the translation (though lots of minor ones), but it still does not meet the inclusion requirements per above. Crum375 15:51, 19 November 2006 (UTC)

[edit] Identification of last victim

Should the last victim to be identified be named in the article? Several news reports have done so, most probably due to the delay in his identification (thirty-three days after the next-to-last set of remains). I'm not sure as to whether or not this is notable, or even frowned upon on Wikipedia. Also, "last victim" (as I added) may be misleading, feel free to reword as appropriate. Fvasconcellos 00:04, 23 November 2006 (UTC)

Good point. I mulled over this one myself and am unsure. I don't think it's necessarily encyclopedic to have it (i.e. the victim's name), unless it becomes very notable (beyond a mere mention in the Brazilian papers). My own criterion would be to include it if it becomes more notable by becoming internationally known (e.g. picked up and reported by the bigger news media in other continents). But I am open to suggestions... Crum375 00:12, 23 November 2006 (UTC)
Agreed. Fvasconcellos 00:28, 23 November 2006 (UTC)
I modified the victim recovery paragraph to reflect the combined effort, culminating with the final victim ID. Hopefully this is acceptable. Crum375 00:31, 23 November 2006 (UTC)

[edit] Primary vs. secondary radar

I made this reversion, as I noted in the edit summary, to use more accurate and correct language. The issue is simple: we know that ATC lost secondary radar on the Legacy target, and we want to convey to the reader, without having a direct quote at the present (but I am hoping we will have a good one at some point) specifically what that loss entailed in relation to altitude information available to ATC.

The 'editorial' language I used, which is based on the wiki-linked references for now, is (editorial explanation in italics):

The Legacy overflew the Brasilia VOR at 15:55, 4 minutes later, and proceeded outbound on UZ6. At 16:02, 7 minutes after crossing the Brasilia VOR, secondary radar contact was lost with the Legacy, which stopped the updated display of the Legacy's altitude on the controller's radar screen.

My specific intention was to convey to the reader, who may not follow the wikilinks and read for him/herself the long explanations there, that once the secondary radar contact was lost, the radar display no longer had updated altitude information reported by the Legacy's transponder. As I noted elsewhere on this Talk page, some ATC radar displays will continue showing the target, even when transponder information is no longer being received, in a 'coast mode', which would continue to provide the last known dataset. Also, some military radars are able to show altitude information even without a transponder signal. I don't want all that detailed info in the article (which would involve speculation anyway); all I think is needed is to explain what we know for sure: that fresh altitude updates were no longer being received from the Legacy. The alternative wording inserted by Psb777 "[Altitude information is provided to ATC by secondary radar, not primary.]" is misleading in at least two ways, IMO. First, it does not explain that, depending on the radar display model and version, ATC personnel could still possibly see the 'coast mode' data, including the last received altitude information, and second, it ignores the possibility that there was some military radar information available to ATC showing altitude despite secondary radar loss (note that Brazil ATC is run by the military). Hence I think my original version is more accurate. Any comments would be very welcome. Crum375 16:44, 25 November 2006 (UTC)

A problem with Crum's preferred version is that it is not grammatically correct. Perhaps "which stopped the display being updated with current altitude information" might suit? That SOME military radar can give altitude information is a FACT but we have NO indication that the very expensive specialised equipment is installed anywhere nearby (or anywhere in South America) so this remains pure speculation and a "joining of the dots" which I think I have seen Crum argue against elsewhere in WP. But, if we are going to join dots, note that simple trigonometry will quickly persuade you how difficult altitude determination with radar is. The verticle angles are tiny over large horizontal distances. Secondary radar is very, very different from primary. Primary is reflection only. The information is bearing and distance. The distance is fairly accurate and is done by timing the interval between transmission and reflection: Errors are in the order of 10 metres. But the bearing is much less accurate as a necessary consequence of the physics: The wavelength of radar requires huge(!) parabolic reflectors to have a narrow beam: Accuracy as close as one degree is difficult. Over 60 miles one degree is one mile variance. One mile of altitude is 6000 feet. That's why altitude information is NOT obtained from PRIMARY radar. Note also that radar aerials revolve about a vertical axis and do not scan up and down for this very reason! When a transponder is "pinged" by primary radar it responds with an amplified signal (thus allowing even wooden or plastic small aircraft to be seen reliably). This is the "ON" mode. In the "ALT" mode the transponder also responds with its altitude BECAUSE, as I have shown, this is VERY difficult to determine. Given that, no commercial (primary) radar even attempts to determine altitude. Note that military personnel provide some ATC in the UK and in Australia and elsewhere. They use transponders to determine altitude. Terminology is poor here: It should never be called "primary radar" and "secondary radar". The "primary response" to a radar pulse is the reflection. The "secondary reponse" to a radar pulse is a transmission from the transponder. There is only ONE radar but TWO responses to it. So "primary vs secondary radar" as a section header betrays a misunderstanding of the technology. Paul Beardsell 06:26, 27 November 2006 (UTC)

Well, it's always possible that my knowledge "betrays a misunderstanding of the technology". Hopefully reading this reference will help readers understand the primary and secondary radar principles and technologies related to Airport Surveillance Radar. That reference, issued by the US government (for the ASR-11 radar system), tells us that compared to secondary radar, primary radar uses much higher power and frequencies, and uses a different (collocated) antenna. If someone has a reliable source contradicting this, please let us see it.
Leaving the issue of radar terminology aside, let me clarify the specific point here. Our goal, as always, is not to say more than we know, based on reliable sources. In this case we simply don't know what military radar capabilities exist, if any, at the Brasilia ATC facility for altitude detection. Hence, logically we can't rule out the possibility that there was (at least some) altitude information available to ATC based on primary radar. Therefore, we can't say that the loss of secondary radar caused the loss of altitude information. In addition, we simply don't know what radar display system and software existed at that facility, hence we can't rule out the possiblity that (stale) altitude information was still visible in the target's data block, despite its going into 'coast mode' (i.e. not getting fresh mode C replies).

Bottom line: we must stick to the minimum we do know, per reliable sources: we know that secondary radar contact with the Legacy was lost, its mode C altitude replies were no longer being received by the facility, and we know that the altitude reported by the Legacy was no longer being updated. Hence my current verbiage:

At 16:02, 7 minutes after crossing the Brasilia VOR, secondary radar contact was lost with the Legacy, thus stopping the updated reporting of the Legacy's altitude on the controller's radar screen.

Hopefully this makes the rationale clearer. Crum375 04:06, 28 November 2006 (UTC)

It seems my understanding as to the separate nature of primary and secondary radar was incorrect - I will pass on the info to the flight instructor who told it to me! Sorry for being so adamant without checking. All the other stuff I wrote is correct. I note the current wording does still deny that ATC had current altitude info from the Legacy. As you might expect, I am happy with that. The only place I have seen it suggested that ATC may have had access to altitude info from primary radar is HERE, on this talk page. Paul Beardsell 10:06, 28 November 2006 (UTC)

Let me try to clarify again. I am not saying there is any evidence whatsoever that there was any updated altitude information available to ATC once secondary radar was lost, but OTOH given the potential military capabilities we have no evidence from a reliable source to rule it out, either. There are also good odds they did still see the last altitude report received from the Legacy in its data block (but again we have no direct knowledge). Hence my current 'minimalist' phraseology. Thanks, Crum375 11:35, 28 November 2006 (UTC)

I understand. But it remains the case that the only speculation that ATC had access to rare and very expensive technology which may have given altitude information from primary radar is the speculation here and by you. That's a joining of the dots, no? Paul Beardsell 20:39, 28 November 2006 (UTC)

I think you are still missing the point. Let's say we know (based on reliable sources) that X is true, and we suspect (based on synthesis of other sources but no direct knowledge) that Y is also true. In that case WP can only say that X is true, since saying Y is true would be speculation or original research, which is not allowed. In our case here, we know that altitude reports from the Legacy (i.e. its mode C replies) were not longer being updated. We suspect that there could have been military primary radar altitude information, but we have no reliable source telling us that. We also don't know, though we suspect, that the Legacy's data block included stale altitude information, and again, we can't say that, since we have no direct source for it. Hence the only thing we can say for sure is that "the Legacy's altitude reports were no longer being updated". If it's not obvious, 'reports' excludes any primary radar altitude determination, and 'updated' excludes a stale altitude value in coast mode. Crum375 21:03, 28 November 2006 (UTC)

[edit] Little attention paid to ATC comms with GOL 1907

Article says: "The report states that there were no radio or radar contact problems with the flight [GOL 1907] until its handoff to Brasilia Center." Seems to me there were no problems technically but the real problem was that ATC, on being unable to speak to the Legacy, did not tell the 737 to take avoiding action. By contrast, the article details the radio communication with the Legacy. Doubtless my edit will be changed but I'm doing one. Paul Beardsell 21:05, 28 November 2006 (UTC)

I agree with Crum's revert on the lack of sources for the statement. At the same time, I agree with the statement, and it's very well documented (just a question of adding the source).--Dali-Llama 21:45, 28 November 2006 (UTC)
While I think the odds are very high the statement is correct, we would need some official source for it, like CENIPA or ANAC, or the official ATC tape transcripts. Just some anonymous source quoted by a reporter is not reliable enough, IMO, for something that could be construed to cast blame on living human beings (the controllers in this case). Crum375 22:25, 28 November 2006 (UTC)

[edit] NTSB report

I only now realize that on Nov 22, 2006, the NTSB issued this report. Although it is supposedly based on the Brazilian CENIPA version, it seems to have more details in it, and possibly some differences. It does complicate the picture, because it's unlcear if we should present a unified combined version or a separate version for each agency. It would have been so much easier for us if they just agreed on a common version, but life's never easy, I guess. Anyway, we need to decide how to present this information. Crum375 22:09, 29 November 2006 (UTC)

I don't think it's in conflict at all. And CENIPA is the final authority in this case. The NTSB might add recommendations to prevent the recurrence of such an incident, but traditionally it's the role of the local authority to issue the final report. The NTSB document was a press release, not a report--there's a difference. Unless we can find conflicts between one and the other, I don't see any need to include an NTSB press release in the article. If the NTSB does issue a report commenting on the CENIPA findings (which is more likely), then it merits inclusion--Dali-Llama 23:36, 29 November 2006 (UTC)
I agree that the local authority is the primary investigating body, but the NTSB is also officially involved in this investigation, as the Boeing aircraft and the Honeywell avionics were US made, the Legacy flight crew was American and (presumably) certified by the FAA, and the Legacy aircraft was US registered and owned by a US based company. Additionally, there is already some discussion in the press about the differences between the NTSB version and the CENIPA version. I was hoping to stay away from that controversy by skirting it for now and just mentioning the NTSB report's existence (and citing it), but based on your comments I may add at least a minor reference to the press article. I do think that as a government agency officially involved in the investigation, the NTSB's report should at least be mentioned and cited. Crum375 23:53, 29 November 2006 (UTC)
I still think this is a press release and not a report. The first line of the documents says this is a dissemination of CENIPA information. What do you think are the conflicts between the CENIPA document and this document? If there are factual conflicts I agree it warrants inclusion. If not, then it's a press release of what essentially is information originally divulged through other means. The Newsday article does not mention any conflict, only that in their view the CENIPA report focuses more on the Legacy than it does on the Gol flight. I'm trying to avoid having two different versions of findings in the future--especially since the NTSB has not released a report, but a press release of CENIPA findings.--Dali-Llama 17:16, 30 November 2006 (UTC)
I agree with you that having both reports is an unneeded complication. Unfortunately we can't just ignore the NTSB version for various reasons. First, it's clearly not a verbatim translation; while some small parts of it appear to be translations, much of it seems to have different levels of detail - some more and some less. Another reason for its inclusion is the controversy in the press: the Newsday article specifically refers to the NTSB document as "NTSB resport" and compares it to the "CENIPA report". I have seen no reference in the press to the NTSB document as a "press release". The NTSB document refers to itself as an "NTSB Advisory", with the title "UPDATE ON BRAZILIAN INVESTIGATION INTO SEPTEMBER MIDAIR COLLISION OVER AMAZON JUNGLE". Just the fact that there are many additional details in the NTSB report vs. the CENIPA, plus the fact that it is contrasted with the CENIPA report in the press is enough reason to include it, IMO. But there may also be actual conflicts, in addition to just more information, when comparing the two versions. One that I found is this: in the seven minutes between 15:51 and 16:02, according to the CENIPA version there was no communication whatsoever between ATC and the Legacy, while the NTSB version says:

At 3:56pm the Legacy N600XL passed BRS level at FL370. There is no record of a request from N600XL to the control agencies to conduct a change of altitude, after reaching flight level 370. The airplane made calls, but there is no communication in which it requested a change of flight level. There is also no record of any instruction from air traffic controllers at Brasilia Center to the aircraft, directing a change of altitude.

It can be understood from the NTSB version that there was some other (unspecified) discussion between ATC and the Legacy, which did not involve an altitude change request, in contrast to the CENIPA version that flatly says that in that time period, 15:51 to 16:02 there was no communication attempts whatsoever by either side. This could be just an unclear phrasing in the NTSB version (and may relate to an earlier or later time window), but it could also be based on some other document which we don't have, which was used by the NTSB to prepare its version.
Bottom line: I think we can't afford to just ignore the NTSB version for now. It is well sourced, both by itself as well as by the press, and does offer many different details, some possibly conflicting. I do hope we can keep the NTSB reference minimized, as it would be very awkward to try to contrast the versions here, and possibly WP:OR. Crum375 18:23, 30 November 2006 (UTC)
Let me add another point re "conflicting findings". In the case of EgyptAir Flight 990, there was a finding by the NTSB which was disputed by the Egyptian government. I am hoping very much (and fairly sure) that in this case we will have a common finding that will be approved by both CENIPA and NTSB, but in the unlikely case there is a conflict, we'll obviously need to present both versions, just as we do when the NTSB itself is split in its own findings (which is not that uncommon). Crum375 18:38, 30 November 2006 (UTC)

[edit] Charging of Legacy crew

I think we should stay away from inflammatory statements by the ex Justice Minister Dias (now the Legacy crew's lawyer), if we can. The mutual mudslinging is non-encyclopedic; WP is not a newspaper or a tabloid. I think we should focus on hard facts as much as possible and avoid the political statements and controversies as much as possible in this accident article. Comments? Crum375 21:03, 8 December 2006 (UTC)

There are two errors that can be made. One is to not include enough material. The other is to include too much. As the sweet spot of getting the amount of material included exactly right is unlikely to be attained, the question becomes: Is it better to have too much or too little material. Obviously, in my view, too much is better. Paul Beardsell 22:24, 8 December 2006 (UTC)
In this particular example: It seems to me that reporting in this article that the Embraer pilots have been charged and FAILING to report their lawyer's comments is a nonsense and exhibits bias. Neither or both. Neither is not tenable. Both, then! Paul Beardsell 22:24, 8 December 2006 (UTC)
From a different angle: There is resistance [esp from Crum] to including anything which will not be in the "final" version of the article. I continue to be opposed to this view. Reasons: (1) If you want to know the status of this topic _now_ you cannot do so as interesting _fact_ is being suppressed. [And by "fact" I mean to include the fact of authoritative speculation, statements to the press by lawyers, reports in newspapers.] (2) It may be many months before the investigation concludes. (3) It may be much longer before the investigators' reports are made public in part or in full. Paul Beardsell 22:24, 8 December 2006 (UTC)
Can we get more view points about this issue? Where should the 'sweet spot', as Paul aptly calls it, be? Favor mudslinging or keep out of it? Please comment below. Crum375 22:31, 8 December 2006 (UTC)
I am not in favour of "mudslinging". The choice presented is too stark. Paul Beardsell 06:35, 9 December 2006 (UTC)
I wouldn't call Dias' comments particularly "inflammatory" — he is after all a lawyer :) — but I don't think reproducing them is necessary either. Perhaps just stating something to the effect of "the charges were criticized by Dias as being (...)"? That being said, this is still an ongoing event, and WP is in fact an encyclopedia; "coverage" of ongoing events by such a medium will invariably be somewhat flawed, and this may be crossing the line into an attempt at journalism. Maybe in the future, when the dust is settled and plenty of verifiable, non-speculative information is available, we will be able to present an entirely balanced and accurate view of facts. Fvasconcellos 22:54, 8 December 2006 (UTC)
Fine by me. Mr. Beardsell? Fvasconcellos 23:07, 8 December 2006 (UTC)
Yes. We should be restrained in our edits but must strive to create a balanced and fair POV. Sometimes these two aims conflict. Examples: (1) A criminal charge is official and therefore is easy to report. The response by the defendants is more difficult to report (sources are less likely to be verifiable, the language is less likely to be balanced) but fairness requires that we do. (2) The interim official report fails to mention any ATC attempt to tell 1907 of conflicting traffic. This omission of the report MUST BE DELIBERATE (for good reason or bad) as it is an essential issue and one that no one can possibly be ignoring. I am glad this article now copes properly (in my view) with both examples. Paul Beardsell 06:35, 9 December 2006 (UTC)
As for the sentencing, a note: If found guilty, they could actually be sentenced to no more than four years in prison: as they were charged with "reckless"(?, IANAL) endangerment, sentence would be as for involuntary manslaughter (maximum of three years) plus one-third due to the aggravating circumstance (action resulting in death). Fvasconcellos 23:15, 8 December 2006 (UTC)
IANAL also, but the CNN article seems to say that the 12 years is the maximum for the stated charge of "endangering an aircraft". Then Dias seems to say that for the maximum, they would have to show it was premeditated, otherwise it would only be 4 years max. I would leave it the way it is for now, unless we get better sources. Crum375 00:22, 9 December 2006 (UTC)
Yes, 12 years is the maximum for willfully endangering an aircraft; they were charged with recklessly endangering an aircraft (so far: in Brazil, "indictment" is the first step in criminal proceedings). Folha and the FP say it better, though in Portuguese. Sorry for pushing this button, but I feel this should be mentioned. Fvasconcellos 01:15, 10 December 2006 (UTC)

I added this link as a reference to the Legacy crew's situation. I think we may need to add a separate section at some point addressing the controversy around the issue of criminalization of aviation accidents (CoAA). I think most aviation accident investigators abhor the concept of CoAA, as it interferes with the investigation of the accident, which is often caused by chain of small errors, often by multiple individuals and organizations. The CoAA can lead to everyone clamming up for fear of prosecution, which will make investigations much harder, sometimes impossible. The goal of virtually all aviation accident investigators is prevention, not prosection. Ironically, the Cachimbo airbase web site has words to that effect also (which I noted elsewhere on this page), as they seem to be in charge of accident investigation too. There is a lot of discussion of the CoAA related to this case published in the media, e.g. here by Flight Safety Foundation, NBAA, IFAPA, etc. What do other editors here feel about this issue? Does it merit a section in this article? Is it 'encyclopedic'? Thanks, Crum375 14:02, 9 December 2006 (UTC)

Usually the discussion of criminalizing aviation accidents is in reference of non-interference with the official investigation (IE: avoiding having the police investigate an accident in parallel with or in lieu of an NTSB or CENIPA). The Flight Safety Foundation's press release echoes this. No one is arguing there should not be a criminal investigation, but rather arguing that there should not be a criminal investigation component to an official aviation inquiry into the crash. So far, there has been no indication of foul play or that the criminal investigation in this case is hampering CENIPA's efforts. On the contrary, I remember the federal police being denied access to the ATC controllers and the CVR/FDR by the Brazilian Air Force [2][3]). If we find substantial evidence that the criminal investigation is affecting the CENIPA inquiry, then it warrants inclusion. At present, the status quo is adequate.--Dali-Llama 00:56, 10 December 2006 (UTC)
I happen to agree with you, because I like the article to focus as much as possible on hard facts and to report official actions or information, or events that directly affect the accident investigation or the relevant parties. I take it to that we agree to keep the wider CoAA issues off the article, at least for now. Crum375 04:39, 10 December 2006 (UTC)

[edit] Chaos over the Amazon

I just read this Newsweek article, and was wondering how much of it, if any, needs to get into the accident article. The ATC work-to-rule slowdown, the flight delays, the political acrimony, all seem to be well established and in many ways related to this accident (although it seems that the underlying conditions were there before). I don't think we need to get drawn into Brazilian controversies and politics in this accident article, OTOH it probably makes sense to at least mention these items peripherally. Comments? Crum375 17:38, 12 December 2006 (UTC)

I think it's an important issue--just not for this article. If someone were to start a "2006 Brazilian ATC crisis" article, then that is a good venue for it. The most we can do is try to argue that the accident was the "trigger" to the cooped-up grievances of the ATC controllers, but even that is argumentative and tenuous. I don't think we need to get into that when we've struggles so hard to keep this a facts-only article solely on the accident and its events.--Dali-Llama 18:20, 12 December 2006 (UTC)
In Monday's London edition of the Financial Times it was reported that the work to rule by ATC was triggered by the investigation into the accident. It also said that the ATC work to rule highlighted deficiencies in equipment levels and equipment maintenance. This last point, if true, is highly relevant. The FT did not quote a source. Paul Beardsell 23:25, 12 December 2006 (UTC)
I would ask Dali-Llama, if we had a 2006 Brazilian ATC crisis article as you suggest, I presume you would consider it reasonable to wiki-link to it from the accident article, with some minimal summarizing verbiage. Can you suggest such language and can we then use it, pending a WP article, by linking to reputable high reliability publications like Newsweek and FT? Or do you feel we should just ignore the issue? I am assuming that you agree that the actual slowdowns and delays were factual, only the nexus to the accident is shakier, right? Crum375 23:38, 12 December 2006 (UTC)
Were I asked I would say, having now read the Newsweek article, it seems that this accident will be remembered as that which exposed the ATC crisis. There can hardly be any informed people around who are not now anticipating the official accident report to blame faulty ATC as the primary cause. Not that we can say that here, yet. But I agree: Best approach at WP is another article on the ATC crisis and, pending more hard info, a "see also" link to there from this article. Paul Beardsell 04:10, 13 December 2006 (UTC)

Oh, absolutely they were factual. I just don't want to give the impression that this was the end-all, be-all cause for the accident. I think the ATC crisis article is very important, actually, and I'd be willing to pitch in if someone was to get it started. I think the ideal verbiage would be something like: The Gol 1907 accident was also a catalyst for air-traffic controller demands in Brazil which culminated in the 2006 Brazilian ATC Crisis.

We could perhaps replace "demands" with issues, to make it a more technical problem as opposed to a labor relations one (you guys tell me which one is more appropriate). I think the language above addresses the ATC issues while staying away from saying that ATC caused the accident (which is a slippery slope in this case).--Dali-Llama 19:30, 13 December 2006 (UTC)

Retrieved from "http://en.wikipedia.org../../../g/o/l/Talk%7EGol_Transportes_A%C3%A9reos_Flight_1907_b717.html"