Landing into a storm

[BoeingBobby]

And why didn't they have tilt capability?

Please address this comment. Where do you come up with this? Was the radar deferred? I don't think so. You can't dispatch with BOTH radar units deferred.

[flight191]

Please address this comment. Where do you come up with this? Was the radar deferred? I don't think so. You can't dispatch with BOTH radar units deferred.

Again, Dave Gwinn's observations, not mine, the thread of discussion is here:

http://forums.jetcareers.com/threads/weather-radar-techniques.35839/page-4

However, item 14 in the casetext.com link (see above) provides this observation:

"14. At 5:59:37, DL 191 was about seven miles northeast of the storm and was requested to turn right to 340 degrees. Between 5:51 and 6:00, the microburst-producing cell (Cell D) had intensified from VIP 1 to VIP 4; the nose of the aircraft was pointed southwest toward this cell until 5:59:37. Except for a period between 5:55:53 and 5:57:19, during which a prelanding checklist was completed, the flight crew was relatively free of in-cockpit duties. During this period the flight crew would have been free to use the weather radar to observe Cell D and to manipulate the antenna tilt to analyze the storm structure and intensity. Since the storm cell had reached a VIP 4 by 6:00 180 , the cell would have contoured on their radar during this period."

[BoeingBobby]

Again, Dave Gwinn's observations, not mine, the thread of discussion is here:

http://forums.jetcareers.com/threads/weather-radar-techniques.35839/page-4

However, item 14 in the casetext.com link (see above) provides this observation:

"14. At 5:59:37, DL 191 was about seven miles northeast of the storm and was requested to turn right to 340 degrees. Between 5:51 and 6:00, the microburst-producing cell (Cell D) had intensified from VIP 1 to VIP 4; the nose of the aircraft was pointed southwest toward this cell until 5:59:37. Except for a period between 5:55:53 and 5:57:19, during which a prelanding checklist was completed, the flight crew was relatively free of in-cockpit duties. During this period the flight crew would have been free to use the weather radar to observe Cell D and to manipulate the antenna tilt to analyze the storm structure and intensity. Since the storm cell had reached a VIP 4 by 6:00 180 , the cell would have contoured on their radar during this period."

I have read the report and just finished reading the transcript. NO WHERE in there does it mention the fact that they were unable to tilt the radar transducer. In fact there are numerous mentions of the ability to have seen the radar echos, and the radars ability to see through the lighter cell in front of it. Your mention of "Camouflage" which is attenuation in radar language.

[flight191]

I have read the report and just finished reading the transcript. NO WHERE in there does it mention the fact that they were unable to tilt the radar transducer. In fact there are numerous mentions of the ability to have seen the radar echos, and the radars ability to see through the lighter cell in front of it. Your mention of "Camouflage" which is attenuation in radar language.

That is Gwinn's assertion. The point that Kapustin took umbrage with was that the tower saw numerous lightning strikes - nine according to Gwinn - and did not report it. He wanted ATC factored into the report. There were numerous instances in this event where information was not relayed, stale or unobtainable. He also wanted to include that as part of the final report. Kapustin dissented but by that time was no longer part of the NTSB.

[flight191]

Ok, dude the plane ahead was a Lear 25. He encountered a pretty nasty wind shear BUT did not report it.

Now, do you want to discuss if the DL 191 crew was wrekless or if they were operating at a high level of competence? That's a whole different discussion!

I do not think they were wreckless becasue they attempted to land through what one minute was a tall cloud, three minutes later a nasty nasty little cell and then shortly later, a gentle rainshower... The stars indeed aligned really bad for them.

And this is exactly Kapustin's assertion. Coupled with the numerous instances of "saw it, but didn't report it" caused an alignment - which you put so clearly - that was not in their favor. This is one of the few instances in the history of the organization where the lead investigator has dissented with the board.

[3WE]

...numerous instances of "saw it, but didn't report it" caused an alignment...

This may be nit picking but I do not agree that the problem was a shortage of information.

They saw the cloud.
They saw the lightning.
They had on-board weather radar.
The discussed that they were encountering a wind shear.

I don't think the weather guy skipping out for lunch, or what this pilot or that pilot WITH THE EXCEPTION OF THE LEAR JET GUY, had or wouldh ave had any real impact on this crash- the thing looked like a small, air-mass thundershower...If the lear guy could have reported that he lost a bunch of airspeed- that might have helped...but this was DFW and they were operating near 100% capacity on a Friday afternoon- planes landing every three mintues...

That meant two things- 1) The lear guy probaby did not have time to report his wind shear- which he succesfully negotiated and 2) the old exercise of "the guy ahead got in just fine".

And- you chose to overlook my Gabriel comment...It might have been some not-so-good stick and rudder execution at a not-so-good time that was the primary cause.

I would rank the slightly wrong aircraft control during the critical time as #1
A sudden, severe downburst from a relatively mundane cloud as #2
And the lack of the PIREP from the lear as #3.

I don't think additional PIREPS or radar reports would have influenced the pilots actions.

[Gabriel]

Side comment:

In most cases, when the pilots find themselves in a performance-critical situation that ends in an accident, the airplane had enough performance to avoid the disaster but the pilots botched the escape.

We have several no-flaps take-offs. The pilots of the DC-9-80 at Detroit made a remarkable job at keeping the AoA at the target, except it was the wrong target. The LAPA 737 in Buenos Aires managed to lift off but the pilot aborted when the stickshaker activated, they were above V2 and above the departure threshold by then. The Spanair MD-80 at Barajas, the stickshaker activated but the pilot didn't lower the nose a bit, with several thousands feet still ahead he had enough runway to abort or could afford the risk of briefly touching down again when lowering the nose, gain a few more knots and complete the take off, but not, he stalled, lost roll control, touched hard down at the side of the runway and kept running away (to the side) from the runway into a ditch with water and some vegetation.

We have AA at Cali, were they lost situational awareness in good VFR conditions and only noted the prominent mountain ahead out of the windshield when the GPWS activated. They immediately initiated a max-performance escape maneuver selecting TOGA and pointing the nose way up. They crashed a few feet below the top of the mountain. The investigation showed that they could have made it if they had remembered to stow the speedbrakes.

Turkish landing in Amsterdam. The AT was commanding IDLE and none of the three pilots noticed neither the AT mode shown in the FMA, the slow bleed of speed well below Vref, and the strange look of the artificial horizon showing much more blue than ususal (nose too high), what is remarkable since the three things are shown in the PRIMARY FLIGHT DISPLAY and one tends to think that one pilot should be looking from time to time at a display that has the words PRIMARY and FLIGHT in its name, especially in an instrument approach in IMC. So, the stickshaker eventually activated. The FO, who was the pilot flying, correctly reacted by pushing down to reduce the AoA and overpower the AT to apply max thrust. However, the captain called "My plane" so the FO removed his hands from the controls. The captain didn't put his hands on the thrust levers (amazing, during a stall recovery) so the AT, which was still commanding idle, retarded the levers. The plane, that had already started to recover from the stall, went more into the stall. The captain failed to lower the nose to exit the stall and applied TOGA again in the last seconds. The engines were just starting to spool up when the airplane hit soft ground at a very high sink rate and high nose-up attitude.

Delta 191 was no exception. The pilot did a very good job at first to raise the nose and add thrust during the "downbrust" and initial tailwind portions of the microbrust, so good that he hardly deviated from the glide slope. Amazingly, what triggered the disaster wan not the strong downdraft but the strong updraft that came after. As it usually happens when you hit a strong updraft when flying slow (approach speeds), the AoA activated for a split second since the updraft suddenly increases the AoA, but then the plane quickly reacts by itself by bending its trajectory upwards due to the increased lift and pointing the nose down due to the longitudinal stability (the plane was trimmed for a lower AoA). The pilot (the FO was the pitlot flying) reacted to that split-second activation of the stickshaker by lowering the nose, which would not be bad except that he grossly over-reacted. He moved the yoke I don't remember what bunch of inches forward, for what he had to apply I don't remember how many dozens of pounds of push force, and effectively lowered the AoA so much that he aimed the nose way down and put the plane in NEGATIVE Gs. So he actively maneuvered the plane to go from follow the glideslope to what would be basically a dive. They had already commenced the Go Around by then and the plane was accelerating, the pilot was already recovering from the dive when they were hit by another up-draft, another split-second stickshaker activation, and another overreaction of the pilot pushing way down. By when they crashed the plane was flying almost horizontally, both in trajectory and in pitch, and had a healthy airspeed that was increasing fast thanks to the TOGA thrust that was still applied. There was nothing performance-wise that would have prevented a successful pull-up and climb out at that point, except that the pilot did not pull up.

Despite all those gross errors in all these accidents, I am not too angry with the pilots for what they did by when they noticed that they were in a critical situation. While the performance was there for them to save the day if they had reacted correctly, they were caught by surprise in a very strange and confusing situation for which they were not well trained.

With different degrees, I am angry with these pilots for getting into that situations in the first place, starting with LAPA who failed to set flaps, check flaps, and abort when the config warning sounded as soon as they advanced the thrust levers for take-off. Second comes Turkish, for the reasons already explained. Third comes Spanair and Detroit, where both also forgot to set and check the flaps but lacked the advantage of a config warning, then Delta where the pilots saw the storm the rain shower, the lighting, and flew directly into all that, and also didn't attempt to abort the the landing and go around early enough (they first gained a lot of airspeed, the FO said "idle" and the captain said "careful, you are going to lose it all of a sudden", and when they did lose it and were hit by the downbrust the FO was still fighting to keep the glideslope, only when the FO started to dive the captain ordered "Go around"), and finally AA at Cali where the pilots lost situational awareness.

Interestingly enough, in all these cases there were factors external to the crew performance. In LAPA there was a very negative culture safety that, for example, encouraged the pilots to verify if an alarm was a false alarm before executing the corresponding procedure. In Turkish, a faulty radalt made the AT "believe" that they were in the flare already so it commanded idle. In Detroit and Barajas, the take-off configuration warning failed to activate as it should. In Cali, there were problems with the Jeppesen database in the FMS that helped the pilots mistake two different fixes. Also, they lacked the advantage of the spoilers' auto-stow function and the enhanced (forward-looking) GPWS, both of which were developed in great part thanks to the lessons learned in this accident. In AA 191, they lacked the PIREP from the preceding plane. Also, the stabilized approach criteria (that would have enforced a go around as soon as the approach became unstabilized), the "no landing into a storm" criteria (which I'm not sure how strictly is applied today), the modern windshear-escape techniques (that includes TOGA and nose-up as much as needed, trading airspeed for altitude, the stickshaker is the limit), and all the paraphernalia of on-board and ground windshear detectors and warning, didn't exist by then and were all developed after and thanks to this accident.

[BoeingBobby]

Delta where the pilots saw the storm the rain shower, the lighting, and flew directly into all that, and also didn't attempt to abort the the landing and go around early enough (they first gained a lot of airspeed, the FO said "idle" and the captain said "carefull, you are going to lose it all of a sudden", and when they did lose it and were hit by the downbrust the FO was still fighting to keep the glideslope, only when the FO started to dive the captain ordered "GO around"), and finally AA at Cali where the pilots lost situational awareness.

Excellent post Gabriel, I agree with your observations and opinions 100%

Flight 191 (Not sure I like the moniker that much by the way)

You STILL never addressed the statement that YOU made about not being able to tilt the radar. NO WHERE in any of your posted links or videos does it mention that they were unable to tilt the radar. In fact, quite the opposite is mentioned in 3 or 4 places.

[Gabriel]

Excellent post Gabriel, I agree with your observations and opinions 100%

You see how sometimes little-plane-low-time-non-IFR-rated pilots can have good judgement...

... those times being when seating in an armchair and typing in an Internet forum with yesterday's newspaper at hand

[TheKiecker]

This may be my favorite Internet Thread ever . Carry on !!

Gabriel that post above was EPIC.

I am sorry I gave you a wedgie .

[3WE]

You STILL never addressed the statement...

You STILL never addressed the question of how AA 1420 could make a landing with a severe thunderstorm occuring on the field when the policy- in really big bold font- is that you don't land in convective weather.

[BoeingBobby]

You STILL never addressed the question of how AA 1420 could make a landing with a severe thunderstorm occuring on the field when the policy- in really big bold font- is that you don't land in convective weather.

As you have seen, pilots do not always follow policy! Oh no there goes that humans f up from time to time thing again!!


What usually happens when one guy f's up is, the rest of us get f'd. New policy since the LCF wrong airport landing just handed down. The landing airport will now always be put into the fix page. The 500' call has now been modified to include the landing runway number as well as "checked" by the PF, and the repeat of the runway number by the PM.

[flight191]

You STILL never addressed the statement that YOU made about not being able to tilt the radar. NO WHERE in any of your posted links or videos does it mention that they were unable to tilt the radar. In fact, quite the opposite is mentioned in 3 or 4 places.

AGAIN, I did NOT make the statement. Dave Gwinn did. I suggest you take up the matter with him. What I agree with in Dave's comments - and does Kapustin - was that the tower saw weather and did not relay it to DL191. As Kapustin said in his comments in the documentary, the pilots made the decision to fly into the storm and crashed but a copious amount of real time information was never relayed to them either from the tower, PIREP or the meteorologists.

"Rudy Kapustin, who has been heading the NTSB investigation of the crash, noted that the meteorologist assigned to Dallas-Fort Worth Air Traffic Control was on a dinner break--his weather radar unmonitored--in the 40 minutes that preceded the crash.

Kapustin added that the weather radar specialist on duty at the Stephenville weather center about 70 miles from the airport was also on a dinner break just before the crash.

But both weathermen testified that although they were aware of the thunderstorm activity in the area before they took their breaks, neither considered the situation serious.

Although reports of thunderstorms near the approach to the runway were relayed to the air traffic control tower at the airport 10 minutes before the crash, Kapustin said these last-minute reports were not used to update the recorded weather message broadcast to pilots on final approach."

The only indicator they had was that AA351 had run into a good shower before they got into the thick of the storm. Kapustin felt they were left to their own devices in terms of the decision to fly into the maelstrom. Had they had all that info available to them would they have still flown in? That is a matter open to debate. Granted, as Gabriel pointed out in his excellent post some of those newer technologies were not available to them in their day but as an ironic consequence of this mishap, their misfortune has prevented other incidents due to heightened awareness, better training and improved technologies.

Oh, and 3WE, I don't disagree with any observations you have made regarding the encounter. The statements you and Gabriel made about the handling by the F/O are spot on. But I still agree with Kapustin about the lack of info relayed prior to the encounter. I know you felt it wouldn't have made a difference but having all the info on hand before making a decision can sometimes be helpful.

[3WE]

191- It seems you are a little fixated on 1) the lack of information being conveyed and 2) the radar tilt thing. 3) some other screwball aviation forum besides this one.

I'm thinking the answer to Bobby should be more like, "ok, perhaps that radar tilt comment is wrong"

Indeed, there's a rule that pilots should be aware of all information pertinent to the flight and there are a few things that could possibly have been fed to 191 that were not.

But have you read up on the Southern airways flameout- that happened at cruise, no less -and involved some "unconveyed weather"?

[Gabriel]

So, I understand that there is a policy not to land with a thunderstorm close to the field. but this policy is maybe not so strictly and evenly applied.

So a question for MCM and BoingBobby:

What's exactly the poilicy, and what data do you use to comply with it?
The second part of the question aims at, how do you know if there is a thunderstorm close to the field? Or how you distinguish the weather condition where the landing is prohibited by this policy from other similarly looking weather?

MCM mentioned that a red return means lots of rain, but not necessarily stormy wind conditions or heavy turbulence. That would be the case with the rainstorms typical of tropical regions (like Miami). And that this weather is neither dangerous nor under the scope of the "no landing" policy. So you get a red return, you see a very dark cloud and a rainshower ahead. Maybe even a lighting here or there. How do you know if you are ok to land?