Sensibility of measurements. Seismometers are instruments that are often extremely sensitive. They register ground movements that are much weaker than those that can be detected by human beings. They even detect “seismic noise”, which is the permanent movement of the ground, which human beings obviously do not feel. Seismometers are combined with very precise clocks (nowadays, the same clock serves many seismographic stations equipped with “GPS receivers”: the clock is a global positioning system [GPS] device) that make it possible to date packets of vibratory energy emitted by seismic sources that spread through the earth in all directions and reach the seismographs. If the velocity of the seismic waves at any point of the earth is known, and assuming that the earthquake has a single source (in space and time – it is essential to make this assumption), it is sufficient, in an ideal situation, to know the times of arrival of a type of wave at four seismic stations in order to calculate the time at which the earthquake originated and the position of its focus (ie, the point to which the source is reduced). Since there are a number of uncertainties surrounding the data used, as with any measurement, and models of the earth are not perfect, there is generally no solution to this set of four equations with four unknowns. This means it is necessary to use much more complex techniques, introduce many more observation data and resort to methods of optimising the differences between observations (measured arrival times of energy packets at the seismographic stations) and the calculated arrival times at the same stations for a given focus. An attempt is made, as it were, to determine the focus of the earthquake (in time and space) as accurately as possible by means of successive approximations, on the basis of accepted methods. One will never know definitely to what extent one is mistaken since the assumption that there is a single focus is only a crude approximation of the physical phenomena that generate the earthquake and the model for seismic velocity dispersion within the earth is only a mediocre approximation of the actual situation. In particular, the earth is assumed to be spherical (which it is not entirely) with the velocity of seismic wave propagation dependent on the radius alone (in other words, we are assuming a spherical earth with radial velocity dispersion), without lateral variations, which is obviously not the case (for example, we find in the vicinity of the earth’s outer s...
