AF447

[JD-EE]

LonardoSecundo,

May I suggest an addition to your remarks here. You must consider the accuracy of the report and the precision of the report. You add them to get the maximum error band. The precision is how many decimal places are presented. The accuracy is related to the accuracy of the data before transmission.

The precision of the raw data is two decimal places, 1/100 of a degree. This is shown on page 46 of the English report and is not shown on the corresponding list on page 49 of the French report. 1/100 of a degree is on the order of 0.6 nautical miles.

The English language report simply declares, "The position transmitted was the aircraft’s FM position which, in normal conditions, is close to the GPS position."

You stated, "The maximum error of the accuracy of the values of position given by the BEA is 0.5 tenths degree, in one way or the other."

That would be on the order of 5 times the 1/100th degree report precision or about 3 nautical miles. Adding the two, which is conservative since we do not know if the position was truncated or rounded in one direction or the other, we get about 3.6 nautical miles. That places the plane on its proper line of flight within the total accuracy we can expect.

Supposing you misstated what you meant the results are different. The accuracy before rounding of the FM position may be close to the 1/1000th of a second GPS reports or even the 100 meters or more the plane FM rounding would not materially degrade the precision of the report. That would mean they're about 2.3 nautical miles left of track. They may have started a turn a few seconds before the report. Or they may have already been in trouble of some sort.

JD-EE

[LeandroSecundo]

Hi,

You stated, "The maximum error of the accuracy of the values of position given by the BEA is 0.5 tenths degree, in one way or the other."

I stated nothing .. I just quoting this site and Google translation.

Eurocockpit - Accueil

BTW.. this link was already in my first message about this analyse.
And it's indication as a english version will be available soon.

Bye.

[blueriver]

Will Fraser wrote:

Prior to a/p disc. the pilots noted turbulence.

Will,

Apologies if I'm wrong, but I thought the pilots reporting turbulence was misreporting in the media and they actually never reported turbulence ??

[Will Fraser]

blueriver

The two pilots noted turbulence in real time, not as a tx. Not that they 'reported' it, they made note of it to eachother, or FP communicated it to PNF. Read what follows, these are 'possibilities'.

[blueriver]

Will Fraser wrote:

The two pilots noted turbulence in real time, not as a tx. Not that they 'reported' it, they made note of it to eachother, or FP communicated it to PNF.

Thanks for clearing that up for me. I missed that bit.

[PickyPerkins]

#3452

#3452 stated:

Data points: no pre-flight extra fuel

This may not be quite right, because the report says on P. 19:
"The on-board fuel weight corresponded to forecast trip fuel of 63,900 kg,
route factor fuel of 1,460 kg,
final reserve of 2,200 kg,
fuel to alternate airport reserve of 1,900 kg and
940 kg additional fuel.

An LMC (last minute change) corrected the definitive load sheet to take into account one passenger fewer without baggage.

"

So, it had 940 kg additional fuel.

Also, the report says on p. 18:
"The estimated takeoff weight was 232,757 kg (11),
for a maximum authorised takeoff weight of 233 t."

So, the a/c was 243 kg under MTOW on take-off, about 0.4% of the total fuel load, with not much of an option to take on more fuel.

The report also says it left the gate over MTOW.

Without the LMC, the TO weight would have been even nearer the MTOW, 91 kg nearer if the missing passenger was a man, 72 kg nearer if a woman.

Looking at the fuel planning, it looks as though the Captain thought it was likely that he might have to deviate for weather.

[JD-EE]

I did indeed err in the attribution. Please accept my apology.

Here is a link to the full article that is more permanent than the one you showed. Eurocockpit - Archives

It appears the original French indicated "0,5 dixièmes de degré" which surely looks like an awkward way of saying 0.05 degrees or 3 minutes of angle. The plane appears to be well within that error band.

JD-EE

[threemiles]

GPS, SALPU, ORARO, Rescue mission

The English language report simply declares, "The position transmitted was the aircraft’s FM position which, in normal conditions, is close to the GPS position."

FM = FMC ?
This implies that ACARS uses the GPS position as a source and not the IRS which may be a few dots off the real position after the loss of nav updates.

SALPU and ORARO position reports
I have found nowhere that these position report were obviously missed. I cannot find that Atlantico said "next report TASIL". SALPU and ORARO arecompulsory reporting points and should have been reported by the crew outside radar coverage. Estimated crossing times from the INTOL report (0133) would have been SALPU at 0148 and ORARO at 0200.
Also Atlantico should have been after these report within a few minutes after they were failed to be reported. It was obviously not.

Though it may be bothering to report every 12 to 15 minutes in a comms hostile environment and with the possible difficulties to get in contact to Atlantico on HF (not seen this in the report, just the inability of other planes to get a response from Dakar which may be due to other reasons as those of us can imagine who fly there) the crew would have normally tried to relay the report though other aircraft on 121.5. Or?

All this sounds theoretical as there seems to be an accepted lack of discipline about comms and reporting in this part of the Atlantic. I would see this as a contributing factor to the delay of the rescue missions.

A compulsory report is what it says. It is absolutely necessary that ground stations can expect that reporting duties are taken serious and a missing report signals something critical. It is then necessary that ground stations become alert immediately if a report is missing and an alert distress phase is initiated.

Even if the relict of propeller flying, which HF is, is not working there are other means to get a report passed on to the controlling center as Iridium phones or SATCOM ACARS to homebase operator and then by phone or SITA through landlines.

The reliance on "modern"technologies as ADS-C and then give up to report as they may not work is not adequate. However ADS-C in this case was present for the Dakar FIR only, not the Atlantico FIR.

I could not imagine that neither airside nor landside behaviour would have happened inside Gander or Shanwick airspace.

[Squawk_ident]

From Xcitation

Having read that the US are giving up the search, I hope that the French g'ment does whatever it takes to recover the FDR/CVR. If we can find the Titanic then we can find F-GZCP. Just because the pingers have failed the data is still there in all liklihood.

About the signal from the BB. I've heard on the 02nd just after the press conference from a journalist that hat attended to it that, from the 14th of Juy new method(s) should be deployed to find the boxes and that the searches would be stopped whatever are the results on the 15th of August.
I haven't heard something new since.


Threemiles

SALPU and ORARO position reports
I have found nowhere that these position report were obviously missed. I cannot find that Atlantico said "next report TASIL". SALPU and ORARO arecompulsory reporting points and should have been reported by the crew outside radar coverage. Estimated crossing times from the INTOL report (0133) would have been SALPU at 0148 and ORARO at 0200.
Also Atlantico should have been after these report within a few minutes after they were failed to be reported. It was obviously not.

ORARO estimate was a discussion that I had a while ago with mm43.
I now think that ORARO estimated at 0200 could not be correct at 0.82 and that it was 0204 as shown by the radar traces and Acars messages. May be that the crew has misread an entry in the CDU.
Because the estimate TASIL was not received by Atlantico, it was very likely calculated by them according to what they could see ont their radar at INTOL and (almost) SALPU. I agree that SALPU and ORARO being compulsory, Atlantico should have (Sel)call AFR447 to confirm the position and request also TASIL again. Like mm43, I think that they did not mind because the traffic on this route was not important and, also perhaps because of their working load. I think that controller(s) are making at the same time VHF and HF COM and that time was lacking for them to properly handle all communications with a given traffic. Or may be a poor communication between different ATC sectors.
Assuming that on this route it was not unusual to be unable to establish contact with traffics, they did not get alarmed. Some pilots used to these routes might enlighten us on this.

[WhyIsThereAir]

In other words: at 2:10, Flight AF 447 is not on the planned route in the flight plan ", as the BEA, but almost 3 Nm to the west of it.

Does anyone know if eurocockpit (or anyone else) has done this same analysis on the other waypoints reported? Unless they can show that all other points are exactly on the flightplan line this final deviation may not be what they are claiming.

Question for pilots; I apologise in advance for not knowing this: when you set the AP to go from ORARO to TASIL, do you key in "TASIL" and the AP knows where it is, or do you dial in the coodinates of TASIL from the flightplan or memory?

[hetfield]

High crash rate for Airbus Air France

Secret calculations aviation circles reveal a devastating result for Air France: Airbus to the French airline crash often from above. An internal report to the SPIEGEL exists, criticized poor safety culture of the company.

Google Translation of todays SPIEGEL

Interner Bericht: Hohe Absturzquote bei Airbussen von Air France - SPIEGEL ONLINE - Nachrichten - Wirtschaft

[Robin42]

Originally Posted by WhyIsThereAir

Does anyone know if eurocockpit (or anyone else) has done this same analysis on the other waypoints reported? Unless they can show that all other points are exactly on the flightplan line this final deviation may not be what they are claiming

The problem is, that the exact values of the other waypoints aren't known to the public until now. But Page 13 of the French version of the BEA report has a Google Maps image of the flightpath, also showing the ACARS waypoints. Zooming up that map and assuming these waypoints had been entered using the same precision, I'd think they line up very well, except for the last one, which is slightly deviating to the West.

[WhyIsThereAir]

Eurocockpit claims that "AF447 was 3 Nautical miles west, well en route for a 10-nm deviation to avoid wx"... So? What do you think?

I don't know that there is enough information publically available to make thinking possible. Simply going by the route map in the report, yes, there does seem to be some deviation in the 02:10 location. I'll believe that it might have been 3nm off the centerline of the track.

However, the "well en route for a 10-nm deviation" seems to be unsupportable, because we have no idea when they begain a turn, we do not know if they are still in the middle of a turn, and of course we don't know (but can perhaps guess) a turn rate.

[LeandroSecundo]

Hi,

English translation available here:
Eurocockpit - Archives

Bye.

[Robin42]

JuggleDan,

well, it's probably only a subtle pitfall prepared for non-native speakers, but, considering the positional messages are actually spaced at 10 min intervals, does the notions "up to the last automatic position report" or "jusque'au dernier point de position automatique" actually includes the last position?

Originally Posted by french copy p. 71

jusqu’au dernier point de position automatique, reçu à 2 h 10 min 34, le vol s’est déroulé sur la route prévue dans le plan de vol,

Originally Posted by english copy p. 68

up to the last automatic position point, received at 2 h 10 min 35 s, the flight had followed the route indicated in the flight plan,

I consider also the following points:

- The F/O had even more experience than the captain on type: 4479 vs. 1747 flight hours, and a well comparable overall experience.
- The F/O had even more experience than the captain re South American destinations (39 vs 16 rotations)
- GPS precision with 12 satellites in sight is around 4 m on the ground.
- ACARS messages reporting GPS failures or severe degradations had not been transmitted

[JuggleDan]

WhyIsThereAir, Robin42,

First of all, thanks for your answers.

The way I see it:
- In technical French, when you speak about a line segment, you can either say "jusqu'à ce point inclus" (end-point included) or "jusqu'à ce point exclus" (end-point excluded).
- If you don't give any further precision, as it is the case here, it most often means that the end-point is included
- Thus, I certainly understand the BEA report as saying that all position reports were sent from the a/c nominal route, up to and including the last message at 0210 .

[surplus1]

Why is it important if the a/c was 2.5 or 3 m west of track or directly on track? What difference would that make re the outcome?

Whether it was on track or 10, 20, or even 50 miles either side of track, whatever happened did so regarless of its precise position. Since we do not know the pricise position of any convective activity that may have been present in the area, or if that was a factor in this accident, the exact position is not relevant.

If a ship hits an iceberg in the ocean it doesn't matter whether it hit it on course or off course. The only thing that matters is what happened to the ship after the impact.

This was open ocean. There is no terrain for the aircraft to fly into by being somewhat off course. The information is irrelevant to the cause.

So is all the detailed information with respect to ocean currents and locations of debris. They may be very useful in locating the point of impact and therefore narrowing the search for the recorders or the remains of the aircraft, but in and of themselves they are not useful in determining the cause of the accident.

[Dutch Bru]

PJ2; SaturnV; on weather and ops ACARS

Flight crews avail themselves of ACARS met reports all the time. In my experience, company dispatch provided timely and up-to-date information on SIGMETS as a matter of flight following, a legal and operational requirement in many states. Most crews would plot the SIGMET'd area on their planning charts then keep a wary eye using all available techniques including radar. That is not where the problem lies. The problem is in teaching how to use the radar "optimally". Likely, most pilots pick up the use of their radar through "osmosis", which does not filter misunderstandings and shortcomings in knowing one's equipment.

Thanks, PJ, but the smart use of wx radar starts with being aware that things are tricky and warrant extra vigilance. My point is that since the AF447 flight crew according to the BEA report did not ask for nor did they receive SIGMET updates by ACARS, they were disadvantaged in their awareness. And I agree with you that SIGMET and other weather related info is and should be regularly supplied through ACARS, certainly on long distance flights. This makes the BEA statement on the absence of such communication with AF447 even more significant.

Saturn V, where did you find the ops message on the weather (your post 3437)? Surely not AF447 related; I stand to be corrected, but I couldn't find that in the BEA report (as there is no mention of SIGMET 8 and 9 for ATLANTICO FIR)

[ArthurBorges]

Just to Add Some More schlitz to the Fan...

I'd started my own translation in a vain bid to beat Eurocockpit, if for little real reason other than idleness and a personal interest in AF, so here it comes -- if this has a stronger dose of sarcastic spin, I stand by it (once checked draft):


AF 447 Deviated West Just Like LH507


Page 71 of the BEA report lists the so-called "first established facts". Their order of presentation is semantically deliberate: words assume fullest meaning when in the right order. True, BEA did have a whole month to think things through.

Here's the demonstration:

"Up until its last automatically-reported position, received at 021034UTC, the flight was proceeding on the route given in the flight plan."

IN CLEARSPEAK: AF447 made no deviations.

"Weather conditions were consistent with what is normal in the intertropical convergence zone in June."

IN CLEARSPEAK: There are always storms there.

"There was a mass of powerful cumulonimbus formations on the flight path of AF447. Some of them may have been a source of significant turbulence.

IN CLEARSPEAK: There were just "huge" storms along its flight path and, as just said, the aircraft made no deviation..

"Several aircraft crossing the area before and after AF 447, at practically the same altitude, changed their route to avoid the cloud formations."

IN CLEARSPEAK: The other planes flying the same route -- the ones that didn't crash -- deviated in order to avoid the "huge" storms.

IN SHORT: All the aircraft that deviated survived; AF 447 that stayed on track, crashed.

So does anybody know who is responsible for not deviating the aircraft they were flying?

If you haven't figured it out, because we know it's complication, CEO Bourgeon gives you a clue in his Figaro interview:

(see translation of Gourgeon/Figaro interview: At issue is Mr. Gourgeon's statement that LH507 hit a weather bump that induced them to increase radar sensitivity; he went on to say that maybe AF447 "wasn't lucky enough" to hit such an heads-up weather bump)

IN CLEARSPEAK: the pilots of AF447 from Rio to Paris didn't know how to use their radar and "weren't lucky enough" not to get killed. Air France will be training mentally retarded pilots who'll losing their lives until they finally learn how to use the beautiful radars we let them borrow. As for his statement "regardless of whether this was the cause of the loss, we have to study all factors and upgrade the processes and procedures," you have to admit that Mr. Gourgeon doesn't talk that way when it comes to the pitot tubes -- because, as everybody knows, the pitot tubes are not, cannot be, and have been prohibited from being, the cause of the loss of AF447.

There you have it: it's all wrapped up! The plane creasehd because BEA and Air France speak with one voice: the aircraft stuck to its route on the flight plan straight into the "very big huge storms" described by Mr. Feldzer.

But is it really the case?

Is it the "established fact" the BEA says it is?

Page 13 of the BEA report show a map indicating transmission times for ACARS messages -- at 10 minute intervals as confirmed on page25.

AT 0200UTC, the aircrafte has not reached "ORARO". At 0210, it has passed ORARO but not reached TASIL.

On page 69, the BEA mentions LH507 that was about 20 minutes ahead of AF447 at the same altitude on the same route. That flight deviated "about 10 nm to the west" in order to avoid an area near ORARO that gave radar echos. Obviously, the Lufthansa B747 was smart enough to use its radar -- and therefore survived. It survived by deviating only 10nm west of the ORARO area. Think about it.

So what did AF447 do "about 20 minutes" later in the same ORARO region, i.e. between 0200 and 0210UTC?

"The flight followed the route in the flight plan," says BEA. End of story.

But before being as certain as the BEA is, you still need two things:

1. The BEA would need a continuous plot of the aircraft's positions, not a set of positions at 10-minute intervals.

2. It would have to be true.

Because, when you monitor somebody every 10 minutes, nothing proves that person's position during the nine unmonitored minutes in between. So, as Mr. De La Palice might have said, throughout each of the nine-minute intervals during which positions were not transmitted, those positions are unknown.

Thus, regardless of the BEA's somewhat hasty "established fact" nothing proves did not change course somewhere near ORARO betwen 0200 and 0210UTC just like LH507.

Quite the contrary: everything shows AF447 did deviate from the flight plan!

This is easy to show on a map alongside the "route (segment) in the flight plan" between ORARO and TASIL.

Between those two points, a straightline between them on a map amounts to the same thing as a direct curved line between them on a spherical globe. In effect, the distance between them is 119nm and the lateral gap between the two lines is negligible over so short a distance.

I'm letting go here because Eurocockpit has finished the job, but if anyone wants the rest, I'm just sitting here at home with no classes to teach till Sept. 1st.

[txl]

Paraphrase of "Spiegel" report on AF security issues

JuggleDan wrote: Thanks for the link. Too bad Google translate gives rubbish there, but I'm confident the the info shall be translated into another language soon.

As requested, I'll try to paraphrase what German magazine "Spiegel online" reported. Here's the original link:

Interner Bericht: Hohe Absturzquote bei Airbussen von Air France - SPIEGEL ONLINE - Nachrichten - Wirtschaft


"High Crash Rate with Air France Airbuses", the headline claims. The report cites "secret calculations" supplied by "aviation sources". AF Airbus aircraft "crash more than average" - meaning Airbus fleet at other operators. It says AF Airbus fleet has a hull loss rate of 1,26 aircraft per one million flights. Average would be 0,3 per million. Of 19 Airbus planes that crashed since 1988, three were Air France and one Air Inter, now an AF subsidiary.

Those numbers, the report claims, add momentum to the debate about possible pilot errors being a factor with AF447. A "Figaro" interview with Gourgeon is qouted where the AF chief rejects speculations the AF447 crew might have flown right into the storm in order to save fuel and avoid delays

The "Spiegel" piece cites a three year old "internal AF report" that allegedly criticizes "insufficient security culture" at the French carrier. The report is quoted: AF is missing "a clear and objective view of achievements in aviation security" (my translation of a German translation of what was possibly a french source, so take with a grain of salt). Eight out of ten AF incidents were caused by human factor, the report is indirectly quoted, like insufficient attention, bad decision making processes and faulty Cpt/FO teamwork.

Analysis of A330/340 incidents supposedly has shown that Captains tend to show "a certain amount of over-confidence or even smugness", the report is further quoted (again, my possibly-lost-in-translation-disclaimer applies).

According to "Spiegel online", AF claims to have addressed these issues since the report was presented in June 2006. Still, AF/KLM Group Chief Jean-Cyril Spinetta is said to have told investors last week that AF is aware of its bad security record (wording by "Spiegel online", not a direct quote).