DU Aviation Experts: I have a couple of questions.

[gateley]

1) Would the air traffic controllers have been able to tell if the plane was coming in too low?
2) How about the instruments -- any warning system for that type of situation?
3) Should the pilot and co-pilot been able to determine the situation visually? In time to rectify it?

I'm puzzled because of the no emergency call from the cockpit, and from what I've heard some passengers say, it was all just last minute and seemed like it was going to be a normal landing.

Thanks for your expertise!

[Agschmid]

1. I feel it's to early to make any assumptions.

If you'd like the flight aware tracking data for this flight is very interesting, not sure if it got pulled, I can't find the link?

There are alarms that warn of ground proximity... But this was during a landing so ground proximity is to be expected.

My "theory" is an issue with an altimeter. But again I feel it's really early for much more speculation beyond that.

Anyone else?

[Arctic Dave]

5. Hard to say.

Altimeter could have done it but I would think the GPS would give accurate readings.

Heck, could have hit a goose and lost power.

[Agschmid]

8. Damn birds/FOD.

[krispos42]

49. I understand that GPS altitude data has a margin of error...

...a bit larger than the horizontal margin of error.

A hundred feet difference in altitude can result in thousand foot difference in touchdown location. Maybe more; the glideslope is only about 3° from the horizontal.


But airports have a system of lenses to show them they are on the correct glidepath. It's a series of white lights over a series of red lights, with directive lenses over them.

If all the pilot sees are the white lights, he's too high. If all he sees are the red lights, he's too low. If he sees red over white, he's on the glideslope and coming in properly.

https://en.wikipedia.org/wiki/Visual_Approach_Slope_Indicator

[burrowowl]

66. One reason some instruments should have

analog back-ups, i.e., the horizontal indicator (Mother Nature does not lie) etc.

[gateley]

7. What about air traffic control?

And if the altimeter is whack, could the pilots have been able to determine it visually? I'm wondering if coming in over water (visually) may have made that more difficult.

Years and years and years ago when I flew, I seem to recall landing based mostly on how it LOOKED to me. Of course, there's a big difference between a Cessna 172 and a 777.

Another question comes to mind, do you think these guys actually land the plane themselves or is it all automated?

[Agschmid]

9. Yes quite a difference!

My experience is also only in a 172/152 so ill let the pros weigh in on the automated question.

[Arctic Dave]

16. I'm not sure they land visually anymore.

I don't think temperature inversion happen in the summer time, do they?

[HERVEPA]

30. It is by no means all automated

[gateley]

34. They're asking that question on MSNBC right now.

Response is that Boeing and Airbus have different philosophies.

Boeing seems to leave more of the flying up to the pilot, Airbus is more automated.

This guy (pilot -- Navy Air? -- and aviation attorney) has said that a large part is how the pilot interacts with the systems.

[HERVEPA]

72. askthepilot.com talks a lot about this issue and the misconceptions

Read his blog for lots of good insight. He's a current pilot.

[gateley]

75. Sounds interesting. I'll check it out -- thanks!

[napi21]

76. Ground proximity would have been turned off because they were landing.

It's always turned off because it would give constant false warnings. What should have been the alarm was the stick would have started vibrating badly to indicate they were flying too slow. I have to assume that was ignored.

The one big question I have is WHY the co-pilot didn't warn the pilot they were too low and slow? That's why there IS a co-pilot.

[Arctic Dave]

2. The planes instruments and GPS tracking would give the pilot all the information.

Altitude, speed, pitch, etc., so they would have known something was amiss.

However, stuff happens.

[Agschmid]

6. Stuff does happen...

This appears to be one of those cases. How many cycles happen everyday, and how many happen safely and accurately. Flying on a plane is a very safe activity, especially when flying a commercial airline.

[NYC_SKP]

3. Not necessarily; Yes; Depends on the circumstances.

ATC don't watch each aircraft as they approach, so there's no guaranty that they'd catch it
There are warning systems and, if they were working, the cockpit would have been awash in noises and warning lights.
Pilots upon approach are hyper aware and this happened in clear daylight, no fog, I'm confident they knew their predicament.

Any number of reasons might explain why they came down too soon, short of the field, but that does seem to be what happened.

[gateley]

20. I know we'll learn more as the investigation continues,

I'm just wondering WHY and HOW this happened. If the pilots are okay, it will be interesting to hear what they have to say.

Breaking News right this minute -- ALL passengers are accounted for! They haven't upped the death toll yet, so I'm keeping my fingers crossed.

Thanks, NYC_SKP! I knew you'd help me out.

[NYC_SKP]

24. Strange indeed. To make it so far and come so close to an uneventful landing...

One wonders if it was a last minute emergency like a loss of power, or a human error, just not paying attention.

There are two times in any given flight that pilot and copilot are vigilant. Take-off and landing.

It's hard to believe that simple negligence would be to blame.

[leveymg]

4. 1) yes; 2) yes; 3) yes - initial report was another aircraft on the runway. Pilot may have tried to

pull up too steeply, causing the 777 to stall (loss of lift over the wings at too low an airspeed for the angle of attack) and crash. That would be pilot error caused by an even worse error by the control tower personnel, if the report is accurate. We will see.

[Agschmid]

11. Do you have a link for this?

[leveymg]

43. MSNBC live interview with a female witness, about 2 hrs ago. No link.

[Arctic Dave]

12. First I heard of that.

But yeah, that would definitely do it too.

[gateley]

15. Thanks -- I remember stalling when I was flying years ago.

I was always nervous about REAL approach and departure stalls -- up in the sky practicing was no problem.

I'm thinking if there was another aircraft, air traffic control would have alerted them and we would have heard that on the radio communication (IF they released that portion and IF that indeed happened and IF ATC was aware). As you say, we shall see. I just get frustrated at some of the questions (or lack of) that the newscasters are asking the experts.

[Agschmid]

21. We're you watching the interview at the hospital?

One question which was asked was "where are your helicopters kept?"... I really did not need to know the answer to that one?!?

[gateley]

23. No -- didn't see that (I was working today so missed all but surreptitious streaming snippets)

What on earth would prompt that question?

[Agschmid]

26. Now that is a valid question! n/t

[malaise]

62. Video here

[leveymg]

64. Yes, you can see the other plane, but it may be on the runway closer to the camera.

The runways are at SFI are parallel with each other, about 750 feet apart. Hard to tell from the video at this distance and angle - it's possible the witness also could not determine whether it was on the same runway. Granted, we haven't heard this anywhere else, so I'm inclined to think at this point she was mistaken.

[malaise]

67. It was not on the same runway

It was waiting to take off

[leveymg]

68. That would have placed it between the witness and the runway where the crash occurred.

I looked at my friend Google Maps, and see the runways are double diagonals. Two intersecting lines of double row concrete runways, with takeoffs happening on the diagonal to landings. Very little chance that an aircraft waiting to take off would be in the path of another on final approach.

Thanks!

[malaise]

69. Depending on the location of the witness

it could look that way. Even on the video of the crash, Hayes and his wife are clearly enjoying the majestic United 747 and then Asiana 214 approaches. United is on a side track but clearly not on the runway.

[Savannahmann]

10. There are two possibilities.

Mechanical or instrumentation error. Pilot error. Ok, three possibilities. Freak of nature. In the PATCO strike shortly after Reagan was sworn in, one of the issues was a big study on several recent aircraft accidents. We had no knowledge of any weather pattern that would cause the accidents, so it wasn't until the late 1980's that we learned about the phenomenon known as Microburst. http://en.wikipedia.org/wiki/Microburst

I'm not saying that's what happened. I'm saying that the way to determine what happened is to gather the information, and then eliminate the impossible. Then whatever remains, however improbable, is the truth.

[Agschmid]

14. Do microbursts only happen with storms?

This is the imagery and it seems to be associated with storms always...

http://en.m.wikipedia.org/wiki/File:Microburstcrosssection.JPG

[Savannahmann]

18. The article says there are two kinds.

Again, I'm not suggesting that microbursts happened in this airline crash. I'm saying that here we were, in the modern world, with Doppler radar going online all over the place, and we had no idea what a microburst was until the late 1980's. We are mortal, and we learn more about our world every year. Just because we know more this year than last, doesn't mean we know it all.

[Agschmid]

19. It's true!

[gateley]

25. How interesting -- didn't even think of that.

And yes, one of my favorite sayings (I attribute to Sherlock but not sure "he" was the originator).

I know we'll get all the info eventually, but I can't stop thinking and wondering... I'm driving myself crazy.

[Callmecrazy]

36. It's called Occam's Razor...

Which states that the simplest explanation is usually the correct one. Among competing hypotheses, the hypothesis with the fewest assumptions should be selected.
So says Wikipedia, so say we all.

[gateley]

42. Oh THAT'S Occam's Razor!!!

I've seen it referenced a zillion times -- guess I should have looked it up! Thanks!

[HereSince1628]

52. The simplest explanation with available information is better, that's not the same as correct

You've also got to be open minded that subsequent discovery of what was unknown at the time the argument was constructed may introduce complicating events that will show an original nice and tight explanation to be perfectly wrong in its details.

That's why there is also a rhetorical meme that goes something like "a beautiful theory is best killed by an ugly fact".

[Warpy]

13. The black box will tell all.

Downdrafts happen.

[gateley]

27. Yes -- and somebody else mentioned a microburst.

We'll just have to wait, but I can never shut my mind up from asking questions.

edit to change microblast to microburst -- no wonder my program didn't recognize the spelling.

[Stinky The Clown]

17. Some thoughts from a total non-expert.

There was a report I heard on TV that the pilot declared an emergency.

San Francisco (SFO) is surrounded by water. Water presents a featureless horizon and can cause depth (and speed) misperception.

[gateley]

31. Someone knowledgeable agreed that it is much different visually (water vs land) --

makes sense.

The reports I've heard (granted, sporadic) have said there was no emergency declared. As usual immediately following an event, we'll get conflicting reports.

I vaguely recall a plane executing a perfect landing years and years ago at SFO, except far, far short of the runway -- ended up in the water. I'm going to go see if I can find anything about it, not sure I'm making it up in my mind.

[Agschmid]

35. This one?

On November 22, 1968, a Japan Air Lines DC-8, named the Shiga[citation needed], operating Flight 2, crash landed on final approach at 9:30 a.m. on a shallow underwater reef at the eastern tip of Coyote Point (three miles short of the runway southeast of the airport). The plane was on a trip from Tokyo to SFO, after making a stop in Honolulu. The pilot was experienced, but apparently misread the instruments on the DC-8, which was less than a year old. There were 107 people on the plane. There were no deaths or serious injuries. The plane was salvaged by Bigge Drayage Company soon after the crash. All luggage and fuel were removed to cut the weight and the plane was lifted onto a barge and taken to the airport for repairs. The cost of repairs was $4 million and the plane re-entered service the following April

[jmowreader]

44. That case produced the "Asoh defense"

The government called the captain, Kohei Asoh, as the first witness in the NTSB hearing.

The first question they asked him was, why did you land in San Francisco Bay?

His answer was, "As you Americans say, Asoh fuck up."

Asoh got busted to First Officer by JAL and flew freight until he retired.

http://www.check-six.com/Crash_Sites/Shiga-SFBay.htm

[gateley]

46. That's hysterical! Thank God nobody was seriously hurt!

And they were able to put the plane back in service! Crazy story.

[jmowreader]

63. Jerry Harvey talked about this in his tape "The Abilene Paradox"

The NTSB, or whatever they had when this incident occurred, interviewed all the passengers. One of them told the government Captain Asoh landed so smoothly that he didn't know anything was wrong until he looked out the window after the plane stopped and saw a sailboat go by.

[gateley]

45. Yes! It must be -- that's right around the time. Ha! Haven't heard of a DC-8 in years!! Good job!

[PoliticAverse]

56. Extensive article (with pictures) on the incident...

http://blog.sfgate.com/thebigevent/2013/07/06/the-japan-air-lines-miracle-water-landing-of-1968-reprinted/

[The Velveteen Ocelot]

22. A lot of speculation right now.

1. ATC probably wouldn't have been able to tell if the descent rate was excessive, at least not visually. Assuming the pilots miscalculated a visual approach (which we don't know), they would already have been maybe 2500-3000' up before starting their final approach, probably at the assigned altitude (or the altitude previously programmed into the flight management system in accordance with a published approach procedure). What might have happened is that they started their descent too late and therefore tried to descend at a steeper than normal rate. There is some information floating around (not sure if accurate) suggesting they were descending at more than 1000 feet per minute instead of the normal rate of about 700 feet per minute. Local radar probably wouldn't have registered this because they would have been too low at that point.

2. Flight instruments and warning systems should have given the crew the information they needed. I read that the 28L glide slope was out of service, but they still had other information, such as the ground proximity warning system. This device would have produced an aural warning like "Don't sink" if it detected an excessive sink rate (normally based on the plane's radio altimeter). Most large airplanes also give altitude callouts when they get close to the ground.

3. Experienced pilots should be able to land the airplane easily using normal visual cues, even without instruments or GPWS warnings. However, there are a variety of explanations for why they landed short, including distraction resulting from (for example) watching for other airplanes on a parallel approach or near the runway. Did the pilot who wasn't flying (called the "pilot monitoring&quot call the attention of the pilot flying to the fact that they were maybe coming in too slow or at an excessive descent rate? That would be a normal procedure.

I'm sure we will find out exactly what happened. The NTSB will take awhile to produce a final report (usually at least a year), but it's likely that the cause of this accident will come out before that.

[Agschmid]

29. Is the flightaware tracking information correct?

That shows altitude and rate of descent?

[The Velveteen Ocelot]

32. Don't know how accurate their info is.

[gateley]

38. One of the experts on MSNBC mentioned how close the SFO runways are --

whereas at other airports there is so much space aircraft can land on parallel runways simultaneously with no problem, so you're right that it may have been a factor.

LOTS of things might have happened, I'm learning. I'll just try to be patient and wait until the investigators issue statements.

[The Velveteen Ocelot]

39. SFO uses PRM (precision runway monitoring) approaches

because the runways are so close together. However, the weather was clear so they would have been doing visual approaches anyhow.

[Mojo Electro]

28. Glide-slope indicator...

http://en.wikipedia.org/wiki/Visual_Glide_Slope_Indicator

"Visual Glide Slope Indicator or Visual Glideslope Indicator (VGSI) is a ground device that uses lights to assist a pilot in landing an airplane at an airport. The lights define a vertical approach path during the final approach to a runway and can help the pilot determine if the airplane is too high or too low for an optimum landing."

Although these days I think they might perform this same function using GPS equipment. But yes there are ways for a pilot to know his glide slope is too steep or too shallow. I don't know if air traffic control would be able to tell, but pilots have ways to determine.

How exactly does airspeed factor in? That I'm not exactly sure of.

[Savannahmann]

33. I think SF still uses the ILS.

Instrument Landing System, a brilliant system to give you both glideslope (usually 3 degrees) and a course heading. Most pilots fly one of the instrument systems, even in good weather. It's policy of many airlines to fly instrument systems whenever available.

[The Velveteen Ocelot]

37. SFO runway 28L has a 4-light PAPI (Precision Approach Path Indicator).

http://en.wikipedia.org/wiki/Precision_Approach_Path_Indicator

Assuming this was working, it should have given them accurate glide path indications. These devices have nothing to do with GPS; what they do is aim lights that change color depending on whether you are too high or too low.

Airspeed is very significant because if the airplane was coming in too slow (that would be surprising considering the level of automation of this airplane, but you never know), it would have been pitched up at an abnormally nose-high attitude - the high angle of attack would have been an attempt to compensate for the low airspeed (the slower you go the higher angle of attack you need to maintain altitude). If the airplane is too nose high it could hit the tail before the landing gear touched the runway.

[gateley]

40. Somebody has posted s/he heard it wasn't working --

but at this point, as we've learned, we'll get conflicting reports.

I recall landing once when I pulled up the nose too sharply -- stall warning! But I had plenty of time to correct -- I guess he didn't. Hard to think of an experienced pilot yanking up on the nose as enthusiastically as I did that time, though.

I'm so interested to hear the investigators' findings. I know it'll be a while, though.

[gateley]

41. I'm thinking many of us will be learning a lot about the factors in the days to come.

[Pilotguy]

47. Some thoughts on the crash today...

The glideslope to runway 28L was out of service at SFO. The glideslope provides the vertical guidance for an ILS or Instrument Landing System. Even on clear days and when cleared for a visual approach it is SOP (standard operating procedure) at many airlines to fly the ILS if a runway is equipped with one.

I am reading some reports that the PAPI for 28L was also out of service at the time of the crash. I'm not sure how accurate those reports are. The PAPI or Precision Approach Path Indicator is an array of red and white lights on the side of the runway that provide visual vertical guidance to landing aircraft.

ATC is capable of and is required by FAA regulations to issue low altitude alerts to pilots when they dip below the glidslope. But they are required to do so only when an aircraft is cleared for an instrument approach. Flight 214 was cleared for a visual approach so no such warning would be required. I believe with the glideslope out of service ATC would not have been able to monitor the aircraft's descent anyway.

With the ILS and PAPI out of service the pilots of Flight 214 had no instrument or visual aids to establish the proper descent to the runway. This really shouldn't be a problem for experienced pilots. Pilots should be able to fly a visual approach without crashing the aircraft.

My speculation:

The pilots of Flight 214 were probably not faced with a similar situation with both ILS and PAPI or VASI out of service in quite a long time and they just misjudged the descent rate. A passenger interviewed on CNN said the pilot advanced the throttles just before impact. If the descent rate was too steep the pilot would get a "sink rate" warning but he may have been too low at that point to save the landing.

One other point: Korean pilots are trained very differently than American pilots. Korean pilots rely much more heavily on the automated features of modern aircraft like the autopilot and the auto-land feature of the B777. Faced with a true visual approach with no ILS or PAPI the pilots were forced to hand fly the approach, something, perhaps, they may not have done in quite some time.

[dkf]

50. Saw this tweet. Is it true?

@LambysPoet: The jet's flight computer would have landed the Boeing 777 without a bump if the Instrument Landing System (ILS) was operational at SFO.

[Pilotguy]

53. Yes...

If the ILS was operational the auto-land feature of the B777 would have been an option for the flight crew. But again, any competent flight crew should be able to hand fly and land their aircraft without crashing it.

[dkf]

54. Thanks.

[gateley]

58. Read today that the guy in the left seat had 17 years experience, which

adds weight to your statement about Korean pilots vs American pilots (relying more on automation).

[Mr. David]

71. I heard something differerent

See post below...

[gateley]

57. Thanks for the interesting info --

CNN had the former head of something (the body to whom the FAA reports) as a guest, and she posited the same possibility, and mentioned that pilots seem to be relying more and more on the automation stuff rather than visually and manually maneuvering.

She also mentioned that not having the apparatuses working really frustrated the NTSB -- it didn't like that at all (not just in this instance, but in others). I just read something saying that these were often shut down for maintenance in clear weather, and that SFO had alerted all carriers that they would be out until August. The former head of ??? did say that the notice was there for everyone to see.

On MSNBC, another pilot said that Boeing is less reliant on automation (or maybe just allows more manual flying?) than Airbus, which is practically totally automated. I thought that was interesting.

[malaise]

74. Two questions related to this

With the ILS and PAPI out of service the pilots of Flight 214 had no instrument or visual aids to establish the proper descent to the runway.

At how many airports are these systems out of service?
two - is this by any means connected to the Sequester?

[canonfodder]

48. Maybe yes yes

ATC people are pretty busy. It may not have been noticed.
There is a barber pole on the PFD, that is not set for landing altitude. It reads altitude above ground regardless of the airfield altitude.
They may have also had an audible cockpit warning.
On the third point. You betcha. Second part, obviously not.
Speculation and second guessing doesn't help. Wait for NTSB preliminary report.

[gateley]

59. Yeah, I have no choice but to wait --

but all these questions fill my head, and DU is always a good resource for anything I want to know.

Heard today it may take months or years -- wow!

[RILib]

51. landed too soon

It sounds like there is an unusable part of the runway at the start and the pilot tried to land there. I think pilot error is overused, but in this case, a likely possibility.

[gateley]

61. I've always had a suspicion that the pilots were sometimes blamed to save someone else's butt --

or to reassure the flying public. Especially if the pilots didn't survive the crash.

That's kind of been dancing around my head, too (wondering if it's pilot error) -- it almost seems like an "oops" kind of thing, you know?

[dipsydoodle]

55. It has already been reported on UK tv news

that issues have already arisen about the legality of the runway itself. The end of it for whatever reason is 150 meters shorter than it should be.

[gateley]

60. Oh -- that IS interesting. Hadn't heard that on American news, of course.

Wonder if it's legit?

[Mr. David]

73. FAA must have known about it

or otherwise it would have not authorized landings on that runway.

[Savannahmann]

65. I am not a Pilot, nor an expert. But I have some experience in the field.

The Radar would show the transponders altitude information, if that was selected as an option on the tower local radar, but not on a scale that would assist them in informing the pilot he was too low. Of course, there is an old school solution, one that isn't used hardly ever anymore. It is called PAR, Precision Approach Radar. With that, a controller uses the radar to manage the glideslope, and course heading of the pilot. It is very accurate, but out of use, it's obviously easier, and cheaper, to use electronic means, like the ILS, Instrument Landing System. With a PAR approach, a controller literally talks the pilot down to about 200 feet above the ground, and if the pilot doesn't see the runway then, bad weather or whatever, a missed approach is called. The controller advises the pilot where he is on the glide path, and course, and issues corrections to get the plane back on the glide path/course. An example, going below glide path, slightly below. Going left of course, slightly left, turn right (new heading at least four degrees from the current one) Then information on how the corrections are working. Slightly below glide path, slightly below. (You're still low, but not getting any lower) Slightly below glide path, coming up slowly. (You're still a little low, but it's getting better) Finally you'll hear. "At decision height" If you can't see the airport go around and try again. By this time, the pilot is normally over the extended lights, and if he can't see the airport now, the weather really sucks.

The GPS approach would alert the pilot that he was low, but not slow, even PAR and ILS would not notify him he was going too slow, that is when the shaker system and planes own systems kick in. Remember the shaker system is the mechanical alarm for the plane going into a stall.

Number 3, yes. The pilot was trying to salvage the approach, but he was doing so from a point too low, and too slow, he should have taken the missed approach much earlier. Asking for one was not the way to go, the published missed approach is always available, and always kept clear. No other aircraft would have been on that path. I don't know what it was for San Francisco, but for the sake of argument, as an example, it would sound something like this. The pilot says executing missed approach. The controller says. "Understand missed approach" Then the controller repeats the missed approach instructions that are published and the standard. "Climb and maintain two thousand feet, fly runway heading, maintain this frequency for further instructions." Once compliance is confirmed. Then the pilot would be directed back into the landing pattern. Usually a turn to get them back into the pattern, followed with an instruction "Contact approach on (frequency)"

The pilot made a mistake, and compounded that mistake by not starting over, by trying to salvage the approach. He was too high, and too fast. He slammed the brakes on by increasing the angle of attack. That's the nose up attitude. Too much, and the plane slows dramatically, and also loses wind across the wings, or lift. Too much lost speed, and the plane no longer wants to fly, but does want to give into gravity. The shaking stick, the controls literally shake in the pilots hands. That's the plane shouting wake up dude, I'm stalling, we're going to fall out of the sky.

The Air France pilots that put the Airbus plane into the middle of the Atlantic ignored the stall warnings, no shaking stick in an A-320, but all sorts of alarms and flashing lights that said Stall. They put the plane in a nose up attitude, high angle of attack, and didn't increase power. The plane literally plummeted down from about 30k feet, and belly flopped into the Atlantic Ocean.

[Mr. David]

70. Latest I heard about this accident is that the pilot flying was inexperienced in this type.

As in 43 hours on Boeing 777's. That's it.

He has 9000 hours plus flying Boeing 747's into SFO as well.

But would it be a entirely different beast as compared to landing with a 747?

The PIC is training him, so it falls under his responsibility for this crash.