The event depicted in this episode are mostly based on the final report. If you wish to find out more, please head to the podcast transcript section of my website where you can download it. Link is in description. Also, this episode depicts loss of lives. If you are sensitive to that, please do not listen.

The story of Air France Flight 447 started two days before the actual departure, during a long stepover for the pilots. Very little is known about the pilots’ schedule prior the flight. All we know was that they travelled two days before and had quite a long break in Rio, and, the outbound flight for Rio went completely uneventful. We know that the captain spent the time with an opera singer in Rio, who was also a passenger of Flight 447, and the first officer travelled with his girlfriend, a physics teacher, whom he was to marry very soon. But it appears as quite unlikely that they actually managed to rest during that time. And rest for us pilots is paramount for operating an aircraft.

We are on the 1st of June 2009. An airbus 330-203 took off from Rio the Janeiro Galeano airport, inbound for a regular scheduled flight to Paris Charles de Gaulle. The airplane was operated by Air France, registered Foxtrot Golf Zoulou Charlie Papa, and had a total of 228 passengers on board. Amongst those 228 people on board were 216 passengers, 9 cabin crew and three pilots. This was a regular flight, scheduled for takeoff at 10pm UTC on the 31st of May, and scheduled for landing in Paris at 9 o’clock UTC on the 1st of June. The flight path was pretty simple, the airplane was meant, after taking off from Rio the Janeiro, to fly along the Brazilian coasts all the way up north and pass above a city called Recife, to which, at this point, it would cross the Atlantic Ocean for reaching Africa, passing via the Cape Verde, and flying over the Sahara. After a short passage over the Mediterranean Sea, Air France 447 was meant to fly over France for a short time and land in Paris on the 1st of June 2009. On board of the aircraft, passengers were for the majority French and Brazilian nationals, but there were also other nationals from other countries.

In order to make the chronology of events as clear as possible, all the time are to be explained in UTC, or the Greenwich Mean Time. Because, as, of course, Air France 447 flew between time zones, in order to improve clarity, we’ll be talking in UTC here.

Air France 447 received its pushback clearance from Rio at around 22.10, clear for starting up engine and takeoff, en route for Paris. Take off was cleared at 22.29, a little late of nine minutes of its normal scheduled departure, and for this flight, the captain was the pilot monitoring and the first officer was the pilot flying. As it is customary on this type of flight, operating by night and being a long haul flight, a third pilot, to whom I will refer to as the relief pilot was on board today, in order to help the two pilots for taking their breaks during the flight and make this more manageable.

Let’s start with the captain, who was a 58 years old male, with a total of 10,988 flight hours throughout his career, 6,258 of which as a captain. He cumulated a total of 1,747 hours on the type A330, and has carried many rotations in the South America sector, where Brazil is located, since he arrived on the division. So he was not a beginner and had an excellent knowledge of the area. The first officer was the youngest on board, and the least experienced pilot too. He was a 32 years old young pilot, with 2,936 hours of flight, 807 of which on the type a330. He was a relatively new pilot, having obtained his private pilot licence in 2000. He was mostly flying on Airbus 320, all along with Air France, having obtained his ATPL, or air transport pilot licence in 2007.

And finally, the relief officer, who would be assisting the First officer for this flight, was a 37 years old male. He had a total experience of 6,547 hours of flight in total, which is quite considerable, and had 4,479 hours of flight behind him on the A330, making him the most experienced on the type. He had, however, performed 39 rotations in the South America sector since arriving on the A330 division in 2002. He was known to be working in the management wing for Air France, overseeing trainings.

Now, Foxtrot Golf Zulu Charlie Papa was fairly new. It was in fact a brand new A330-203, whose serial number was 0660, having entered service in April 2005. It cumulated at the moment of the incident more than 18,870 flight hours and had flown for 2,644 cycles. It was reported that the aircraft had an issue with the Radio VHF1 by the captain on the last landing, but this won’t have any impact on the story that I’ll be talking about today. Other than this, everything else was reported to be in good working conditions, and no issues were noted by the pilots at the time of take-off.

The Airbus A330 is an excellent plane, whose safety records are excellent in general. Airbus, before we start on anything, is the main competitor of Boeing, the American airplane manufacturer whom we talked about in the episode of Palm 90. Airbus is a European consortium. What that means is that, the head office of Airbus is located in Toulouse but most European countries are assembling parts of those airplanes, as this is one of the creations that came with the establishment of the European Union. Airbus, for example, is famous for the 320 type, which is one of the bestselling airplanes in history, as well as the 380, which is the double-deck, biggest aircraft ever. However, an Airbus aircraft is a bit more specific to fly, as it has been a pioneer in manufacturing aircraft using a fly-by-wire system, of which we will talk about in a bit, as well as this detail: you fly an Airbus with a side stick, and not with a yoke, which is very, very different. For example, when you fly an aircraft with a yoke, both yokes are moving altogether, but that’s not the case with a side stick. But… it would be really difficult to explain what actually happened to this flight without mentioning a few technical details about meteorology and avionics in general.

Let’s start with the basics. About weather and turbulences. As you guys know, Earth turns around the sun and spins on itself. But, as Earth spins on itself, it spins at a certain angle, which is what creates is responsible for seasons on the planet. What that means, besides changing temperatures and the landscape where you live, is that, when it’s winter in northern hemisphere and we’re about to cook Turkey for Christmas dinner, it’s actually summer in the south, and they go to the seaside, eating cold turkey, after having opened the gifts. So that’s probably a small detail, but it has its importance, as Paris is located in the Northern Hemisphere whilst Rio is located in the Southern Hemisphere. As a result of this geographical detail, Air France 447 was meant to cross a zone that is called the Intertropical Convergence Zone, or ITCZ. And crossing the ITCZ isn’t much complicated on paper, but it’s always tricky. It’s basically the region on Earth where the heat of the southern hemisphere and the cold of the northern come into a clash. It’s usually quite a bad moment to pass with severe turbulences. With usually rain and other fun things like that. And now, since we talk about rain and turbulences…

Remember that when it rains, well, it’s wet. Even worse when you don’t wanna ruin your makeup.  But, when you’re at a normal temperature, then it’s gonna make your clothes wet, and that’s it. But inside the ITCZ, it’s another story, even worse at 39 thousand feet, or a high altitude. Because of the current of winds, water droplets could be forced backwards or to fly off in different directions, which, as a result, will make them colder and colder. And, of course, what’s happening when super cold water hits something at a high altitude? Well, it freezes instantly. This could hopefully be heard, as it usually means that some sort of noises like hail falling could be heard on the fuselage on the aircraft whilst on flight, which could be not really reassuring, but in the end are not really dangerous. The danger, on the other hand, is when the water droplets hits the pitot probes. And can clog it.

And this point will lead me to talk about this: speed. You guys have noticed when I spoke about Palm 90 that I spoke in knots, and not in kilometres per hour. Why is that, you may wonder, but it’s for this main reason: when you’re driving a car… or a train, for those who can, you are evaluating your speed in kilometres per hour (or miles per hour, depending where you live), because this is the distance that, in an hour, that you’ll reach, assuming that you will keep a constant speed. What that means, let’s say that you drive at 50 kilometres per hour, then in an hour, you’ll be fifty kilometres away on this surface. Perhaps not in a distance of fifty kilometres geographically because of obstacles or mountains on your road, but on that surface you’ll be fifty kilometres away. Now, for an airplane, as well as a boat, first, it’s important to consider this: an airplane does not fly on a surface but on a fluid. So, as a result, if an airplane is at a speed of 500 kilometres per hour, well, depending on the wind or turbulences, then in an hour, the airplane may be closer or farther.

Now, as you guys remember, speed and altitude on an airliner are two very important factors that allow us to calculate our aerodynamics and making sure the airplane is suitable for flying and, therefore, would not stall. For that, we are measuring the pressure coming through two different sources, provided by two different devices: the pitot probes, and the static ports.

Let’s talk first about the Pitot Probes. These are attached to the aircraft, generally before the static ports, and they look like small needles or gunbarrels. Or under the wing, on the Cessna that I fly, on the pilots side. How they work, it’s pretty simple: in flight, it captures a sample of air pressure that passes through the “needle”, or the tube, and, the faster this sample of air is captured, the faster the airplane flies. The thing is, the A330 has, like most airliners, several sets of probes, one for the captain, one for the first officer, and another one. Capturing the speed is essential, this is why, the pitot probes do not work only alone to calculate the speed. They also work with the static port.

Now, a static port is also something different. It’s a small parabole attached within the fuselage of the aircraft, generally not very far away from the flight deck. So why it’s called the static ports, it’s pretty simple: it measures the static pressure around the aircraft, which gives, along with the pitot probes, data that are calculated in an internal computer, called the Air Data Module, or ADM, And this all send all the information about the airspeed, and, ultimately, altitude, to the pilots.

So now, okay, you guys will surely tell me, what if the probes are clogged? Well, as both sources of data, from both the pitot probes and the static ports, are paramount to bring up an accurate airspeed, then if one of them is obstructed or malfunctional, this will indeed bring an inaccurate information to the aircraft. So it will bring up on display changes of speed that do not necessarily make sense, and this was something that already happened in the past, especially between 2008 and 2009. Air France Pilots already faced similar situations, and this could be a bit tricky to diagnose when on flight because, sometimes it could be minor, so showing a few knots discrepancy, but sometimes it can show a big variance in speed and thus, if we don’t have any correct airspeed or clear indication of altitude, then operating an aircraft can become very quickly perilous.

This problem was known by pilots, and was referred to as IAS Douteuse. IAS means, Information Air Speed and douteuse, in French means suspicious, or questionable, so I will refer to it in here as unreliable airspeed. This problem was reported by Air France to Airbus, the manufacturer, who started to look into it, and suggested an upgrade in the pitot probes, so those pitot probes could be safer as they could heat up better. On top of that, Air France issued a safety bulletin where it was explained how to deal with that. In addition to that, the crew received training on type 320, which is smaller compared to the 330, and, when it came to stall, received only a training for stall in low altitude.

Now, let’s talk about general avionics for a bit. Airbus airplanes and some Boeing airplanes, amongst other characteristics, are equipped with a fly-by-wire system. And this fly-by-wire system comes with many amazing benefits, such as general flight protections and, of course, since this is a lighter system to carry, it reduces the weight of the aircraft. And, no, Airbus did not invent this system, this was a system that already existed in military aircrafts such as fighter jets, but, Airbus was one of the pioneer to bring this system to civil aviation.

So, fly-by-wire, as I mentioned, what does it means? Well, in order to understand it, let’s talk about my Cessna. To manoeuvre the aircraft, when I use the yoke, it’s all a system of gears and wheels that turn to make the aileron move, and this in either pitching up, rolling, or going down with the aircraft. But this Cessna isn’t a fly-by-wire aircraft. In an A330, when the pilots want to manoeuvre their aircraft, they inputs the move they wants to do on their sidestick. Basically, when they orders their aircraft to use, pitch or roll, they’re gonna input the move and the system with analyse it, according to the available data it has. Data that generally come from the static ports, the pitot probes, and other sensors on board. Based on this, the ADIRU, or Air Data Inertial Reference Unit, which is the name of the computer, will analyse if the move the pilot wants to do makes sense, and if it does, the aircraft will move accordingly.

One of the main benefits from this system is that, you cannot place the aircraft in a dangerous situation such as a brutal yaw or pitch, or you cannot make an excessive roll on your aircraft, or fly at a speed that would be too low. The system will prevent you to do so and will block your input. For those who play Flight Simulator and have tried in the Airbus 320 Neo to make similar situations, you must have seen that trying to make a 320 stall is very complicated, if not nearly impossible when the system isn’t disturbed. So, the ADIRU does calculate everything from the data it has available. And this isn’t one but three computers, and all three need to agree in order to proceed with the move, and this agreement is called Normal Law. But… what if one computer, after an input from one of the pilot, disagrees?

Well, then this is called the Alternate Law, or Alternate Law2, depending on the severity. Or direct law, in the very worst-case scenario. When on alternate law, an airplane loses most of its protections. It could be something happening, for example, when one of the sources of data brought to the ADIRU is obstructed, such as the pitot probes being clogged or malfunctioning. As a result, the autopilot is automatically disconnect since, passing generally on Alternate Law indicates a problem and leaves the pilot to manage the flight by himself, generally to try to troubleshoot and, of course, keep the aircraft flying, until the aircraft can get back on normal law. One thing is to note also here is that, the A330 cannot be manually set on direct law. At least at this time, and I don’t think this had changed since.

The fly-by-wire system has a lot of advantages, one of them being that it is far lighter than a normal gear system and tends to become more global nowadays in modern aircrafts, as more recent Boeing models are also using fly-by-wire systems, even though this system has many detractors. Usually, people say that this system leave the airplane to have the final word on the crew, but as I explained, this is not entirely true. It’s more like a safety nest for pilots.

Now, there is also the golden rule that us pilots are following at all time, which are Aviate (or, keep me in control), Navigate (tell me where you go), and Communicate (tell me what you do). But communicate isn’t always possible, especially when it comes to flying over an ocean. You guys know, earth is a sphere, and when you’re above the water, and there is no ATC close to you, the next one communication point is sometimes below the horizon. You cannot connect to something that is below the horizon. So this is why, when flying over an ocean, a special training is required because, you’re on your own, technically. But you’re not completely on your own. Nowadays, pilots uses the ADS-B, this is also why you guys can also use website such as FlightRadar24, and track any airplanes across the planet. The ADS-B is a system that gives information anytime about where an aircraft is, but back in the days, the ADS-B was a system that Air France pilots trialled. If you guys wonder what ADS-B means, it stands for Automatic Dependent Surveillance- Broadcast. Either way, this system, also trialled on this flight, failed to be activated by the pilots due to a formatting problem when they input the codes on the system, and this will unfortunately have very dramatic consequences in the search for this flight.

Anyway, the pilots were now in the flight, expecting for the turbulences ahead, and everything went relatively uneventful, as they were soon to cross the ITCZ. From what we know, as the cockpit voice recorder started shortly after midnight, and Air France 447 was flying at this stage at Flight level 350, or roughly 35,000 feet. At this moment, the captain was in the cockpit along with his first officer, and the relief pilot was on his break, waiting for the captain to take his one. And the aircraft was flying normally, even though, it started being rocked by a few turbulences. And the aircraft was on autopilot, which, for now, makes it an uneventful flight.

Captain: It’s the equator.

First officer: Okay

Captain: You understood I suppose

First officer: I thought so. 

Captain: I like feeling where we’re going

First officer: Yeah

Captain: what did he say?

First officer: Well…

Air Traffic Control: Air France four five nine three five zero squawk ident under radar control

Captain: Squawk Identification Air France four four seven.

First officer: it’s not ours. Was that for us?

Captain: Yeah.

Air Traffic Control: Air France four four seven go-ahead.

First Officer: you see it was Air France three five nine

Captain: Error Air France four four seven squawk ident.

Air Traffic Control: Okay Air France four four seven, copied squawk ident for Air France four five nine, maintain three five zero, under control Air France four five nine.

Captain: Well get there, we’ll get there!

First officer: Ah, forty five nine, four four seven, forty five nine…

Air France 447 was operating during the circadian low region. What that means, it’s that it’s usually the time, late during the night, that the human brain is usually less alert. At that moment, the captain, whom we just heard, and the first officer, were in command, and were mostly discussing about the weather and random other things. At this stage, the two pilots on the flight deck were facing a pitch black weather, with the moon in their back, and were just monitoring the autopilot operating the aircraft. The thing was, they received earlier warnings about severe thunderstorms ahead, so, they tried to have a look, increasing the range of their radars on board, to try to see what they had ahead. They were in the stage of the flight where there were less paperwork to carry out, and they were mostly monitoring the autopilot and checking if everything was okay.

Captain: Okay will do.

First officer: We’ll soon be asked to climb surely

Captain: Yeah. And now he doesn’t want to? He doesn’t want to?

First officer: no, no, he doesn’t want to. I’ve checked it’s the right code.

Captain: he doesn’t want to.

First officer: so we’ve got a thing right ahead.

Captain: Yeah I saw that. Standard plus twelve.

First officer: Yes, yes still otherwise we’d… we’d have a problem most probably higher up.

Captain: Ah yes. That went up slightly earlier at three forty eight. The… the optimum.

The thing was, between 1.45 and 2 am, the captain and the first officer noticed that they entered a cloud layer and, as a result, started discussing about a strategy to adopt should the plane enter severe turbulences. The first officer wanted to change flight level to climb two levels, so to reach flight level 370, or about thirty-seven thousand feet. With the ITCZ ahead, this reaction made sense in a way from the first officer who knew that, with the cloud layer, it would be even harder to make any move and, turbulences would quickly start rocking the airplane.

So, turbulences in themselves are of no danger for any aircraft as I explained in the Air Florida flight 90 episode. The main risk is in fact injuries for the crew member or passengers if the turbulences strikes the aircraft too hard, but, as the speed the aircraft was and the condition the A330 were flying, it was absolutely of no danger and climbing would mean for the aircraft, avoiding the turbulences, or at least, perhaps not avoiding them but making these less severe, if they were taken from a higher altitude.

Anyway, at the same time the first officer suggested that it would be better to climb, the captain also recognised that it was clearly not possible at the moment to do so. This seemed to be something that preoccupied the first officer, as it was shown that some anxiety was present in his requests. But the captain was unresponsive to his requests. In fact, he completely dismissed it, but did not explain the reason why. As I mentioned before, and even though we do not have the clear information of what was on his instruments before, he seemed in fact actually relaxed by the crossing of the ITCZ. He did it several times in the past, and it has been assumed that, according to what he saw on his screen, he did not feel that there seemed to be a major drama to cross it this time. In fact, he was certainly expecting to have a bit of turbulences, but… overall, the usual. It would be just another flight. And, in fact, the crossing of the ITCZ was, at least, during that night, normal. There was bad weather, but this was completely expected and nothing to be truly scared about.

Either way, at 2 am and 01 minutes the captain wanted to take his break. The airplane was to cross soon the ITCZ and slight turbulences were already rocking the aircraft. As the plane was now stable, he asked for the Relief pilot to come back and, as the 330 has a small compartment behind the cockpit where the crew can take their breaks, the procedure was that, the copilot would remain pilot flying unless otherwise has been agreed. As the relief pilot entered the cockpit, the copilot, who was in command, asked him:

First officer: did you sleep?

Relief pilot: So-so…

Captain: You didn’t sleep?

First officer: he, he said so-so, so-so…

Captain: Well, then I’m out of here.

Relief pilot: I… I was dozing in fact. Are you okay?

As the procedure required him to do, the first officer immediately briefed the relief pilot on what was happening in flight. The fact was, the procedure imposed by Air France back in the days stated that, as soon as the pilot monitoring was to go on his break, the relief pilot would come automatically as the new pilot monitoring unless agreed otherwise. So, for now, the first officer would remain the pilot flying, so keeping an eye on everything, if everything was operating normally, and as a result, he briefed the relief pilot, who sat at the seat of the captain. Also an interesting factor that would also certainly contribute to the accident: we do not exactly know when the relief pilot started his break, but, the thing was, as the captain called him, he said that, he hasn’t been sleeping, but instead, he was dosing. As you guys remember, we’re operating on the circadian low time, and, this would perhaps have had certain consequences in what’s to happen now.

First officer: Okay… well the little bit of turbulences that you just saw we should find the same ahead, we’re in the cloud layer. Unfortunately we cannot climb much for the moment because the temperature is falling more slowly than forecast so what we have is some REC MAX a little too low to get to three seven.

Relief pilot: Yeah.

So now you may say, okay, the captain is out of the flight deck, and they know that it’s gonna become soon difficult to fly due to the weather. The thing is, and I think this was what motivated the captain to actually take his break now to come back at a moment later in this flight, he certainly believed his two pilots more than capable to carry out the crossing, considering the experience of the relief pilot on the aircraft, for dealing with the turbulences. Moreover, the aircraft, besides bouncing a little, which was something completely normal for the area as the weather was particularly bad during that night, was stable, and did not have any major issue that might require the attention of the captain during the crossing. So this decision is perfectly justified and, in a way, makes sense.

Anyway, now that the captain was gone, it was just random and piloting chat between the two pilots in front. The relief pilot took the place of the captain, and we also know that he set his seat on Idle, probably to make himself more comfortable and he needed some more legroom, and both were still monitoring the aircraft flying on autopilot. Not long afterwards, the relief pilot switched the gain on the weather radar from “calibrated” as it currently was during cruising to “max”, making the aircraft even more sensitive to weather change and, to give the possibility to the two pilots to have a clearer view as to what was ahead of them, should they need to change their strategy. It was still pitch black outside, and rain started to fall. At 2 am and five minutes and 55 seconds, as turbulences were becoming a but more intense, the first officer make a decision: informing the cabin behind of the spectacle ahead.

First officer: we’ll call them in the back to tell them anyway because…

Flight attendant: Yes?

First officer: yes, it’s me, in front, just so you know, in two minutes there we ought to be in an area where it will start moving about a bit than now, you’ll have to watch out there.

Flight attendant: Alright, are we going to sit down?

First officer: well I think that it might be a good to tell you?

Flight attendant: yeah okay, I’ll call the back. Thanks a lot!

First officer: yeah I’ll call you when we’re out of it.

Flight attendant: okay!

But they were cruising, once the handover was complete, rocked by turbulences, but nothing unusual really happened. The captain was now in his break, and the two pilots were chatting and monitoring the aircraft, with the first officer still in commands of the aircraft, flying as planned, with absolutely no change agreed for the crossing of the ITCZ, they planned to fly exactly as they were flying to cross the zone, which has already been crossed by many aircraft’s at the same period of time. But eventually… it started smelling funny in the cockpit.

First officer: ah… you did something to the AC?

Relief pilot: I didn’t touch it.

First officer: no but to the AC? What’s that smell now?

Relief pilot: … it’s… it’s ozone.

First officer: its ozone, that’s it… we’re alright.

Ozone? What’s ozone? Well, you surely heard about the ozone layer? The fact that the aircraft was now flying really high started to overwhelm the AC and as a result, some strange smell came in the cockpit. But nothing dramatic. I believe our relief pilot will explain it better than me.

Relief pilot: no it’s er… it’s the air with an electrical charge.

But it kept raining… and the rain turned to hail, making this sound inside the aircraft, of the small tiny pieces of ice hurting the fuselage. But inside the flight deck, nothing really changed. And the monitoring kept going as normal…

First officer: there I’ve taken it down a bit…

First officer: there you are.

Relief pilot: It’s free.

Relief pilot: Go on.

First officer: here, Go ahead.

First officer: Do you us to put it on ignition start?

But right as the first officer started to ask this question, literally out of the blue, was heard that sound Cavalry charge indicating a sudden disconnection of the autopilot. This is not something that is supposed to happen out of the blue, especially in the middle of turbulences, since this is generally the first indication of a problem that needs to be troubleshooted. But it gives the priority to the pilot flying to get control of the aircraft and the pilot monitoring to try to find out what’s going on. At least in this case. You can still hear it if you disconnect the autopilot manually, but this sound is very loud, and immediately indicated to the first officer, or at least, pilot flying, that he needed now to take the commands again. To which, in fact, he immediately acknowledged:

First officer: I have the controls.

Relief Pilot: Alright.

But for now, the two pilots were suddenly surprised by the sudden disconnection of the autopilot. The actual reason why the autopilot started disconnecting was because, in the pitot probes, as I explained earlier, some small ice crystals started to form, as the rain was falling around the aircraft. But the reason was, one of the pitot probes was now completely clogged, bringing to the ADIRU system an inaccurate information, and it was more a measure of safety the aircraft was applying by disconnecting the autopilot, as the information received by the onboard computer became inaccurate. But that… it was just the beginning of the problems.

But, the pilot flying would face a very sudden situation, when the autopilot disconnected, as the aircraft violently rolled on the right, and that needed an immediate intervention of the pilot in order to keep the aircraft afloat and in trajectory. Later simulations have highlighted that the roll was in fact due to the accidental inputs of the pilot flying, surprised by the sound suddenly popping to his ears. And if you ever flew an aircraft, you guys may know how hard it is to keep good tracks of altitude when you do not have the help of the autopilot. But like I said, it was pitch black outside, and the two pilots were facing a problem that would be tricky to find out: at this stage, with one pitot tube clogged, the information on the display were already inaccurate, and the fact that the wrong information of speed were displayed brought also a recalculation of the altitude of the aircraft, showing a loss of three hundred feet.

But, this loss was in fact… not a loss. It was an error from the system based on the false information it had, now ready to mislead the pilot flying. And of course, the first officer, thinking that his aircraft lost altitude due to the disconnection of the autopilot and perhaps because of the accidental roll he ordered the aircraft to do, started pitching up the aircraft. But, anyway, good news, the pilot flying managed to counter the roll…

First officer: Ignition start.

Now, what was happening on the instruments was quite peculiar. And since the airplane was passing through a layer of cloud in the midst of turbulences, it was impossible for the pilots to rely on any other information but their instruments.

Let’s start with the flight director. Or FD. The flight director consists of two bars, generally displayed in green above the artificial horizon (the famous blue and brown zone that moves depending on whether the aircraft is rolling, pitching or going down and that gives us the indication as to what we are actually doing). It’s basically a prediction as to what the aircraft will be when the move the pilot executes on his side stick is completed. It is pretty much an accurate synthesis as to what the altimeter, the speed indicator, and other many settings of flight are giving. This system making the summary, when you are inputting a move, as to how your artificial horizon will head once the move is completed. As the system had one of the pitot probes missing, it started adapting itself to the lack of information… and quickly disappeared from the screen.

At the same time, as only one of the pitot probes was working, the speed samples were immediately recalculated by the on-board computer. At this moment, until now, the aircraft was flying at a normal speed of about 278 knots, which was a normal speed. But due to the change of information as those were now corrupt by the clogging of the probe, the samples of air were slightly different, and as a result, the speed was recalculated as being at 60 knots, which was highly improbable for the pilots. On top of that, the calculation of altitude dropped by a mere 400 feet. So indeed, he did what any pilots would have done, and climb, unfortunately misled by his instruments.

60 knots is generally, depending on the wind, the takeoff speed of my Cessna. For a A330, it’s above the taxiing speed. So it didn’t make any sense. Oh yeah, and by the way, I forgot to mention. This was just the beginning of troubles. Because very soon will come the biggest surprise. Which is precisely this: (stall warning)

Relief Pilot: What is that?

Now, the perfect chaos was taking place before the pilots eyes. Not only the flight director was gone, the autopilot was now off and unable to restart, it happened in the middle of the flight when they were not expecting it… but they had no valid information of speed and altitude was questionable, but none of them seemed to have noticed the drop of altitude. On top of that, the Flight director went completely missing from the indication display, but nothing was said aloud about the things that brutally did not make any sense, because everything, from the C-chord and the master caution and master warning that were now fully lit, the first stall warning that got heard… and the altitude that is strange, with the speed now that do not make sense… you remember that story of the circadian low region, and the relief pilot who was dozing?

First officer: We haven’t got a good…

First officer: We haven’t got a good display…

Relief Pilot: We’ve lost the, the, the speeds so… engine thrust A TH. Engine lever thrust

First officer:…of speed.

At this point, unfortunately, not only, even though the first officer was mostly pushing the sidestick down, the aircraft’s altitude remaining increasing, with a speed indication slightly increasing, giving to the pilots the false impression that they were accelerating. As he said this, the speed indicator climbed slightly to 73 knots, barely moving from a few knots, but this was genuinely misleading. In fact, at this stage, it was a recalculation from the central computer considering that the pitot probe brought a bit of air, but this was completely false. In fact, the aircraft had its vertical speed increasing, and had now climbed from Flight Level 350 to 362.

We were now in a challenging situation. It was now at 2 am, 10 minutes and 20 seconds when the aircraft called an ECAM message, that popped up on the pilot screen. This message read, “Alternate Law2”. This was in the middle of severe turbulences. The c-chord was still in the ears of the two pilots, indicating that a master caution was advised, and the confusion became tenfold more intense in the cockpit. The change of fly-by-wire law could have been perceived by the pilot flying, as the aircraft, no longer in autopilot, would have become more sensitive to his inputs.

Relief pilot: Alternate law protections—(low/lowlo).

First officer: (engine lever?)

Relief pilot: Wait we’re losing…

Relief pilot: Wing anti-ice.

Relief pilot: Watch your speed.

Relief pilot: watch your speed!

Thinking that the aircraft was now dramatically pitching up, the first officer kept pushing his side stick in order to try to stabilise the aircraft and try to make sure the aircraft wouldn’t pitch down, but the reality was different. The aircraft was now in its most dangerous situation, as the speed was decreasing due to the fact that it kept climbing, and the necessary lift to generate to remain flying in the sky would soon be about not to be enough as the speed was badly decreasing. The pilot probes were still frozen. It showed now an inaccurate and improbable speed of 100 knots. At the same time, the aircraft kept climbing, and the movements of the sidestick of the first officer made the airplane having the vertical speed slightly decreasing… But, again, this was an illusion. And the aircraft kept climbing. Up and up. In fact, it was about to cross Flight Level 370, and enter in a high-altitude and put itself in a very perilous situation. Even though the first officer put inputs to try to decrease the altitude, the plane was still pitching.

And the second clue was more subtle, as it was now up to the relief pilot to try to suddenly troubleshoot the big issue the plane was now facing, was the fact that another ECAM message popped on their screen, indicating that a maximum speed should not be above 330 knots. So that could be an indication that the aircraft would be in fact to enter in over speed.

First officer: Okay, okay okay I’m going back down.

Relief pilot: Stabilise.

First officer: Yeah.

Relief pilot: Go back down.

Relief pilot: According to that we’re going up.

Relief pilot: According to all three you’re going up so go back down.

So, now… if you’re wondering why an airplane is losing speed when it pitches up and gain speed when it goes back down is all about simple physics. Take a ball and launch it in the air. The air will slow the ball as air, as well as gravity, will try to catch it again. In physics, it’s called swapping your potential energy for kinetic energy. It’s basically the same for an airplane: for climbing, it needs to increase its thrust in order to keep its altitude and stabilising at a higher altitude. Unfortunately, the two pilots here, whilst the aircraft was climbing, did not really change anything in the thrust. They were hoping that, going back down will keep the kinetic energy enough to counter their potential energy. A simple bet.

But at this stage, it was more for the two pilots all about trying to understand what had now happened. But the fact that the flight director kept coming and leaving the artificial horizon, was already enough to diagnose an IAS douteuse. But, they seemed to have missed the information. On top of that, the workload of the pilot flying was now dramatically increased by the fact that he needed to keep control of the aircraft, and for that, he deactivated the Thrust lock perhaps in order to search for clues. But the stress was certainly at its most intense point. But, the flight director may have been disappearing again, something strange now appeared on the airspeed, right next to the artificial horizon: the display of speed came back, showing now a speed of 230 knots. A fairly low airspeed. Realising what was happening, it would make sense that the pilots would be wondering, is it the correct airspeed now? One thing is sure, 230 knots, true or false, would not be enough in high altitude.

First officer: okay.

Relief pilot: You’re at…

Relief pilot: Go back down!

First officer: it’s going we’re going (back) down.

Relief pilot: Gently.

Relief pilot: I’ll put you in in A T T…

Relief pilot: what’s that…

First officer: We’re in… yeah, we’re in climb.

The stress quickly increased since the autopilot got disconnected for the two pilots who were now facing a situation they had a hard time to understand, and it is possible that they could have been, at this stage, overwhelmed by the ongoing situation, lacking evidence to understand the situation into which they were into. Between the continuous c chord, the airspeed that did not make any sense, the flight director that keeps coming and disappearing and the altitude multiple times recalculated, none of them seemed to understand. So, perhaps out of the stress they were facing, the relief pilot called the captain more than seven times, pressing the ... button, but something, outside of the aircraft, was now happening: the pitot probes got cleared and now were perfectly deiced. Now, the new problem was, how comes that the aircraft, that had its pitot probes working, was still showing an inaccurate airspeed?

It was now 46 seconds that the incident started and the pilots already faced one stall one warning, a continuous C-Chord, a master caution, and the environment became incredibly stressful for them, when, 46 seconds ago, everything was just completely normal. In such a very short time, the situation turned completely upside down and they were now desperate to find out why.

And… When something becomes inaccurate, you stop trusting it. Unfortunately, when the situation started, the airplane was at flight level 350, or 35 thousand feet, and was now passed flight level 370, or thirty-seven thousands feet. And now the airspeed, as the pitot probes started functioning normally, was recalculated and showed an airspeed of 121 knots. Still very inaccurate. Perhaps at this moments the two pilots realised in what mess they actually were: in an airplane far too high, with absolutely inaccurate speed indications, with a vertical speed that was slowly decreasing but an aircraft still climbing. There was only thing that could save the aircraft at this moment: descending, in order to avoid the stall and regain speed. But the fact that the pilot flying kept pushing back in order to gain speed, certainly in a natural reaction of confusion or stress, brought the situation now nearly impossible to resolve and would eventually doom this flight.

But at this stage, the aircraft was far too high and at a low speed. What that meant, in terms of physics, was pretty simple: the airplane, having betrayed the pilots due to the fact that the probes got frozen, lost a lot of speed since the pilots, thinking they were going down, were in fact going up, reducing considerably the speed of the aircraft, having swapped their kinetic energy for potential energy. Now, the aircraft started buffeting, vibrating in every possible ways. This was… indeed, the indication of bad news. The aircraft, in terms of physics, entered now a high-altitude stall. And, yes, theoretically, it could have been saved. It could have, if only the pilots figured out the real situation they were into.

Relief pilot: Where is he eh?

Relief Pilot: oh fuck…

Feminine voice: Hello?

First officer: TOGA

Feminine voice: yes?

Relief pilot: Above all try to control the lateral touches as little as possible eh.

Feminine voice: Hello?

First officer: I’m in TOGA eh.

As the first officer declared that the thrust was now in TOGA, the full airspeed was now correctly displayed, since the pitot probes were now completely clear of any ice. At this moment, the airplane was flying at a speed of 180 knots, which is barely above the takeoff minimum speed. The first officer had the flight director back, indicating temporarily that the aircraft was now heading down, but this seemed very weak. 

Now the two pilots were startled, because the calculated airspeed and the displayed airspeed became suddenly accurate, by the fact that the pitot probes were clear, this showed their real speed: a very weak 183 knots, which was, truly an unbelievable situation. At the same time, the aircraft was now at 37,924 feet, and already six different stall warning have been heard in the cockpit. The first officer now placed the thrust levers on TOGA position, applying full thrust position, and the vertical speed started slowly decreasing, even though the airplane was still nose up, and not descending at all. In other words, the airplane was now in a fully developed stall, and getting out of it would require very crucial and critical decisions. Decisions that, unfortunately, our two pilots would have to take very, very quickly.

Relief pilot: Is he coming or not?

Both First officer and Relief pilot: Damn, quick!

Relief pilot: But we’ve got the engines what’s happening (…?)

Relief Pilot: Do you understand what’s happening or not?

First officer: I don’t have control of the airplane any more now.

First officer: I don’t have control of the airplane at all!

Relief pilot: Controls to the left!

At this time, by saying controls to the left, the relief pilot requested control, which didn’t seem to be fully acknowledged by the first officer as he didn’t reply anything but was instead absorbed by the situation and accidentally slightly rolled the aircraft to the left. This was another indication of a profound confusion in the cockpit. From now on, until the end of the flight, the flight directors completely disappeared, this seriously increasing the stress of the two pilots. The situation they faced was now of an increased stress, and the stall warning, in the background, kept ringing. And they certainly realised it: vertical speed was going down, airspeed kept falling but they were probably wondering, should we rely on this, nose was up, this indicated that the airplane was, at this stage, more falling than flying, in a stall that became harder to recover. But now you may be wondering: okay, what about the warning?

Have you guys ever been in a very, very stressful situation? The fact is, and it’s human, the human brain becomes poorly productive and completely disorganised when it comes to stress and, all of us have different ways to react to a very stressful situation, and this reaction is called the fight-or-flight response. The normal reaction we usually have when we are stressed is, going back to our natural instinct: will we fight the problem, or will we try to escape? And if we fight the problem, are we able to be clear-minded enough to give an appropriate response? Now, of course, all the indications of this flight were becoming inaccurate, and the pilots had absolutely no idea of what was going on, because they were all surprised by the intensity of the events that happened, and the total lack of understanding of what they were in fact facing and had to deal with. Now, regardless of the fight-or-flight response, when facing a stressful situation, the first thing that your brain instincly does, is to try to reduce the sensorial information, to go in a sort of lockdown mode. And the first thing that goes away when you’re in a scary situation… is your hearing.

So everything became confused. Yet, this was about to change, as the captain re-entered the cockpit. But at this same time, considering the fact that the nose of the aircraft was up and, therefore, unable to catch properly the flow of air to give any accurate indication of speed, the main computer started showing inaccurate information, until the point, after having showed a 38 knots speed (which is a taxiing speed and not even a takeoff speed), ceased showing any information… for the reason that none of the information made any sense for the computer. The aircraft was in a position that was barely believable and not studied to be into.

Relief pilot: Hey, what is that?

First officer: I have the impression that we have the speed.

Captain: Er what are you (doing)?

Relief pilot: What’s happening? I don’t know I don’t know what’s happening!

First officer: We’re losing control of the aeroplane there.

Relief Pilot: We lost all control of the aeroplane we don’t understand anything, we’ve tried everything… er…

Captain: So, take that, take that!

Relief pilot: Take that, take that!

Relief Pilot: Try to take that.

First officer: I have a problem, it’s that I don’t have vertical speed indication.

Captain: Alright.

First officer: I have no more displays

Relief pilot: We have no more valid displays.

First officer: I have the impression that we have some crazy speed, no? What do you think?

Now that chaos was digging deeper and deeper, the captain stood behind his two pilots also trying to figure out what was happening. Since he hasn’t seen what started the chaos, his understanding of the situation was based on absolutely nothing, and it was hard to make a diagnosis on the fact that the aircraft entered a high-altitude stall. Even though he heard the stall warning and saw the master caution warning, he entered the cockpit faced with certainly the most unprecedented situation of his career. To try to understand what was happening, it would have been important to question the two pilots as to how the situation started, but they now faced a situation that was chaotic and needed immediate attention so there was no time for that. Unfortunately, at this stage, it was far too late to have the airplane recovered. Either way, up until this stage, the first officer wanted to extend the flaps of the aircraft, thinking that the aircraft was going incredibly fast but, this wasn’t the case and the relief pilot prevented him to do so. On top of that… the airspeed, as the first officer said, I have no more valid displays”, completely vanished from the screen, leaving the first officer an error message instead of where the speed would be displayed.

Relief pilot: no.

Relief pilot:No, above all, don’t extend the…

First officer: no? Okay.

Relief pilot: don’t extend!

First officer: So we’re still going down.

Relief pilot: We’re pulling.

Relief pilot: What do you think about it? What do you think we need to do?

And at this moment, a Priority Right is heard, meaning that the first officer gave commands of the aircraft to the relief pilot.

Captain:There, I don’t know, there, it’s going down.

First officer:There you are.

First officer:That’s good, we should be wings level, no it won’t (not).

Captain:The wings to flat horizon the standby horizon.

Relief pilot:the horizon…Speed?

First officer:Okay.

Relief pilot:You’re climbing.

Relief pilot: you’re going down down down

Captain: (I’m) going down.

First officer: Am I going down now?

Relief pilot: Go down

Captain: No you climb there

First officer: I’m climbing okay so we’re going down.

Captain: you’re climbing.

First officer: Okay, we’re in TOGA.

First officer: What are we here?

First officer: On alti what do we have here?

Captain:…it’s impossible.

First officer: In alti what do we have?

Relief pilot: What do you mean on altitude?

First officer:Yeah yeah yeah I’m going down, no?

Relief pilot: You’re going down yes.

Captain: Hey you.

Captain: You’re in…

Captain: Get the wings horizontal.

Relief pilot: get the wings horizontal

First officer: That’s what I’m trying to do now.

Captain: Get the wings horizontal.

First officer: I’m at the limit with the roll.

At this stage, the two pilots both simultaneously tried to regain control of the aircraft by putting moves on their side stick, creating a dual imput warning being heard aloud.

Captain:The rudder bar.

Captain:Wings horizontal …go on gently gently.

Relief pilot: We lost it all at… left!

Captain: hey, er…

Relief pilot: I’ve got nothing there.

Captain: eh… What do you have? No wait.

First officer: We’re there we’re there we’re passing level one hundred!

Relief pilot: Wait me I have I have the controls eh.

First officer: What is… how come we’re continuing to go right down now?

Relief pilot: Try to find what you can do with your controls up there.

Relief pilot: The primaries and so on…

Captain: (do anything…)

Captain: it (won’t do) anything.

First officer: We’re at level one hundred!

First officer: Nine thousand feet!

Captain: Careful with the rudder bar there.

Relief pilot: Climb climb climb climb!

When the relief pilot ordered the first officer to climb, at this point, the first officer said, “but I’ve been at maxi nose-up for a while” and this was in fact the clue that helped the captain understanding the real situation, that the plane was now in a deep stall, to which it was now too late to ever recover from. But either way, whilst the two pilots triggered another Dual input signal, it was too late to save the aircraft.

Captain: No no no don’t climb!

Relief pilot: So go down.

Relief pilot: So give me the controls the controls to me. Controls to me!

First officer: Go ahead you have the controls we are still in TOGA eh.

Captain: So wait... AP OFF.

First officer: Gentlemen!

Captain: Watch out you’re pitching up there.

Relief pilot: I’m pitching up.

Captain:You’re pitching up.

Relief pilot: I’m pitching up!

First officer: Well we need to we are at four thousand feet.

Captain: You’re pitching up.

Captain: Go on pull!

First officer: Let’s go pull up pull up pull up.

First officer: Oh my god, we’re going to crash!

First officer: This can’t be true.

Relief pilot: Come on!

First officer: But what’s happening?

Captain: Ten degrees pitch altitude.

And it was on those final words that at precisely 2 am 14 seconds, three hours and forty five minutes after takeoff, that Air France Flight 447 terminated its route in the Atlantic Ocean, hitting belly-first the ocean at a vertical speed of 108 knots and a horizontal speed of 107 knots. It immediately and violently destroyed the aircraft in the impact. All the lives of the two hundred and sixteen passengers and twelve crew members were immediately lost.

In the wake of the first of June 2009, the families of the passengers waiting at Roissy-Charles-de-Gaulle were informed that the aircraft was delayed, originally scheduled for landing at 11.15 in Paris, Local time. At 1 o’clock, the families were taken by Air France who informed them that the aircraft was missing and, no updates were given considering that the airplane did not issue any mayday call or distress signal. I recall this, as, it was nearly 4 AM Paris time when the airplane crashed in France and Air France started communicating on the fact that the aircraft issued automatic message of issues and failures during its fall. The first debris were found five days later… but unfortunately, as I said, the ADS-B system on board of the aircraft that was in trial back in the days failed to be activated, bringing the researches even more complicated. It was a needle in a haystack.

Now, it took a long time for the black boxes to be found, since, as I mentioned, there was no precise indications as to where the aircraft impacted the ocean and crashed but just elusive clues. And the fact that the French Government and the Brazilian Government both send militaries on site in order to help any effort of searches within the first thirty days, where at least a sound coming out of the black box could be located did not bring anything successful, and for an extensive period of time, many analysis of the various debris were carried out, but did not bring any final conclusion as to what might happened during that night. Some families were able to have the bodies of the passengers returned back for funerals, but not everyone in board had been found. In the cockpit, for instance, only the captain’s body was recovered. I remember this day as a very tragic day in France, from a personal experience, I was still a teenager, I was fourteen, and, I was in secondary school when it happened.

I also remember that France, on the evening of the 1st of June 2009 was playing a football match against another country and there has even been a minute of silence made for the families and the victims. On a personal level, I haven’t lost anyone in this flight, but this plane crash in fact pushed me to become a pilot myself. Air France will always be an airline that will hold a dear place in my heart, and I felt personal sorrow for the victims and their families. This is why I wanted to finish this series by dedicating this episode to Captain Marc Dubois, First officer David Robert, and First officer Pierre-Cedric Bonin, as well as the entire flight crew and all those passengers whose unfortunate sacrifice made today aviation safer, and have a thought for their families and say that, we will never forget them.

Either way, researches for the aircraft were unsuccessful a month after they started, and both the French and Brazilian Governments continued anyway, but less intensely, because the signal emitted by the black boxes became weaker and could not be captured. But in was only in April 2011 that parts of the aircraft were finally found deep under water. Straight afterwards, the French government ordered the black boxes, as well as bodies remaining and the debris to be brought to the surface and the BEA, the French body for the investigation of plane crashes was commissioned for the investigation, and, after having announced that the black boxes, despite their time under water, were finally readable, and they gave in July 2012 their final conclusions in the final report that you can also download from my website if you wish to find out more about the crash.

And very quickly, it was pointed that the crash was caused by a series of events that started by the obstruction of the pitot probes, creating a temporary inconsistency of the airspeed measurement as a result, and caused the autopilot disconnection and the switching between Normal Law to Alternate Law. It also pointed inappropriate control imputs that destabilised the flight path, a lack of any links by the crew between the loss of indicated airspeed and the appropriate procedure, the late identification by the relief pilot of the deviation of the flight path and the insufficient corrections applied by the pilot flying, the fact that they did not identify the stall warnings and their lack of reaction when the plane started to stall.

As a result, a plethora of recommendations were issued for all Air France, the EASA, (the European body supervising Air Safety), and the manufacturer Airbus. Amongst other things, what was recommended was to have made mandatory the installations of emergency locators on board of aircrafts, making mandatory the use of ADS-B which would have had facilitated the researches of the aircraft should these had been activated prior departure, as well as training for pilots for high altitude stall, reconfigurations of laws of the Fly-by-Wire system, and ensure those are also well understood. It also recommended a better re-display of the Flight Director, as this brought issues on this flight, and also required that the conditions in which, on approach to stall, the presence of visual indications combined with the aural warning should be made mandatory.

Air France had all the pitot probes of all its aircraft replaced by better pitot probes right after, as well as many air simulations training on unreliable airspeed in summer 2009 and high altitudes flight in alternate law, and other trainings. As for the aircraft, it was supposed to have its pitot probes updated on landing in Paris. Airbus reviewed the Unreliable speed indication procedure and the EASA brought many measures to improve aviation safety, mostly about the pitot probes obstructions, the autopilot reconnection, and many assessments have been carried out on simulator to find out how the major incident problem would be solved.

Now, I hear you. On this season, I spoke about both Palm 90 and Air France 447 and you guys probably think: what about me? Well, first of all, I am not a 330 pilot, so I cannot say if I would have done better or worse, but considering this entire situation, I believe that I’d have been misled the same way. Now, let’s think about this: many recommendations have been adopted following air crashes and this is why today, more than ever, in the outcome of those flights and those crashes, it is safer than ever to fly an airplane. It is unfortunately throughout mistakes that we learn. Now you know that, Airbus represent a big proportions of aircraft’s in the air and, you may take an Airbus next time you fly somewhere. Now you know that this Airbus cannot easily stall, and even if it does, it is protected by the fly-by-wife system, made with enhanced performances since Air France 447 crashed. Pilots are used now and trained to deal with high-altitude stall thanks to this. It unfortunately takes sacrifices to achieve something brilliant. It is sad. But that’s the way it is.

Now, I would like to thank Dustin, for having composed the beautiful main theme that you guys have heard at the beginning and in the end of each of the series episode. A massive thank you to Brian who gave his voice to the captain, to Matt who gave his to the first officer, or the pilot flying as well as Nick, who was the relief pilot. And a special thanks to my wife Saveria who gave her voice to the flight attendant. And I would like to thank you guys for having listened to my first season, which was a new trial for me, and I’m happy to have done this. Season 2 will come next year, but in the meantime, if you want to know more about me, please check my website on the description, subscribe to my channel and hit the bell to receive all notifications. If you wish to talk directly to me, consider following me on my Instagram, where I post all my updates as well as news for the season 2 coming next year (I should be announcing the dates of release by next winter, as I’ve got a busy calendar), or if you guys want to check my gallery of are interested in my books. Also, I recently opened a patronage program on my website, if you guys wish to have updates or listen to the future episodes in advance. Please go, feel free to check this out, and let me thank you for having listened to me for this summer!

As we conclude this episode, we’re left in the contemplative silence that follows the tragic story of Air France Flight 447, a modern enigma that confronted not just the aviation industry but also our collective sense of vulnerability. A plane with state-of-the-art technology, vanished in the vastness of the Atlantic, leaving behind a trail of questions and a sea of sorrow. The eventual discovery of the wreckage and the years of meticulous analysis have provided answers, yes, but they’ve also challenged our understanding of man’s mastery over machine. In the grand scheme of things, Flight 447 serves as a poignant reminder of the fine line between human judgment and technological dependency. As you absorb the gravity of this episode, I invite you to share your own reflections in the comment section below. Could the lessons of Flight 447 shape the future of aviation or even our relationship with technology itself? Until our next rendezvous with the enigmatic and the inexplicable, keep your minds alert and your hearts compassionate. See you in next year in next season, fellow seekers of the unknown.