Tamara Hunjak

Stable isotope analysis of the karst hydrological systems in the Bay of Kvarner (Croatia)

Here we present the results of the first systematic analysis of the stable isotope composition of the karst hydrological systems in the Bay of Kvarner. Gaussian mixture modelling, time series analysis and autoregressive integrated moving average (ARIMA) modelling were applied using the stable isotope compositions of the karst groundwater. This study revealed that the recharge is dominated by winter precipitation, the dual-porosity system is dominated by baseflow, the hinterlands of the individual springs have different degrees of karstification and the springs within the Rječina River catchment have higher recharge elevations than the springs in the Bakar Bay catchment.

Characterization of the Gacka River basin karst aquifer (Croatia): hydrochemistry, stable isotopes and tritium-based mean residence times.

The Gacka River basin aquifer is a highly-developed karst system, located in the Croatian Dinarides. It is mostly composed of permeable Jurassic and Cretaceous carbonate rocks, and clastic sedimentary rocks of Paleogene age. Gacka River provides high quality water for the town of Otočac and several villages; together with the neighboring Lika River, the water is used for the Hydroelectric Power Plant at Senj on the coast. About 10 perennial and over 20 seasonal springs are located at 450 to 460 ma.s.l. (above sea level). Three major springs (Pećina, Majerovo and Tonkovića) provide 57% of the mean annual river flow. Similarities between the average groundwater temperatures as well as between the average specific electrical conductivity values (9.0°C-328 μS/cm, 9.6°C-350 μS/cm and 8.9°C-312 μS/cm) of the springs imply that they are fed from aquifers with similar mean residence times (MRTs). The mean δ(18)O contents of Majerovo, Tonkovića, and Pećina are around -10.1‰, -9.2‰ and -8.9‰, respectively, revealing differences in the mean recharge area elevations. Compared to the temporal amplitude of the(18)O signal of precipitation, the (18)O signal variations of the springs are substantially attenuated because the recharges occurring at different times are well mixed within the aquifers. This indicates MRTs of more than just a few years. The average tritium contents of Pećina, Majerovo and Tonkovića are 5.48 TU, 6.13 TU and 6.17 TU, respectively. Serially connected exponential-plug type unsteady lumped-parameter models run on an annual time scale resulted in rather satisfactory matches between the observed and calculated tritium contents for all studied springs. The models revealed similar MRTs (and corresponding reservoir volumes) for Pećina, Tonkovića and Majerovo of 12 years (470 Mm(3)), 12 years (1,190 Mm(3)), and 12.2 years (1,210 Mm(3)), respectively. Plug flow conditions dominate in about 90% of the total aquifer volumes.

Stable isotope composition of the meteoric precipitation in Croatia.

The precipitation is the input into the water system. Its stable isotope composition has to be known for the proper use and management of water resources. Croatia is not well represented in the Global Network of Isotopes in Precipitation (GNIP) database, and the geomorphology of the country causes specific local conditions. Therefore, at the Stable Isotope Laboratory (SILab), Rijeka, we monitor the stable isotope composition (δ18O, δ2H) of precipitation. Since δ18O and δ2H are well correlated, we concentrate the discussion on the δ18O distribution. Together with GNIP, our database contains 40 stations in Croatia and in the neighbouring countries. Their different latitudes, longitudes and altitudes give information of great detail, including the influence of the topographic structure on the precipitation in the south-eastern part of Europe, as well as the complex interplay of the different climate conditions in the area. Within a few hundred kilometres, the stable isotope values display a significant change from the maritime character in the south (mean δ18O around−6 to−8%‰) to the continental behaviour in the north (mean δ18O around−8 to−11%‰). Depending on the location, the mean δ18O values vary with altitude at a rate of approximately−0.2%‰/100 m and−0.4%‰/100 m, respectively. Also the deuterium excess has been found to depend on location and altitude. The data are being used to construct a δ18O map for the entire area.

Comment on the paper "Definition of the river Gacka springs subcatchment areas on the basis of hydrogeological parameters" by Lukač Reberski et al., published in Geologica Croatica 66/1, 39–53 (2013)

This paper by Lukač Reberski et al. (in the following referred to as paper I) presents measurements of hydrological and hydrochemical parameters as well as results of an isotopic analysis of the spring waters and the precipitation. As the authors of paper I mention, the isotopic analysis has been performed in Graz and in Rijeka. We want to point out that the data for the year 2008 resulted from a cooperation project which was headed by Rijeka University and supported by the International Atomic Energy Agency (IAEA Vienna); the corresponding isotope results were preliminary and shown only in an unpublished and not citable report. In spite of a confi dentiality agreement, the authors of paper I took these data without our consent. This breach of trust and the somewhat sloppy data handling has some unfortunate consequences: The isotope data shown in Fig.7 appear to scatter between the Local and the Global Meteoric Water Lines, LMWL and GMWL. This was used to argue a direct correspondence between precipitation and spring recharge. Firstly, note that this LMWL is plotted incorrectly in Fig. 7; the correct LMWL (eq.5 in paper I) had to be lower by about −1‰ (i.e. shifted to more negative δ2H values) if compared to the plotted line. Therefore, the majority of data points are actually located slightly above the correct LMWL. The fi nal analysis of the 2008 isotopic data further corroborates this small but noticeable effect; as a consequence, the d-excess d = δ2H − 8۰δ18O is somewhat higher for the spring waters than that of the precipitation, supporting the conclusion that recharge happens mainly during the cold season (c.f. also MANDIĆ et al. (2008)). Furthermore, the absolute altitude values of the recharge areas (c.f. the end of section 7.3 in paper I) cannot simply be inferred from the measured isotopic ratios; the available data can only provide (approximate) values for the altitude differences. The reason is that the derivation of absolute values requires some reference data, e.g. the long-term mean δ18O of the precipitation at the spring altitude (more precisely, taking into account the mean residence time of the ground water as well as the infi ltration into the catchment); or a reliable value of the absolute altitude of at least one of the recharge areas (assuming that the soil surface and thus the infiltration is of approximately the same nature for all catchments). Without further discussion neither is really known; for example, depending on the season, the δ18O values of the collected and analyzed rainfall in the Gacka spring region in the years 2006 and 2008 vary between approximately −4‰ and −13‰. The authors do not provide any clue for the reference data they used. The publication of the fi nal analysis of our results, together with an estimate of the mean residence time of the ground water as obtained from tritium model calculation, is presently under preparation.