Indicated vs actual altitude

Timbeck2

Final Approach
Joined
Nov 4, 2015
Messages
9,161
Location
Vail, Arizona
Display Name

Display name:
Timbeck2
I have dual G5s, altimeter for the sector I was flying selected and they were giving traffic on me to another aircraft at 7,700' when my indicated altitude was 7,500'. My avionics guy is in Asia right now otherwise I'd ask him. Pitot static check passed with flying colors, no leaks at all.

What could be the problem?
 
Mode C is always in pressure altitude. Your altimeter and the ATC radar had different ideas as to what the altimeter setting should be. The error is just as likely at either end.
 
I’d also evaluate what you show when on the ramp, local alt setting in place. Then how happy ATC is when flying at a known altitude.

The other guy, no FF, or IFR, may have an issue. Who knows what the altimeter is set at.
 
Was the sector (or airport) altimeter reading you used the equivalent of 200ft higher than the standard 29.92? Something like your altimeter set to 29.72. I figured they then correct it back so that the traffic callout is back in altimeter speak for the other plane, not pressure altitude. If your altimeter offset doesn't jive with the 200ft then there is probably something else even more complicated. Otherwise its what Ron said. They reported Pressure Altitude.
 
From my experience, ATC has always worked with pressure altitude. They're the ones giving us the Kollsman setting.
I would welcome any heads up for another aircraft passing within 500' of me, your reported altitude is easily obtained by requesting their altitude reading and your primary altimeter set at 29.92. Any discrepancy should be looked into.
 
My understanding is the same as Bob's. ATC sees the altitude your encoder transmits.

Was the current baro 30.12?
 
What's the source of the baro-corrected altitude, for the encoder?

I'm guessing it's your new G5, so I just looked up the install manual for the G5, and at a quick glance, it looks like it has an RS-232 connector that can provide an output of either pressure altitude or baro corrected altitude; a slightly different configuration setting will give you one vs the other. I mention that just in case the installer goofed in selecting the config setting. (OTOH, I just took a quick look at the manual, and I might have misunderstood.)
 
The altimeter bounced around between 29.94 - 29.89
Not enough to explain a 200ft error. Heck, even if you were on alternate air in the cabin it shouldn't be 200ft...but might be worth checking.
 
Next time you fly, call STC for a transponder check. They will say hey have radar contact and what altitude they show you at. Hey will adjust what you show for non-standard baro setting. See if they show you at the correct altitude. It could be an anomaly, or the controller misspoke. Don’t assume something is wrong with one occurrence like you mention.
 
I have dual G5s, altimeter for the sector I was flying selected and they were giving traffic on me to another aircraft at 7,700' when my indicated altitude was 7,500'. My avionics guy is in Asia right now otherwise I'd ask him. Pitot static check passed with flying colors, no leaks at all.

What could be the problem?

Does your transponder display the pressure altitude in the cockpit?
 
I don't think the govt gets everything right, but I would think there is a safeguard against what you're suggesting. The safety implications are pretty serious for an error on ATC's end. Maybe I'm naive though.
The safeguard is that ATC gives the altimeter setting they believe is right to everybody they are controlling. It actually makes little difference if that is the "right" value or not, just so everybody is using the same reference point.

Further, ATC qualifies the traffic calls unless they have gotten confirmation from the pilot as to what actual altitude he thinks he's at.


Even though Tim says the encoder is new, I'm betting that's where the trouble is. From the tenor of Tim's post, this is a new discrepancy. What has changed? The encoder has.
If he maintains his plane for IFR, the encoder and altimeter should have been checked at the time it was installed.

I'd never EVER take one report from ATC about my transponder as an indication that I have a problem. And 125 feet is the allowable error on the aircraft side (note that the encoder only sends alittudes in hundreds anyhow), so 200' errors are not unexpected.
 
Can I ask a dumb question (and therefore learn something!). I thought it worked the following way:

1. Pilot periodically adjusts the Kollsman offset based on nearby ATIS/AWOS reports. In theory his altimeter reads altitude above sea level with not too much error. This error is known when the pitot static check is done and the tech checks the altimeter against various (pressure) simulated altitudes.
2. The Kollsman offset DOES NOT effect the transponder's altitude encoder. So the transponder only knows Pressure Altitude (as if it had a Kollsman offset that is fixed to 29.92).
3. When a transponder replies with altitude it sends the pressure altitude as part of its squit/reply. This is also tested when a pitot static check is done.
4. ATC can only receive pressure altitude (Mode C, mode S probably has GPS altitude info?).
5. When ATC reads back your altitude or uses your altitude to report to another plane they adjust the pressure altitude back to msl. If they didn't, during High and Low pressure systems the error could be pretty large.
6. At or above 18,000 everyone just uses 29.92 and therefore pressure altitude.

...is that even close to right?
 
The safeguard is that ATC gives the altimeter setting they believe is right to everybody they are controlling. It actually makes little difference if that is the "right" value or not, just so everybody is using the same reference point.

It seems like it could make a BIG difference on a plane on an instrument approach.
 
LIttle bit of a side track here.... I thought G5’s were not certified as replacement for airspeed or altitude, just replacement for slip/skid or Attitude; and it’s common for a difference between the G5 and primary gages for the non-primary functions.

I’m sure a bunch of guys will now post “mine’s rock-n accurate”.
 
It seems like it could make a BIG difference on a plane on an instrument approach.
Are you instrument rated? The altimeter setting on the approach is often a different number entirely.
 
Are you instrument rated?

Yes, though a long time since I was current. In fact, in another lifetime I was a pretty competent CFII.

The altimeter setting on the approach is often a different number entirely.

Good point. I was wrongly remembering using the altimeter setting provided by Atlanta Center to do the GPS approach into Copperhill, TN. Thinking about it now, when I did that IFR there was no approach into Copperhill, so I was limited to the MVA when looking for the airport in IMC. Now that there is an approach there, it does specify which two altimeter settings are usable.

Thanks for the correction/clarification. When teaching types of altitude, I would point out that other than when on the ground, a pilot rarely knows his or her true altitude, due to non-standard lapse rates, and that true altitude is irrelevant enroute so long as everyone below 18,000’ is flying indicated altitude and everyone above is flying pressure altitude. I concede that and see that’s the point you were making.
 
Last edited:
Yes, the various errors including lapse rate and just what's allowable mechanically in the altimeter are figured in the various minimum altitudes. In a few cases (very cold, very high pressures), the FAA has additional procedures.
 
Is this a certified airplane? If so you still have your steam altimiter? What is the altitude difference between the two?
 
Yes, the various errors including lapse rate and just what's allowable mechanically in the altimeter are figured in the various minimum altitudes. In a few cases (very cold, very high pressures), the FAA has additional procedures.

I'm pretty sure these guidelines have not changed dramatically since we referred to them as 91.171, and 91.172.

I've never heard of a lapse rate.


https://www.law.cornell.edu/cfr/text/14/appendix-E_to_part_43
 
I've never heard of a lapse rate.

I know we taught lapse rate, as relates to temperature and pressure, in Private Pilot ground school.

The idea is that pressure dropping about 1” and temperature 3 1/2 degrees every thousand feet are just averages. Just like your adjusted sea level temperature and pressure are unlikely to be 59°F and 29.92” at any given moment, it’s unlikely that temperature and pressure drop at the average rate with increase in altitude.

Though temperature lapse rate has effects on air mass stability, the main takeaway is that indicated altitude rarely if ever correlates to true altitude, but most times it doesn’t matter, given adequate terrain clearance.

As a minor correction, I said upthread the time a pilot knew his or her true altitude was on the ground. That was generally true, with minor exceptions, prior to the introduction of GPS, which does triangulate and display true altitude very accurately.
 
GPS does NOT display true altitude very accurately, in general. First off, the accuracy in the vertical direction is poor to begin with (compared to horizontal accuracy) based on the geometry of the satellites you can see. Second, the GPS "altitude" Is based on an ellipsoid model that doesn't really pretend to be useful for terrain avoidance unless it's referenced to the actual geoid (which is done for the cases of instrument approaches).
 
GPS does NOT display true altitude very accurately, in general. First off, the accuracy in the vertical direction is poor to begin with (compared to horizontal accuracy) based on the geometry of the satellites you can see. Second, the GPS "altitude" Is based on an ellipsoid model that doesn't really pretend to be useful for terrain avoidance unless it's referenced to the actual geoid (which is done for the cases of instrument approaches).

Thanks for that.

Googling leads me to a vertical accuracy of 10 to 20 meters (35 to 70 feet). It's all relative, but I'd call that at least "fairly accurate", but perhaps not "very accurate". Not accurate enough for approaches, perhaps, but still a lot better than we ever had before GPS.
 
It's not 70 feet. It's closer to 100 feet (non-WAAS) from the ellipsoid reference. There can be as much as 400 feet error to the actual elevation from that. Maybe 500 feet is "very accurate" to you. It isn't to me.

Even when flying with WAAS with a proper geoid reference, you're getting about 8 meters. That's about twice what is acceptable for a precision approach.
 
It's not 70 feet. It's closer to 100 feet (non-WAAS) from the ellipsoid reference. There can be as much as 400 feet error to the actual elevation from that. Maybe 500 feet is "very accurate" to you. It isn't to me.

Even when flying with WAAS with a proper geoid reference, you're getting about 8 meters. That's about twice what is acceptable for a precision approach.

WAAS GPS must be within 25 feet 95% of the time. According to Garmin it is more typically less than 3 meters (~ 10 feet). If you live near a center facility presumably it would be even more accurate as I believe this is where WAAS antennas usually are.

Since this is within the height of the airplane I'd say that's about as accurate as we can practically make use of without trimming trees.
 
The "WAAS" antennas are 22,500 miles up in geosynchronous orbit. But we weren't necessarily talking about WAAS as I explicitly stated.
Further, WAAS does *NOT* fix the other errors I described. Unless you're on a surveyed approach, you don't actually know the relationship of the GPS altitude to height above the geoid.
 
The "WAAS" antennas are 22,500 miles up in geosynchronous orbit. But we weren't necessarily talking about WAAS as I explicitly stated.
Further, WAAS does *NOT* fix the other errors I described. Unless you're on a surveyed approach, you don't actually know the relationship of the GPS altitude to height above the geoid.

I wasn't clear on my context- I agree altitude on a non-WAAS GPS is not that accurate as it relates to the original question.

I didn't realize that apart from approaches, WAAS navigators are not actually required correct to MSL. I had read somewhere that all save the early ones do anyways though but I can't find a reference (maybe because that is wrong).

It looks like in North America a WAAS GPS will usually read about 50-100 feet high if they do not correct. I used the calculator below to figure Denver's geoid vs WGS84 difference of -55 ft and in Denver my WAAS GPS should tell me the height of the antenna to within about 10 feet (or maybe better, we're just south of the Longmont reference station). Very neat. I'll have to pay attention to this and see what it ends up being in practice on the runway.

https://www.unavco.org/software/geodetic-utilities/geoid-height-calculator/geoid-height

Also the WAAS antennas are on the ground, if I'm understanding your terminology correctly. The correction data from them is uplinked to the geosynchronous payloads and then sent back down to our units. I fear I am being pedantic...
 
Last edited:
The "WAAS" antennas are 22,500 miles up in geosynchronous orbit. But we weren't necessarily talking about WAAS as I explicitly stated.
Further, WAAS does *NOT* fix the other errors I described. Unless you're on a surveyed approach, you don't actually know the relationship of the GPS altitude to height above the geoid.

Ron, I don't know where you get your data from, but the vertical accuracy for October 2019, the worst monitoring station site was an accuracy of 5.18 meters. WAAS has even better performance. You can read the https://www.nstb.tc.faa.gov/reports/2019_Q2_SPS_PAN106_v1.0.pdf for the most recent report on the standard non WAAS service and https://www.nstb.tc.faa.gov/reports/waaspan69.pdf for the most recent WAAS report. Also here are the most recent report card on non WAAS GPS https://www.nstb.tc.faa.gov/reports/ReportCards/2019 10.pdf .

Geometric Altitude is a GPS altitude referenced with respect to the WGS84 spheroid, but generally when GPS altitude is displayed by GPS navigators, it corrected for the local geoid using a database so is much closer to a true altitude. ADS-B Out transmits the geometric altitude and if you want to avoid hitting an obstacle, you really want to use GPS or geometric altitude rather than MSL or a baro corrected altitude. AFAIK terrain and obstacle databases use values referenced to the WGS84 ellipsoid. WAAS provides HPL and VPL limit alerts to pilots and most navigators will display HFOM and VFOM accuracy values, so pilots have a means of judging the accuracy of the vertical at a point in time.
 
I may be wrong on this, but depending upon the altitude encoder, don't they transmit the altitude in "steps," rather than being able to transmist ANY integer? For instance,the test spec on an AR850 is +/- 125 feet. The display of pressure altitude on my ESG is always rounded to the nearest 100 feet, IIRC. Could that be part of the scenario Tim experienced?
 
The

The Garmin GTN and GNS display GSL on the terrain pages or as a parameter on the Map view. These are WGS84 corrected to the local geoid.

.Garmin GSL.jpg
 
The "WAAS" antennas are 22,500 miles up in geosynchronous orbit. But we weren't necessarily talking about WAAS as I explicitly stated.
Further, WAAS does *NOT* fix the other errors I described. Unless you're on a surveyed approach, you don't actually know the relationship of the GPS altitude to height above the geoid.

The antennas are on the earth's surface for WAAS at the ground station locations. They are surveyed to very precise locations in terms of altitude and latitude-longitude. They provide measurements that are used to correct the values provided by the GPS positions. These corrections are transmitted via up link stations to the geostationary satellites used by WAAS and the height of the WAAS satellite is not a factor other than it may also have a GPS position of its own that is added to the mix of other GPS satellites. The WAAS satellites also transmit a GPS position that may be used as part of the overall GPS receiver vertical and lateral position determination. I don't agree with your last statement, at least for all the certified GPS systems I am aware of. As far as I can tell, the only value that is provided as referenced to the WGS84 Spheroid is the geometric altitude broadcast by ADS-B Out.
 
The

The Garmin GTN and GNS display GSL on the terrain pages or as a parameter on the Map view. These are WGS84 corrected to the local geoid.

.View attachment 79605


Awesome, I could not find that reference despite searching through the manual. Where is that located?

So what is the altitude the controller is seeing on his screen with an ADS-B aircraft? Is the primary reason to use the altimeter setting given by the controller then to ensure everyone is flying the same altitudes relative to other aircraft (that is, with the same error)?
 
John M, just search the PDF of the Pilot Guide for the term GSL.

Controllers see the MSL altitude based on the baro setting corrected pressure altitude.
 
I may be wrong on this, but depending upon the altitude encoder, don't they transmit the altitude in "steps," rather than being able to transmist ANY integer? For instance,the test spec on an AR850 is +/- 125 feet. The display of pressure altitude on my ESG is always rounded to the nearest 100 feet, IIRC. Could that be part of the scenario Tim experienced?

A parallel encoder only provides altitude in 100 foot increments, a serial encoder can provide greater precision, like 10 feet, but the reported altitude is rounded based on the type of system broadcasting it. Mode A/C transponders report the altitude to the nearest 100 feet. Mode S transponders and ADS-B Out report the altitude to the nearest 25 feet if the encoder provides for that precision or less.
 
Can I ask a dumb question (and therefore learn something!). I thought it worked the following way:

1. Pilot periodically adjusts the Kollsman offset based on nearby ATIS/AWOS reports. In theory his altimeter reads altitude above sea level with not too much error. This error is known when the pitot static check is done and the tech checks the altimeter against various (pressure) simulated altitudes.
If you're flying VFR outside controlled airspace, yes, that's correct.

If you're flying IFR or flying VFR in class D or higher (or getting VFR flight following, at least in Canada), then ATC gives you a standard transponder setting for the control zone or ATC sector on first contact. You need to keep that and not manually change your setting by tuning into nearby ATIS/AWOS. That way, ATC knows that everyone they're talking to is using the same altimeter setting (as far as traffic separation goes it doesn't matter if it's wrong, as long as it's wrong the same way for everyone). They'll give you periodic altimeter updates, often when you cross into another sector.
 
Back
Top