Glideslope intercept on an LNAV+V approach

Precision Final Approach Fix. You know it as glidepath or glideslope intercept, but we don't have enough acronyms ;)
It's changed under new TERPs (8260.3C). It is now the Precise Final Approach Fix and applies to the beginning of the final approach for all IAPs, except for those that don't have a fix marking the beginning of the final approach segment. (On airport VOR and NDB without FAF.)
 
It's changed under new TERPs (8260.3C). It is now the Precise Final Approach Fix and applies to the beginning of the final approach for all IAPs, except for those that don't have a fix marking the beginning of the final approach segment. (On airport VOR and NDB without FAF.)
...as opposed to all the imprecise final approach fixes we have come to know and love :D
 
The altitude busts at LAX had to do with a failure to comply with published crossing restrictions -- not altimetry error. Insofar as the LNAV+V "advisory glidepath" is concerned, I have not seen a case where it takes an aircraft below the non-precision FAF crossing altitude in hot weather. I will be flying this approach with that equipment later this week and will follow up on what I observe.
 
Precision Final Approach Fix. You know it as glidepath or glideslope intercept, but we don't have enough acronyms ;)
Thats what I figured. Just hadn't seen it as "PFAF" before.

EDIT: I just realized I have seen "PFAF" in print before. I forgot
 
Last edited:
The altitude busts at LAX had to do with a failure to comply with published crossing restrictions -- not altimetry error. Insofar as the LNAV+V "advisory glidepath" is concerned, I have not seen a case where it takes an aircraft below the non-precision FAF crossing altitude in hot weather. I will be flying this approach with that equipment later this week and will follow up on what I observe.
Yeah. There was an Advisory Circular that came out back then "clarifying" that you have to comply with the minimum altitudes at fixes on the final approach course beyond the FAF. It specifically said that, when its hot out, you will go through the glide path and have to get back down to it from above. I think that was around the same time some of the Notes to AIM 5-4-5 b. 4. got added. If its pretty hot out your altimeter should be telling you that you are lower than expected at the FAF if you go over it on an ILS glideslope. If the +V "pseudo" glidpath is derived from pressure then things should look normal, the +V glidepath will be artificially "steepened." If it's derived from WAAS then you should get the same effect as flying a glideslope.
 
The altitude busts at LAX had to do with a failure to comply with published crossing restrictions -- not altimetry error. Insofar as the LNAV+V "advisory glidepath" is concerned, I have not seen a case where it takes an aircraft below the non-precision FAF crossing altitude in hot weather. I will be flying this approach with that equipment later this week and will follow up on what I observe.

While yes, the busts were because the pilots were not complying with the published altitudes, that's not the whole story. The reason they weren't complying is because they were following the glideslope down, outside the FAF, in hot weather. Meaning their indicated altitude was below the published altitude, on glideslope, which is exactly the situation being discussed.

If the Avidyne IFD is using some temperature compensation algorithm to avoid this problem, I would be very interested.

However, if not, the effect should be easy to replicate and can be done by answering one question as you fly - if you are exactly at glidepath intercept altitude on a GPS approach, do you intercept glidepath right at the FAF? On a warm day, I have found this not to be the case - rather, intercept occurs a bit prior - 1/4 mile maybe. Note that there is a lot of "how precise is the pilot flying" involved here. If the pilot is 50-100 feet off of altitude, it could easily mask this effect.

Yeah. There was an Advisory Circular that came out back then "clarifying" that you have to comply with the minimum altitudes at fixes on the final approach course beyond the FAF. It specifically said that, when its hot out, you will go through the glide path and have to get back down to it from above. I think that was around the same time some of the Notes to AIM 5-4-5 b. 4. got added. If its pretty hot out your altimeter should be telling you that you are lower than expected at the FAF if you go over it on an ILS glideslope. If the +V "pseudo" glidpath is derived from pressure then things should look normal, the +V glidepath will be artificially "steepened." If it's derived from WAAS then you should get the same effect as flying a glideslope.

Exactly the effect I'm seeing. The +V glidepath is generated by WAAS, at least on the popular Garmin GPS units. In that respect, it's identical to an LPV and an ILS as far as lack of temperature compensation goes.
 
While yes, the busts were because the pilots were not complying with the published altitudes, that's not the whole story. The reason they weren't complying is because they were following the glideslope down, outside the FAF, in hot weather. Meaning their indicated altitude was below the published altitude, on glideslope, which is exactly the situation being discussed.

I'm confused here. How is glideslope affected by hot temperature? Your altimeter is set to local setting to compensate for it, and glideslope is just a beam from the ground?
 
I'm confused here. How is glideslope affected by hot temperature? Your altimeter is set to local setting to compensate for it, and glideslope is just a beam from the ground?
The glideslope is NOT affected by the temperature. That's the point. The airplanes indicated altitude is affected by the temperature.
 
The glideslope is NOT affected by the temperature. That's the point. The airplanes indicated altitude is affected by the temperature.

But isn't that why we get the local altimeter setting before we commence our approach...?
 
But isn't that why we get the local altimeter setting before we commence our approach...?
The further you get from the local altimeter setting (altitude), the greater the error.
 
The further you get from the local altimeter setting (altitude), the greater the error.

I thought that's why some approaches are NA or "add x ft to y" if local altimeter setting cannot be obtained?
 
But isn't that why we get the local altimeter setting before we commence our approach...?
The altimeter setting is corrected for temperature only where it is read and at the same elevation as the airplane. They were probably using the KLAX altimeter setting.
 
The reason they weren't complying is because they were following the glideslope down, outside the FAF, in hot weather. Meaning their indicated altitude was below the published altitude, on glideslope, which is exactly the situation being discussed.
We're on the same page here. I contend, though, that at most airports, there will not be crossing altitudes outside the precision FAF, and descending on the glide path is not a problem in those cases. Of course, one cannot descend below a published crossing altitude in any case, and the altitude at the published precision final approach fix should always be checked to ensure the false glideslope has not been captured. I don't believe it is necessary for aircraft to descend to the published glideslope intercept altitude if, for example, they are on a 20-mile final tracking the glideslope and there are no crossing altitudes along the way to the final approach fix.
 
The altimeter setting is corrected for temperature only where it is read and at the same elevation as the airplane. They were probably using the KLAX altimeter setting.

I would imagine there needs to be pretty severe weather with some drastic inversion layers if that becomes a problem when flying an instrument approach. You're not that far from the "reference point" after all?

Thinking about it, LAX approach over KONT would likely be a great candidate for such a drastic delta between the reference point and current temperature in a short distance.
 
But isn't that why we get the local altimeter setting before we commence our approach...?
Yes of course. What is happenning is this. Its a standard day. Altimeter is 29.92. Thats because sea level pressure is 29.92. You are at sea level. If you take the altimeter and take it up 1000 ft it feels a pressure of 28.92 so tells you your altitude is 1000 feet. When it is cold out, it finds a pressure of 28.92 at a lower "real" altitude, say 900 feet. It still tells you that you are at 1000 feet because thats the pressure it feels. On a hot day it finds a pressure of 28.92 at 1100 feet up. The airplane is at an actual altitude of 1100, but the altimeter says 1000 because thats the pressure it feels
 
Yes of course. What is happenning is this. Its a standard day. Altimeter is 29.92. Thats because sea level pressure is 29.92. You are at sea level. If you take the altimeter and take it up 1000 ft it feels a pressure of 28.92 so tells you your altitude is 1000 feet. When it is cold out, it finds a pressure of 28.92 at a lower "real" altitude, say 900 feet. It still tells you that you are at 1000 feet because thats the pressure it feels. On a hot day it finds a pressure of 28.92 at 1100 feet up. The airplane is at an actual altitude of 1100, but the altimeter says 1000 because thats the pressure it feels

Yes, but it needs pretty dramatic temperature deltas (delta between point of observation and point of reference = your altimeter vs. AWOS station) for it to actually make that big of a difference. Thinking about it a bit, LAX basin is a place where this can happen (today LAX was 28C and ONT was 41C).
 
Yes, but it needs pretty dramatic temperature deltas (delta between point of observation and point of reference = your altimeter vs. AWOS station) for it to actually make that big of a difference. Thinking about it a bit, LAX basin is a place where this can happen (today LAX was 28C and ONT was 41C).
If I'm following this right, it's not the temperature delta between those two. It's the temperature deviation from standard which is the calibration point.
 
Yes, but it needs pretty dramatic temperature deltas (delta between point of observation and point of reference = your altimeter vs. AWOS station) for it to actually make that big of a difference. Thinking about it a bit, LAX basin is a place where this can happen (today LAX was 28C and ONT was 41C).

If I'm following this right, it's not the temperature delta between those two. It's the temperature deviation from standard which is the calibration point.

@Cooter, that's correct. It's not about difference in temperature between locations, it's difference in temperature versus standard. Hotter than ISA, and you're higher than expected. Lower than ISA, lower. The difference between indicated and true altitude is essentially zero sitting on the ground at the airport. As you go higher above the airport, that's where the difference increases drastically. And it IS enough to be substantial - at 5000 feet above the airport elevation it would not be unusual to see a 200 foot difference with just small deviations from ISA. Large deviations cause larger errors.

@mtuomi, I recommend the Pilot's Handbook of Aeronautical Knowledge, page 8-4, for a very brief synopsis of this scenario.
 
Last edited:
If I'm following this right, it's not the temperature delta between those two. It's the temperature deviation from standard which is the calibration point.
Yeah. If planes are flyin close to each other with different altimeter settings there can already be a situation that's putting them closer than 1000 feet. Ain't no thang, it uses up a little of the acceptable margin. They're 950 feet apart instead 1000, big deal. Maybe there's a little calibration error in the altimeters in each airplane. May move them both up or down a bit, may move them farther apart or it may move the up guy down a little bit and the down guy up a little. If they're using different settings, what's the pressure gradient. Is the down guy using the one that pushes him up a bit while the up guy is on the glideslope which doesn't give a damn what the temperature and altimeter is. Now add in the "in hot air you're higher than you think" effect, it's no surprise that bells and whistles were going off every now and then because separation deteriorated to the point of pilot deviations being processed.
 
Last edited:
@Cooter, that's correct. It's not about difference in temperature between locations, it's difference in temperature versus standard. Hotter than ISA, and you're higher than expected. Lower than ISA, lower. The difference between indicated and true altitude is essentially zero sitting on the ground at the airport. As you go higher above the airport, that's where the difference increases drastically. And it IS enough to be substantial - at 5000 feet above the airport elevation it would not be unusual to see a 200 foot difference with just small deviations from ISA. Large deviations cause larger errors.

@mtuomi, I recommend the Pilot's Handbook of Aeronautical Knowledge, page 8-4, for a very brief synopsis of this scenario.

What you mean is difference in ISA lapse rate, not difference from ISA temperature? If it's ISA+20 on the ground and ISA+20 at 5000ft, your altimeter reads correct. The altimeter setting takes care of this? Only if your lapse rate is different than standard will you encounter this error?
 
What you mean is difference in ISA lapse rate, not difference from ISA temperature? If it's ISA+20 on the ground and ISA+20 at 5000ft, your altimeter reads correct. The altimeter setting takes care of this? Only if your lapse rate is different than standard will you encounter this error?

This is one of those things thats tough to visualize and it's counterintuitive. Look at that chart posted above, that's the only way I can "see" it. But to answer your question, no. It's anytime the temp is largely deviated from standard. I think of it as the altimeter being calibrated for a certain temp, and therefore, indirectly, a lapse rate. Like someone else said above, the air column is compressed in cold weather. Thats why there are cold weather corrections whenever you're landing below 0°C (or whatever it is).
 
This is one of those things thats tough to visualize and it's counterintuitive. Look at that chart posted above, that's the only way I can "see" it. But to answer your question, no. It's anytime the temp is largely deviated from standard. I think of it as the altimeter being calibrated for a certain temp, and therefore, indirectly, a lapse rate. Like someone else said above, the air column is compressed in cold weather. Thats why there are cold weather corrections whenever you're landing below 0°C (or whatever it is).

Got it now. Roughly 4% per 10C per 1000ft. And the reason to apply that at certain temperatures is that the error becomes larger than approach design allows for.
 
Got it now. Roughly 4% per 10C per 1000ft. And the reason to apply that at certain temperatures is that the error becomes larger than approach design allows for.
That would be about the cold air scenario and separation from rocks. Can you calculate what the altitude error would be at say 8000 with the temperature at 100F?
 
That would be about the cold air scenario and separation from rocks. Can you calculate what the altitude error would be at say 8000 with the temperature at 100F?

That thankfully doesn't happen very often...

Quick off-the-cuff math, around 16% difference.
 
Checkout this other thread here. https://www.pilotsofamerica.com/community/threads/what-were-they-thinking.105560/
Go to the link of the accident. He set his altimeter to read the altitude his GPS said he was at while enroute, not the reported altimeter setting. Then he descends and his altimeter is 300 feet off when he gets down to the field. Good illustration of how the differences are more significant higher up that column of air.

EDIT: GPS is very innacurate vertically. Temperature didn't necessarily cause the error.
 
Last edited:
Checkout this other thread here. https://www.pilotsofamerica.com/community/threads/what-were-they-thinking.105560/
Go to the link of the accident. He set his altimeter to read the altitude his GPS said he was at while enroute, not the reported altimeter setting. Then he descends and his altimeter is 300 feet off when he gets down to the field. Good illustration of how the differences are more significant higher up that column of air.

That's got nothing to do with this issue though, GPS altitude is extremely unreliable without WAAS.
 
We're on the same page here. I contend, though, that at most airports, there will not be crossing altitudes outside the precision FAF, and descending on the glide path is not a problem in those cases. Of course, one cannot descend below a published crossing altitude in any case, and the altitude at the published precision final approach fix should always be checked to ensure the false glideslope has not been captured. I don't believe it is necessary for aircraft to descend to the published glideslope intercept altitude if, for example, they are on a 20-mile final tracking the glideslope and there are no crossing altitudes along the way to the final approach fix.
That 20 miles would be okay provided you are on a published segment of the approach. If not, you shouldn't be given the approach clearance.
 
That 20 miles would be okay provided you are on a published segment of the approach. If not, you shouldn't be given the approach clearance.
Yeah. Most of the approaches to 24/25 at LAX are straight in from about 35 miles out. That, along with the temperature inversions and density of traffic in the LA Basin, it's no wonder this reared its ugly head there.
 
Yeah. Most of the approaches to 24/25 at LAX are straight in from about 35 miles out. That, along with the temperature inversions and density of traffic in the LA Basin, it's no wonder this reared its ugly head there.
Indeed so. But, I was responding to Harold's hypothetical of a less complex ILS with no step-down fixes prior to the PFAF.
 
EDIT: GPS is very innacurate vertically.

Vertical accuracy for raw GPS unaided by WAAS is just not that bad, unless you compare it to lateral GPS accuracy. The 95% error number for the quarter ending in June 2017 is only 4.042 meters or 13.26 feet, which is pretty good except that the lateral accuracy for the same period was 2.751 meters. A baro altimeter has less precision and the value is uncompensated for temperature. So even if the baro altimeter has a precision of 10 feet, it is measuring a pressure and converting it into an altitude based on a standard model of the atmosphere. The atmosphere rarely ever cooperates with the standard model. 2000 feet above a point where the altimeter setting is determined with a 10 degree C temperature above ISA will be 70 feet off, with a precision of 10 feet and an accuracy of 70 feet. The GPS will be more accurate. An altitude encoder will report a hundred foot error to radar via the transponder.
 
The other thought that comes to mind is that Garmin doesn't do the "+V" without WAAS...are those glidepaths barometric or do they use WAAS? If the latter, they're not subject to temperature.

Same question for the L/VNAV glidepaths, for that matter.

Garmin now supports +V outside of the WAAS service volume on the latest version of their GTN systems. There is virtually no risk in using non WAAS for an advisory glidepath, all you are doing is following a suggestion on the rate down to the MDA. The glidepaths are geometric based on GPS lateral and vertical positions. Those in Hawaii can now use the GTN to fly the LNAV+V, even though they are outside of the WAAS vertical guidance service volume. Avidyne supported this before Garmin.

The path based on GPS is not subject to temperature and inside the WAAS service volume, the WAAS vertical path may be used in lieu of Baro-VNAV for the LNAV/VNAV line of minima. The TSO C146 navigators support this. AC 90-105A describes that in US airspace, WAAS may be used for the vertical with LNAV/VNAV. A few LNAV/VNAV approaches will have a note "WAAS VNAV Not Authorized". This is because the vertical integrity at these locations is not good enough to meet the criteria. Example see TJSJ RNAV (GPS) RWY 8 approach in San Juan, Puerto Rico. Otherwise the WAAS vertical glide path is constrained to meet the LNAV/VNAV Vertical CDI requirements.
 
Vertical accuracy for raw GPS unaided by WAAS is just not that bad, unless you compare it to lateral GPS accuracy. The 95% error number for the quarter ending in June 2017 is only 4.042 meters or 13.26 feet, which is pretty good except that the lateral accuracy for the same period was 2.751 meters. A baro altimeter has less precision and the value is uncompensated for temperature. So even if the baro altimeter has a precision of 10 feet, it is measuring a pressure and converting it into an altitude based on a standard model of the atmosphere. The atmosphere rarely ever cooperates with the standard model. 2000 feet above a point where the altimeter setting is determined with a 10 degree C temperature above ISA will be 70 feet off, with a precision of 10 feet and an accuracy of 70 feet. The GPS will be more accurate. An altitude encoder will report a hundred foot error to radar via the transponder.
Ok. So it looks like that guys decision to set his altimeter to read the altitude GPS said he was at while enroute instead of the altimeter 'setting' and finding the altimeter 300 feet off down at the airport was more about temperature than GPS vertical error.
 
Ok. So it looks like that guys decision to set his altimeter to read the altitude GPS said he was at while enroute instead of the altimeter 'setting' and finding the altimeter 300 feet off down at the airport was more about temperature than GPS vertical error.

It's not about GPS accuracy per se, it's about consumer electronics showing whatever they want as GPS altitude. I just walked around this building and it shows +-60ft variations in altitude (and it's flat).
 
It's not about GPS accuracy per se, it's about consumer electronics showing whatever they want as GPS altitude. I just walked around this building and it shows +-60ft variations in altitude (and it's flat).
Yeah. I'm sittin here on the couch and my vertical speed according to ForeFlight on an IPad is 100ft per min. Anyway, we don't know high high he was or what the temperature was, just that he was enroute and it was Minnesota in April. Temperature may or may not have accounted for some of that 300 feet.
 
What you mean is difference in ISA lapse rate, not difference from ISA temperature? If it's ISA+20 on the ground and ISA+20 at 5000ft, your altimeter reads correct. The altimeter setting takes care of this? Only if your lapse rate is different than standard will you encounter this error?
Lapse rate is the change in temperature with altitude. The issue here is the change in pressure with altitude, which is a temperature-dependent thing. The higher the temperature, the more the atmosphere (including the pressure levels) expands, so a given pressure level is found at a higher (true) altitude. That's why you will actually be flying higher than your altimeter indicates.

Altimeters are calibrated at standard conditions, so only on an ISA day does the altimeter read the true altitude when you're well above the altitude of the station where the altimeter setting was taken from. Colder than standard, and the altimeter will read higher than true altitude (you're flying lower than you think); higher than standard, and the altimeter will read lower than true altitude (you're flying higher than you think).

Earlier in the thread someone posted a pictorial from one of the handbooks that shows this very clearly... will try to find the exact post.

Edit: it's post #23.
 
Yeah. I'm sittin here on the couch and my vertical speed according to ForeFlight on an IPad is 100ft per min.
I'm sure I'm not the only one who has had ForeFlight's bitchin' betty start screaming "sink rate! sink rate!" while in level flight.
 
Back
Top