How ditching should be done

Ditching is always a high risk option, but this US Airways A320 crew made the right decision under the prevailing circumstances.

They then showed the world how ditching should really be done.

They reminded us that this exercise can be completely survivable, given piloting skill and the right weather and visibility conditions.

Pilots are informed how ditching should be done, but practically nobody actually trains for it, even in a simulator.

The reason why training is not offered is rather obscure, but it’s probably a combination of two facts: it hardly ever happens nowadays – especially to jets – and the probability of passenger and crew survival is – subliminally – considered low, so it’s not reckoned to be worth it.


Propeller-driven aeroplanes normally have a better chance of ditching successfully simply because they can touch down more slowly on the water.

But there has been one almost completely successful jet airliner ditching in the last ten years:

That was a Garuda Boeing 737-300 that suffered engine flame-out in a rainstorm in November 2002 and ditched on a river near Jogjakarta, Indonesia. Only one person died in that event.

There have also been two successful business-jet ditchings in the last decade, both in North America.

In April 2003 a Dassault Falcon 20 ran out of fuel and ditched in the Mississippi near St Louis. Everyone survived that.

And in July the same year a Canada-based Cessna Citationjet ditched with very limited pitch control because of a trim failure, but set down in Oak Harbour, Victoria, near the shore.

The two pilots – the only people on board the Citationjet – swam ashore.

In 2005 a turboprop aircraft – a Tuninter ATR72 – ran out of fuel and ditched in the Mediterranean north of Sicily. There were 16 fatalities out of the 39 people on board. A high-wing aircraft like the ATR series gives less time for evacuation, because the wings don’t provide flotation until the fuselage has disappeared beneath the surface.

And ditching in the sea, as opposed to on a river, is usually more risky because large waves and swell are more likely.


But what of this US Airways A320 accident?

What did the pilots face?

From the time the birds hit the aircraft’s engines the crew could see that they had insufficient power to remain airborne.

And they quickly became pretty sure they couldn’t glide back to LaGuardia – the captain’s first instinct - or divert to Teterboro airport, because they knew that if they attempted that and got it wrong, they would crash-land in a built-up area, risking all those on the aircraft and also lives on the ground.

So almost instantly the crew turned left toward the Hudson river, looking out for any opportunity to put down on a large, flat unobstructed piece of land.

No such land space existed within the distance the pilots knew they could stretch their glide, so they committed to ditching as the best option, and briefed the cabin crew to prepare for it.

When the birdstrike happened, the aircraft was at a height of about 3,000ft, approximately 900m.

From that height, they settled into a rate of descent of about 1,000ft per minute, leaving flaps and gear up initially, and deploying them shortly before touchdown on the water. The crew had about three minutes to prepare for ditching having made that decision, and touchdown took place about 3.5min from the birdstrike.

The considerations for ditching are basically these:

  • Aim to touch down with the aircraft speed sufficiently high to retain full control, because stalling and letting a wing drop would be disastrous.
  • If one wingtip or engine touches the water before the other does, the aircraft swings violently and breaks up.
  • On final approach to touchdown, selecting some flap (about 20deg) allows a slower impact and ensures the aircraft attitude is not too nose-high.

If the nose is too high, the tail strikes the water first and can break off, letting water into the hull.

Here is a pictorial guide to some of the considerations when forced to ditch:



This crew obviously got the aircraft speed and attitude absolutely right.

And they are rightly receiving praise for having very cool heads.



Subscribe to our e-mail newsletter to receive updates.

7 Responses to How ditching should be done

  1. Mark Carlisle 16 January, 2009 at 5:54 pm #

    “…the probability of passenger and crew survival is – subliminally – considered low, so it’s not reckoned to be worth it. ”

    I would agree that the above mindset seems to be the prevailing one but, nonetheless, survivable ditchings seem to far outnumber unsuccessful ones. In fact, has there ever been a disastrous ditching of a jet transport? (Excepting the attackers-in-the-cockpit Ethiopian affair)

  2. David Nicholas 19 January, 2009 at 10:41 am #

    The likelihood of a successful forced landing on water is greatly enhanced by having time to prepare and good water surface conditions. In this instance the 3 and half mins or so from birdstrike to ditching was more than adequate for the pilots to react to the loss of power, the cabin crew to prepare the passengers, the water was calm, visibility and light conditions good, and probably the best imaginable commander in the cockpit. His decisive action rather than an almost instinctive attempt to reach Teterboro was perhaps a consequence of his background as a military (F4) pilot, accident investigator and glider pilot. His experience would have helped him dismiss stretching the glide in a marginal attempt reach a runway in the midst of metropolitan New York. I read that the co-pilot stated to the captain after rescue that this had never been done before (or words to that effect). He may have truly believed this, and if this is a common view then perhaps its as well that both pilots were not under this misapprehension. This, I would submit, was an event where experience triumphed over training, and is deserving of inclusion in training courses against the inevitable day when this happens again.

  3. Blake Reid 19 January, 2009 at 10:45 am #

    Your info on the CitationJet ditching is incorrect. It was in Washington State not Victoria. Being registered as N996JR leads one to think that it was probably not based in Victoria either. A search on the internet found this info.

    “The airplane is registered to the Tango Corporation of Minden, Nevada”

    “About 15 minutes after takeoff from Victoria, while climbing through FL160 (en route to FL330), the airplane abruptly nosed over to approximately a forty-degree nose-down angle. The pilot disconnected the autopilot, throttled back to idle, and attempted to re-trim the airplane. He reported that the cockpit trim indicator was in the full forward (nose down) position and that neither the manual nor the electric trim actuators would respond to his inputs. After numerous configuration changes and unsuccessful attempts to regain full pitch control, the pilot elected to ditch the airplane in the water at Penns Cove, 200 yards from shore. Both occupants swam ashore.

    PROBABLE CAUSE: “The loss of airplane pitch control (trim runway and mistrim condition) resulting from a failure in the airplane’s electric pitch trim system. Factors that contributed to the accident were the manufacturer’s inadequate design of the pitch trim circuitry that allowed for a single-point failure mode, and the absence of an adequate failure warning system to clearly alert the pilot to the pitch trim runaway condition in sufficient time to respond in accordance with the manufacturer’s checklist instructions.”"

  4. Jim Johnson 3 February, 2009 at 3:16 am #

    Mr. Learmount: Would you please write about “passenger preparation” for diching? I always read the safety info in the seatback poscket, and it seems that there are 3 or 4 different “brace positions” drawn. Some look awful to me, iseeming to invite broken necks as heads snap forward upon impact. What is the best brace position according to safety experts?

  5. David Learmount 3 February, 2009 at 9:04 am #

    The best brace position varies according to how close the seats are together and whether your seat faces the direction of travel (some, but not many, face aft).

    If you are in a charter or low cost carrier and the seats are very close, the brace position shown will normally involve placing your forearms on the back of the seat in front of you and nesting your head between them against the seatback. That is because most adults, in that confined space, cannot put their face right down on their lap with their hands behind their head, which is the ideal position.

    Whiplash harming the neck should not occur from the brace position because that affects people (in cars) who are sitting upright at impact with their chests restrained, so only their head can move. Passengers in aeroplanes are restrained at their waist level only, so if the passenger was sitting upright at the moment of a dramatic longitudinal deceleration the danger of whiplash would affect the whole of the upper body and head.

    Feet, meanwhile, should be placed such that your shins are basically vertical.

    Anyone seated facing aft is advised to sit with their back and head against the seatback, and arms and hands protecting their face and torso.

    None of this guarantees freedom from impact injury, it just reduces the risk of it.

  6. arthur leete 4 October, 2009 at 9:57 pm #

    with a modern jet airliner, some say the shear pins will cause the engines to break off when hitting the water so the plane can be ditched level

    but the pilot ditched into the hudson landed tail down, apparently expecting that the engines would dig in and cause the plane to break up if he landed level

  7. John Laming 12 November, 2010 at 11:34 am #

    Despite direct evidence that these events have occurred in the past, it beats me why operators always find excuses for not allowing their pilots to practice certain past incidents in the simulator. These events include serious unusual attitude recovery on instruments in IMC – and not just the easy ones with 45 degrees angle of bank and 20 degrees nose high or low.

    Or – the final ditching approach in IMC from 1000 ft above sea level. That takes very careful instrument flying since mis-handling resulting in high rate of descent or less than ideal touchdown attitude, can be disastrous.

    Or – loss of power on all engines at high altitude necessitating profile planning and execution of a forced landing pattern. Or – practice at landing with loss of all flight control systems and using only the throttles for pitch and roll.

    The excuse often given is that these are so rare events it is not worth wasting simulator time, despite these events requiring pure flying skill. With accent by airlines and manufacturers on full use of automation, automation dependancy becomes inevitable. This is not a new finding of course, and has has been documented in numerous research papers for many years. Basic stick and rudder skills fade with time.

    The above events requiring manipulative skills however are life saving in their importance yet todays airline pilots are ill-equipped to cope. LOFT exercises for example, are time consuming especially when simulator time is at a priority.

    Yet, there must be surely more value in accenting pure flying skills rather than an hour or more on autopilot playing LOFT CRM and talking to phantom flight attendants, phantom ground engineers, and ATC?