Regular levelling of the tide benchmarks has two objectives. The first is to verify the reference (zero value) of the sea level measured by tide gauges at the station (and, subsequently, conserve it). It should be remembered that this reference is both local and arbitrary (chosen by convention). In France, it is often the hydrographic datum (article in French). The second objective is to determine if the site is stable locally using an analysis of successive levellings between the tide benchmarks. The latter generally constitute a network of between 3 and 10 benchmarks situated around the tide gauge. Some benchmarks are designed such that they can be observed using spatial geodesy techniques, sometimes continuously (an example being the permanent GPS stations), or by absolute gravimetry. The analysis of these observations provides information on the geocentric stability.
The results of the test measurements ’in situ’ are directly incorporated into the pages corresponding to the tide gauge in question. The test missions are generally the subject of a report in which the results of the use of the calibration measurements (Van de Casteele diagram) are published. Here is an example: test of the tide gauge at Socoa (Saint Jean-de-Luz) in May 2007.
These are measurements of the water level which are carried out simultaneously with a tide gauge and an instrument that is used as a standard, often an electric sound/light sensor. When the measurements are performed over a complete tide cycle (12h25), they can be used to produce a Van de Casteele diagram. The shape of this diagram is highly instructive; it can be used to assess the performances of the tide gauge and to detect any operational defects. This type of test is recommended by COI/UNESCO (Cf. Manuel 1, 1985). Research carried out recently by SHOM/GRGS and the University of La Rochelle shows that the use of this kind of comprehensive test should be generalised to modern tide gauges (Cf. Martin Miguez et al.).