9
$\begingroup$

As I understand, GNSS uses the WGS84 model of the earth to pin-point points in 3D space. However, once the GPS coordinates of a point in space have been determined (say a GPS waypoint), do it's coordinates ever change due to correction of errors in pin-pointing/triangulating the position? If so, how often would that be expected to occur?

That also begs the question, how are GPS coordinates of a fixed point in space found: by calculations and plotting the point in the WGS84 model or using satellites (just like the coordinates of the moving aircraft)?

For example: Say point A has the GPS coordinates 20.21457N 46.87453W. Will these numbers, once calculated, always point to the exact same point in actual space or will some sort of error correction ever cause these numbers to change (in order to continue pointing to the same intended point in actual space)?

My intuition says they don't change, because [considering] the actual earth and the GPS model of earth aren't changing in their shapes so a fixed point in 3D space with relation to each model should stay the same. Does anyone know any differently?

$\endgroup$
8
  • 3
    $\begingroup$ Oh, the earth is changing constantly.... Slowly, but change it does. $\endgroup$
    – Lnafziger
    Apr 3 at 1:39
  • 1
    $\begingroup$ When you say point in space, do you mean coordinates on the surface of the earth, or a satellite's point "in space"? $\endgroup$ Apr 3 at 2:13
  • 3
    $\begingroup$ I'm almost sure you mean WGS84 but i didn't want to just change it for you. $\endgroup$
    – Steve V.
    Apr 3 at 16:48
  • 2
    $\begingroup$ I guess you did say specifically "do it's coordinates ever change due to correction of errors in pin-pointing/triangulating the position", that's clear enough. Some answers seem to have missed that. In other words you are saying they are fixing an error to make the reported position more correctly match the position on the WGS84 model. Might be more clear to insert the word "published", i.e. "do its published coordinates ever change". $\endgroup$ Apr 4 at 12:37
  • 1
    $\begingroup$ I'm still waiting for someone to find 120º N... $\endgroup$ Apr 9 at 17:42

5 Answers 5

5
$\begingroup$

The coordinates of a fixed point on earth do change, but they do change very slowly.

WGS84 is a coordinate system. It is based on an ellipsoid that approximates Earth's shape.

The surface of the ellipsoid is the vertical datum of coordinate system. Heights in WGS84 are measured with respect to the ellipsoid ("height above ellipsoid" or HAE). It is important to note that the surface of the ellipsoid does not coincidence with the surface of the earth nor with the mean sea level.

Horizontally, position is determined as two angles. First the latitude, the angle between the equator, centre of the ellipsoid and the position and the longitude, which is the angle between the projection of position on the equatorial plane,the centre of the ellipsoid and the prime meridian.

The location of the prime meridian is chosen to be in Greenwich, close to the prime meridian of the Airy 1830 ellipsoid. (The difference is ~100 meters,reason for the offset has to do with local gravity direction offset and how time UT1 is defined).

Within the WGS84 coordinate frame, every position can be expressed exactly. GPS coordinates are by default expressed is WGS84 coordinates, but some GPS receivers allow output to be configured for different coordinate systems.

But the shape of the earth is not constant; due to continental drift and tectonic plate movements, coordinates of "fixed" positions slowly change. It is agreed that:

The International Reference Meridian and Poles and, hence the WGS84 datum, are stationary with respect to the average motion of the Earth’s crustal plates. As a consequence, all individual locations are in motion relative to them.

Source: thegreenwichmeridian.org

According to that website, in the UK, latitude and longitudes are changing by approximately 2.5 centimeters per year.


In addition to changes in continental positions, GPS itself is changing as well. GPS is not perfect, like every measurement system is a limited accuracy. So when you fix a GPS receiver to a position and record the position output over a longer time,you will notice that the position output is wandering around. It is somewhat random, but very strongly correlated over an approximately 12 hr period.


From the comments I see that you are wondering if positions of fixed points in aeronautical databases do change over time. The answer is: they do. It is rare, but sometimes waypoints do get moved, errors are corrected and routes and procedures do change. Hence the requirement to keep your GPS database up to date.


An example: due planned works near the threshold of a runway, the published threshold of that runway may be displaced temporarily to ensure a safe clearance height above the works. This will result in an update to the aeronautical database.

If that calculation of the new threshold was wrong, and the updated threshold point is not on the centreline but on the runway edge, another update is required.

If the update is not applied (because it was not picked up by the database provider), and subsequently the pilot relies on both the GPS being accurate and the database being correct, and continues to land based solely on GPS guidance despite this being a non-precision approach and the visibility is too low to see anything, the aircraft may land on the edge of the runway and crash.

A far fetched scenario? It may sound like it, but it actually happened.

$\endgroup$
6
  • $\begingroup$ Great answer! So even if the physical/geographic point on the earth's surface hasn't changed yet, the GPS coordinates can still change as corrections to previous satellite-based coordinates are realized. $\endgroup$
    – PapaMike99
    Apr 4 at 21:49
  • $\begingroup$ @PapaMike99, I don’t think that is what is being said. Errors are corrected, but the coordinates of a fixed location on Earth is not changing. A waypoint that is ‘moved’ is not at the same original location. Also, it is the navigation database that must be kept up to date. $\endgroup$
    – Timbo
    Apr 5 at 4:01
  • 1
    $\begingroup$ @PapaMike99 not necessarily satellite-based coordinates. As an example, the position of many points in the database were never measured, but instead calculated w.r.t. to measured points. If a calculation error is made and this is discovered, it will be updated in a next release of the database. So a waypoint in version 1 of the database may have different coordinates than the same waypoint in version 2 of the database. I've added an example. $\endgroup$
    – DeltaLima
    Apr 5 at 7:35
  • 1
    $\begingroup$ @DeltaLima, excellent example! To be clear to PapaMike99, the coordinates of the runway threshold were incorrect as provided by the ANSP in the AIRAC supplement. The actual coordinates of the threshold did not change. This wasn't a change with the WGS-84 coordinate system, but an errant calculation that led to incorrect coordinates being published. $\endgroup$
    – Timbo
    Apr 5 at 15:28
  • $\begingroup$ Re -"So when you fix a GPS receiver to a position and record the position output over a longer time,you will notice that the position output is wandering around. It is somewhat random, but very strongly correlated over an approximately 12 hr period." -- I didn't realize till just now that the deleted spam answer apparently copied this content directly. $\endgroup$ Apr 5 at 15:35
4
$\begingroup$

Conceptually, yes they change over time with respect to locations on the earth.

Older mapping used a global model that assumed a static coordinate system, tied to the globe in one place (prime meridian). But over the entire globe, continents have relative motion of centimeters per year. It doesn't take too many years for some places to have enough motion for that to matter. There have been construction problems due to ignoring this drift.

Australia is moving at something like 7cm/year. Points mapped in the '90s might be a couple of meters away from where a current mapping might place them.

In the future, many places will start using reference frames tied to a specific tectonic plate and calculations will map that to a global lat/long for a specific date.

$\endgroup$
4
$\begingroup$

Does the GPS coordinates of a point fixed in space change with time?

The GPS uses a coordinate system which is Earth-centered and Earth-fixed (ECEF). Coordinates are fixed to the average Earth (defined a bit later), which moves in space. Thus the GPS coordinate system also moves with Earth. If a point is fixed in space, it moves with respect to the GPS coordinate system, and its coordinates change continuously. I guess you're not asking for this trivial case, rather for the case the point is fixed with respect to Earth.

This steadiness is relative, for instance the Eiffel Tower has apparently a fixed position on Earth. But actually Paris is part of the Eurasian tectonic plate, which slowly drifts over the top of Earth mantle.

enter image description here

(Source).

To illustrate with an example, this plot is the progressive easting component of the drift at a GPS reference station in Bahrain:

enter image description here

(Source)

Static vs. dynamic coordinate systems

Thus the tower, like any landmark on Earth, moves with respect to other tectonic plates. Whether this motion should appear on maps is usually a State choice, and in the past the choice was simply to consider the local plate as fixed, even the plate deformations were simply ignored, because the physical network of geodetic markers was not accurate enough to show a difference.

Coordinate systems can be static or dynamic:

  • Static means the system moves with the local plate, measures done in this system are done against geodetic marks moving with the plate. The measure is constant when the plate moves, the only condition is to measure a feature located on the same plate.

  • Dynamic means the coordinate system is relative to a mean Earth: Plates have velocities in various magnitudes and directions (the cause of volcanoes and Earthquakes). Velocities also vary within a plate as it deforms. The mean Earth has a motion which is the average of all plates motion.

The GPS coordinate system is one fixed to this mean Earth, the mean Earth is considered fixed and all plates move and deform with respect to it. It means landmarks have GPS coordinates changing with time unless they move exactly at the mean velocity. This change is a good indicator of the tectonic drift.

Example and consequences of GPS coordinate drift

At 45° of latitude, 10 cm east-west are represented by about $\small 3.3 \times 10^{-7}$ degree. Coordinates are usually not available with this precision. But drift accumulates and with time it becomes apparent even to smartphone GPS.

As visible on the picture above, Australia is a country where tectonic drift is one of largest, about 7cm per year. The government used to create local maps in static coordinate systems fixed to the plate (up to GDA94). In spite of intra-plate deformations, coordinates remained constant. In recent years, the generalization of accurate GPS receivers revealed a significant discrepancy between GPS (WGS84) and GDA94 coordinates. This motivated Australia to reduce the divergence with an interim static system (GDA2020) and to later adopt a dynamic coordinate system. GDA94 coordinates were reassigned to GDA2020 locations distant from 1.8m in average.

GPS coordinate system (WGS84)

WGS84 system is not based on latitude, longitude and altitude coordinates. Instead it defines points in usual 3D coordinates XYZ (the green axes below):

enter image description here

(Source).

XYZ axes are Earth-Centered and Earth-Fixed (ECEF). The center is the center of mass. At any time the XYZ coordinates can be converted to latitude $\small \varphi$, longitude $\small \lambda$ (purple line) and ellipsoidal height $\small h$ using the ellipsoid shown. This ellipsoid is a good simplification of Earth oblate sphere at sea level.

Altitude is a matter a bit more complex, as it has been also with traditional surveying, the vertical reference being in general a distinct system, with its distinct benchmarks. Altitude is the height above the mean sea level. Water surface, discarding other factors like currents and winds for simplification, always align on a surface of equal gravity. Gravity in turn is mostly influenced by local density. Thus the surface of equal gravity is irregular, and so is the sea level. There are differences of about 200m in the distance to the center of mass.

The ellipsoid is close to the mean sea level, but is a perfect oblate sphere. The WGS84 coordinate system includes a model of gravity, EGM2008, describing, among other things, the height of the surface of equal gravity corresponding to the mean sea level, the geoid. The model provides the geoid height with respect to the ellipsoid. This height is added to the ellipsoidal height of the point to determine its altitude MSL.

How do we use a traditional coordinate system

Everybody use “datum” (like in North American Datum 83) as a synonymous for “coordinate system”, but there is a difference:

  • A coordinate system is a mathematical set of models and transformation formulas. It is useless to field people who need to determine the coordinates of some landmark. For example, in a traditional system, latitudes are counted with respect to meridian 0 by Greenwich observatory. That's irrelevant when someone determines the coordinates of a bridge in Chile.

  • In contrary a datum is the realization of the system, traditionally materialized by geodetic survey marks and monuments. These marks are organized in networks, with a first order of a few but very well surveyed marks, using different methods. The corresponding coordinates are published. A second order network build from the first one, densifies it. Possibly finer networks are created, so that the final user (a surveyor) can directly aim at close physical marks to determine the coordinates of a new location by triangulation.

By construction these datums are fixed, the coordinates of the marks are not updated, any location determined using such datum is final.

How does that scheme translates to the GPS coordinate system? Where is the WGS84 datum?

WGS84 realizations: Frames

For the GPS, the first order network is made of 10 GPS reference receivers, in red on this image:

enter image description here

(Source).

These stations are surveyed with great care and by multiple means very-long-baseline interferometry, satellite laser ranging, DORIS (and GNSS, after the frame initialization).

The second order network, the one used by everybody, is made of satellites, associated to the reference stations this way:

  • Satellites are launched and satellite tracking stations (black starts above), which positions are determined with respect to the reference frame, start tracking their orbits.

  • Each orbit is measured to determine the orbital elements describing the satellite timing and motion. As the orbital elements are measured using the tracking stations coordinates, the orbital elements are derived from the first order network of the red stations.

  • The orbital elements are uploaded by satellite control stations to the satellites. Each satellite continuously broadcasts its orbital elements and the current (atomic) time. This broadcast is technically equivalent to broadcasting the current XYZ coordinates of the satellite in the GPS frame. This also means these coordinates are fixed to the mean velocity of the tectonic plate.

A user with a GPS receiver passively receives at least four satellites. From the orbital elements and the atomic times they broadcast it determines the satellites XYZ coordinates, and from them, by multilateration, its own XYZ position in the GPS frame.

Displaying latitude, longitude and altitude values is a matter of coordinates transformation as seen above. As the GPS receiver is unlikely to be or to remain at the same velocity than the GPS frame, with time the coordinates it computes drift.

GPS frame realignment on the current ITRF

The mean tectonic drift is subject to permanent study in the context of the ITRS, the international terrestrial reference system, another ECEF coordinate system. Initially the GPS frame is aligned on the latest realization of the ITRS: The IRTF (F for frame). While the ITRS follows its own path with numerous GPS stations recording the drift of all plates, the WGS84/GPS frame stands on only 10 reference stations, indeed distributed on different plates, but the plates are subject to deformation. At some point in time, the ITRF and the WGS84 frames are de-aligned and it is desirable to re-align the WGS84 frame.

The realignement consists in updating the XYZ coordinates of the 10 reference stations. This change triggers in cascade, via the tracking stations and the satellites, a change of all GPS receiver coordinates.

The latest change, WGS84 (G2296), occurred in January 2024. The difference was only a few centimeters. The new coordinates of the stations:

enter image description here

(Source).

Since the GPS exists, its frame has been updated several times:

enter image description here

(Source).

The process is somehow similar to leap seconds and years used to realign UTC on actual solar time.

The first update (G730) led to a mean position change in the order of the meter, by far the largest change. It denoted the improvement in the accuracy of the reference stations coordinates.

From the epoch to the current velocities

While the frame is updated at intervals, it's still possible to get a better position than the one derived from the latest update. The solution is to know the current velocities of the reference stations, and to compute the displacements since the last update time (which is called the epoch of the frame). In the previous table for epoch 2024, velocities are found in the last columns.

Coordinates means nothing without a coordinate system

For a given latitude and longitude, unless you know which coordinate reference system (CRS) was used, e.g. NAD83 or WGS84, you can't say which place it indicates (errors of 100m or more are possible). The same coordinates, when used in three different coordinate systems points to distant locations:

enter image description here

(Source: Datums and Map Projections, Iliffe, Lott)

Knowing the CRS is not enough for certain uses. The CRS is the mathematical model, but coordinates are measured using the datum, realized at a given epoch. Different realizations leads to different coordinates, albeit the difference is not as important as between different systems.

$\endgroup$
0
$\begingroup$

You are confusing the coordinates given by the GPS with the physical coordinates of a point. The physical coordinates will never change unless the earth moves, as in an earthquake, for example.

The GPS coordinates always have a margin of error, which the GPS itself will tell you. The margin of error will be within around 5 to 60 feet (higher with basic GPS, lower with additional info like WAAS). The margin of error can be affected by:

  • Satellite geometry
  • Signal blockage
  • Atmospheric conditions
  • Receiver design features/quality
  • Clear view of the sky
  • Obstacles like buildings or trees
$\endgroup$
8
  • $\begingroup$ While this is a good factual answer, I don't believe that it actually answers the OP's question about whether the GPS coordinates for a fixed point in earth ever change. $\endgroup$
    – Lnafziger
    Apr 4 at 4:21
  • $\begingroup$ @Juan I understand the physical/geographic coordinates and GPS coordinates are different (as they're based 2 different model earths), but I'm talking about the GPS database. Are you saying that positions of stationary points (a WPT or APT) are also determined through satellites or are they pin-pointed on the GPS model on paper (not literally)? If the coordinates are calculated, not triangulated, then they shouldn't change unless for changes in the earth's shape - as I understand. $\endgroup$
    – PapaMike99
    Apr 4 at 7:21
  • $\begingroup$ @PapaMike99 Not different because of any model, it's because one is based on surveying and the other by approximation of calculation. There is no "GPS database". There is no pinpoint of GPS models on paper. There are satellites and their info is used to calculate where you are, with a margin of error. Triangulation is calculation. $\endgroup$ Apr 4 at 18:38
  • 1
    $\begingroup$ @PapaMike99 The coordinates of points in the GPS database are based on surveys of the land and use of highly previse advanced GPS receivers. The more advanced the receiver on the aircraft the more accurate it will be. The current top-of-the-line accuracy tech on civilian aircraft is WAAS. It will generate correction data for receivers equipped to get that info and correct to very small tolerances, but only if there is a nearbly operable WAAS station. $\endgroup$ Apr 4 at 22:00
  • 1
    $\begingroup$ "The current top-of-the-line accuracy tech on civilian aircraft is WAAS", SBAS are available on all smartphones and GoPro, WAAS in the US, EGNOS in Europe, GAGAN in India, etc. The more advanced differential GNSS are ground-based augmentation systems used to land. $\endgroup$
    – mins
    Apr 7 at 17:53
0
$\begingroup$

Let me give an attempt at this one. So you have it slightly backward... "once the GPS coordinates of a point in space have been determined (say a GPS waypoint)." There are no GPS waypoints. There are RNAV waypoints defined by WGS84 coordinates. The coordinate system logically comes first. A location's lat/long is not based on GPS. Rather, GPS uses the same coordinate system.

Do coordinates change "due to correction of errors in pin-pointing/triangulating the position?" Technically, WGS84 does get updated. That's part of the Office of Geodesy at the National Geospatial-Intelligence Agency. https://earth-info.nga.mil/index.php?dir=wgs84&action=wgs84

For aviation purposes, this is far more precise than ever needed. In fact, in the US procedures designed by the FAA were based on NAD83 (horizontal datum) and NAVD88 (vertical datum). Since GPS was developed by the DoD, it was always based on WGS84. Likewise, your navigation database is based on WGS84. So consider this... procedures were designed based on a slightly different coordinate system and that is still considered "good enough." Note the FAA says this in the TPP legend: enter image description here

Now that Performance Based Navigation has become widespread, any RNAV procedure with a navigation specification (NavSpec) is based on WGS84. So the procedure design, the aircraft's navigation database, and the GPS receiver are on the same reference system.

$\endgroup$

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .