Another Successful Ditching

  • E-Mail this Article
  • View Printable Article
  • Text size:

    • A
    • A
    • A

Last weeks ditching of a Cessna 310 13 miles off the coast of Hawaii's big island, had the local press tittering about the pilot's steely nerves and what an amazing thing it all was. Ho-hum. Wake me up when the movie is over. The fact is, ditchings happen all the time and the vast majority of them are successful, with everyone surviving. (Maybe more interesting is how the fuel calculations went wrong. As the 310 flies, 13 miles is about five minutes; less in the descent. We'll see what pithy comments the NTSB has when then investigation is complete.)

The accident reminded me to review the data I've been collecting for years on ditching accidents. I got interested in the topic when Ross Russo and I were organizing the Cayman Caravan 20 years ago. We were trying to brief pilots on how to equip for overwater flights, but to do that, you have to understand the nature of the risk. Without going into great detail, for warm water ditchings—by warm, I mean where survival in the water is measured in hours, not minutes—your chance of survival is well into the 80th percentile, based on about two decades worth of statistics I collected.

Most people who fly overwater and contemplate ditching worry about getting trapped in the airplane and going down with the ship. They shouldn't. It almost never happens. The uninjured egress rate is more than 90 percent in most waters, although a little less in deep ocean bluewater and markedly less in deepwater ditchings at night. The overall survival rate—that is, after egressing successfully—is 88 percent for general aviation aircraft. That's based on a study of about 220 accidents over nearly 30 years.

Where you ditch has a greater bearing on survival than what airplane you're flying. Rivers and lakes are the best bets, bays and estuaries next and open ocean the worst. But, frankly, the data is too sparse to draw any but the coarsest statistical conclusions.

Viewing the very good video of the Hawaii ditching, I can't tell if the engines are running or not. This raises a sticky aeronautical and survival judgment. At 13 miles out, the pilot would probably have been in sight of land. So does he push on hoping to make it on fumes, or cut his losses, decide to ditch, but do it while he still has power? I vote for the latter. Ditching is challenging enough without the additional variable of trying it for the first time deadstick. But if it was done deadstick, it just goes to show how forgiving a water landing is, even in 6-foot seas.

Remember the 1991 ditching of an Air National Guard HH-60 off Montauk, New York? It was depicted in The Perfect Storm. Having failed to connect with a C-130s re-fueling drogue because of extreme turbulence, the command pilot decided with 10 minutes of fuel left he wasn't going to die trying. The helo ditched with power and all but one crewman survived, following one of the most harrowing sea rescues in Coast Guard history.

Like all military helo crews, the Guardsmen were trained in water egress which is challenging for helicopters because they always turn over. Fixed wing airplanes sometimes do, but often don't. Either way, it doesn't seem to effect the egress rate. Also, the basic egress survival tactic is the same, and it's simple: Know where the door latch is by feel, and keep your seatbelt tightly fastened until the airplane comes to rest. Pop the door and only then release your seatbelt. Seems simple enough, but a lot of people don't know it.

So this latest ditching and dramatic video serve as reminder of how survivable water landings are. If a crash impends and you have the choice or a lake or river over trees and rocks, the water is a high percentage option.

Comments (47)

Thanks Paul for this revealing compilation! Would you have any insight on high vs. low wing and fixed gear vs. retractable?

Posted by: Sven Girsperger | October 17, 2011 3:06 AM    Report this comment

Helicopters do not always turn over. If the sea state is below 4, and they are equipped with Emergency Flotation Devices, most will stay upright at least long enough for an orderly egress. If one is flying overwater without fixed or emergency floats, then you don't want it to stay upright as it would be very dangerous to exit while the rotors are still turning.

Posted by: George Cunningham | October 17, 2011 3:47 AM    Report this comment

Thanks Paul. I vaguely remember to have read an article of yours appr. 10 y ago where you mentioned "auto rough" single mode. Very informative with fixed vs retractable, high vs. low win info. Do you have a link to this? My personal preference sequence for a glider (or forced) landing is: brown field, green field, water In Spain i saw fuselage and wing spars stuffed with foam or empty water bottles.. Makes a nice survival platform. Problem seems to survive after ditching

Posted by: Albrecht Lepple-Wienhues | October 17, 2011 3:56 AM    Report this comment

Retract vs. fixed and high and low-wing data is inconclusive. Not enough accurate data to draw anything really meaningful from the accident pattern. Or at least I couldn't.

I seem to recall a subset of 150, 172s and 182s having the same general egress rate as everything else, but these tended to be lake, river and inshore ditchings.

The 210s, Saratogas and other retracs are more likely to be flown on long overwater segments so the exposure to bluewater conditions is greater.

As for helos, in the accidents I reviewed, none had flotation. All inverted. In fact, the recent water crash in New York also inverted, as the pictures clearly show. I don't know what the state of the industry is now, but I don't see many with flotation, including in the Gulf oil trade. That may be different now. I don't know if it's the weight or cost or what, but operators don't seem to install the floats much.

If it doesn't have floats, whether you want it to or not, because of the high center of mass, a helicopter will roll over in the water. Not sure if that's true of small helos like the Robi's, but I suspect it is.

Posted by: Paul Bertorelli | October 17, 2011 4:27 AM    Report this comment

Here's a touch choice: Ditch in cold lake/river water (40 degrees) or in the tops of 90-foot pine trees?

If I lost my engine over tree-covered mountains I think I would try to land in a lake near shore or in a creek or river, which might be shallow.

Any comments or suggestions for this scenario?

Posted by: Jim Dunn | October 17, 2011 7:52 AM    Report this comment

Jim, Don't run out of gas in the first place. Brilliant decision making (at the end of a trail of bad decisions) is still dumb and still ends up in a wrecked ship. Any crash has too many variables so the best choice is don't crash.

Posted by: Mark Fraser | October 17, 2011 8:37 AM    Report this comment

Thanks, but my question is really directed at engine failure, not fuel exhaustion.

Posted by: Jim Dunn | October 17, 2011 8:41 AM    Report this comment

"Helicopters always turn over" is a bit dogmatic. Small helicopters that aren't designed for ditching do turn over, but grown-up ones like the old S-61 are designed not to, and may not, if the conditions are right. See Wikipedia "British Airways Helicopters" for a photo of one that didn't.

Posted by: John Stanning | October 17, 2011 9:31 AM    Report this comment

A side point: don't inflate your life-jacket until you're out of the plane. Seems obvious, but people do it. For example, in the crash of an Ethopian Airlines 767 off the Comoros in 1996, many passengers were thought to have survived the crash, despite the violence of the semi-controlled ditching, but drowned because they inflated their life jackets in the cabin and couldn't get out when water came in.

Posted by: John Stanning | October 17, 2011 9:42 AM    Report this comment

when I fly over water(cold), in my C-172, I wear a mustang flotation suit. not an immersion suit but that's what I have from my whale watching days. I am a bit concerned about the possibility of extraction because of the positive boyancy. comments?

Posted by: Neil Angus | October 17, 2011 9:53 AM    Report this comment

The S-61 is an amphib and was designed to *land* on the water. That's different than ditching. I'm sure the S-61's big sponsons help.

In none of the accidents we reviewed--about a half dozen for helos--did the aircraft remain upright. As I recall, most of them were oil patch aircraft in the Gulf, one was a medical helo. No survivors out of that one. In fact, I went through ditching training at STI with the company providing the replacement crew. I think we can reasonably say most helos invert when ditched.

Jim, it's fielder's choice on the trees versus the water. An FAA inspector I know once challenged me on the assertion that water is more highly survivable. So I went back and tried to develop data on tree and rough ground landings, but it was ambiguous at best. Couldn't sort the crashes from the intended landings and there weren't that many of them. The injury rate was higher.

But for the water landings, I know the egress rate is between 80 and 90 percent and the overall survival rate is in the 80s. Very few victims go down with the airplane, the fuselage remains relatively intact and there was no incidence of fire in any of them.

My intuition tells me you couldn't say the same about landing in trees, hence the reasonable conclusion that water is more survivable. But in fact, it could be very close.

Posted by: Paul Bertorelli | October 17, 2011 9:58 AM    Report this comment

my question on the 310 running out of gas is was it poor fuel management? and if so, why not run on one engine and conserve?

Posted by: Antonello Davi | October 17, 2011 10:56 AM    Report this comment

Of course, if you crash in the tree tops, S & R might see your plane more quickly. Plus, does the ELT work underwater more than a few minutes? I don't know if they're submersible, and I doubt they can transmit through much water depth. So if you're out in the boonies, the trees are likey more survivable, especially since you have a better chance of recovering gear from the plane, or maybe even starting a fire to attract attention. Maybe not true of 90 foot trees. FWIW, I stuck a hang glider between two big Douglas-firs about 30 years ago. As I recall the hardest part was getting out of the harness. But probably not very similar to a plane crash, certainly much slower.

Posted by: David Chuljian | October 17, 2011 11:34 AM    Report this comment

It was a long time ago, but I seem to recall being taught to pop open the door before touchdown. Makes sense to me to avoid the chance of the door being jammed by distortion of the fuselage from impact.

I would certainly have done so in my 310, as there are two latches.

Not sure what to do in a 340.


Posted by: Edd Weninger | October 17, 2011 12:13 PM    Report this comment

Not sure about a C-310 but in the turbine aircraft I flew in the Navy your range was better on two engines at a higher altitude. In other words, you got better mileage on 2 engines at FL350 than you did on one engine at FL250.

Also on a WestPac cruise in the Indian Ocean c1980 an SH-3 ran low on fuel and ditched under power. It is made for water landings and inflated it's collar, but it turned turtle. Off Oman seas were always heavy, up to 20 feet. Fortunately the crew all survived without injury. I don't know if it turned turtle before or after the crew got out.

Posted by: Jim Dunn | October 17, 2011 12:28 PM    Report this comment

"Makes sense to me to avoid the chance of the door being jammed by distortion of the fuselage from impact."

This is common ditching training advice and probably makes sense. However, there's no evidence that jammed doors actually happen much, if at all. That's the going down with the ship scenario. It could happen, just doesn't seem to. The more likely scenario is water pressure holds the door shut.

One upside of opening the door is also a downside. An open cabin door will flood the cabin more quickly, allowing the pressure to equalize to fully open the door. The downside is float time will be a little less and you'll have a little less time to get your head in the game after impact. That can take more than a few seconds and maybe half a minute.

Last, cold water gasp can be a survival issue in fast flooding. It's the tendency to inhale sharply when exposed suddenly to cold water. Our trainers warned us about this.

Worth noting is that in many single-engine ditchings,, the bow wave stoves in the windshield and floods the cabin quickly and near totally. That's where gasp could be an issue. Pax should be briefed.

Posted by: Paul Bertorelli | October 17, 2011 12:55 PM    Report this comment

Open vs. close door debate... I read years ago that the door is critical to the structural integrity of an airframe and helps keep the airframe intact upon impact. Just a thought...

Posted by: Phillip Potter | October 17, 2011 2:08 PM    Report this comment

Paul, Have you seen my email about our Fuel Timer app on iTunes? This pilot could have used it :) See the link below. Would love to have you review it:

Posted by: James Keane | October 17, 2011 2:32 PM    Report this comment

I think the most important criteria is to attain optimum impact conditions. As the captain of a DC3 which successfully ditched into Botany Bay (Sydney, Australia) after an engine-failure on take-off, I put into practice skills learnt in the RAAF flying P3 Orion aircraft. One of the criteria on these was ditching SPEED. Try to ditch at 10 kts above the zero thrust stall speed, less than 100fpm ROD, with land flap. Another important fact that I advise pilots is to know where the circuit breaker is for the landing gear warning horn as you don't want this blaring in the final few seconds.

Posted by: Rod Lovell | October 17, 2011 4:07 PM    Report this comment

Tell us more about your ditching, Rod.

Posted by: Paul Bertorelli | October 17, 2011 6:55 PM    Report this comment

Hi Rob. I wrote an article for the July 1996 edition of Australian Aviation. Copy of it here

Posted by: Rod Lovell | October 17, 2011 9:30 PM    Report this comment

Oops! How do I post a link?

Posted by: Rod Lovell | October 17, 2011 9:32 PM    Report this comment

Not sure who Rob is, but because of anti-spam code, you can't paste a link in whole. To get it to work, paste it in, then remove the www and replace with a asterisk or other character. Readers will figure it out.

Posted by: Paul Bertorelli | October 18, 2011 5:51 AM    Report this comment

There is no such thing as a "successful ditching"; it's still a case were a perfectly good aircraft was wasted because of really bad pilotage. Bad pilotage and a wasted aircraft is not "successful" just because you walked(swam) away from it.

Posted by: Mark Fraser | October 18, 2011 7:15 AM    Report this comment

Sorry Paul for the mixup. Whoa there Mark. I find your comments ill-informed and derogatory. You're saying Sully did not carry out a "successful ditching". Anyway, here is my story. Perhaps you would have handled my situation differently. Type in or paste this

Posted by: Rod Lovell | October 18, 2011 8:26 AM    Report this comment

The recent 'almost to Hawaii' ditching had several factors that were unique... 1. the Coast Guard had just practiced such a rescue the day before and was very sharp. 2. The pilot had tried that ferry the day before and turned back due to winds after about 5 or 6 hours. (Slow learner?) and 3. There was apparently a severe case of "get-there-itis". That company has lost several pilots in similar circumstances over the past few years. They seem to run out of fuel and altitude somewhere near their destination with alarming frequency. It can happen to any of us, and Paul has a good point about the frequency overall, but when a pattern develops with one operation, one has to wonder about it. Just food for thought.

Posted by: Bill Straw | October 18, 2011 10:04 AM    Report this comment

Rod, I was referring to the title of THIS story; calling it a "successful ditching". It's meant to be blunt when 100% good aircraft run out of gas over water.

US Airways Flight 1549 was a damaged aircraft and the pilot made the best off-airport landing decision possible to minimize loss of life on the ground as much as increase his chances of survival. Water is never a "first choice" for a forced landing...

Posted by: Mark Fraser | October 18, 2011 10:16 AM    Report this comment

FYI to get the link easily from Rod Lovell I deleted everything after ~rodlovel/ -got the home page, and clicked on DC-3 VH-EDC on the left.

Sounded like a very successful ditching to me, despite the BASI report, and quite similar to the Hudson landing with just seconds for decisionmaking required. Stressful just to read...congrats on saving all souls aboard, though it sounded like just another day, a walk in the park for the Aussie pax!

I don't fly over expanses of water enough for worry, usually have the shore in sight, but I do wonder how my bubble canopy would fair if submerged, latched or unlatched. Maybe large airbag-type balloons that pop out of the wing lockers upon sudden impact and I could stay dry, eating my sandwich while awaiting rescue...

Posted by: Dave Miller | October 18, 2011 1:19 PM    Report this comment

"There is no such thing as a "successful ditching";

On further rumination, you are absolutely correct. I shall change the head to: Pilot Survives Unsuccessful Ditching: All Escape Without Injury.

That work better for ya, Sparky?

Posted by: Paul Bertorelli | October 18, 2011 6:14 PM    Report this comment

Loosing a 100% perfectly good aircraft is not a "successful" flight. Please change the title to "pilot survives in spite of himself" and that will be accurate.

Posted by: Mark Fraser | October 18, 2011 6:48 PM    Report this comment

Nope, because any reader paying attention will know that this blog is about general survival rates in ditching incidents, not finding fault with why the airplane wound up in the water in the first place. Different story.

Fact is, you don't know squat beyond the initial report. Based on that, you're willing to assign blame. I'm not.

You need your own blog where you can pass judgment and comment on the frailty of human nature.

Posted by: Paul Bertorelli | October 18, 2011 8:12 PM    Report this comment

Paul, you are right about the thread being on ditching in general, but it might be worth a look to evaluate the causes at some point. Pressure to 'get there' where ever 'there' may be, is a major cause of ditching and other inidents and worth a comment at some point Ina new story line.

Posted by: Bill Straw | October 18, 2011 9:29 PM    Report this comment

"But it might be worth a look to evaluate the causes at some point".

Couldn't agree more. And that point is after the accident report is done and we have competently collected information. But I do not have information on this specific accident other than speculation . Know-nothing conclusions are just that. There's always time to fry the pilot later.

We have been writing about get-home-itis and fuel exhaustion for years. You know all about it. You know exactly how to avoid fuel exhaustion. So do I. Yet writing about it hasn't changed a thing. For the last 20 years at least, fuel exhaustion incidents happen twice a week.

But I venture to guess you don't know about the overall ditching stats I posted, in terms of egress and survival rates. These could be useful to know because everyone is exposed to little slices of over water flying and could face the prospect of ditching in a river, lake or bay.

The fuel judgement part is a separate topic, which I think you suggest. Fraser's contention was the head didn't fit the piece. It does. He just wanted a different piece.

For what it's worth, in general, ditchings are more the result of mechanical/engine failures than fuel exhaustion. I don't recall the distribution exactly, but it's significant. If bluewater ditchings are considered as a separate category, you see a higher percentage of fuel exhaustion and also failure in plumbing and pump failures in ferry tanks. And also more than a few unknown causes.

Posted by: Paul Bertorelli | October 19, 2011 4:33 AM    Report this comment

Paul, Survival rate is NOT an issue if you land your aircraft. Point being that this flight was(at best) a coin-toss, intentional high risk for both aircraft and pilot.

The pilot purposely gambled and he lost. That's fine, but The lesson for GA pilots is DON'T GAMBLE. Living after an intentional gamble is Johnny Knoxville stuff, not the stuff of 21st century flying.

Posted by: Mark Fraser | October 19, 2011 7:54 AM    Report this comment

"The pilot purposely gambled and he lost. That's fine, but The lesson for GA pilots is DON'T GAMBLE."

If this statement made sense, you wouldn't get out of bed in the morning. By your reasoning, every GA flight is a gamble. Everytime you take off, you're depending on a $9 connecting rod nut and a hundred other parts over which you have no control.

The truth is, hundreds of transocean ferry flights are conducted every year. Some come to grief for various reasons. The pilots who fly these flight take precautions to reduce the risk, but they are still high risk. But just because you consider it too risky, doesn't mean another pilot will and that he is wrong doing so. Don't lay your fears off on everyone else.

In this case, you don't even know what or how the pilot planned. What were forecast and actual winds? What was internal fuel capacity? Did he have a ferry tank? Did it feed correctly? What was his calculated reserve at PONR? Had he done the route before? Could there have been a fuel leak?

All you know is the outcome. You don't know the why, thus you are unqualified to determine why the accident happened. You're just laying your judgement and fears on him.

I prefer to have some post-accident data before doing the same.

Posted by: Paul Bertorelli | October 19, 2011 11:05 AM    Report this comment

Mark, the distinction between a successful "ditching" and a successful "flight" is rather straight forward is it not? Paul's article addresses the former. Nowhere within the article do I find a statement purporting the latter...

Posted by: William McClain | October 19, 2011 11:22 AM    Report this comment

"But I venture to guess you don't know about the overall ditching stats I posted, in terms of egress and survival rates. These could be useful to know because everyone is exposed to little slices of over water flying and could face the prospect of ditching in a river, lake or bay."

Certainly true for me, Paul, and the information surprised me. I can't imagine that there are many pilots who haven't experienced "automatic rough" when flying over large expanses of open water or inhospitable terrain and wondered about the survivability of a ditching or a treetop crash and how to best increase one's chances. Plan A for me is to do those things that I think are reasonable and prudent before and during a flight to reduce risk and have a successful flight. Plan B for me is to recognize that I'm capable of making mistakes and exercising poor judgment and to learn as much as I can about how to survive my stupidity.

Thanks to Rod for his story of a successful ditching (I'd love to hear from others who have done this) and to Jim Dunn for his question about water vs. trees.

Posted by: Bob Davison | October 19, 2011 11:42 AM    Report this comment

I DO ferry planes across the ocean regularly and I can attest to the prep for bad events and the careful planning most of us do. The issue many of us ferry pilots have with the event that triggered this discussion is a professional one and goes to management. The stats on ditching in general and Paul's thread here are good points of interest and valuable to all pilots. There is some specific info on the event near Hawaii and more will come out later, but I'm respectfully bowing out of this discussion and will wait for the one one "Management, pressure and greed in GA" at a later date.

Posted by: Bill Straw | October 19, 2011 12:04 PM    Report this comment

Paul, It is 100% certain that he planned to fly beyond the certified limits of aircraft. He splashed BEYOND the certified range of the aircraft. That means it is NOT the fault of the aircraft nor the design.

That makes me 100% qualified to say with 100% certainty that the pilot's judgement caused the loss. Q.E.D.

Posted by: Mark Fraser | October 19, 2011 12:14 PM    Report this comment

"Paul, It is 100% certain that he planned to fly beyond the certified limits of aircraft. He splashed BEYOND the certified range of the aircraft. That means it is NOT the fault of the aircraft nor the design"

You are being dense, Mark. It shows what you don't understand about long-distance ferry and the FARs, for that matter.

This airplane was a 1978 310R. The most internal fuel it could carry would be 207 gallons, although it could have been less, depending on options out of the factory. In still air, that's good for about 1400 miles, tops. To dry tanks.

The distance between California and Hawaii is about 2300 miles so it seems almost certain he had ferry tanks. But this we do not know. With ferry tanks, "certified limits" goes out the window, as Bill Straw will tell you.

I was partners in a Mooney and my partner did a trans- Atlantic in it. He flew an unusual route, from St. Johns to the Azores, a 1200-mile overwater leg. We put in a huge ferry tank--maybe 100 gallons. He had five or six hours of fuel reserve, easy. All nice and legal, with a ferry permit.

We--and especially you--don't know how this airplane was equipped or fueled. It probably had ferry fuel, but we don't know how much and we don't know if the pilot had a fuel transfer problem or what.

Posted by: Paul Bertorelli | October 19, 2011 1:07 PM    Report this comment

These ferry flights happen all the time and sometimes with all the careful planning in the world, they still go wrong. That's a fact. As Bill notes, there may be other external human factors we also don't know about yet. You should learn more about this before commenting on things you don't have a good grasp of.

This is why I am careful not to spout off about accident causes based on my own knee-jerk reaction. I think any pilot deserves the benefit of the doubt and a conclusion of the investigation.

Posted by: Paul Bertorelli | October 19, 2011 1:09 PM    Report this comment

"With ferry tanks, "certified limits" goes out the window"

And the PIC then assumes 100% responsibility. After 1400 miles you can't blame the aircraft. Accident report will be pilot error.

Posted by: Mark Fraser | October 19, 2011 2:23 PM    Report this comment


I watched the vid, I'm wondering why he didn't pull power way back, as in shutting down one engine? Then he could of flown in ground effect, giving him plenty of time to reconsider fuel consumption versus distance flown!

Too much work really, what do you go to Hawaii for, anyway?

He really wanted to take a cool dip, what else is Hawaii for?

Hopefully there's still female attendants (nurses), in the barf hotels (hospitals).

Wonder if he got a Lei?

I bet the crossover fuel diagram looked like a hieroglyphic!

Anyway, a sinking C-310, fittingly, is more symbolic of the ripped-off economy of the USA, and the government's agencies that are deleting our freedoms, and putting GA in the doldrums! At least we got a C-130 left, but for how long?

Posted by: Ron Brown | October 20, 2011 1:24 AM    Report this comment

And now for a comment of (I hope) a completely different nature...

Here's a big THANK YOU to Paul for having openly addressed ditching survivability statistics.

Why am I so grateful for this info? It's because I have had a life-long paralyzing fear of water, or more particularly, fear of drowning. I fly an aircraft that doesn't glide well and as a result my over-water flights are extremely conservatively planned. I don't cross water unless I have sufficient altitude to glide to the other side. Not the usual "reach half way and glide the other half". Nope, I want a 100% gliding distance reserve. Yes, despite years of working to overcome this irrational fear it still grips tightly to my psyche.

The information Paul has provided is going a long way toward easing my mind. Again, thank you, Paul.

Posted by: Eddy Riggs | October 20, 2011 9:30 AM    Report this comment

You're welcome. I have a similar fear...of mountains.

Posted by: Paul Bertorelli | October 20, 2011 10:56 AM    Report this comment

Eddy, If you fly a rental, then venture as far from landing sites as you wish. Modern advances (flight following) mean that chances are that you will live are very good.

My dense (and self-centered) remarks were as an aircraft owner.

Posted by: Mark Fraser | October 20, 2011 12:33 PM    Report this comment

Regarding helicopters rolling over when ditching. Normal ditching procedure is to roll the machine away from the door you hope to egress just as you touch water. This is done to stop the rotor blades. No one wants to sit in a sinking ship with the blades spinning. I believe the perfect storm pilot did this after having hovered in another area while passengers jumped into the 70+ foot swells. I believe only helicopters with flotation land upright.

Posted by: E Eaton | October 21, 2011 6:56 AM    Report this comment

Add your comments

Log In

You must be logged in to comment

Forgot password?


Enter your information below to begin your FREE registration