13
$\begingroup$

I am not a pilot and have come to this site to ask a question about the measurement of height. This is as a result of the comments made as a result of a question in the Physics Stack Exchange forum "Does barometer improve gps accuracy?"

I feel that there will be more of you on this site who have real time experience with flying and so can answer my question which is:

Which is more accurate for height measurement: GPS or barometer?

I have noted that those engaged in geocaching often have GPS units with built-in barometers to improve height accuracy.

One of my friends who is a pilot responded as follows:

A light aircraft relies on a barometer that is set to the standard 1013 millibars when in cruise at height (or the regional QNH when lower) but they set to the local reading for the airfield (QFE) when planning their decent. Commercial aircraft also have ground proximity radar but they also have a standard set of old fashioned flight instruments in case the electronics fail. My old aircraft GPS did indeed read out height as well but you were told never to rely on it!

$\endgroup$
  • $\begingroup$ (Not relevant to the actual question, just a note on your friend's comment) I think pretty much only gliders use QFE (Q-codes suck by the way. Why can't non-US aviation learn to use natural language?), iirc it's the setting that makes your altimeter show zero at ground level for the particular airport. It's easy to set when you're on the ground (just twist until it reads zero), and gliders often return to the same airport. Most other aircraft use the QNH (also known as altimeter setting) for the region and simply add the ground elevation to whatever AGL altitude they're shooting for. $\endgroup$ – falstro Jun 9 '16 at 8:52
  • $\begingroup$ @falstro Thanks for your comment. Does it hint at the fact that I should be flying with my friend in his light aercraft? :-) $\endgroup$ – Farcher Jun 9 '16 at 9:17
  • 3
    $\begingroup$ Usually we measure altitude. A barometer provides only a barometric altitude based on pressure so the actual altitude may vary. On the other hand the differential GPS can provide an altitude with a millimeter precision, but it takes time. Altitude/height concept is complex (image taken here) $\endgroup$ – mins Jun 9 '16 at 9:49
  • 11
    $\begingroup$ Your precise height above the ground is not nearly as important as that everyone agrees on the height. Even if one value is more accurate than another, everyone within an airspace should use the same measurement technique. $\endgroup$ – abelenky Jun 9 '16 at 11:56
  • 1
    $\begingroup$ @Lnafziger I have heard of some of the different definitions of height. For me the height that matters is the one that makes sure that the pilot does not unknowingly hit the ground.or any other aircraft. $\endgroup$ – Farcher May 18 '17 at 13:27
7
$\begingroup$

There is a pretty comprehensive description of the issues at this link.

Briefly - the conversion of pressure to altitude makes certain assumptions about the "standard atmosphere" - that is, it works best when the temperature and the lapse rate (change in temperature with altitude) exactly matches that from the model. In the real world, this is almost never the case. For example, on a hot day the air will be less dense, so the weight of a column of air will be smaller - and so when the pressure seems to have dropped by say 200 HPa, you will underestimate your altitude. The example calculation in the link shows that when the surface temperature on a dry day is 42°C, the error at 10,000 feet could be as much as 800 ft - which is considerably bigger than the typical error in a GPS (although GPS is less accurate in the vertical direction, you should routinely be able to get height within about 50 feet when you have an unobstructed view of the sky, and ionospheric activity is not unusually high). In principle it is possible to make corrections for these things - but a simple mechanical pressure-based altimeter probably doesn't.

On the other hand - if everybody else is flying by barometer, you might be "right" about your GPS based altitude, but "wrong" compared to other planes in the vicinity.

So it is advisable to use the same method as everyone else, so you don't crash into each other. On the other hand, when it comes to your final glide, the GPS is more likely to get you home safely on a hot day.

$\endgroup$
7
$\begingroup$

For surveying glacier movement, or a mountain range's growth/decay, GPS is more accurate.

For airplanes: refresh rate, reliability, and simplicity matter more. When it comes down to accuracy, ground augmented GPS provides ILS CAT I accuracy using LPV. So alongside refresh rate, and accuracy, comes the question of application. For airplanes in cruise, where separation matters, reliable devices are best.

GPS signals can be jammed. As can be seen in the news lately in California and Egypt. And GPS signal can be lost due to ionospheric interference. So, any equipment that doesn't rely on external sources is safer.

$\endgroup$
6
$\begingroup$

Directly comparing accuracy of GPS, barometric altimeter and radio altimeter makes little sense, because they each measure different quantity.

  • GPS measures ‘geometric altitude’. This is the actual height above the reference ellipsoid, in units of length. It is fairly accurate, but there are external reasons it might be unavailable.

  • Barometric altimeter measures ‘pressure altitude’. It is given units of length, but it really is just a barometric pressure, expressed in funny scale. It only corresponds to altitude when the atmospheric conditions match the “International Standard Atmosphere,” which they of course usually don't.

    The biggest error is due to pressure, which might be corrected using the altimeter setting. With this, the altimeter reading is better approximation of actual altitude, but since temperature also changes the rate of pressure change with altitude, it still only matches the geometric altitude exactly on the ground at the airport for which the altimeter setting is.

  • Radio altimeter measures ‘height’ above ground level.

Now in aircraft, each of these quantities has different use:

  • The geometric altitude from GPS is, together with a topographic map, used in EGPWS to alert pilots that they are flying too low. Compared to older GPWS that only used the radio altimeter this has the advantage that warning can be given when the aircraft is high enough above ground directly below, but approaching higher ground where it won't be.

    GPS is good for this, because when it works, it always gives the geometric altitude without need for setting that might be incorrect. However, it is a new technology that not all aircraft have and it might be unavailable for external reasons (that can't be mitigated by adding redundancy).

  • The barometric altimeter is used for aircraft separation. Here, the exact altitude does not matter, what is important is whether the aircraft are at the same altitude or not. Barometric altimeter is excellent for this, because it is old technology, so all aircraft since the early days have it, and it is simple technology, so it is reliable.

    Above certain altitude (the ‘transition altitude’, 18,000 ft in the USA, but varies in other parts of the world), all aircraft have the altimeter set to the standard value of 29.92 inHg/1013 hPa. This high, there is no concern of separation from terrain, so everybody just uses the same setting to keep things simpler and errors less likely. The pressure altitude can easily differ from the geometric by couple thousand feet, but nobody cares, because it is only important to know whether the other plane is above, below or same altitude and this serves that purpose well.

    Below transition altitude, the altimeter setting from nearest airport is used, so the altimeter reading better approximates the geometric altitude for purpose of separation from terrain. However, the altitudes still only match at the altitude of the airport from which the setting is. Above (and below) the error increases. In cold weather, you can easily be 10% lower height (above ground) than the altimeter says. This is simply handled by adding sufficient margins to the published minimum altitudes.

  • The height measured by radio altimeter is used in the GPWS and EGPWS systems to warn pilot of flying too close to the ground and for announcing the height during final approach so the pilot can better judge the landing without having to look at the instrument. However, due to terrain irregularities, radio height is not much use outside these two specific cases. Also, radio altimeters usually only indicate up to 8,000 ft in large aircraft and 1,500 ft–2,500 ft in smaller aircraft if installed at all (most GA planes don't have one).

The radio altimeter is accurate to a few feet. The barometric altimeter has to be accurate IIRC to 75 ft. Even simple GPS receiver should be able to do that too, but since the values can be easily off by a few thousand feet at high altitudes, they can't be mixed. So barometric altitude is always used for traffic control.

$\endgroup$
  • $\begingroup$ "...if installed at all (most GA planes don't have one)." If a GA plane doesn't have a radar altimeter, then how does its GPWS work? $\endgroup$ – Sean May 22 '18 at 3:09
  • $\begingroup$ @Sean, EGPWS is GPS-based. This has the advantage that it can warn you of high ground ahead, which the radio altimeter can't. Older GA planes usually don't have GPWS either; the newer ones would have it integrated in the avionics suite that has GPS. $\endgroup$ – Jan Hudec May 22 '18 at 21:36
2
$\begingroup$

Most likely GPS is more accurate, after all, it can calculate your location in three-dimensional space within a few feet. However, it's not really that usable, since GPS measures altitude from the geoid. That said, new GPS-receivers should be able to correct the difference between the geoid and actual earth surface.

Baro-altimeters are also quite accurate, but only when the atmosphere is at ISA-conditions. Baro-altimeters are calibrated to ISA-atmosphere, and when any condition varies from this, you get an error. Some of those, for example temperature error, can be compensated for, but removing all potential errors is a tedious task.

If I was geocaching and wanted the most accurate elevation information for my location, I would use an old-fashioned geographic map with elevation information.

$\endgroup$
  • 2
    $\begingroup$ Just a comment: The GPS determines the altitude on the ellipsoid. Then using a table of geoid undulations it approximates the altitude (and vertical direction) on the geoid. The actual AGL cannot be determined without using the equivalence of a detailed map (which the GPS device may have or not). See this for more details. $\endgroup$ – mins Jun 9 '16 at 10:08
  • $\begingroup$ @mins thanks for info. I had it backwards then. Relying on nav-lessons from flight school.. $\endgroup$ – Sami Jun 9 '16 at 10:09
  • 2
    $\begingroup$ Do you have anything to back up the claim of the 3D accuracy? I was under the impression that satellites are more overhead than sideways, thus I would guess the lateral accuracy is higher than the vertical one. $\endgroup$ – falstro Jun 9 '16 at 10:10
  • $\begingroup$ @falstro Nothing solid. Just the fact that most modern receivers use 6, 8 or even over 10 satellites to determine it's position. This increases the accuracy, and diminishes, to some extent, the effect of satellite positioning. But you're right, vertical accuracy is smaller. $\endgroup$ – Sami Jun 9 '16 at 10:18
  • 1
    $\begingroup$ @falstro: Determination of oceans height is done using GPS (from another satellite, like Topex-Poseidon, and radar) and reaches 1 cm precision. Modern geodesy is done by GPS, and maps are accurate. $\endgroup$ – mins Jun 9 '16 at 10:31
2
$\begingroup$

This depends on the accuracy of the barometer and the weather, since the weather influences the air pressure. So even if you're standing still at the same altitude the barometer will show you different readings. This is why a pilot constantly has to adjust his altimeter. An altimeter is designed to measure the altitude above a fixed level. This is why pilots have to calibrate their altimeters before take-off with ATC.

The GPS system, on the other hand, works with an accurate clock and the known positions of the satellites. The satellite sends out packages which are time stamped. The receiver receives these time-stamped packages and uses its own calibrated clock to determine the sequence. Comparing this to the know positions of the satellite at certain positions allow the use of triangulation formulas to solve for the position of the receiver.

The GPS system is owned by the US military and they have this graph about the accuracy improvements over the years. source: http://www.gps.gov/systems/gps/performance/accuracy/URE.pdf

So to answer your question, it is determined by the equipment you use, but cost wise it is cheaper to get an accurate reading from a GPS since it is also less influenced by the weather.

$\endgroup$
  • 4
    $\begingroup$ "The receiver [...] uses it's own calibrated clock to determine the sequence.": The receiver would need to be fixed and use an atomic clock for this to work. Actually 1/ the receiver clock is not precise and 2/ its speed is unknown. Pseudo-range name is used for that reason. It evaluates what could logically be its local time from at least frames from 4 satellites and by successive approximations guesses the offset between its local clock and the GPS clocks. Elegance of the GPS system is only the satellites need to have a precise clock. $\endgroup$ – mins Jun 9 '16 at 11:36
  • 4
    $\begingroup$ Interesting - the graph doesn't indicate if it's showing error range for horizontal location (lat/long), vertical location (altitude), or both. I'd be careful assuming that this error range also includes altitude, as GPS "accuracy" is usually (in a non-aviation sense) perceived to mean surface location. $\endgroup$ – FreeMan Jun 9 '16 at 12:04
  • 2
    $\begingroup$ @FreeMan. Good catch as it is none of them. But the embedded frame explains SIS URE. It's an error in the pseudo-range, due to satellite data itself. This assumes a perfect receiver in perfect view conditions (DOP), and no ionospheric perturbation. Said otherwise it's an assessment of the "space segment" only. $\endgroup$ – mins Jun 9 '16 at 13:08
  • 3
    $\begingroup$ In my experience with actual GPS units, altitude data are quite erratic and at least appear to be less accurate than lateral measurements. $\endgroup$ – David K Jun 9 '16 at 13:30
  • 2
    $\begingroup$ @DavidK: It's normal vertical data is less accurate. To be equivalent the receiver would need to have access to satellites at say 45° below the horizon (mentally exchange height and latitude of an aircraft and see how satellites now help in determining height, but are less efficient to determine the horizontal location). $\endgroup$ – mins Jun 9 '16 at 13:44

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.