What is the expected precision of the built-in GPS altitude of the iPad Mini 4. At times I observe a difference of many more than 1000 feet, while GPS lat/lon is accurate. In ForeFlight the GPS accurate indicates 10 feet or better (green font). I am not sure if the accurate is laterally only or 3D. It’s not a ForeFlight issue, since the same wrong altitude is indicated by another GPS app. Apple support was useless. Connecting to GTX-345 via Bluetooth, of course, is accurate. There is nothing else wrong with the iPad that I can see, except for the GPS altitude. The GPS altitude is not always off. There is no consistency. Sometimes it is accurate, sometimes not. I am wondering whether to return it.
Thank you.
What are you comparing it against? GPS apps don’t know the QNH on a day so a GPS altitude will never be the same (interestingly those apps that display a vertical profile will have this error vs any airspace) and I also believe GPS altitude uses a different geo datum so there will also be a slight difference there (calibrator types may know the exact details of that?)
I would try some 3rd party GPS app; something really simple.
1000s of feet difference between a GPS and an altimeter cannot be explained by a QNH or some other atmospheric profile difference, unless you are on top of Everest.
Also, at any airport, the GPS altitude (once geoid corrected) should be within feet of the airport elevation, because that elevation is true AMSL, not related to QNH.
The sort of geoid correction which is IME typical is 150-200ft, in Europe. Could Lucius be located where it is much more? Like many, his profile is totally empty 
I mentioned that GPS altitude is off by more than 1000 feet to exclude QNH as a factor. I also have seen it 1800 feet off. I have seen it decreasing linearly with pressure altitude, and off by 1800 feet upon landing at my home airport, which is at sea level. Turning location services on and off corrected the GPS altitude immidiately. I have used a very simple GPS app to compare against and the error is the same between different GPS apps, so not a Foreflight software bug, but rather the library returning the wrong value. I have updated to the latest iOS version.
So either my GPS chip is broken in an insidious way (the library also returns that location is accurate within 10 feet), or iOS is broken, or both.
Am I having complete brainfade?
Surely the whole point of the QNH system is that vertical points in space will always show the same altitude; that’s what stops us bumping into mountains and masts.
So surely GPS should show the same altitude at an altimeter, provided the QNH is set on the altimeter? There is no need or reason for the GPS to know anything about atmospheric pressure, because that’s not how it works.
Obviously the same cannot be said for flight levels, but a WAAS GPS should be reasonably close to a pressure altimeter (set to the correct QNH) if both are working correctly.
I write this with hesitation, because I know that @Balliol has forgotten more about Altimetry than I have ever known, so I am confident that he is right and I am wrong, but I am missing something somewhere in the discussion.
Timothy wrote:
Obviously the same cannot be said for flight levels, but a WAAS GPS should be reasonably close to a pressure altimeter (set to the correct QNH) if both are working correctly.
On the ground (at the QNH datum), yes. In the air there is the usual issue of a pressure altimeter indicating the wrong altitude if the temperature distribution is different from ISA.
Balliol’s point about the geoid is more interesting. I’ve never really understood all of the geoid business. AFAIU a certified aviation GPS will correct the “raw altitude” to match the WGS84 geoid. Don’t consumer grade GPSs do this?
You can find geoid variation maps online – e.g. here.
It doesn’t anywhere near explain the 1000ft+ error.
IFR GPSs are required to do this; I believe it is implemented by interpolating a look-up table. Consumer units don’t do it AFAIK. I think the better aviation handhelds do it, but I haven’t checked this. WAAS/EGNOS doesn’t help; you still have to do the correction (and anyway EGNOS reception is not that reliable especially on handhelds).
For the UK I always have to adjust the GPS altitude by 160ft when doing my movie subtitles with the GPS position and latitude By the time I get down to Greece it needs a further 50ft change but I have forgotten in which direction… I am using an old SIRF2 GPS for that.
cannot be explained by a QNH or some other atmospheric profile difference, unless you are on top of Everest.
Remember that during winterflying the temperature effect on pressure altitude is 4% per 10° C ISA deviation. In the US it’s presently very cold. Could be ISA minus 20°C – so flying at 10.000ft would result in minus 800ft = 9200ft. GPS altitude would more or less show that difference.
Which is not the fault Lucius reported here, because he finds this discrepancy also after landing at MSL.
( Lucius’ profile says he is at KRNT Renton / Seattle.)
Is your iPad WiFi only or with a SIM card? The reason I ask is that on the iPad the GPS lives on the same chip as the SIM / mobile phone. Therefore, if you – like me – have a WiFi only iPad, you will get the altitude info somehow derived from the WiFi access point you’re using. Once you move away from this, all bets (and altitudes) are off.
That’s a very good point. All these mobile device firms are covertly capturing WIFI MAC numbers and their GPS coordinates, for WIFI assisted GPS and who knows what other purposes, so if you are sitting at home on your WIFI, the device could well be showing a really accurate lat/long, from a previous acquisition. What happens to the altitude is a good question….
