Is the GPS a perfect navigation aid? It is almost, but not quite. The accuracy of the system when its signal is emitted on the frequency L1 reaches 20 meters. There are other sources of errors that can introduce inaccuracies in the final position, ranging from one meter to hundreds of meters. These sources of errors are as follows: refraction in the ionosphere, refraction in the troposphere, GPS positioning precision and multipath phenomena.
Error #1: refraction in the ionosphere
The ionosphere is an envelope of charged particles (ions) that surround the Earth at an altitude of nearly 20 km. The carrier wave of the GPS signal must enter this layer on its path. The fact that this layer is not neutral, at its charge, causes a disturbance in the speed of the electromagnetic wave that propagates. The magnitude of this inaccuracy is related to the wavelength and density of charged particles in the medium through which the density is obviously unknown and variable in time and space. The time taken by the GPS wave is changed to an unknown duration, called the ionospheric delay. The evaluation of the distance between the satellite and the station will be distorted, the accuracy is thus reduced by this first phenomenon. In the case of a very agitated ionosphere, during a solar storm for example, the evaluation of the ionospheric delay will only be approximate and the measurement of the position imprecise.
Error #2: refraction in the troposphere
In the same way, the propagation time of the GPS wave is affected by the water vapor content of the lower layer of the atmosphere (from 0 to 10 km altitude): the troposphere. It is necessary to know this content accurately all along the path of the wave. In practice this is very difficult, if not impossible. Indeed, the delay caused is more complicated than a simple ratio of proportionality with the percentage of water vapor. This problem is all the more important as the weather conditions and tropospheric thicknesses differ between two stations. This error of position will be found more particularly on the vertical component, the horizontal errors more or less compensating for the fact that the satellites cover almost all directions the horizon. There is research on instruments to directly measure the water vapor content along the path followed by the GPS wave; but they are in the experimental phase.
Error #3: GPS positioning precision
It is obvious that if there is an error on the position of the transmitting satellite, this error will have a direct effect on the position displayed by the receiver. The distance between two stations (bottom line) The orbit of the GPS satellites can be calculated very precisely, but it is made public by the US military with an accuracy of about 200 m, over 20000 km it gives an error of 10 cm on a baseline of 10 km! This error is disabling for areas requiring high accuracy.
Error #4: Multipath errors
Multipath errors or reverberation errors also affect the good location of a GPS. They can be factors of technological types (example: tunnels, buildings …) but also natural (mountains). Obstacles can then deflect the waves emitted by the satellites. We can simply imagine that a signal intercepted by a mountain has difficulty reaching the receiver (see diagram below) and sometimes even cannot reach its final destination.
These phenomena are among the most difficult to grasp. It is clear that any reflective object placed in the vicinity of the antenna of the GPS station, can return part of the signal from the satellite on this antenna. Just as a mirror creates a self-image even when looking into it, the reflector creates an image of the GPS antenna. It is the position of this virtual antenna that is then likely to be measured in place of the real antenna. Moreover, as the satellite moves in its orbit, the angle of incidence on the reflector changes, and the image moves accordingly. So it’s finally the position of a mobile virtual antenna that we measure! Given the complexity of the corrective calculations that should be made, there are really no remedies for the problems of multipath. A “shielding” of antennas against parasitic reflections is always possible, but it can only be partial since it is necessary that the real signal reaches the antenna. The only solution is to try to avoid multitrafts (ie parasitic objects) as much as possible, which is not so easy when you consider that the ground itself is a potential reflector!