ILS Capture from Above False GS

[Founder]

I just flew to Antalya Turkey, and there they always keep us high on approach, we were at 8000 ft when the GS started coming, we later on descended and caught the GS from above at 3000 ft.

I started thinking during the approach about the possibility of capturing a "false loab", I'm not too sure about the spelling on that one but anyway, I remember from my ATPL theory that the GS transmitter is not only sending the "standard 3 degree" glide slope but also some false loabs above and below the 3 degrees...

Has anyone ever managed to catch one of these false loabs? my thoughts are because almost everytime when we're flying in Turkey we're capturing the GS from above...

[zerozero]

As part of your SOPs, are you required to check your altitude when passing the Final Approach Fix inbound?

Look at this ILS. You should verify your altitude on the GS at EISEN as 1500'. Then you're guaranteed to be on the proper GS.

http://www.ntsb.gov/publictn/2006/AAB0606_files/image004.jpg

Second, with a groundspeed of approx. 140KIAS your sink rate should be about 700-800fpm.

Have a look at the following note. The first false glideslope is at 6 degrees and *reversed*. That's a big warning right there. The second false glideslope is at 9 degrees and properly sensed. That's three times steeper. If you need a sink rate of 2100-2400fpm just to maintain the glideslope then something is definitely wrong!

ILS (Instrument Landing System) (http://www.allstar.fiu.edu/aero/ILS.htm)

False signals may be generated along the glide slope in multiples of the glide path angle, the first being approximately 6º degrees above horizontal. This false signal will be a reciprocal signal (i.e. the fly up and fly down commands will be reversed). The false signal at 9º will be oriented in the same manner as the true glide slope. There are no false signals below the actual slope. An aircraft flying according to the published approach procedure on a front course ILS should not encounter these false signals.

Finally, the proper spelling is "lobe".

:ok:

[Pontius]

I've seen the false glideslopes but never actually captured one. As you are no doubt aware, they can exist in multiples of the glideslope angle, depending on how many false lobes are 'powerful' enough to be detected by your ILS receiver. So, with a 3 deg G/S you may see a false G/S at 6 deg, 9 deg etc, although I've only ever experienced the 6 deg variety.

If, as in your Turkish scenario, you will be approaching the G/S from above the only thing you can do is to ensure it's the correct path by doing the 3 times table. If you're on the G/S at 20nm and 12000' you know you're on the false G/S. I add this only by way of additional information, rather than trying to 'teach' you something you undoubtedly already know.

False G/S and an early capture of the altitude when in FLCH (Open Des or whatever it's called :)) and then having to quickly reduce the MCP altitude so you can keep the descent going to capture the G/S from above certainly make for 'interesting' approaches. Sometimes these things can't be helped, especially with hills around. Now add in a 'foreign' accent or two and up goes the workload.

You would think those chaps and chapesses, with big brains, who design these things could put in an aircraft filter of some kind to stop the receiver 'seeing' the false G/S. They will, of course, always exist at the transmitter but I'm quite sure a simple 3 times table machine could be added to the aircraft. Yes, I know it would have to adjust to allow for other G/S angles but certainly not beyond the wit of man, especially as the G/S angle is written on the Legs page (I assume Jambes page on the 'Bus) .

Edited to add: DOH! Note to self: don't post at the same time as other people are adding proper stuff :}

[pilotmike]

You would think those chaps and chapesses, with big brains, who design these things could put in an aircraft filter of some kind to stop the receiver 'seeing' the false G/S. They will, of course, always exist at the transmitter but I'm quite sure a simple 3 times table machine could be added to the aircraft.
The lobes don't 'exist at the transmitter'. Rather, they are the effect of Superposition, or in layman's terms, a mixing of the original signal with another 'echo' signal. And no, there is no simple solution, as the receiver simply receives the frequency to which it is tuned, and it is not a simple task to accept some of the signal as 'good' and to reject some of it as 'bad'.

It is rather similar to the acoustic effect of mixed delayed sounds arriving from a PA system from multiple speaker sources, say at a fairground or an air display. When close to one speaker, there is no ambiguity as the strong signal from the speaker dominates. However, move away to a place where sounds from multiple speakers can be heard with similar intensity but with diffferent delay paths, and the muddled sound can be so confusing as to be completely unintelligible at times. Given that the muliple signals are indeed present, you should be completely clear that there is virtually nothing that you can do, as a remote observer or 'receiver of the signal' to separate the signals to regain the intended intelligibility.

Echo deconvolution or removal is a huge subject, and is an inexact science. Seismology is one field where much work has been done, and the Cepstrum, with its multiple Fourier Transforms, is a potentially very good way to mathematically deconvolve a mixed signal to reveal the intended source signal. However I don't believe it will be arriving in aircraft ILS receivers any time soon. For now, the altitude check at aproximately 4NM, as cited by zerozero, is our best defence against false glideslope lobes.

[forget]

Excellent Air New Zealand video here

YouTube - NZ60 Erronious ILS Incident, Apia (http://www.youtube.com/watch?v=GelRBhJ4gmI)

PS. The video is in three parts.

[tom775257]

Yes, I have been PNF on a jet that captured the 6 degree lobe (Airbus 320). I called go-around; to be honest it is bloody obvious (unless you are completely behind the aircraft and very unstable as happened with the captain flying it). But I learnt a lesson - don't expect someone who is very behind an aircraft to fly a go-around correctly!

[Founder]

Very interesting reading...

But regarding the ROD, if you're capturing from above the ROD is probably gonna be around 1500 FPM or maybe more already so it might take some time before you actually realize that its the wrong lobe that has been captured...

I'm just speculating here, but I would say that capturing from above and from high altitude could be a great source of danger... and it's bad procedure from the ATC to keep aircraft high like that...

well well, thanx for the great input, it's been interesting =)

[tom775257]

Well yes, you might descend at a high rate to intercept, but if the aircraft has achieved glide slope capture and is still going down at 1400 feet/min something is very wrong - plus the attitude is wrong and it looks wrong if visual.

Regarding ATC, if you are kept high, it isn't great work from them; but you can always ask for delaying vectors etc. to give you more time. The pilots are to blame if you push on regardless while hot and high.

The airline I am currently flying for is a big advocate of the discontinued approach. On a tangent, most airlines teach us in the sims go arounds from decision altitude, not from further out on the approach. Infact it is far easier at a higher altitude just to level off, start going up and reconfigure and have another go at an approach, than sticking in TOGA thrust.

[Sciolistes]

I just flew to Antalya Turkey, and there they always keep us high on approach, we were at 8000 ft when the GS started coming, we later on descended and caught the GS from above at 3000 ft.
Be aware that the normal limit of the GS is 10nm. However, even in my meager experience I have seen several undulating captures within 10nm. As PF I now don't engage APP until Flap 5 (some drag to reduce unwanted acceleration) and at a reasonable altitude. I'm usually managing the descent in V/S shortly before capture within the distance limits.

[Pontius]

Thanks PilotMike. I didn't realise it was all so complicated and that probably explains why the big-brained people haven't done it. There's me thinking my 3 times table machine would solve world hunger and cure cancer :{

Echo deconvolution

I'm going to try and remember that phrase and get it into dinner party conversations :ok:

[goeasy]

I had one recently, when tail wind kept us high with ATC speed reductions. Instaed of chasing GS, we levelled off at 4000'. Then ATC requsted us to descend to 3000', but at same time Acft detected false GS ABOVE us and started climbing frantically!!! :confused:

Good lesson... disarm GS if you arent interested in capture... before the surprises. :ouch:

[mike rondot]

Attempting ILS glideslope capture from above was an automatic instrument rating test fail item when I was a boy in the mlitary. It was considered dangerous for good and sensible safety reasons.

[zerozero]

Here's an interesting bit of trivia--at least for the American pilots.

The GS signal is "flight checked" by the FAA only from the FAF inbound down to DA.

In other words, if you couple the autopilot to the GS outside of the FAF (in the USA) you do so at your own peril.

So, keeping in mind what 'mike rondot' says about joining the GS from above, I would submit, if you chose to join the GS from above, *outside* of the FAF, it should only be used for *advisory* purposes and NOT be used to control your descent.

I'd recommend making that descent on V/S to altitude pre-select, until the aircraft is fully established within the service volume of the ILS.

[chuks]

You have lobes on the localiser too, you know! I remember once watching the system lock onto one of those and try to capture it. Didn't work very well; it just turned in, lost the signal and then went into basic "heading and altitude" mode pointed completely in the wrong direction. Our mistake was to have "approach" armed as we flew in "heading and altitude" mode, forgetting about possibly catching a localiser side-lobe. The "LOC" annunciation went from white to green and the airplane turned off the intercept heading while we (somewhat new to EFIS) did that "What is it doing now?" number.

When you are established high on the localiser then you might want to use "navigation" mode, when you track the localiser but do not leave the pre-set altitude, until you are close to the final approach fix at the correct altitude. You won't just have the machine trying to fly an impossible angle but it will not get you to the fix at the correct altitude.

If you have "approach" pre-selected then the machine won't know the difference between a false LOC or GS signal and the real one so that it will just lock on and try to fly it once that course guidance signal comes close to centre: one more example of how people are smarter than computers! Of course Situational Awareness comes into this; it is no good just sitting there expecting the machine to do all the work for you.

[SIDSTAR]

Airbus has a simple procedure for capturing GS from above. You set VS of 1500 to 2000 fpm depending on how far off slope you are and this will bring you down to the GS when it captures (GS*) and you set GA altitude.

However, many lazy unthinking pilots forget to set an altitude ABOVE current altitude when starting the procedure resulting in an altitude capture at just the wrong time before GS intercept.

Never forget your 3 times tables and you won't go far wrong in any descent. Unfortunately many of today's young uns can't seem to do this without a calculator which is slightly inconvenient on the approach!

Published GS height/altitude over the OM or equivalent position (typically 4 to 5 DME on newer ILS installations should always be checked and should be in your company SOPs.

To the guy with the Captain who was "surprised" I'd say, "think of every approach as an approach to a GA" (and every takeoff as a takeoff to a reject) and you'll rarely be surprised. When you're going think of stopping and when you're stopping think of going!

[IO540]

I've been vectored above the GS a few times, most recently at Bournemouth the other day. Presumably I would have captured the 6deg one had I carried on. I was visual so disengaged the AP and proceeded manually.

[bfisk]

Well, there is no FAF for a precision approach, the final approach starts at the FAP, which is wherever you happen to be when you start down the glideslope. You are the, by def, on final approach.

USA does have it's own non ICAO method of establishing instrument approach procedures (TERPS vs PANS OPS), but that does not mean that you are "at your own peril" with the glideslope coupled at any time, as long as you are not below the relevant altitude (MSA, intial approach altitude or whatever depending on how far out you are).

Coupled or not, the autopilot is NEVER responisble for terrain separation, you, the pilot, are. GS, VS or whatever descend mode you may be using.

[zerozero]

The last sentence is absolutely true.

But anytime you're using a signal beyond the "operational service volume" (do you prefer that terminology?) you do so at your own.....risk.

Perhaps "peril" is a little too dramatic for you.

[A37575]

Many years ago I was flying a 737-200 to Manila. Due noise abatement we were kept high at 4000 feet until localiser intercept which was around 8 miles where the glide slope should have been on slope at 2400 ft. We expected full scale fly down at LLZ intercept but instead the glide slope indicator on both captain and F/O ILS showed only slightly high by less thah half a dot. Cloud tops were 3000 ft and base 1200ft. Despite full flap and idle thrust we stayed on the apparent glide slope which gave a well over 1000 fpm rate of descent. Ground speed showed no wind. We broke visual around 1500 ft and the runway perspective showed we were definately high with glide slope indication just slightly high. The T-VASIS was useless with red and white lights in unlikely combinations.

Landed Ok and sent report to Manila ATC. A year or so earlier, an Air Manila (?) Boeing 707 from Guam had crashed short of the runway in perfect weather. A couple of fatalities I think, and the captain swore he was on the ILS glide slope right down to impact. In other words he blindly followed his instruments (even though the sink rate was out of this world).

The ILs was tested shortly after we had complained, and the calibration aircraft reported a perfectly flyable glide slope around seven degrees as well as three degrees. This was the result of poor maintenance and it was just by luck we picked it up because of the initial high vectoring. If other aircraft had had similar experience they did not report it. It also explained that the 707 pilot was probably telling the truth a year earlier when he said he followed a steady glide slope to impact.

Asked by the investigators why he didn't go around from the obviously unstable approach, he replied he thought the noise of all engines spooling up from idle thrust would only frighten the passengers! At the time of impact the 707 had full flap, speed brakes extended and idle thrust with a sink rate of around 2000 ft per min. Obviously the captain was a proponent of the saying "Real Men don't go around."

The T-VASIS? The all over the place T-VASIS was caused by severe earth tremors a few days earlier which moved the light boxes. When asked why the T-VASIS was not turned off when it was obvious that damage to the boxes had occurred, ATC advised they were abiding by a standing directive that the T-VASIS must be switched on for all jet arrivals. The Directive omitted to say the VASIS must also be serviceable.. .

[Graybeard]

In the early 1970s, a DC-8 freighter captured a false LOC beam into Cold Harbor(?), Alaskan Aleutian chain, and hit a mountain.

GB

[mike rondot]

There is no "choice" about intercepting ILS glidepath from above. Or, there was not when I was flying. It was a prohibited manoeuvre then and surely it still is. As I said, for good and sensible reasons it was considered dangerous practice.

Maybe a current IRE would care to quote chapter and verse.

[DutchBird-757]

With our company SOP's we are NOT allowed to capture the GS from above. If it's looking to head that way then call go-around or ask for vectoring to come back around and try again. My opinion is that if you are kinda 'forced' to capture from above than you might be flying a rushed approach and seriously consider breaking it off to try again after a quick DODAR session. Nothing wrong with that IMHO.

I've had false LOC captures before (particulary at MAD) but no false glideslope's yet. ZRH is one of those places that tries to keep you high and fast as well. So that's in our descent briefing and what action to take if your mentioned scenario happens.

[BelArgUSA]

I cannot recall an incident of being on a false glide slope.
Above the glide slopè, that is. Many years, many countries.
One thing I always did, was to maintain a "normal profile" for descent.
That is to be at some 10,000' some 30 NM away from runway, at 250 KIAS...
To be at 6,000' some 20 NM away, start flaps, and be 3,000' some 6 NM away.
Then follow the glide slope, and x-check altitude at the OM or FAF.
And see that all that does not bring you below terrain.
No need to fly acrobatics to get to a glide slope.
xxx
I can recall intercepting GS for LAX 24R/25R some 70-80 NM out (profile).
However would first maintain the profile with my distance/height rule above.
Glide slope itself...? Would wait until some 10 DME to ARM the GS.
Looks like I did ok... Now retired and safely in my rocking chair...
xxx
:8
Happy contrails

[Vee1Kut]

To the Op....most autopilots won't capture from above...so were you hand flying?

[Kiltie]

One of my Boeing TREs told me it is ok to arm and capture a glideslope from above. I have discouraged this practice on the flight deck of every aeroplane I have flown for risk of false glideslope capture so begged to differ. He stood his ground saying it was not an unsafe practice.

Would any other trainers care to comment?

[PantLoad]

There is no perfect antenna. If it's supposed to be omnidirectional or directional or whatever....it's not perfect. Whatever antenna you have, it'll have undesired lobes. (In this case, we're talking about the
glide slope transmitter antenna.)

BelArgUSA is totally correct in his method. Not only will this circumvent
tracking one of these undesired lobes, it also usually keeps you from hitting mountains.

Your company's SOPs should address this issue. If not, the manufacturer of your aircraft should have a procedure for such a situation.


Fly Safe,

PantLoad

[A Squared]

In the early 1970s, a DC-8 freighter captured a false LOC beam into Cold Harbor(?), Alaskan Aleutian chain, and hit a mountain.

Nope, in the early 1970's a DC-8 hit a mountain near Cold Bay, because the captain was wandering around, a long way from being anywhere close to the localizer, knowing he was not on the localizer (or anywhere close) well below the sector altitude, and an altitude that he should only have been at if he was established in the inbound course. There may have been some difficulties with the navigation radios, but that was because they were below the mountains in a place where they never should have been.

Read the report and especially take a look at the map with the aircraft path compared to the transition route and approach. They weren't even in the same ballpark.

accident report (http://www.airdisaster.com/reports/ntsb/AAR74-06.pdf)

[nitpicker330]

Cold Bay has a back course LOC approach so he may have read the course bar incorrectly?? Dunno

As to capturing G/S from above, for goodness sake fellas its not something to aim for all the time!!! But we all should now how to capture from above, what our particular Aircraft can/cannot do and still maintain a safe operation......

Hence my Airline trains for it in the Sim and where it happens during line training WHILST still maintaining SOP's with regard to stabilized criteria..

[FullWings]

I've seen plenty of false LOC indications but not very many for the GS.

Our SOPs don't restrict whether you capture from above or below.

I'd have thought being on a 6 or 9deg glideslope would make the subsequent approach 'interesting' in terms of stability, unless your aircraft type was designed for it, like those that go into LCY (5.5deg).

As other posters have said, there are many ways to check the authenticity of a GS (RoD, height vs. distance, etc.). It doesn't strike me as particularly dangerous (unlike a false LOC), as it will be difficult to follow and actually give you increased terrain clearance. Probably not a good idea for sub-CAT-I Ops, though...

[plain-plane]

false glides...
I have only seen them briefly when overflying the airport, in a very slow old turbo prop...

I have never been too worried about false glides, my descent planning take me right to the point where i intercept the correct G/S.

Also I would expect of most of my fellow pilots that before arming the autopilot/ FD system, they pause for one second and think are the parameters looking good... and then monitor the descent rate following the G/S.

Where I fly most G/S are normally only certified to 10 NM, which makes the math very simple... for your standard 3 DEG G/S expect to be just 3000 above the airport...


So yes I have intercepted "many" glides from above, and are "happy" to do...

[A Squared]

Cold Bay has a back course LOC approach so he may have read the course bar incorrectly?? Dunno

It would be hard to imagine how a simple left/right confusion CDI confusion would place the airplane at the location and altitude of the crash site. I linked the NTSB report in my previous post. If you're interested take a look at the map in the appendix. It shows the crash location and the aircraft's path to it and also shows the localizer BC course.

[MungoP]

Capturing a false glide slope is not that rare once away from the relatively low lying ground most pilots are accustomed to. A classic example is Kabul... it's not uncommon to be vectored in from the very high MSAs surrounding KBL and given a late turn to intercept the localiser and a descent that will take you down onto the glide-slope.. (pull & roll) you need to be calculating your height above the a/f (5,800ft apx) so 16 miles out times 3 = 4,800 + 5,800 = 10,600 ft if you're on the correct lobe.. not that tricky unless you find yourself getting kicked around in a night-time blizzard, struggling to hear the checklist over an indecipherable ATC instruction and partial radio failure due to ground jamming of cell-phone signals.. if it's then end of a long 6 sector day then it can become quite challenging.

[thermostat]

In today's world, ATC for whatever reason will from time to time keep you high till you find yourself above the GS. The Airbus FCOM directs the pilot to wind the alt alert above the aircraft altitude, (to prevent a level-off at ALT*), select 1500 fpm down, arm the ILS button and capture the slope. This works well if your ground speed is 140 kts.
Since the groundspeed and rate of descent are linked and affects each other, a good method to determine the rate down is to note the ground speed then add a one to the first two digits. e.g. GS 150, ROD 1600 FPM. GS 160, ROD 1700 FPM. GS 170, ROD 1800 fpm, NOT to exceed 2,000 fpm. The above table represents a 6 degree slope (a 3 degree angle to capture a 3 degree GS). It works very well indeed and takes the terror out of the exercise. Try it, you'll like it.
p.s. When capturing from above with the alt alert above the aircraft altitude, make darn sure the ILS / GS button is active (on), if not you'll fly through the GS. If you check the front of the Jepp approach manual you'll find the table of GS/angles/rate of descent. Research is king.

[Ollie Onion]

Its no problem although not ideal to intercept from above, happens every now and then. Those occassions are probably a 50/50 split between ATC induced and Pilot induced. I find it best to just correct as early as possible and I just don't mess around. I arm the LOC and GS, lower the gear, select a V/S 0f -1800 fpm and get myself onto the GS as soon as possible. The times I have seen it become a total mess is were there is only a half ass effort to capture with the pilot then just continuing all the way down high on the slope just within stable limits, I have never understood why you don't just correct early and be done with it.

[nitpicker330]

Yes Ollie wise words indeed, some Pilots leave it way too late to correct safely.

Its not something we aim for day today but we should all know the correct way to recover back down to the GS if needed. Hence my company gives the required info in out FCTM. As they do with Rejected Landings and Sidestep to parallel runways on approach. :ok: And we practice in the Sim where things like this should be trained. :ok:

[nitpicker330]

Hey mods I just realized there are 2 basically identical threads running in here at the same time!!! Can you merge them please?

[FullWings]

If you're doing a barometric RNAV intermediate approach to an ILS, there are often times when you will *have* to intercept the GS from above, due to temperature error, even though the altitude constraints at the intercept point are theoretically on the glideslope.

A good example of this is the RNAV transition to RW15 at GIG (SBGL).

In fact, any hard altitude restriction nominally coincident with an ILS GS will produce this scenario in >ISA temps. We tend not to notice because we are usually intercepting from below and GS capture takes precedence over ALT hold and FMC constraints, so the descent will start before you actually get to the waypoint/distance on the chart.

[clunk1001]

Hey mods I just realized there are 2 basically identical threads running in here at the same time!!! Can you merge them please?

No, THIS is the real thread, the other one is Echo deconvolution ;)

[Capn Bloggs]

To be at 6,000' some 20 NM away, start flaps, and be 3,000' some 6 NM away.

I'm surprised you made it to your rocking chair following a profile like that! :)

9k at 30nm, 5k at 20nm, 3k at 10nm is good for me/GS.

[blind pew]

Captured and flew the 6 degree glideslope during conversion onto DC9 51.
FD app, mod turbulence and icing on to 05 at GVA.
ROD wasn't always the clue as the alps are known for odd winds and shear -once lost 25 knots at 1200 ft on departure (with huge temp inversion), Heavy and rising terrain at night - not nice.
Outer marker check got it but initially thought it was an altimeter setting problem as flying QNH but had grown up on QFE.
Closed throttles, ordered FD off and intercepted the correct one.
broke cloud around 500ft.
Might sound hairy for the times but the company I was entering had abilities that did exist in my previous employer.(all of the training guys were ground attack/ fighter jet jockeys).