As it was mentioned above in the Chapter 7.1, the real sea level does not follow exactly the level surface defined as geoid. In the geodetic systems, the geoid is defined by the potential value of this level surface. The sea level can differ from this even by two meters. The difference is characteristic as a time-average from place to place, while temporal variations can be also detected.
Because of all of this, the definition of ’sea level’ or ’mean sea level’ is quite complex. The sea level is measured by mareographs: they record the water level at the point as a function of the time. The null points of the mareographs are ad hoc placed vertically. The real geopotential values are rarely determined therefore the read values at different mareographs – however their connections can be analyzed statistically – are not in direct connection. As it was above mentioned, the temporal trends of the read values are affected by the global – and in case of the inner seas: the local – sea level changes and the regional crustal movement of the area the mareograph is placed in. The first affects the real values while the second modifies the geopotential value of the local null level. The sea level is defined by the elevation data read at all mareographs, together with the horizontal position of the instruments as well as the vertical situation of their null level. The sea level can be interpreted for different time intervals (epochs), as the temporal average of all measuring points (e.g. mean sea level of 1905-1910).
In the practice, the leveling network is connected to the mean sea level at one (or more) pre-defines point. For example, the height network of the Austro-Hungarian Monarchy was set to a mareograph that was working (now abandoned) at the Molo Sartorio in Trieste (now Italy) at a given epoch (Figs. 41 & 42). The military cartography of the late Warsaw Pact used a null level connected to the Kronstadt mareograph near Leningrad (now: Sankt-Petersburg, Russia). The null level of the height system of the European Union is connected to the Amsterdam mareograph. In case of land-locked countries, such as Switzerland, the Czech Republic or Serbia, any land base point can be used as starting elevation data. In this case, of course, this value is not zero. For example, the also land-locked Hungary selected one of the eight fundamental elevation base points of the former Monarchy at Nadap. Besides the old point, a new one was set up in 1951, connected to the new elevation network (Fig. 43), the ’Hungarian zero’ level is positioned beneath this point, with a distance given by tenth of millimeter accuracy (Nadap base level).
In the topographic maps, besides the horizontal reference (the geodetic datum), also the vertical system, the so-called vertical datum should be given. Usually this refers to a null level of a mareograph and, if applicable, the epoch. In case of Hungary, the ’Adriatic system’ (connected to Trieste) and the ’Baltic system’ (connected to Kronstadt) are the examples, and as it was mentioned, both are realized as (different) heights from the Nadap base point. In vertical terms, the vertical difference between different vertical datums can be interpreted as constant for practical use. Because of the ocean streams, the salinity differences and the evaporation surplus of the Mediterranean Sea, the level of the Baltic sea is physically higher than the one of the Adriatic Sea. The difference is 67.47 centimeters, this value should be subtracted from the elevations given in the Adriatic height system to get the elevation in the Baltic system. When Hungary started to use the Baltic level instead of the Adriatic one, in the later (mainly at the end of 1970s) issued touristic and hiking maps, majority of the summit elevations were decreased by a meter. As the decrease is less than one meter, according to the rounding rules, the decrease does not occur in every case.
Similarly to the horizontal geodetic datums, the realization of the vertical datums can be made by a system of base points. The vertical datum is characterized by the physical location of the base points, as well as their fixed elevation data. In some national systems, where the regional crustal movements are very high (e.g. in Scandinavia), the average annual uplift or subsidence values are also indicated. The elevation of the surrounding terrain points can be determined from a nearby base point by local survey technology. If the vertical point network consists of many points, it is divided to sub-networks, according to the leveling technology of their points. The accuracy of the network, however, is mostly determined by the creation method and accuracy of the first-order vertical network (Fig. 44).
The elevation correction – the simple difference making – between the different vertical datums, especially the sign (direction) of the shift, should be accomplish with special care in case of construction of such objects (bridges, tunnels), whose endpoints are in different countries using different geodetic datums. Nowadays, the European height standard is connected to the Amsterdam mareograph. The local elevation differences from this level are shown in Fig. 45.
Finally it should be mentioned that as the three-dimensional data collection techniques are spreading, the unification of the divided horizontal and vertical databases and networks is expected. The cause of the still-characteristic division is mainly the different methodology and accuracy in the physical-geodetic realization of the horizontal and vertical references. The geodetic use of the GPS this difference is decreasing and eliminating, the determination of the horizontal and vertical positions is unified, based on the same physical theory and geodetic practice.