Flare height illusions when landing in fog
Thread Starter
Flare height illusions when landing in fog
I have just had the pleasure of reading the book "Yellow Belly" by John Newton Chance. It was first published in Great Britain in 1959. It is the personal story of a novelist who went to War to fly but it is also the story of every man who learnt to fly in the last Great Air war that we shall know. The author became a flying instructor throughout the war.
In the book he describes being in an accident in which he was landing an Airspeed Oxford at night in a shallow layer of fog. The year was 1942. He was doing a check flight on a student when fog rolled in over the Lincolnshire aerodrome where they had been practicing night circuits. They flew a circuit by timing and turned in on final.
This is what he wrote: Quote:
"We turned in, keeping the dim line of little lights ahead. They disappeared in sudden patches where the fog was thicker, but we were on the line and were not going to let go, or we would have to go up again. So it had to be this time or perhaps not at all.
We came down very steadily. It could have been a model exercise - because everything depended on this one being right. The method of landing by flares was to sink down towards the second flare until it appeared to be a little way up the port window, then you throttled back and the aircraft stalled gently on to the ground. We came in a bit high over the first flare but dropped down till the second glowed through the white gloom just in the right position up the side window. I throttled back.
The bottom dropped out of the world. There was an appalling sensation of falling miles and miles through impenetrable blackness, and during the fall I saw what had happened as the actual flares, smoky red ghosts through the fog, showed their true light. There was a layer of fog four yards high over the flare path and, through refraction, the flares were shining as light bubbles on top of this blanket. I did not see the real flares until we stalled into the fog.
There was one hell of a crash and we were flat on our belly. I remember using the disaster to demonstrate crash action, but what reflex caused this I have no idea. "Petrol off! Switches off! Out!" The aircraft did not catch fire though we got well away from it as the fire engine and ambulance came rumbling on through fog." Unquote.
During my own flying career on aircraft from Tiger Moths to Boeing 737's I have landed in foggy conditions a few times. But I was never aware of the existence of false illusions of runway lighting at night when landing in shallow fog and which can fool the pilot into rounding out prematurely. I certainly have experienced false visual horizons caused by heavy concentrations of water on the windshield but what I read in this book of fog illusions caught me by surprise.
Have an Pprune readers struck this phenomena particularly in Cat 3 landings in fog?
In the book he describes being in an accident in which he was landing an Airspeed Oxford at night in a shallow layer of fog. The year was 1942. He was doing a check flight on a student when fog rolled in over the Lincolnshire aerodrome where they had been practicing night circuits. They flew a circuit by timing and turned in on final.
This is what he wrote: Quote:
"We turned in, keeping the dim line of little lights ahead. They disappeared in sudden patches where the fog was thicker, but we were on the line and were not going to let go, or we would have to go up again. So it had to be this time or perhaps not at all.
We came down very steadily. It could have been a model exercise - because everything depended on this one being right. The method of landing by flares was to sink down towards the second flare until it appeared to be a little way up the port window, then you throttled back and the aircraft stalled gently on to the ground. We came in a bit high over the first flare but dropped down till the second glowed through the white gloom just in the right position up the side window. I throttled back.
The bottom dropped out of the world. There was an appalling sensation of falling miles and miles through impenetrable blackness, and during the fall I saw what had happened as the actual flares, smoky red ghosts through the fog, showed their true light. There was a layer of fog four yards high over the flare path and, through refraction, the flares were shining as light bubbles on top of this blanket. I did not see the real flares until we stalled into the fog.
There was one hell of a crash and we were flat on our belly. I remember using the disaster to demonstrate crash action, but what reflex caused this I have no idea. "Petrol off! Switches off! Out!" The aircraft did not catch fire though we got well away from it as the fire engine and ambulance came rumbling on through fog." Unquote.
During my own flying career on aircraft from Tiger Moths to Boeing 737's I have landed in foggy conditions a few times. But I was never aware of the existence of false illusions of runway lighting at night when landing in shallow fog and which can fool the pilot into rounding out prematurely. I certainly have experienced false visual horizons caused by heavy concentrations of water on the windshield but what I read in this book of fog illusions caught me by surprise.
Have an Pprune readers struck this phenomena particularly in Cat 3 landings in fog?
Well the aircraft flares in Cat 3 landing, not the pilot so ... no?
I can't ever remember landing in a few meters of fog. A couple of feet, yes, a thousand feet, yes, but not a couple of meters - especially in the dark. (Radiation fog generally forming towards the morning.)
I can't ever remember landing in a few meters of fog. A couple of feet, yes, a thousand feet, yes, but not a couple of meters - especially in the dark. (Radiation fog generally forming towards the morning.)
When landing in shallow fog, assuming the layer to be at least up to flare height, it is possible to suffer illusions.
Some effects involve the perception that the far point of what is seen (the visual segment) is related to pitch attitude.
The first contact with lights / ground should be at the cockpit cutoff, the visual segment is from the cutoff to the far point of what can be seen (slant visual range). The extent of visual segment generally relates to the density of the fog, with a homogenous fog - consistent density with altitude / location, then the visual segment should grow with decreasing altitude.
An increasing visual segment has been related to over-pitching during the flare, high round out, but with potential to correct (and some examples of the opposite effect).
With varying density, or with relatively shallow layers, then the visual segment can reduce; an extreme could be seeing most of the runway during the approach, then at the fog top experiencing a closing visual segment. This might induce a pitch down reaction or late flare hoping to see more.
Shallow 'stable' fogs are generally more consistent, the reality is that fog has a wide range of structures. Beware proceeding traffic stirring up a stable fog, good or not so.
Fogs still forming tend to have layers of different density varying with altitude, 'layered stratus'. These can result in a combination of opening and closing visual segments, and reversals, slowly, with subtle change in what can be seen.
Dispersing fogs vary more with location, approach and along the runway, where effects are similar to entering and exiting mini cumulus clouds, rapidly changing visual segment and (often most disturbing) the ambient light levels, bright sun to dull, lights appearing disappearing instantaneously; potential for illusion.
A range of studies of manual landings in cat 3 indicate a tendency to over-flare in the higher visibilities, then correct, and under-flare in lower visibility with less capability for correction.
The flare performance to a significant degree is associated with height judgement (ht / ht rate), opposed to conventional flare theory based on attitude.
Some effects involve the perception that the far point of what is seen (the visual segment) is related to pitch attitude.
The first contact with lights / ground should be at the cockpit cutoff, the visual segment is from the cutoff to the far point of what can be seen (slant visual range). The extent of visual segment generally relates to the density of the fog, with a homogenous fog - consistent density with altitude / location, then the visual segment should grow with decreasing altitude.
An increasing visual segment has been related to over-pitching during the flare, high round out, but with potential to correct (and some examples of the opposite effect).
With varying density, or with relatively shallow layers, then the visual segment can reduce; an extreme could be seeing most of the runway during the approach, then at the fog top experiencing a closing visual segment. This might induce a pitch down reaction or late flare hoping to see more.
Shallow 'stable' fogs are generally more consistent, the reality is that fog has a wide range of structures. Beware proceeding traffic stirring up a stable fog, good or not so.
Fogs still forming tend to have layers of different density varying with altitude, 'layered stratus'. These can result in a combination of opening and closing visual segments, and reversals, slowly, with subtle change in what can be seen.
Dispersing fogs vary more with location, approach and along the runway, where effects are similar to entering and exiting mini cumulus clouds, rapidly changing visual segment and (often most disturbing) the ambient light levels, bright sun to dull, lights appearing disappearing instantaneously; potential for illusion.
A range of studies of manual landings in cat 3 indicate a tendency to over-flare in the higher visibilities, then correct, and under-flare in lower visibility with less capability for correction.
The flare performance to a significant degree is associated with height judgement (ht / ht rate), opposed to conventional flare theory based on attitude.
Checkboard, the unaided manual Cat 3 landings were part of a research programme investigating pilots' capabilities after fail passive AP failure, to allow use of lower RVRs. This also considered FD HUD.
The trials looked at separate aspects of the RVR required to make a decision - above DH, and possibly a lower value in continuing with a manual landing below DH; the excluded option was for manual GA below DH.
The first use of adjusted regulation was UK CAA clearance for the Avro RJ with 50ft DH.
Generally Cat 3 operations were more consistent; enabling lower RVR because of less variability in the fog, opposed to Cat 2 where conditions are more associated with fog formation and dispersal or other reductions in visibility, snow, dust. In these, Cat 2, it is possible to decide to land and then find that the visibility was not valid for manual continuation as the conditions changed. Greater exposure time from higher DH and more lateral deviation to be considered.
Operational minima have a significant safety margin for the authorised approach, although Cat 2 has the lesser margin, and in extreme where it is possible to decide to land and then be wrong !
The trials looked at separate aspects of the RVR required to make a decision - above DH, and possibly a lower value in continuing with a manual landing below DH; the excluded option was for manual GA below DH.
The first use of adjusted regulation was UK CAA clearance for the Avro RJ with 50ft DH.
Generally Cat 3 operations were more consistent; enabling lower RVR because of less variability in the fog, opposed to Cat 2 where conditions are more associated with fog formation and dispersal or other reductions in visibility, snow, dust. In these, Cat 2, it is possible to decide to land and then find that the visibility was not valid for manual continuation as the conditions changed. Greater exposure time from higher DH and more lateral deviation to be considered.
Operational minima have a significant safety margin for the authorised approach, although Cat 2 has the lesser margin, and in extreme where it is possible to decide to land and then be wrong !
Join Date: Jan 2006
Location: US
Posts: 2,205
Likes: 0
Received 0 Likes
on
0 Posts
Landed in HPN (White Plains) back in 1979-1981. Air Florida taxiing out....just the vertical stabilizer was visible. Ground fog from the lake just NW of the airport. Common early morning problem in the spring/fall. We could see the approach lights all the way down...until 10-15'? FO is looking down outside his side window saying "lower, lower, lower....." I could see the runway lights but had zero depth perception.
Join Date: Jan 2006
Location: US
Posts: 2,205
Likes: 0
Received 0 Likes
on
0 Posts
Many implementations of single HUD are extensions of Cat 2; a similar concept with reduced minima enabled by technology, beware false credit, must consider the overall operation.
Some situations are like 'being wrong' in Cat 2; technology gets you into a situation, but with HUD failure the operation depends on human recovery. Single HUD requires a head down pilot for monitoring and GA. The head-up PF has to revert head-down, and/or change of control for a GA, need to focus on the new task and reinterpret conventional instruments.
The reduced minima for manual landing with HUD may be not be sufficient to continuing a manual landing without HUD. The change of view, new mental task, takes time. Although an aircraft is capable of a low GA, the time-consuming mental process could increase the height loss.
There is anecdotal evidence that pilots do not 'look through' the HUD, decisions are made on fewer cues than specified - we perceive what is expected, and landings are made with FD opposed to using the visual scene as normally assumed - because FD is an easier mental task and usually results in 'good' landings.
The manual Cat 3 trials used a HUD (Mono HUD); a single eye system using the then current technology from helmet mounted displays (1978), but with a simpler format than shown. There were no computed flight path or flare aids, no FD or GA mode.
There were advantages when monitoring the progress of an auto or manual landing, and for flying a GA after auto failure; but never really cost effective when considering true Cat 3.
Beware what some authorities have approved - check the assumptions: HUD can generate 'illusions', false security from technology, and false sense of human capability.
https://rochesteravionicarchives.co....t-demonstrator
Click Documents for tech description; download pdf, note different formats and other evaluations / uses.
As tested in BAC 1-11 https://cdn.rochesteravionicarchives...og/2_D0549.pdf
Video:- https://rochesteravionicarchives.co....civil-aircraft
Expand to full screen. Fifth approach ~ 4:30, Cat 3b 150m 50ft DH, these were the trial test limits for manual takeover and landing.
HUD shown was not as used; other autoland clips including DC-9.
Some situations are like 'being wrong' in Cat 2; technology gets you into a situation, but with HUD failure the operation depends on human recovery. Single HUD requires a head down pilot for monitoring and GA. The head-up PF has to revert head-down, and/or change of control for a GA, need to focus on the new task and reinterpret conventional instruments.
The reduced minima for manual landing with HUD may be not be sufficient to continuing a manual landing without HUD. The change of view, new mental task, takes time. Although an aircraft is capable of a low GA, the time-consuming mental process could increase the height loss.
There is anecdotal evidence that pilots do not 'look through' the HUD, decisions are made on fewer cues than specified - we perceive what is expected, and landings are made with FD opposed to using the visual scene as normally assumed - because FD is an easier mental task and usually results in 'good' landings.
The manual Cat 3 trials used a HUD (Mono HUD); a single eye system using the then current technology from helmet mounted displays (1978), but with a simpler format than shown. There were no computed flight path or flare aids, no FD or GA mode.
There were advantages when monitoring the progress of an auto or manual landing, and for flying a GA after auto failure; but never really cost effective when considering true Cat 3.
Beware what some authorities have approved - check the assumptions: HUD can generate 'illusions', false security from technology, and false sense of human capability.
https://rochesteravionicarchives.co....t-demonstrator
Click Documents for tech description; download pdf, note different formats and other evaluations / uses.
As tested in BAC 1-11 https://cdn.rochesteravionicarchives...og/2_D0549.pdf
Video:- https://rochesteravionicarchives.co....civil-aircraft
Expand to full screen. Fifth approach ~ 4:30, Cat 3b 150m 50ft DH, these were the trial test limits for manual takeover and landing.
HUD shown was not as used; other autoland clips including DC-9.
Complete speculation by an SLF here. But if you thought you were looking at two adjacent flares, when in fact the uneven fog had obscured one in between, you might be a lot higher than you thought you were.
