For the contrail to stay perfectly straight, the air would have to be extremely stable and this unusual. I've never really tried to correlate the appearance of contrails with the turbulence we're experiencing, but now that you've brought it up I'm going to be taking notice. Makes sense to me that turbulent air would mess us the contrail.

I've never had an instance where passenger input led to the discovery of a real problem, but I know it's happened. I did have one case where the flight attendants called up to tell us about strange engine noises. I was flying the MD-88 at the time, which has engines mounted on the aft fuselage. This was when I was flying the shuttle between NY and DC and the FAs we had did those flights all the time, so we took their concerns very seriously. But the instruments looked fine and it made no sense to divert or escalate the issue (the flight is only about 40 minutes).

After landing, we talked to mechanics who examined the engines and found some damaged turbine blades. The plane was taken out of service for maintenance. We made a point of contacting those flight attendants to let them know that their report was useful and greatly appreciated. Those engines might have gone a long time before any real issue came up, but they were a candidate for airborne problems and it was nice to avoid that.

I've never had an instance where passenger input led to the discovery of a real problem, but I know it's happened. I have a vague memory of one instance where passengers brought a hot floor to the attention of the crew. It seems they could feel the heat through their feet and it turned out to be a fire in the cargo compartment. The crew didn't yet have any indication of a problem. They investigated the complaint and diverted.

I had one case where the flight attendants called up to tell us about strange engine noises. I was flying the MD-88 at the time, which has engines mounted on the aft fuselage. This was when I was flying the shuttle between NY and DC and the FAs on those flights were very experienced and flew that plane all the time, so we took their concerns very seriously. But the instruments looked fine and it made no sense to divert or escalate the issue (the flight is only about 40 minutes).

After landing, we talked to mechanics who examined the engines and found some damaged turbine blades. The plane was taken out of service for maintenance. We made a point of contacting those flight attendants to let them know that their report was useful and greatly appreciated. Those engines might have gone a long time before any real issue came up, but they were a candidate for airborne problems and it was nice to avoid that.

How often is it pilot error vs mechanical failure that brings a plane down?

Just to clarify your question a bit; if you are referring to an actual crash of commercial aircraft, it is seldom the result of pilot error alone. It is generally the result of an "act of God" and/or a mechanical failure which is sometimes compounded by pilot error.

So probably your question is better phrased: How often is pilot error a significant factor in an air crash?

I add the word significant because the NTSB (and their peer organizations) tend to take a comprehensive approach to assessing the causes of accidents. It's not uncommon to see the cause of an accident be a mechanical failure, brought about by extreme weather and/or a maintenance issue, compounded by pilot error which is the result of poor training and/or procedures.

If by "commercial" aircraft you mean airliners, then maybe (but I'd still disagree). If you're talking actual commercial flights then it's overwhelmingly pilot error imo.

I'll let OP weight in rather than asserting my non-expert opinions on the matter. What I was attempting to say (perhaps not well), is that while pilot error is often a contributing factor, and frequently given as the "primary cause," it is seldom the only cause. In other words, there are not many incidents that involve a well-trained crew crashing a perfectly good aircraft in normal conditions.

I think a timely example of this is the AF 447. This was the flight which crashed in the South Atlantic, and for which the final report is due out soon. The primary cause will likely be pilot error. However, the crew flew into a severe storm (weather) for which their equipment was insufficient to help them avoid (manufacturing issue). The pitot tubes (which provide key information like speed) iced over (mechanical issues) leading to an extreme number of faulty readings (misinformation) for which the pilots had never trained for (training issue). The design of the fly-by-wire system made it difficult for one pilot to know what the other was doing (design issue). Ultimately it was the job of the crew to make it through the storm, something they should have been able to do, but as with most air accidents, there were a significant number of things that worked against the crew, each of which dramatically increased the chance of pilot error.

Ok so maybe I couldn't avoid asserting a few opinions. I'll let the expert take it from there.

I thought the vast majority of commercial flights were conducted by airlines? I read a stat on AOPA's site that said something like 85% of all US commercial pilots worked for airlines. I'm struggling to find it again though.

If you have a potential for a high cross-wind landing, do you have a longer final?

There is almost always some kind of chain of causation. But my point is that there are few accidents that correct action by the crew would not have averted. Off the top of my head I can think of AA191 (DC10 in Chicago), the miracle on the Hudson, the UA DC10 that crash landed in Souix City, the JAL 747 into Mt. Fuji, the 737s with uncommanded rudder deflections, maybe the PSA that midaired with the Cessna in San Diego, maybe the BA 777 that flamed out on final approach into Heathrow last year.

I see that your question sparked some interesting posts from HouseRulz and ElliotR, both making some excellent points. One of the main things is that you can't very often pick just one reason for a crash; it's usually a case of contributing factors and quite often an "error chain" that begins with something small.

But I get the sense of your question. Was the accident caused by something mechanical (e.g. the jackscrew in the horizontal stabilizer fails in an MD-80 resulting in an uncontrollable aircraft), or did the pilot do something that actually caused the accident (e.g. forgot to put down the landing gear)? [BTW, besides mechanical and pilot error, you have to add weather as a major cause.]

An example of real pilot error would be not setting takeoff flaps. This is clearly an error that could result in an accident. But you don't find many accidents which are so clear-cut. Usually it's more like AF447, where a bad situation (unusual attitude due to faulty indications in severe weather) was compounded by incorrect pilot response to the situation. The situation wasn't caused by pilot error, but the accident wouldn't have occurred if they had responded correctly.

One of the problems with the term "pilot error" is that it encompasses several things. You have actual errors (incorrect flap setting), poor judgment (flying through a Level 5 thunderstorm), lack of experience (e.g. handling icing conditions), and, lastly, lack of skill (incorrect response to inputs or, even worse, failure to maintain airspeed). If you look at almost any accident, you can find some way to put one of these "errors" forward as a contributing factor and in some cases it's the root cause.

So, I haven't really answered your question and that's because there is no simple answer. It's an interesting discussion but I can't give you a number here.

No, it's not necessary to have a longer final, but it is a good idea to get stabilized as early as possible to get a better feel for the crosswind during the approach. In good conditions (VMC with calm winds) we might accept a tight turn onto final for a visual approach and not be stabilized (i.e. landing configuation and on speed) until just above 1000' agl. With gusty winds and/or strong crosswinds, I would definitely prefer to be established on a straight-in approach.

This wasn't you was it?

http://aviationblog.dallasnews.com/a...ay-rate-i.html

Sorry for a super delayed response, I forgot about the whole thing. It appears I mixed accidents (I've read wikis for near all of the biggest plane crashes but apparently my memory isn't very sharp). For some reason I thought that the Valujet that crashed into the Everglades was because of icing on the wings. I guess that was some other crash. Sorry and thanks for all the great responses!


I recently came across a newspiece on a Finnish news site that had a video about crazy looking crosswind landings in some storm at the UK. I can't find the video anywhere but this serves the purpose as well: http://www.youtube.com/watch?v=9PdUd...eature=related Why are they choosing to land on winds like this? Some of those look pretty extreme, is it really safe?

If you ever somehow end up flying to Helsinki let me know if you need a tour guide. Our airport's been voted the best in the world a couple of times!

capt. w0x,

why do accident investigations always go out of their way to include the nationalities of the passengers?

A good few of those landings looked to be official Boeing footage of 777 certification, part of which would be to a series of crosswind landings. In order to achieve this, Boeing will fly to an airport where they can be guarnteed crosswinds.

Birmingham airport in the UK (EGBB) chose to go to a single runway configuration a while back with the runway at 90 deg to the prevailing winds. I assume they did this because that runway was the longest and had the most potential for further lengthening compared with the other runway. It makes for crosswind landings being de rigueur though. I'd be interested to hear what pilots who regularly fly into the airport think of the decision.

Do they? I know that in the U.S. they always mention how many Americans were on board any foreign crash, but I don't think I've noticed a complete breakdown of the nationalities.

Does a place exist where you could rent a flight simulator similar to used in pilot training to test it? I was re-reading the thread and came across the talk about how in mythbusters they tried the 9/11 scenario of whether someone with no experience could land a plane succesfully with radio instructions. Damn it would be epic to try that.

edit. Also: would be an epic prop bet, you could even record and broadcast that. One time please exist

You could probably find someone to rent you some sim time, but it would be expensive. Many airlines that own simlulators rent them out to other air carriers. World Airways pilots use Delta sims in Atlanta and, for a brief time a couple of years ago, Delta rented sim time from United at their Denver facility. Of course, those are other airlines and I'm not sure an individual could rent time, but it's a business and if you cough up enough money I'm pretty sure you could get the time.

Hey WOXOF....what does "flight following" mean?

"Flight following" is a service that controllers provide to VFR traffic on a workload-permitting basis. En route controllers, as opposed to Approach/Departure control and local control (aka "tower"), have as their primary function the separation of flights on instrument flight plans.

When a VFR pilot contacts one of these controllers (e.g. Washington Center) and requests flight following, the controller might respond with "unable" if he's saturated with IFR traffic. In most cases, the controller will assign the VFR traffic a discrete squawk and then provide traffic advisories through his airspace. But remember, the controller is mainly concerned with separation of IFR traffic and may not even see some VFR traffic (particularly those not equipped with a transponder*), so it is still the VFR pilot's responsibility to "see and avoid."

When the VFR traffic approaches the boundary with the next sector, the controller will often attempt a handoff to the next controller and give the pilot the new frequency, just as he would with IFR traffic. If he can't effect the handoff, he might tell the pilot to squawk VFR (1200) and try the next controller on his own. Also, of course, the next controller might have a high workload and decline flight following.

[*In some airspace (e.g. around major metropolitan areas) use of a transponder is required, so theoretically the controller will see all traffic. But his main job is to separate IFR traffic.]

W0X0F seems like one of your colleagues had a really close class today in Boston.

This story has a video with ATC audio.

Also a pax interviewed said one of the pilots told her they were within 300 feet. That would qualify as VERY close for an A319 configured to land, wouldn't it?

After watching countless episodes of Mayday, I have come to the conclusion that most airline accidents are caused by something to do with the pitot tubes.


Check the pitot tubes!

http://i.imgur.com/wKt9a.jpg -- Doubt if real but entertaining either way

Yes, that's very close. Runway incursions have been a hot topic for the past several years and this is why.