Nevenka Ozanić,
Croatian Waters
51000 RIJEKA, G. CIOTTA 17B, CROATIA
Phone: +385/51/339755
fax: +385/51/336947
E-mail: nozanic@voda.hr
Josip Rubinić B. Sc.E
Croatian Waters
52000 LABIN, ZELENICE 18, CROATIA
E-mail: josip.rubinic@ri.tel.hr
Abstract: The
cryptodepression of the Vrana Lake on the island Cres (Croatia) is a specific
natural phenomenon of 220 million m3 of fresh water on the island’s karst in
the immediate nearness of the sea. The average water level of the Lake for
analyzed 1929-1999 period is 13,13 ma.s.l. (Meters above sea level) and the
deepest bottom are 61.3 m bellow mean sea level. The Lake is the only source of
water supply of the islands Cres and Loiknj. The noticed trend of declining of
the water level at the end of eighties caused certain worries about further
pumping and keeping the balance of the Lake’s system. Because recharge and run
off happen through not localized subterraneous karst ways their direct
measurements and balancingitoring of dynamic changes in the water level of the
lake in different hydrological conditions. The significant influence of basic
water balance parameters was determined as well as the influence of pumping on
the lowering of the water level of the lake for approximately 2 m. An average
annual inflow of 0,588 m3s-1 was determined according to measured average annual
rainfall of 1064 mm and average annual evaporation quantity of 1161 mm. This
paper discusses the hydrological and hydrulical aspect of functioning of the
lake system as a part of the island karst aquifer, and th are not possible.
Therefore the elements of water balance and the pattern in the behavior of the
lake’s system have been established by hydrological methods through the mone
related problems of protection of the Vrana Lake from excessive water
consumption and pollution, i.e. disturbing of its highly sensitive ecosystem.
Keywords: islands aquifer, Vrana Lake, karst hydrology and hydraulic
The Vrana Lake is situated on the
island of Cres, belonging to the group of Croatian islands of the Adriatic Sea -
a marginal sea of the Mediterranean (Southern Europe) (Fig.1). With its average
area of approximately 5,75 sq.km. and volume of approximately 220 million cu.m.
of water of exceptionally high quality on otherwise dry area of the island of
Cres (the total area of the island is only 405,78 sq. km.), it represents an
invaluable resource in both the environmental and economic sence. The lake is a
kryptodepression, with the water level between 16,70 m a.s.l. (1938) and 9,11 m
a.s.l. (1990) - on the average 13,13 m a.s.l. in the period from 1929 to 1995.
The lake bottom is at 61.3 m below the mean sea level. The average annual
rainfall in the area of the Vrana lake is 1064 mm, the mean annual air
temperature is +14.8 oC , and the annual evaporation from free water surface is
1161 mm. The falling trend of the water level in the lake in the eighties, never
recorded before, caused a great concern among proffesuinals and the public
regarding the future of the lake, and initiated the beginning of complex
research work. In the recent years this trend has been stopped; however the need
for protection of the lake is still present in the sense of defining the regime
of water pumping from the lake. Because of its position and size, the issue of
the origin of the lake water has been attracting the attention of scientists for
a lond time. Already in the past century contrary assumptions appeared regarding
its recharging - from the mainland (Lorenz, 1852) or the local island catchment
area (Mayer, 1873). The results of all hydrological investigations carried out
so far, both earlier and recent (Cecconi, 1939; Petrik, 1957; Rubinic and Ozanic,
1992; Bonacci, 1993; Ozanic and Rubinic, 1994; Bonacci, 1995), althought not
identical in figures and conclusions, speak in favour of the latter assumption.
Mechanism of lake’s aquifer functioning in way of hydraulic relation between
salt and fresh water was analyzed in Ozanic and Rubinic, 1995. In Bonacci, 1996
hypotetical hydraulic model with assumption Lake’s recharging from the
mainland was analyzed. The results of these hydraulical investigations confirm
hydrological conclusions, so the problem of functioning and protection of Vrana
Lake must be consider as part of island karst aquifer i.e. balance relation
between salt and fresh water in according to the Ghyben-Hertzberg Low (Linsley,
1964).
However, due to the specific properties of the Vrana Lake which has no directly measurable imflow nor runoff, the hydrological analyses of functioning of the lake system based on measuring of water balance parameters of the lake system (water level monitoring, precipitation regime, water pumping from the lake, evaporation from lake surface) have made a considerable contribution to the general level of knowledge on this natural phenomenon and its protection. Due to its size compared to the comparatively small island enviroment, the lake containing 220 million cu.m. of potable water of exceptionally high quality affects significantly the dynamics of functioning of the entire aquifer. Except its position and size, the Vrana Lake is specific due to the fact that inflow into the lake, as well as the outflow, is accomplished by so far unidentified underground ways. No significant permanent springs have been noticed on the lake, except two temporary springs with the yield up to 0.005 cu.m./sec. Compared to the mater in the mainlend area beyond the direct influence of the sea, salinity of the lake water is somewat higher, ranging between 62 and 92 mgl-1. In the wider zone of the Vrana Lake - on the coast of the island of Cres several springs, vrulja and coastal sources have been noticed, as well as a zone of diffused fresh water outflow. Due to the constant outflow regime, some of these phenomena may be related to the Vrana Lake, or its aquifer, as possible privilleged directions of its discharging. Water level observations have been carried out on a daily basis since 1928, completed by rainfall observations on the catchment area and directly along the shore. Fig. 2 shows the trends of recorded values of mean annual water levels in the lake, annual rainfalls and annual quantities of water pumped out of the lake (started in 1952). An expressed trend of water level decrease in the lake is noticed, in particular in the eighties. The trend of water level decrease in the entire period of analysis, 1929 to 1995, was 0,04 m per annum, while in the period from 1985 to 1990 it was even 0,48 m per annum. This trend caused the concern of the public and of the professionals for the destiny of the lake and possible disturbance of its balance, and possible breakthrough of sea water into its system. On the other hand, it may be seen that water pumping from the lake, which started in 1952, shows a clear growing trend, with the annual average for the 1964 to 1995 period of about 65 000 cu.m. per annum. During the recent decade, the average pumping rate from the lake is 0,072 cu.m./sec., with the maximum during the summer season, up to 0,160 cu.m./sec. Fortunately, such drastic trend of water level decrease that could have endangered the balance of the fresh water aquifer of the Vrana Lake, was not continued - the rainfall conditions wereb improved, and due to the war situation in Croatia in the early nineties, the touristic traffic was reduced, resulting in reduced water pumpings from the lake, compared to pre-war maximum values.
Multiple regression analysis of the annual changes of the lake status, in dependance on the magnitude of relevant hydrological parameters for the period from 1980 to 1995, when the most complete input data were available, gave a very acceptable regression dependance with the linear correlation coefficient k=0,96 (Equation 1), wich may be seen in Fig.3. The resulting equation is: Hn=2,7369 On+1,12836 In-4,0069 Cn+0,23208 Gn-0,31284 H(n-1)+0,3718 (1) In the above equation: Hn - annual water level change, On - annual rainfall, In-annual evaporation, Cn-annual pumping, Gn-calculated losses through sinking determinated on the basis of the sinking curve (Ozanic, 1994) and H(n-1)-mean annual water levels in the previous year - all expressed in (m) with regard to the mean lake level. Such analyses, along with detailed analyses of sinking from the lake system (Bonacci, 1993) and (Ozanic and Rubinic, 1995), provided the preconditions for elaboration of the complex mathematical model of functioning of the lake system of RANA which was used to compute the mean monthly inflows into the Vrana Lake. According to the computations, in the period from 1929 to 1995, the mean annual inflow into the Vrana Lake system was 0,588 cu.m./sec., and its value varied from 1,144 cu.m./sec. (1960) and 0,273 cu.m./sec. (1938). The average value of inflow from the direct catchment area was calculated as 0,393 cu.m./sec., and figure of 0,195 cu.m./sec. refers to the rain falling directly on lake surface. The area of the orographic catchment area sourrounding the Vrana Lake oscillations and to the analysis of regional runoff, it comes out that the average catchment area satisfying the Vrana Lake recharging balance is only about 24 sq.km. (Ozanic, 1996). This size of catchment area is not clearly determinable, changing due to the karst character of the terrain in dependance on hydrological conditions.
As already mentioned, the Vrana Lake is the only source of water supply of the islands of Cres and Loiknj, and at the present requirements the annual quantity of water pumped out of the lake reaches 2,3 million cu.m. As the calculated annual average inflows are 18.5 million, and the evaporation losses at the mean water level are of the order of 6,7 million cu.m., and the losses through sinking from the lake system are 11,7 cu.m. it is obvious that pumping, although globally the smallest element of the water balance, disturbs the balance of the Vrana Lake system and influence lowering of the water level. This lowering of the mean lake level occurs to the case when the sinking losses are reduced in the magnitude of the increase of the pumped quantity. Namely, a functional relation has been determined between the quantity of sinking and the water level in the lake, and its gradient in domain of the observed water levels per each meter is approx. 0,028 cu.m./sec. Until 1995, the total quantity of water pumped out of the lake reached 42 million m3 which would, at the mean water level account for 7.3 m. The analysis showed that such water consumption, in spite of the expressed trend of water level lowering, caused, instead of the said 7.3 m, the actual lowering of the lake level by approximately 2 m. Therefore, it is obvious that this water loss was compensated by the change of hydrological conditions in the lake-first of all by reduction of the losses from the aquifer due to his lower level in relation to the sea, and partly due to the increased recharging from the underground part of the aquifer, caused by lowering of the lake level (Ozanic and Rubinic, 1994). Fig. 4 shows the trend of recorded annual water levels in the Vrana Lake, and of values, obtained by mathematical simulation on the model of RANA, assuming there was no pumping, but also under the assumption that in the period of analysis pumping was even more intensive than the actual mean annual rate of 0,072 cu.m./sec., i.e. that they were double - 0,150 cu.m./sec., and 0,250 cu.m./sec., which is equal to the earlier forecasts (Petrik, 1961) of possible maximum allowed pumpings rates. Althought the simulations of the Vrana Lakeb behaviour in cases of its intensified use have carried out on the basis of extrapolation of the determined regularities of behaviour in the real - higher water levels in the lake, and are therefore largely approximative, they still show that more intensive pumping of water from the lake may result in excessive consumption which, in turn, could cause disturbance of its dynamic balance with the sea. According to analysis (Ozanic, 1996) minor planned increases of the pumping rate of the order of 20 to 30 percent for the needs of the island population are not yet alarming. The formed fresh water lense of the island aquifer, by its higher pressure, on the principles of the Ghyben-Hertzberger Low, maintains balance with the sea and prevents its breakthrough into the lake. As the bottom of the Vrana Lake is even 61 m below the mean sea level, and with respect to the theoretical balance ratio (1m of fresh water lense above the sea level corresponds to 40 m of fresh water below the sea level) in that case the breakthrough of sea into the lake would occur at the lake level of 1,5 m a.s.l. Such breakthrough into the lake would be a disaster for the lake system because the thermal currents in the lake, as well as winds, would cause salination of the entire lake, without any possibility of desalination later on. However, in reality, as the Vrana Lake is in the karst environment the limit of the contact of fresh water and the sea is not distinct, but involves a wider fresh and sea water mixing zone, which is considerably closer to the lake. Therefore, the above-mentioned theoretically possible lowering of the lake level should not be allowed, and increased pumping of water from the lake must be subject to constant control, and limited to the intensity that will not endanger the balance of the lake.
Vrana Lake is part of the complex system of island’s karst aquifer, which functions in accordance with the principals of dynamical balance of the fresh water and the sea. In this paper are given the main characteristic of its hydrological system as also the analysis of problems of pumping from the lake. This paper shows that we can analyzed aquifer behaviour by hydrological modeling when their natural hydraulic give limitation of excessively exploitation. It has been determined that the observed lowering of the water level in the Lake during the eighties was the result of the coincidence of extremely dry hydrological conditions and increased consumption of water from the Lake (pumping). Existing pumping of average 0,072 m3/s does not disturb Vrana Lake to its exhaustion. Such development, in perspective, with unavoidable increase of water consumption standards, results in new requirements for water pumping out of the lake, thereby bringing new risks of disturbing the existing water quality. In accordance with the conceptions we have found out it is obvious that the former estimations of permitted 0,250 m3/s were too optimistic. It has been established that rapid trend of falling of water level was caused by more intensive pumping from Vrana Lake and driver hydrological conditions.
[1] Bonacci,O., 1993. The Vrana Lake Hydrology (Island of Cres-Croatia). Water Resources Bulletin AWRA, 29/3: 407-414.
[2] Bonacci,O., 1996. Hidrololki radovi na Vranskom jezeru-III faza programa. (Hydrological works on the Vrana Lake-IIIth Phaze). Hrvatske vode-VGO Rijeka, Rijeka, Croatia.
[3] Cecconi, A., 1939. Il Regime Idrauliko del Lago di Vrana. Annali dei Lavori Publlici, Capodistria, pp. 14-65.
[4] Linsley, R.K., Kohler, M.A., and Paulhus, J.L.H., 1972. Hydrology for Engineers. McGraw-Hill, New York, 508 pp.
[5] Lorenz, R., 1859. Der Vrana See. Petermans Georg. Mitt 1, Gotha.
[6] Mayer, E., 1873. Der Vrana See aut der Insel Cherso. Mitt. Geogr. Ges. 16, Wien.
[7] Ozanic, N., 1994. Hidrologija Vranskog jezera na otoku Cresu. (Vrana Lake Hydrology on the Cres Island). M.S. thesis, Faculty of Civil Engineering - University of Zagreb, Croatia.
[8] Ozanic, N., 1996. Hidrololki model funkcioniranja Vranskog jezera na otoku Cresu. (Hydrological Functioning Model orf the Vrana Lake on the Cres Island). Ph.D. thesis, Faculty of Civil Engineering - University of Split, Croatia.
[9] Ozanic, N. and Rubinic, J., 1994. Analiza hidrololkog repma Vranskog jezera na otoku Cresu. (Analysis of the Hydrological Regime of the Lake Vransko jezero on the Island of Cres.) Hrvatske vode, 2/8: 535-543.
[10] Ozanic, N. and Rubinic, J., 1995. Hidrololki koncept funkcioniranja Vranskog jezera na otoku Cresu. (Hydrological Concept of the Vransko jezero Lake Functioning (the Island of Cres.) Proc. 1st Croatian Conference on Waters, Dubrovnik, pp.159-167.
[11] Petrik, M., 1957. Hidrololki repm jezera Vrana. (Hydrological Regime of Vrana Lake.) Krac Jugoslavije, 1: 109-151.
[12] Petrik
M., 1961. Raspolopva koliY a vode u jezeru Vrani na Cresu. (Aveilable Quantity
of WAter in Vrana Lake on the Cres Island.) Gra0ninar,13/4: 93-99. Rubinic, J.
and Ozanic, N., 1992. Hidrololke karakteristike Vranskog jezera na otoku Cresu.
(Hydrological Characteristics of the Vransk o Lake on the Island of Cres.)
Gra0ninar 44/8: 521-530.
Fig. 1 Location map
Fig. 2 Annual values of rainfall, water levels and water pumpings from the Vrana Lake
Fig. 3 Ratio of recorded and calculated water levels according to multiple regression analysis
Fig. 4 Annual water levels in real conditions, withought pumping, and with pumping of 0,150 m3s-1, and 0,250 m3s-1