UDC 528.371 (497.12) 
VE TICAL DATUMS OF 
LEVELLING NE OR S IN 
SLOVEN 
dr. Božo Koler 
FAGG-Oddelek za geodezijo, Ljubljana · 
Receivedfor publication: Feb. 11, 1994 
Abstract 
The author presents vertical datums of levelling networks in 
Slovenia (Austro-Hungarian levelling, I. and JI. levelling of 
great accuracy) i.mdproblems arising when defining the 
vertical datum of the Austro-Hungarian levelling. 
Keywords: levelling network, medium sea leve!, Slovenia, 
swface reference plane, vertical datum 
l. INTRODUCTION 
In defining altitudes of points it is of vi tal importance to determine in advance a comparative surface from which heights of points above sea level are calculated. 
Thus the vertical datum of a levelling network is given by the medium sea leyel in a 
certain tirne span defined as a surface reference plane. The medium sea level is 
defined on the basis of results of yearlong observations of the sea level oscillation by 
tide gauges represented by the arithmetic mean of the sea level at one point of a sea 
shore in a longer tirne span. The medium sea level is not constant neither at the 
same point at different tirne periods nor at different places of the sea shore. The 
medium sea level variance is a result of a constant oscillation of the sea leve! due to 
various causes. 
Periodic oscillations of the medium sea leve! are the result of the attractive force of the Sun, Moon, and Earth. Irregular oscillations of the medium sea level are 
due to activities of wind, atmospheric pressure, differences in temperature and 
quantity of salt in the sea water. Greater influence on oscillation of the medium sea 
leve! have periodic influences e.g. the medium sea level oscillates more or less 
periodically. To obtain reliable data on medium sea level which is represented by the 
vertical datum of a levelling network constant observations of oscillation of the sea 
level are needed for at least 18,6 years due to nodes of Lunar trajectory. 
2. NORMAL BENCH MARK 
ccording to agreement the altitude of the medium sea level e.g. the surface 
eference plane equals zero. The position of the reference surface plane is 
defined by a vertical distance form the so called normal bench mark which is 
stabilized on an area regarded as geologically stabile one (Stefanovic 1955). 
Geodetski vestnik 38 (1994) 1 
2.1 Connection of the Austro-Hungarian levelling netwm:k t~ a normal bench mark 
The normal bench mark for connection of the Austro-Hungarian levelling network to the surface reference plane is the bench mark at the Sartorio pier in Trieste, 
ltaly. The height of the normal bench mark in Trieste was calculated by dr. Farolfi. 
The altitude of the normal bench mark was determined on the basis of one year 
observations of the oscillation of the level of the Adriatic Sea in 1875 and the result 
is 3,352 ± 0,01 m. The tide gauge on the Trieste pier was set up in 1869 yet the first 
data on the Adriatic Sea leve! registration for 1875 were published in 1877. In 
Austro-Hungarian monarchy the height of the normal bench mark was defined on 
the basis of one year obs~rvations because the wish in Europe in those years was to 
connect medium levels of the Medierranean with the north seas, and to d.etermine a 
unique normal bench mark for the whole of Europe. As it was found out that the 
medium level of the Mediterranean is for 13 cm lower from the medium level of the 
norih seas it was decided not to determine a unique normal bench mark for the 
whole of Europe but each country should keep its normal bench mark (Zeger 1986). 
A s carly as 1904 Sterneck was the first to doubt the reliability of the reference 
J-\..surface plane determination beeing gathered only on the basis of a single 
year registration of oscillation of the sea level. So he compared heights of the basic 
bench marks of tide gauges in Trieste, Pula and Dubrovnik which were defined on 
the basis of yearlong observations of sea level oscillation and on the basis of the 
Austro-Hungarian levelling network integration. To be able to compare heights of 
basic bench marks of tide gauges he defined the height of the normal bench mark in 
Trieste anew for 1901. He defined the medium sea level on the basis of data of 
registration of the sea level oscillation for the period 1875-1879 (without 1877) 
and from 1901-1904. The new defined altitude of the normal bench mark equaled 
3,2621 ± 0,0099 m. For both heights of the normal bench mark stands they were not 
defined on the basis of uninterrupted observations of the sea level oscillation ( at 
least 18,6 years). In addition both data differ as to the position of the surface 
reference plane for 8,99 cm. According to Sterneck's calculations the old 
Austro-Hungarian altitudes of bench marks do not refer to the medium leve! of the 
Adriatic Sea in Trieste but to a comparative level surface which is 8,99 cm lower than 
the surface reference plane defined on the basis of eight years of observations of 
oscillations of the Adriatic Sea level (Zeger 1986). 
The problem of defining a vertical datum of the Austro-Hungarian levelling network was also dealt with other surveyors and geophysicists. The professional 
literature gives also the following values of the position of the comparative surface as 
to the reference surface plane which should be taken for the vertical datum of the 
Austro-Hungarian levelling network: 
o 8,93 cm - this value was determined by Kasumovic by the aid of observing 
oscillation sea level at Bakar and refers to the year 1933 ( Kasumovic 1950). 
o 10,57 cm the value was obtained from data Kasumovic published later and 
they also refer to the tide gauge in Bakar and year 1933. The two mentioned 
data differ since Kasumovic defined the medium sea level for 1933 by taking 
the tide gauge constant defined for 1948. 
Geodetski vestnik 38 (1994) 1 
• 13,83 cm the value was calculated by Bilajbegovic from observations of the 
sea level oscillations in Bakar for 1971. 
• 18,5 cm - this value was defined by Istituto Talassografico Trieste from 
observations of the tide gauge in Trieste for 1969 (Bilajbegovic, Marchesini 
1991). 
The question bound to arise is how comes to differences among medium levels defined by individual observers. Inspecting the above mentioned data we notice 
the individual values were defined at different tide gauges (Bakar, Trieste) and for 
different years (1901, 1933, 1969, 1971). Bilajbegovic explained the differences 
among the mentioned values by the difference between linear trends of oscillation 
of the medium sea level in Trieste and Bakar. In addition in 1991 he published that 
the medium sea level in Trieste was defined too low for 6,056 cm (Bilajbegovic, 
Marchesini 1991). The mentioned value was defined on the basis of observations of 
the oscillations of the sea level in the tide gauge in Bakar and data of levelling 
between the basic bench mark of the tide gauge in Bakar and the normal bench 
mark in Trieste. · 
2.2 Com1ecticm of the I. levelling ol' great accuracy (I. NVN) to non:nal bench mark 
hen surveying and adjusting the levelling network of the I. NVN on the 
territory of the ex Socialist Federative Republic of Yugoslavia there was no 
stabilized normal bench mark. So the altitudes of the I. NVN were connected to the 
normal bench mark of the Austro-Hungarian levelling which means we have the 
same vertical datum for the Austro-Hungarian levelling network and the network of 
the I. NVN. 
2.3 Defining the vertical datum ol' the II. levelling ol' great accuracy (II. NVN) 
The levelling II. NVN network was first connected to the normal bench mark, which was stabilized on the territory of the ex Socialist Federative Republic 
of Yugoslavia. The normal bench mark was stabilized at a geologically stabile 
area in the main part of the ex Yugoslavia (city of Maglaj). On the east coast 
of the Adriatic Sea there are seven tide gauges. The oldest are the tide gauges 
in Bakar and Split (in the port), set up in 1929. After the second world war tide 
gauges were set up in Split at the Cape Marjan (1952), Dubrovnik (1954), Rovinj 
(1955), Koper (1962) and Bar (1964). Table 1 shows data on the uninterrupted 
registration of the sea level oscillation for individual tide gauges till 1989 and 
data on including tide gauges into the levelling of great accuracy . 
. 
Geodetski vestnik 38 (1994) 1 
TaMe 1 
Unintenupted 
Year of including tide gauge into levelling 
Tide gauge in registration 
since 
of great accuracy 
Koper 1962 1964, 1972 
Rovinj 1955 1957, 1964, 1972 
Bakar 1954 1957, 1964/65, 1970/72 
Split - the port 1954 1957, 1964/65, 1970/72 
Split - Marjan 1954 1957, 1962/63 
Dubrovnik 1954 1957, 1962/63, 1970/71 
Bar 1965 no available data 
he vertical datum of the n. NVN e.g. the medium sea leve! on individual tide 
gauges are defined for 3.7.1971, from data of the registration of the sea leve! 
oscillation from 1962,2 till 1980,8. On all tide gauges except the one in Bar the 
medium sca level was defined on the basis of a registration of sea level oscillation for 
one whole period 18,6 years (Bilajbegovic et al. 1989). 
In 1962 and 1963 the normal bcnch mark in Maglaj was connected by a precise levelling with tide gauges in Split and Dubrovnik. Later also other tide gauges 
were connected by a precise levelling (1964-1965) and they were included into the 
surveying of the IL NVN. Into the II. NVN surveying only the tide gauge in Bar was 
not included as the basic bench mark of the tide gauge was not connected with the 
levelling network of the II. NVN. The altitude of a normal bench mark is determined 
by a mutual adjustment of the measured altitude differences among tide gauges and 
the normal bench mark from 1962/63, 1970-1973, and by altitude differences among 
the basic bench marks of tide gauges determined from tide gauges observations. 
3. CONCLUSION 
Altitudes of points, used in Slovenia, are given on the basis of a vertical datum of the Austro-Hungarian levelling network. These altitudes are determined from 
the survey of the I. NVN and connections of levelling networks of lower orders to 
bench marks of the I. NVN. Ifwe are to decide for a new NVN survey on the 
territory of Slovenia the most suitable area for stabilization of the normal bench 
mark has to be chosen and decision has to be taken as to which basic bench marks of 
tide gauges the normal bench mark and with it the NVN network is to be connected. 
There is the tide gauge in Koper yet it remains to b.e checked whether it is in 
accordance with standards and regulations about registration of the sea level and 
data on a constant monitoring sea level changes. In my opinion the vertical datum of 
the new NVN network ought to be defined also on the basis of observing the sea 
level oscilations on tide gauges set up in Rovinj, Bakar and Triest. Naturally this 
should be discussed with appropriate offices in Italy and Croatia. 
Geodetski vestnik 38 (1994) 1 
References: 
Bilajbegovič, A. et al., 1989, II. Nivelman visoke točnosti Jugoslavije - svezak 1, Geodetski fakultet 
Sveučilišta u Zagrebu, Zavod za višu geodeziju, Zagreb, p. 3-4. 
Bilajbegovič, A., Marchesini, C., 1991, Jugoslavenski vertikalni datumi i preliminarno povezivanje 
nove jugoslovenske nivelmanske mreže s austrijskom i talijanskom, Geodetski list, Vol. 45, 
Zagreb, No. 7-9, p. 233-249. 
Kasumovič, M., 1950, Srednja razina Jadranskog mora i geodetska normalna nula Trst, Geodetski 
list, Vol. 4, No. 10-12, p. 243-256. 
J(asumovič, M., 1959, O srednjo} razini Jadranskog mora i njenom utvrdivanju, Geodetski list, 
Vol. 13, No, 7-9, p. 159-169. 
Stefanovic, M., 1955, Normalni reper nivelmanske mreže, Geodetski list, Vol. 9, No. 1-2, p. 3-14. 
Zeger, J, 1986, Historische Entwicklung des Praezisionsnivellements in Oesten·eich, 
Oesterreichische Zeitschrift fuer Vennessungswesen und Photogrammetrie, Jahrg. 74, Heft 4, 
S. 250-263. 
Review: Marjan Jenko 
Ivan Štupar 
Geodetski vestnik 38 (1994) 1