80
$\begingroup$

The "miracle on the Hudson" was extremely unusual - dual engine failures and successful off-field landings in commercial A320-size aircraft are very rare. The event has been dramatized and the crew (and captain "Sully" in particular) have been widely praised as heroes.

The crew certainly did an excellent job ensuring that all those aboard lived to tell the tale. But was it really exceptional? Aren't all commercial pilots trained to handle emergencies? I don't want to belittle the fantastic job the crew did, but I wonder whether they did something no other crew could do or whether this is something most crew would be able to handle if called upon.

Here is a more precise question:

If a random commercial airline crew were placed in the same situation as the crew of US Airways Flight 1549 on 15 January 2009, how likely is it that they would achieve an outcome with no fatalities?

I understand that to a certain extent this is a matter of opinion, but I hope people can provide some objective comments about what situations crews are trained for and whether there are other comparable examples with similar or differing outcomes.

$\endgroup$
  • 5
    $\begingroup$ How 'random' are we talking? A random U.S. mainline crew? A random U.S. crew? A random crew from anywhere in the world? The results will not necessarily be the same. As far as comparable examples, there are very few (dual engine failure immediately after takeoff is extremely rare in airliners.) Likely not enough to produce any kind of numbers with meaningful statistical significance. $\endgroup$ – reirab Sep 13 '16 at 5:46
  • 33
    $\begingroup$ In the NTSB report, they note that they had several flight crews fly this scenario in a simulator. Of those that ditched the aircraft, rather than trying to turn back to the airport, all did so successfully. $\endgroup$ – Michael Hampton Sep 13 '16 at 6:47
  • 2
    $\begingroup$ If im not mistaken, they experimented with a number of crews in a simulator, and gave them differing instructions ("attempt to return to a rrunway", "Attempt to ditch on the hudson" etc). $\endgroup$ – Jamiec Sep 13 '16 at 8:07
  • 4
    $\begingroup$ Possible duplicate of How, exactly, did US1549 land on the Hudson? $\endgroup$ – Jason C Sep 13 '16 at 13:02
  • 22
    $\begingroup$ @supercat: IIRC, Captain Sullenberger mentioned in an interview that most simulators cannot simulate ditchings precisely because of the lack of such data, and also probably more simply, because nobody programmed them to. Simulators typically have fairly precise and accurate aerodynamical models as well as models for fuselage, airframe, engine, and control surface behavior and stresses, onboard systems, etc. They also have models for ground behavior, rolling and braking resistance on different surfaces etc. Yet, they simply don't have hydrodynamical models of the fuselage acting as a boat hull. $\endgroup$ – Jörg W Mittag Sep 13 '16 at 23:28
79
$\begingroup$

This would be an extremely difficult question to answer without being opinionated, but I'll give a try. As part of the investigation, the NTSB, along with other agencies carried out a number of tests on simulator, where pilots were tasked with either landing or ditching the aircraft.

There were three different scenarios contemplated:

  1. Normal landings on runway 4 at LGA, starting from an altitude of 1,000 or 1,500 feet on approach;

  2. Attempted landings at LGA or TEB after the bird strike, starting both from zero groundspeed on takeoff from runway 4 at LGA and from a preprogrammed point shortly before the bird strike and loss of engine thrust, and;

  3. Ditching on the Hudson River starting from 1,500 feet above the river at an airspeed of 200 kts.

In the first case, all the landings were successful, which is not surprising. However, in the second case, landing was successful in (slightly) more than half the cases, but only where the response from the pilot was immediate (i.e. there was no delay between loss of thrust and decision to return to airport).

Regarding the second flight scenario, 20 runs were performed in the engineering simulator... Five of the 20 runs were discarded because of poor data or simulator malfunctions. ... In eight of the 15 runs (53 percent), the pilot successfully landed after making an immediate turn to an airport after the loss of engine thrust.

In the one case where a (35 sec) delay was included, the landing was not successful.

One run was made to return to an airport (runway 13 at LGA) after a 35-second delay, and the landing was not successful.

It has to be noted that in all these cases, the pilots were already briefed about the manoeuvre they were about to perform.

The third series of tests were conducted to see whether the pilots were able to achieve the target flight path angle of -0.5$^{\circ}$. Only one Airbus test pilot was able to come near that value. From the report:

Regarding the third flight scenario, a total of 14 runs were performed in the engineering simulator in which pilots attempted to touch down on the water within a target flightpath angle of -0.5°, consistent with the structural ditching certification criteria. Two of the 14 runs were discarded because of poor data.

In 11 of the 12 runs, the touchdown flightpath angle ranged between -1.5° and -3.6° (the touchdown flightpath angle achieved on the accident flight was -3.4°). In 1 of these 12 runs, a -0.2° touchdown flightpath angle was achieved by an Airbus test pilot ...

From these simulations, it appears that the flight crew took the correct decision to ditch in the river. Also, we have to take into account that taking decisions in real world under extreme stress is quite different from taking them in a simulator. The report notes that,

The one simulator flight that took into account real-world considerations (a return to LGA runway 13 was attempted after a 35-second delay) was not successful. Therefore, the NTSB concludes that the captain’s decision to ditch on the Hudson River rather than attempting to land at an airport provided the highest probability that the accident would be survivable.

Another thing to note is that there was no simulator training to the pilots, either as part of the airline or manufacturer program. In fact, the dual engine failure scenarios were presented only at high altitudes.

Ditching scenarios were not included in either the US Airways or Airbus simulator training curriculum.

A similar incident happened in case of the Scandinavian Airlines Flight 751, where the pilots successfully landed the aircraft without any fatalities.

One also has to consider the other issues in ditching in water (like the associated visual illusions). From these, we can form the opinion that a sizeable minority of airline crews would have ditched the aircraft successfully on the river after bird strike so as to enable rescue. As for zero fatalities, it would be near impossible to say that with any certainty, as there are too many factors in play here.

$\endgroup$
  • 15
    $\begingroup$ Is "not successful" a jargon term, does it imply how unsuccessful the attempt was? Or does it cover anything from "didn't get a round of applause from the passengers" through to "less than 50% of the aircraft came to a halt on the runway", or "didn't reach the runway"? $\endgroup$ – Steve Jessop Sep 13 '16 at 9:41
  • 4
    $\begingroup$ @SteveJessop -- "not successful" = "didn't make the runway" in this case -- you generally don't worry about being hot and high on final if you're in a glider, and if you are -- a forward slip fixes that in a jiffy. $\endgroup$ – UnrecognizedFallingObject Sep 13 '16 at 11:37
  • 27
    $\begingroup$ This is an excellent answer, but I'd like to add one point. Before the Hudson case, ditching was considered the least favorable option for a forced landing. So, as a pilot, I admire and respect the cold and fast analysis of the situation and the decision to land the plane in the river. I've flown a similar flight in the sim, and the aircraft handling part is easy, the decision making is not. $\endgroup$ – Sami Sep 13 '16 at 20:41
  • 4
    $\begingroup$ @Sami It's also notable that most pilots of light GA aircraft that I've spoken to still feel that ditching is the least favorable option in situations like this despite some good evidence to the contrary (a documented egress rate of over 90% if the aircraft went into the water under control). The airline folks have taken a valuable lesson from this incident, hopefully it sinks in for the rest of us too. $\endgroup$ – voretaq7 Sep 15 '16 at 16:09
  • 3
    $\begingroup$ @voretaq7 I that a successful ditch attempt aims to avoid "sinking in" $\endgroup$ – Bassinator Sep 17 '16 at 2:04
35
$\begingroup$

It's important to note that there were many elements of both skill and luck involved in the US Airways 1549 "miracle" landing - the NTSB report and NTSB deliberations make for an interesting day of study.

We can divide your question into two different segments:


1 - Could another crew have executed this ditching after losing both engines?

The answer here is "Probably" - transport-category aircraft are designed for ditchings, and crews receive some training on ditching procedures (though as the NTSB noted it's not a normal component of simulator training). In many ways a ditching is just a special subset of engine failure training in which your landing will be on water rather than dry land, so pilots are in part training for this eventuality from the beginning of their flight training.

The decision to ditch on the river is a judgment call - Sully chose it rather than trying to make a runway at nearby airports that he wasn't sure he could reach without crashing into a populated area.
If another captain were presented with the same situation they may elect to make the same decision. Speaking as someone who occasionally flies over that same river in a much smaller aircraft and at a much lower altitude it's a reasonable choice to make.


2 - Would another ditching in a different aircraft have similar results?

This is where luck starts to enter into the equation, and the incident starts to earn its "Miracle" label.

US Airways 1549 was operating on an aircraft equipped for extended over-water flight, even though that was not required for this particular flight. This means there were life vests onboard, as well as slide-rafts designed to be used in the event of a ditching in open water.
It's debatable how much this survival equipment contributed to the successful outcome: The NTSB found that many passengers didn't take their flotation devices, and because of airframe damage during the ditching some of the slide rafts were not accessible or usable during the evacuation, but the equipment was considered a positive element in survival factors for this incident.

The weather was also fairly favorable (while temperatures were below freezing it was a fair-weather day: No rain or high winds), and the location of the ditching placed the aircraft close to ferries and other vessels able to quickly effect a rescue of the passengers and crew, so they did not spend much time in the weather or water. (This was noted in the NTSB deliberations: Technically the aircraft could have travelled further down-river, but doing so would have put it out by the Verrazano Narrows bridge, further from potential rescue vessels).

$\endgroup$
  • 19
    $\begingroup$ A related part of the "miracle" that you hint at: they landed close to ferries and rescue boats, preventing many cases of severe hypothermia, yet didn't hit anyone on the busy river. $\endgroup$ – Zach Lipton Sep 13 '16 at 15:35
17
$\begingroup$

"Ask the Pilot" author Patrick Smith addressed this question. I'll include excerpts of his post here:

As the public has come to understand it, Sully saved the lives of everybody on board through nerves of steel and superhuman flying skills. The truth isn’t quite so romantic...

[The] nuts and bolts of gliding into water aren’t especially difficult. The common sense of water landings is one of the reasons pilots don’t even train for them in simulators... and nowhere in the public discussion has the role of luck been adequately acknowledged. Specifically, the time and place where things went wrong. As it happened, it was daylight and the weather was reasonably good; there off Sullenberger’s left side was a 12-mile runway of smoothly flowing river, within swimming distance of the country’s largest city and its flotilla of rescue craft. Had the bird-strike occurred over a different part of the city, at a lower altitude (beyond gliding distance to the Hudson), or under more inclement weather conditions, the result was going to be an all-out catastrophe, and no amount of talent or skill was going to matter...

Nothing they did was easy, and a successful outcome was by no means guaranteed. But they did what they had to do, what they were trained to do, and what, presumably, any other crew would have done in that same situation. And let’s not forget the flight attendants, whose actions were no less commendable. Thus the passengers owe their survival not to miracles or heroics, but to less glamorous forces. They are, in descending order (pardon the pun): luck, professionalism, skill, and technology.

There’s little harm in celebrating the unlikely survival of 155 people, but terms like “hero” and “miracle” shouldn’t be thrown around lightly. A miracle describes an outcome that cannot be rationally explained. Everything that happened on the river that day can be rationally explained.

$\endgroup$
14
$\begingroup$

It is impossible to answer your precise question with any authority, it would be pure speculation. What we can answer is what would happen if you took some non-randomly chosen crews and put them in a simulation of the events which occurred to Flight 1549.

The problem with this, is that every single one of those pilots knows what to expect, they know that they are about to get a simulated bird strike which will cripple both engines. They can plan in advance what course of action to take, and execute it immediately. Sully did not have this luxury, and there was a lag between the event and taking action.

Documents indicate nearly two dozen emergency simulations were flown by experienced aviators, including an Airbus test pilot, at the manufacturer's headquarters in Toulouse, France. Four out of four attempts to return to the closest La Guardia runway were successful, according to the safety board's summary. There were nine additional simulated attempts to land at La Guardia, either at different runways or under a scenario in which the plane was more severely disabled. Of those, three were successful.

But only those pilots who immediately decided to turn back toward La Guardia after the simulated engine problems made it. The board's document concludes that such a scenario fails to "reflect or account for real-world considerations such as the time delay" in recognizing the bird strike and to "decide on a course of action."

Source WSJ - Miracle on the hudson

If you are interested in more regarding the findings of the NTSB the report is public.

$\endgroup$
  • 1
    $\begingroup$ As the data cited in aeroalias' answer shows, answers to this question need not be pure speculation; some data is available to give an indication of what we might expect $\endgroup$ – J Walters Sep 13 '16 at 12:20
  • 2
    $\begingroup$ @JonathanWalters the "precise question" I referred to started If a random commercial airline crew were placed in the same situation.... there was nothing random about the crews selected for simulations. Hence it would be speculation how a random crew would fare. $\endgroup$ – Jamiec Sep 13 '16 at 16:55
12
$\begingroup$

Wikipedia used to maintain a list of commercial airlines ditchings on water, but I cannot find it. I did find related accidents on the US Airways 1549 article though:

1963 Aeroflot Tupolev ditching on Neva River: Probably the most similar thing because it was total engine failure (starvation) and it ditched on a cold river, narrowly missing a bridge. All 52 people aboard survived. This was in 1963 in the Soviet Union.

Japan Airlines Flight 2: In 1968, this flight accidentally landed on water short of the runway. All 107 people aboard survived.

Pan Am Flight 6: In 1956, this flight experienced engine trouble and eventually ditched in the Pacific Ocean. All 31 people aboard survived. (Thanks to @CodyP for pointing this one out.)

I swear there was 1 more, in 1940 or 1943 or around there. Some commercial prop plane ditched in the Pacific Ocean and everyone survived. Cannot find the article, though.

(EDIT: I believe I found it. Honolulu Clipper. It was actually a military flight in 1945 transporting some troops. 2 out of 4 engines failed and it ditched on the Pacific, but everyone (37 people) survived and was rescued later. What makes it especially impressive is that the ditching happened around midnight local time.)

EDIT: finally did find the list here. It lists successful and unsuccessful ditchings on water by commercial aircraft.


What's needed is to compare the successful list with a list of ditchings on water where some or everyone died. Off the top of my head, I can only think of 1 right now: That flight that was hijacked over the Indian Ocean, starved of fuel, and ditched. I think about 100 died and 50 survived, but interestingly, I remember the Air Crash Investigation episode claiming that most people survived the crash only to inflate their life vests early and thus trap themselves in the fuselage.

There are plenty more crashes over the ocean, such as that Swiss Air flight that had a fire aboard, but they shouldn't count unless the pilots maintained controlled flight and attempted to ditch. Total loss of control due to fire or explosions should not count.

So to sum up, we can compare lists, though they are not very big. Weather and sea conditions at the time of ditch should be taken into account too.

And I also swear I remember this, maybe someone can confirm. Sully got awards saying things like "unique air achievement" but tried to explain it was not unique. I could have sworn he mentioned the successful Aeroflot Tupolev ditching on the river Neva on the Tonight Show with Jay Leno, but it was glossed over.

Luck undoubtedly plays a role when encountering water. Hydrodynamic forces are undoubtedly chaotic where a tiny variance can produce cascading feedback, or lack thereof. I think your question is in the right spirits because you mentioned both luck and pilot skill playing a role. Sully had a lot of glider experience, for example.

$\endgroup$
  • 3
    $\begingroup$ Don't forget about Pan Am Flight 6, that had several engine failures and was forced to land on the ocean. The only fatalities were fourty-four cases of live canaries. $\endgroup$ – Cody P Sep 14 '16 at 15:46
  • $\begingroup$ I remember an Asian flight ditching in shallow river and the tail breaking off killing at least one flight attendant $\endgroup$ – Antzi Sep 14 '16 at 16:58
  • $\begingroup$ The second-to-last paragraph should be And I also swear I remember this, **maybe someone can confirm.** Sully got awards saying things like "unique air achievement" but tried to explain it was not unique. I could have sworn he mentioned the successful Aeroflot Tupolev ditching on the river Neva on the Tonight Show with Jay Leno, but it was glossed over. but I don't have the reputation to make such small edits now that the site has graduated. (Fixes rememebr and Tuploev.) $\endgroup$ – a CVn Sep 15 '16 at 8:52
  • $\begingroup$ @MichaelKjörling Thanks, made those edits. $\endgroup$ – DrZ214 Sep 15 '16 at 9:27
  • 1
    $\begingroup$ From your list, the cases of Garuda Indonesia 421 (one death out of 60 aboard) and Laoag 585 (19 deaths out of 29) are the recent cases that are probably the closest to AA 1549 — both involved losing engines at a relatively low altitude (on approach and shortly after takeoff, respectively.) $\endgroup$ – Michael Seifert Sep 16 '16 at 18:24
7
$\begingroup$

Ditching in water after engine failure is a standard emergency drill for pilots. It's not required, but I'd expect major airlines to train their senior pilots for this. Over mainland US, it's unlikely to ever be relevant - it's having to ditch in the centre of a major city which makes this particularly exceptional. However it's the only possible answer for an airliner flying over the sea.

A quick Google suggests that most recent water landings have involved substantial loss of life. However it's probably fair to say that none of the airlines involved have a high reputation. For higher-quality airlines, not only will the pilots be better trained, but the planes will be kept in better condition so the situation is less likely to arise.

For a comparable example though, the Gimli Glider was very much a one-off. Had the first officer not served at Gimli, he almost certainly wouldn't have considered it at as a landing location. And had the captain not been an experienced glider pilot, the aircraft almost certainly would not have reached Gimli or any other suitable landing location, and the landing would not have been survivable. Whereas the Hudson incident was generally reproducable on simulators by skilled pilots, attempts to run the Gimli Glider incident on simulators only resulted in crashes by other pilots.

$\endgroup$
  • 7
    $\begingroup$ Actually, ditching is not part of most simulator scenarios. $\endgroup$ – GdD Sep 13 '16 at 10:43
  • 7
    $\begingroup$ According to the accident report (page 60), pilots were given a Powerpoint presentation about ditching at some point. There was no simulator training for it. $\endgroup$ – chirlu Sep 13 '16 at 11:14
  • $\begingroup$ Another example where variables matter is Air Transat flight 236, which glided 120km after a fuel leak caused dual engine shutdown over the North Atlantic during a transatlantic flight - it successfully landed in the Azores. The variable being that the flight was abnormally routed much farther south on its track, if it had been on a normal track it would never have made the Azores or any other airfield. $\endgroup$ – Moo Sep 13 '16 at 12:03
2
$\begingroup$

There are no "miracles" in aviation, or anywhere else on this planet, for that matter. What Sully did was something all pilots are trained to do, ditch an aircraft in water if you have nowhere else to land.

The reason why Sully is celebrated is because he executed the ditching procedure flawlessly and in so doing, ensured the safety of his entire crew and all pax. Part of it was luck, the vast majority was skill. Remember as well that there are aircraft that are designed to operate exclusively off water -- they are called "seaplanes."

My beef with the movie is that it is extremely disrespectful to the NTSB investigators and from that point of view, historically inaccurate in a gratuitous manner. Tom Hanks should be ashamed of that.

$\endgroup$
  • 20
    $\begingroup$ he executed the ditching procedure flawlessly This statement is factually incorrect: Sully's landing was far from perfect (a number of issues were called out in the NTSB report - key among them are excessive sink rate & angle when contacting the water and failure to fully complete the ditching checklist). Ultimately this didn't have a material impact on the outcome, which was without question almost the best we could ask for in such a situation, but "flawless" is a vast overstatement. $\endgroup$ – voretaq7 Sep 13 '16 at 12:52
  • 3
    $\begingroup$ The question defines miracle as "But was it really exceptional?" and it should be noted that the word "miracle" comes from popular usage of the phrase used to denote this particular incident. As such, the first and third paragraphs are, at best, comments, and the middle paragraph isn't differentiable from other answers already given. Please consider adding more detail to the meat of your answer, and removing, or at least de emphasizing, portions irrelevant to the question. $\endgroup$ – Adam Davis Sep 13 '16 at 13:56
  • 10
    $\begingroup$ @JuanJimenez The criteria for "flawless" is not an outcome-based one, it is a consideration of all the performance factors leading up to that outcome, as well as the outcome itself (and whether the criteria for perfection are achievable in specific circumstances). As an example a pilot can have a beautiful "greaser" landing after a lousy approach - that outcome doesn't make their performance "flawless", nor does it guarantee a similar outcome for the next pilot who makes similar errors. $\endgroup$ – voretaq7 Sep 13 '16 at 14:34
  • 1
    $\begingroup$ @JuanJimenez I agree with voretaq7. A Formula 1 driver winning a race does not mean his driving was flawless, nor does breaking a world record mean the athletic performance was flawless. Flawless means "without any mistakes or shortcomings" in the Oxford dictionary, not "successful" or "reached all the desired objectives." Perhaps you could replace "flawlessly" with "exemplary" or "successfully" $\endgroup$ – Cody P Sep 13 '16 at 19:03
  • 6
    $\begingroup$ @JuanJimenez Perhaps the word you are looking for is "successful". At the end of the day, it was the distinction between a successful and unsuccesful ditching that saved lives, not whether the ditching satisfied someone's criteria of "flawless". $\endgroup$ – David K Sep 15 '16 at 14:17

Not the answer you're looking for? Browse other questions tagged or ask your own question.