Localizer intercept after the glide path...what to do?

Hello everyone, I have been flying for almost 150hrs now on VatSim and I usually manage to fly from A to B without major embarrassments. However, sometimes a controller will direct me to the localizer very late, or give me a heading that intercepts to localizer just after the top of the glide path and I always get confused. Should I start a descent before capturing the localizer in that case (and if yes what is the best way to do it – change to V/S mode, -800ft/min or so?) or is it better to just go around and try again?


Generally speaking, controllers should aim for you intercepting the procedure (in most cases that will be the final approach course) in such a way that you will have at least 1 NM of straight and level flight for precision approaches and 2 NM of straight and level flight for non-precision approaches before you intercept the descent profile when they vector you. So if you are turned in such a way that you intercept closer to the vertical profile than that, the controller messed up - but you as the PIC can decide if it’s still fine or if you need to abort the approach attempt (just tell ATC you’ll intercept too close to the descent profile and they’ll likely just turn you around for another attempt - if you maintain a good situational awareness, you might even realize it early enough to allow ATC to adjust your heading accordingly for you to still make the intercept at an appropriate distance from the descent profile). However, you are not allowed to descend below the cleared level until you are established on the lateral part of the procedure (and then only to the next charted minimum level on the procedure or along the procedure’s glideslope/-path), so do not start your descend before you are established - in your example - on the localizer (someone doing that almost caused me to fail my S3 CPT - luckily the examiners agreed that it was a pilot error and overlooked the resulting loss of separation with an aircraft on the parallel approach :sweat_smile:).

As for the late turns: when you are on a base heading, you should be prepared to receive a turn to an intercept heading which can also come relatively late, especially if it’s busier. The best advise here is to start dialing in your heading while receiving the instruction (so when you hear “callsign, turn left…” you already start turning the heading knob to the left). In my experience, what causes a lot of people to overshoot, is that they listen to the instruction, read it back, and only then start to turn their heading knob, which often means that they only initiate the turn around 15 seconds after the original instruction. But if you start turning quickly and it’s still too close, it’s probably ATC who messed up.

Thanks a bunch David, that helps me a lot! So basically, what I understand is that I can try to help the controller by turning as early as possible, maybe also decrease my speed to 180 or less before the turn to final to give him/her and myself more time? But I should not descend before establishing on the localizer, or try to catch up with the descent profile by descending faster. I could however warn the controller when I expect to come in too close, in order for him to correct my heading and prevent a go-around.
I’m certainly guilty of writing down, then reading back and only then executing instructions because I’m afraid of forgetting what I’m supposed to do if I don’t. But I’ll try to turn that around, at least for the turn to final. Thanks again!

Yes (but obviously not before you actually get the instruction to do so).

If you are not under speed control, that can help, yes. However, when you are under speed control, it is very likely that you will kept at a higher speed (as an example: here in Germany, controllers should not instruct below 200 knots before 12 NM final unless necessary for separation and it is very usual to keep people at at least 220 knots until they are established on the approach procedure), so this shouldn’t be the only tool for you to rely on.
Also, controllers should know your turn radius for a given speed (rule of thumb: GS/100=turn radius in NM), so unless they mess up, there should be enough space for you to complete the turn if you execute it promptly. Conversely, reducing your speed too much in this situation might result in your turn radius decreasing so much that you it becomes the cause for you ending up with less than the minimum straight and level segment.

You should not descend before establishing on the lateral part of the approach procedure, correct. But unless ATC specifically tells you to maintain an altitude until you intercept the vertical profile, once you are established on the lateral profile, you can descend with the procedure (charts will tell you minimum altitudes for each leg).

You can increase your rate of descent to catch up with the vertical profile, that’s perfectly permissible as long as you as the PIC deem it to be safe. In fact, ATC may even ask you if you are able to intercept the glide slope from above if they vectored you in too tightly, if you overshot and ATC wants to know if they can turn you back immediately even though you would end up above the normal profile, or if you didn’t descend as fast as ATC expected.

Yes, if you feel that it will be too tight, let ATC know so they (together with you) can figure out a solution. However, if you need more than the 1 NM straight and level (or 2 NM if it’s a non-precision approach), you’ll have to let ATC know well in advance and not when you’re already on an intercept heading.

Perfectly reasonable! Just keep in mind that many ATC instructions (particularly in a terminal environment or on the ground at an airport) can be really time-critical and in general, on busy frequencies (of which we have quite a lot on VATSIM as controllers usually cover much more airspace alone than any controller ever would IRL), ATC needs prompt readbacks because they usually have a couple other instructions already lined up that they need to give to other aircraft, so promptly reading back and complying with instructions can be very important.
I personally have a hardware MCP which allows me to turn a physical heading knob while reading back an instruction, but that can obviously be tricky with the default HOTAS setup many simmers have where you need to turn the heading knob with the mouse. Something you can do here would be to have the mouse already on the heading knob in anticipation of the turn, start turning in the instructed direction while ATC is giving the instruction, then once ATC has finished talking, immediately read back, and only after that finish turning the heading knob (unless you can get it set up completely while ATC is still giving the instruction, of course) - that way you already start the turn and continue turning while reading back.

Excellent and thanks for that last tip, I will certainly try that the next time! Now I don’t want to drag on about this subject but there was one instance where I was flying on the downwind leg and the controller asked me to do a ~150 degree U turn to capture the localizer. In this case I did manage to respond quickly and again ended up after the beginning of vertical profile.
Later I thought: maybe I should have waited just a little bit, say half a mile or so, before beginning the turn in order to give myself some more room. Is that allowed? Or should I ask the controller to extend the downwind leg? But then I can never execute the turn in time anyway.

The two most common turns you can expect from downwind is either one 150 degree turn to directly put you onto the intercept heading (you need to be a fair bit past abeam the FAP/FAF, usually 2-3 NM after it, for this turn to work) or one 90 degree turn onto a base heading followed by another 60 degree turn onto the intercept heading. The reason here is that controllers should aim for a 30 degree (or less) difference between the intercept heading and the approach course (maximum allowable difference is 45 degrees). Just throwing that in here because that knowledge can help you anticipate what exact heading you might get next.

In any case: if ATC gives you that 150 degree turn and you don’t drastically change your speed throughout the turn, but still end up high on profile afterwards, most likely ATC messed up.
Now, the problem on VATSIM is that you have a stark difference in pilots. Some start their turn immediately while you’re still giving the instruction, others may take a full minute or even more to start the turn (unlike IRL where you can generally assume that instructed turns will be initiated promptly, especially in a terminal environment), and while you can’t predict a pilot’s reaction time, the vast majority falls within the group that takes a bit longer to react, so some controllers may simply have gotten used to it and issue the turn a tad early in anticipation of the pilot taking at least 15-20 seconds to start the turn, so if you react promptly, it may put you high on profile - but again, that’s ATC’s fault, they should never give an instruction when immediate compliance would cause problems (on VATSIM that means finding a middle ground where the pilots who take a bit to react can still reasonably intercept the lateral part of the procedure while pilots who react quickly don’t end up high on profile).
In that regard, good energy management can also be very helpful - you constantly see pilots barrelling down their base heading with 250 KIAS or even more, which is of course fine if they are not under speed control but it also gives them less room to react. Slowing down to 200-220 knots shortly before you expect the base/intercept heading (if you are on a downwind, e.g., you’d want to be at that speed roughly when passing abeam the FAP/FAF) will give you a bit more time and since you are not changing your speed during the turn, ATC will still be able to accurately anticipate your turn radius.

Great stuff and thanks again David, it makes a lot things clear for me. Speed management is something that I have only recently payed more attention to. The thing is that the A20N by FBW, which I usually fly, will automatically activate approach phase and reduce speed to ~210 knot quite late in the approach, I think mostly 12 miles out or so. Which is probably realistic and very efficient in real life but I guess on Vatsim, especially on busy approaches it makes sense to reduce speed a couple of track miles earlier on many occasions. Also good to know that speed 180 is generally considered too slow unless on final approach.

I always fly with the navigraph Approach Graph that I am planning on a separate screen. This graph shows the planned altitude to find the glide path by waypoint, and in my case also shows the position of my plane. When I notice ATC will send me in late I start an additional descent of a few 100 ft, depending on which point I am supposed to reach the GS. This is aligned with ATC instruction that you are allowed to go for the ILS. Speed is another story - I recently failed a landing because ATC did not instruct free speed and I was way too fast on the final. Since, I proactively request free speed when needed (with an Airbus 320 210 max when entering GS, reducing speed as needed - I slow to 180 at 6DME and continue to slow as I add flaps).

Right, that sounds like a good option, and one that will probably work out fine on Vatsim, but in real life that would only be allowed once your established on the localizer if I understand David correctly.

When ATC tells you you are cleared to the ILS approach I interpret that you can make a last-minute (seconds actually) adaptation to the altitude if you are risking to end up above the GS. I bet this happens in real life also. This small change should never get you close to another flight unless ATC brings you too close.
What many pilots do not do is follow their approach in detail hence seeing when they are too high happens when they missed already. One controller told me that the instruction to go for the ILS assumes you are doing your best to get in the GS.
Mind you - my statements only hold for very small corrections when close to entering the GS. We are not talking about changes that would create a risk for other flights.

This is wrong. You are never allowed to descend below the cleared altitude until you are established on the lateral part of the approach procedure - even 100ft can result in a loss of separation. And yes, this small change would regularly get you too close to other traffic, particularly at airports with parallel approaches where procedural separation won’t exist until both aircraft are established on the lateral part of the approach procedure and separation before is almost always maintained through level separation.

As I said before in this thread: if you would intercept the localizer behind the FAP (or at whatever point you would intercept the glideslope if ATC clears you onto the ILS at a higher level than the platform altitude), ATC messed up big time. But the PIC also has a responsibility to their flight, so they need to keep an eye on what’s happening and request a different heading early on or go around.
If you intercept early enough for the lateral part, but are still too high at that point then it depends: did you descend at a reasonable and predictable rate or did you suddenly slow your rate of descent drastically for some reason? If the former is the case: ATC messed up. If the latter is the case: the pilot messed up.
In that regard

is definitely true, but only within reason. If you’d need to descend at 4000fpm to still make the vertical profile, that’s an unreasonable expectation by ATC and you should reject the clearance outright when you get it - just tell them “unable, we need more track miles” (and already have a sufficient number in mind in case ATC wants to know how many exactly).

David, TX for the follow-up. I have asked a UK-based professional controller (I will keep the controller anonymous). Feedback is: If controller clears you for approach e.g., “turn left heading 290, cleared for the ILS approach”, the PIC could descend to reach the GS. If controller says “turn left heading 290, report established on the localizer”, PIC should maintain level. Sounds logical to me - what do you think?

Okke, did you read my reaction to David about what a real-life controller said? A pity David did not respond to that comment, maybe it went unnoticed?

So, I’ve looked a bit further into this and it seems like there is no singular definitive ICAO doc that precisely answers the question or anything. However, Eurocontrol’s SKYbrary has a pretty concise answer in this article (which suggests that there simply is no precise rule, but a relatively clear answer can be inferred from various other rules).
Essentially: when not on a published procedure (i.e. when on radar vectors, as in OP’s example), you don’t make any altitude changes unless ATC tells you to. As soon as you are on the published procedure (in this example, the localizer) again, you can descend according to the procedure - if you feel like you’ll end up too high that way, ask ATC for a descend. In some countries the mitigation for potential confusion on the ATC side is to either explicitly include an instruction to maintain the last assigned altitude with the approach clearance or that controllers not actually give the approach clearance until the aircraft is established on the lateral part, but these are more like country-specific deviations to the standard than something to always rely on regardless of where you are as a pilot.

Likewise, I learned it the way I described above from multiple real world controllers here in Germany and the SKYbrary article seems to confirm that way. It’s also really in the best interest of the pilots and all other people on an aircraft (take a look at this CFIT, e.g.) to not change the level without explicit authorization when not on a published procedure.

Great follow-up. I was unsure if it passed.

I have another opinion - according to FAA regulations the terminology “cleared for the ILS RWAYXX approach” is clearance for a published procedure. This includes the adjustments to capture the GS.

Out of the last 100 flights controllers managed to get me on the localizer without the need for any deviation, except in one case when I was too high and had to steer down to the GS when in the LOC. The controller did not notice this adjustment and frankly, I do not believe the controller can see it (GS versus slightly above GS).

I agree it is always better to ask the controller if you notice that you are too high when reaching the LOC. If it is very busy that may not work as you will have just a few 10s of seconds to react.

Finally, I noticed several times over the last weeks that controller-reported QNH and both MSFS and metar-taf.com QNHs differ. Just now by 2hPa, so approximately 60 ft off. This may explain why the GS is missed on some (rare) occasions.

And this is usually the way it’s done and the way it should be, but sometimes it’s not feasible due to either terrain in the direction you are coming from (and thus a higher MVA which necessitates a higher intercept altitude when guided to the approach via radar vectors) or due to traffic below (e.g. on an airport with parallel runways where traffic being guided to the adjacent runway may be passing below). There are also cases where airspace design makes it undesirable as controllers are trained to keep their IFR traffic within protected airspace (i.e. classes D, C, B, and A - all the airspace classes where all aircraft require clearance and ATC knows and controls everything that’s going on) wherever the structure allows for it, which - at least here in Europe - is the case for almost all bigger airports; particularly when it’s busier and the final gets longer, the airspace design might mean that aircraft intercepting at the platform altitude would enter class E (or maybe even G) airspace.

Although 60ft should not be enough to have an impact if controllers follow the rule of 1 NM straight and level (and while not as important on VATSIM, particularly as many pilots don’t even know about it - a controller here recently got positive feedback for bringing a pilot in with a simultaneous localizer and glideslope interception :sweat_smile: - IRL controllers will usually try to make absolutely sure there’s at least 1 NM by aiming for 1.5 or 2 NM as company SOPs sometimes forbid pilots from flying an approach if this segment is missing or too short and I believe some aircraft’s autopilots are also not able to capture the glideslope if it hasn’t fully established on the localizer beforehand).
However, the discrepancy is usually due to the update frequency of the weather information on these services. For example, METARs in Germany are published every half hour, at xx20 and xx50, but it isn’t rare for the xx20 METAR to appear in the controller client at xx40 (and then stay until like xx10). It’s also not updated for all controllers simultaneously, so one may already have the latest weather information while it takes 5 more minutes to update for another. The in-sim pressure value may differ beyond that as the weather engine tries to come up with reasonable weather patterns based on very localized weather information that is potentially dozens of minutes old (particularly in places like the US, where the METAR is only updated once every hour, or places that only publish METARs during hours of operation); some more advanced weather engines like the one in MSFS will use additional data to construct realistic weather, but meteorology is sadly not an incredibly precise science with an immense amount of variables, so any extrapolation is prone to error, especially if you try to make it as detailed as you need for a flight sim.

David, this is very interesting additional information, giving extra understanding.
Nice, as I am also planning to go for the controller (I passed the S1 theoretical exam), alas the waiting list for training is taking time (9 months).
One extra point for Otte, and unsure if controllers are always aware: The speed restrictions on the published approaches are there for a reason i.e. to allow the correct turns. Therefore, as a pilot, you should always study the published approach. I recommend a Navigraph full subscription if as PIC you wish to be well prepared and able to change rapidly to changing circumstances.

In some instances controllers ask speeds that are too high - mostly for separation - but this has its consequences on the turn radius. A few times the controller forgot to allow free speed, resulting in the plane being way too fast to make the turn to the localizer. It is complicated to do everything correctly - both for pilots and controllers.