ISOTOPES AS NATURAL TRACERS IN THE WATER CYCLE: THE CARPATHIAN BASIN
Isotopes as natural tracers in the water cycle: examples from the Carpathian Basin
István Fórizs
Laboratory for Geochemical Research, Hungarian Academy of Sciences
H-1112 Budapest, Budaörsi út 45., Hungary,
Abstract: Environmental isotopes are very good tracers for studying the subsurface water flow systems, and to determine the origin of water or pollutants. The stable isotope composition of precipitation linearly correlates with the air temperature, so the water infiltrated during the Ice Age (>10000 years ago) into the subsurface aquifers is characterized by an isotope composition different from that infiltrated during the Holocene (>10000 years) providing a tool to study groundwater mixing processes and vulnerability of water resources. The stable oxygen isotope composition of the Danube River shows seasonal variability giving a possibility to determine the Danube water transit time from the river to the production wells installed along the riverbank. The stable isotope composition of groundwater together with age determination can be used for paleoclimate reconstruction.
Introduction
Environmental isotopes are the best tracers in the natural water cycle, and at the same time they are environment friendly, because introduction of artificial tracers into the flow system is not necessary. These isotopes can be grouped by different point of views, e.g. 1) stable or radioactive; 2) incorporated in the water molecule or not incorporated in the water molecule. Those incorporated in the water molecule gives information about the water itself (origin of water, water-rock interaction, mixing, evaporation, etc.), while the others give information about the solutes (including origin and fate of pollutants) or physical conditions during the infiltration of water.
The major stable environmental isotopes: 1H, 2H (D), 3He, 4He, 6Li, 7Li, 10B, 11B, 12C, 13C, 14N, 15N, 16O, (17O), 18O, 20Ne, 22Ne, 32S, 34S, 35Cl, 37Cl, 79Br, 81Br, 86Sr, 87Sr.
The major radioactive environmental isotopes: 3H (T), 14C, 37Cl, 39Ar, 85Kr, 129I, 222Rn, 226Ra, 230Th, 234U, 238U.
Among the above isotopes the most commonly used ones in hydrogeology are 2H/1H, 18O/16O, 13C/12C, 15N/14N, 34S/32S (stable isotopes); T, 14C (radioactive isotopes).
Terminology
Stable isotope composition is always expressed in a special way as (delta) value, what is a dimensionless number in parts per thousand:
Rsample - Rstandard
= 1000 [‰],
Rstandard
where R denotes the ratio of the heavy to light isotope (e.g. 2H/1H or 18O/16O) in the sample and international standard, respectively. For the major material types international standards with known isotope compositions are used, e.g. for water the international standard is the Vienna Standard Mean Ocean Water (VSMOW) distributed by the International Atomic Energy Agency. For instance oxygen composition of a water sample is reported as 18O value in [‰]VSMOW, and in the case of other elements it is done in similar way (D, 13C, 15N, etc).
The stable isotope composition is usually measured by gas-source mass spectrometers, less frequently by solid-source mass spectrometers.
Concentrations of radioactive isotopes are reported in different way regarding different isotopes. For examples: 1) tritium is reported in tritium unit, TU, where 1 TU = 1 tritium among 1018 hydrogen atoms; 2) 14C is reported in percent modern carbon, pmC, where “modern carbon” is defined as 100% 14C activity of a wood grown in fossil-CO2-free environment (just before the CO2 emission from firing coal and crude oil increased by magnitudes, for more details and other isotopes see [1]).
The content of radioisotopes are usually measured by scintillation methods, but recently the accelerator mass spectrometry is getting to be more widely used as before.
The water cycle
The majority of the water is accumulated in the oceans. Clouds are formed from the vapour coming mainly from the oceans and in less amount coming from the terrestrial areas by evaporation and transpiration. Liquid (rain) and solid (snow) precipitation fallen to the surface partly gets back to the oceans through rivers, and partly infiltrates into the ground (a very small part evaporates). The infiltrated groundwater moves from the recharge (infiltration) area through the flow path to the discharge area, where it gets back to the surface. These flow paths vary greatly, in some of them water spends only few days while in others tens of thousands of years. During the whole way of the water both the water molecules and the solutes have characteristic isotope composition determined by the physical and chemical processes and the rock-water interaction. If we know these characteristics then measuring the isotope composition of a water sample and/or its solutes, we can tell a lot about them.
Isotopic characteristics of the precipitation in the Carpathian Basin
There is an indirect linear relationship between the stable isotope composition of the precipitation (D, 18O) and the mean annual or monthly temperature of the air (T) on global scale and local scale as well (see e.g. [1] pp. 64-65, or [2]). The decreasing temperature comes together with decreasing 18O and D values, with other words as the temperature of the air decreases the precipitation becomes more depleted in heavier isotopes of hydrogen and oxygen.
In the Carpathian Basin the first study on the T-18O and T-D relationships was made by Deák [3] who measured the stable isotope composition of precipitation collected at Abádszalók (Great Hungarian Plain) between 1977 and 1988. He found that δ18Omonthly=0.37*Tmonthly–12.8‰, where T is temperature in ºC (Fig. 1). Palcsu et al. [4] have got similar correlation for precipitation collected in Debrecen (Hungary) in the years 2001 and 2002: δ18Omonthly = 0.28·Tmonthly – 11.6‰. These relationships are not far away from the global relationship: δ18Omonthly = 0.338·Tmonthly – 11.99‰ ([1], p. 64).
Fig. 1 The linear relationship between the mean monthly air temperature and the δ18O value of the precipitation collected at Abádszalók Meteorological Station (Hungary) between 1977 and 1988, [3].
According to this relationship the stable isotope composition of precipitation at a location varies seasonally. The precipitation in wintertime is isotopically very light, while summertime it is rather heavy. Deák measured δ18O values between 2‰ and 22‰ [3].
Considering the mean annual stable isotope composition of precipitation, if the climate changes it changes as well mostly according to the above relationship. As a result of this rule, there is a great difference between the stable isotope composition of today’s precipitation and that of the Ice Age, when the temperature was considerable lower (for details see chapter “Groundwater”).
The stable isotope composition of the infiltrating precipitation is rather close to the mean annual isotope composition of the precipitation. In the Carpathian Basin the bigger portion of the precipitation falls in spring and autumn and infiltration occurs mostly in these seasons. Summertime, when the precipitation is isotopically heavy, the infiltration is not characteristic (it occurs rarely), therefore the average stable isotope composition of the infiltrated water is expected to be characterized by a little bit more negative δ18O value than that of the precipitation. And indeed, Deák has got average δ18O value of 9.1‰ for precipitation of 12 years (1977-1988) and 9.3 ±0.4‰ for shallow groundwater in Hungary [3]. It is worth mentioning that the difference between the two values is within the uncertainty of the measured numbers. So we do not make big error if we use the same δ18O value for the fallen and for the infiltrated precipitation in the Carpathian Basin for paleoclimate reconstruction.
Groundwater
Subsurface water resources are very important from the point of view of drinking water supply. The Carpathian Basin has a unique morphology; high mountains in an almost perfect circular form surround the basin. Water infiltrated in the mountains moves through the rock fissures and in the sedimentary layers to the lower parts, to the lowlands. These flow systems are discharged by the rivers, or on the Great Hungarian Plain there are places, where the old water gets to the surface forming shallow lakes or gets close to the surface. In both cases the majority of the water evaporates and the result is salt accumulation on the surface or in the soil. The transit time or the flow through time of the regional flow systems in the Carpathian Basin is usually few tens of thousands of years (20-40 ka). The intermediate and local scale flow systems have shorter, sometimes much shorter flow through time.
The latest glaciation (Ice Age) ended cca. 10000 years ago, when the temperature changed a lot, raised more or less to the present level. According to this climate change there is a great difference between the δ18O values of the groundwater infiltrated in the Ice Age and the groundwater infiltrated during the latest 10000 years (Holocene). In Hungary the δ18O value of the Holocene infiltrated groundwater varies between 9‰ and 10‰ depending on the geographical area, while that of the Ice Age infiltrated groundwater varies between 11‰ and 14‰ depending on the time of infiltration and also on the geographical area.
Usual problem is the overexploitation of groundwater, when the amount of abstracted (exploited) water is higher than what the aquifer can provide, and as a result, the potentially or actually polluted near surface water flows down to the aquifer and mixes to the deeper water. In the Carpathian Basin in many cases this deep groundwater was infiltrated in the Ice Age, while the near surface water is a young Holocene water, so the stable isotope compositions of the two types of water differ, and monitoring the 18O value of the exploited water allows us to catch the arrival of the down-flowing water far before the arrival of any pollution front.
A good example for this phenomenon is the Újkígyós Regional Water Works situated on the Great Hungarian Plain. The production wells had been installed into the Maros/Mureş alluvial fan close to the villages Újkígyós and Szabadkígyós [5]. In the original conditions the area is a regional discharge area of the flow system in the Maros/Mureş alluvial fan. The piezometric head of the deeper groundwater was higher than those of the shallow ones, so the deeper groundwater, which is an old, Ice Age infiltrated groundwater, could get to nearby the surface, where it mixed with the infiltrating modern precipitation. The start of the exploitation changed the conditions a lot in the depth range of 0-80 meters (see Figure 2). The piezometric head (water pressure) in the exploited aquifers decreased so much that the shallow groundwater (in some places polluted by nitrate) could seep down and mix to the ascending old groundwater.
Fig. 2 The 18O value of the groundwater vs. depth in the area of the Újkígyós Regional Water Works in 1992.
Below 100 m down to 400 m the 18O values of the groundwater varies between 12‰ and 13‰. This is the original old groundwater (20-30 ka according to radiocarbon age determinations). This water mixes with the descending shallow groundwater resulting in 18O values between 10‰ and 12.2‰ in the depth of 20-80 meters. The multiannual mean 18O value of the infiltrating precipitation is 9.3‰.
Rivers and riparian resources
Riparian water resources along the river banks also have great importance, because many people are supplied with drinking water from this kind of water resources including, for example, the inhabitants of Budapest (almost two million people), capital of Hungary.
The catchment area of the Danube River upstream from the Great Hungarian Plain is mostly the high elevation Alps, so the Danube water is isotopically lighter (mean 18O value is 11‰) than the local meteoric water (9.3‰) providing a tool for tracing the flow of river water into the riparian aquifers and its mixing to the background water, latter one is infiltrated from precipitation.
Fig. 3 The 18O time series of the Danube River at Budapest and of the Halásztelek-5 (H5) production well on the Csepel Island (south of Budapest) in the years of 1998-2001.
The stable isotope composition of the Danube River has special feature. It has a seasonal variation, which is almost the opposite to that of the precipitation. The most negative 18O values occur usually in May-July (see Fig. 3), because the precipitation in the high elevation Alps fallen in wintertime (in the form of snow) melts and gets to the Danube in this part of the year. In wintertime on the high elevation parts of the catchment area the air temperature is continuously below the freezing point, so the precipitation is snow, which does not get to the river in this time, so the Danube river at Budapest has a water originating only from the shallowest groundwater of the lower part of the catchment area. This groundwater is recharged from the whole year precipitation, therefore the 18O value of the Danube River in wintertime is more positive than in summertime.
This seasonality in the stable isotope composition of the Danube River provides a good tool to calculate some hydrodynamic parameters of the riparian flow system. Production wells close to the riverbank exploit mostly river water (close to 100%). If we measure the time series of the stable isotope composition of the exploited water and compare it to that of the river water, then we can calculate the transit time of the river water to the production well. Figure 3 shows a good example, where the time series of the 18O value of the H5 production well has a delay comparing to that of the Danube. This delay is shows the transit time. Beside this delay, the amplitude of the 18O curve of the H5 smaller, which is related to the dispersion phenomenon. Applying appropriate mathematical formulations describing the dispersion, we can calculate the transit time with an uncertainty of ±1-2 days [6].
The average 18O value of the Danube for the period of 1998-2002 at Budapest is 10.9‰, which differs from 18O value the infiltrated precipitation in Hungary. On the Csepel Island (a river island south of Budapest) the shallowest groundwater is exploited for providing drinking water. This water is a mixture of river water and background water, latter is an infiltrated precipitation. In this area the 18O value of the infiltrated precipitation is between 9‰ and -9.5‰, which differs considerable from the mean 18O value of the Danube water. Using a two component mixing model based on the 18O value of the shallowest groundwater of the Csepel Island we can calculate the mixing ratio of ‘river water’/’background water’ [7]. These data are very important for the validation of the three dimensional hydraulic model of the area under construction at the Budapest Technical University.
Paleoclimate applications
When we know the relationship between the temperature and the stable isotope composition of the infiltrating water, then we can use it for paleotemperature reconstruction. Measuring the stable isotope composition of groundwater of a regional flow system (any flow system with a long enough flow through time), we can calculate the air temperature of the infiltration area for the time of infiltration. If we measure the age of water by radiocarbon (or other) method, then we can reconstruct the air temperature variation of the infiltration area in the past. Since temperature is one of the main parameters of the climate, in this way we can infer climate change happened in the past.
Figure 4 shows the stable oxygen isotope composition vs. calculated radiocarbon age of some groundwaters of deep aquifers on the Csepel Island (river island south of Budapest). It is interesting to notice that samples with the most negative δ18O values are not the oldest ones. Both younger and older samples have less negative δ18O values. Although the radiocarbon age determination has big uncertainty (uncertainty originating only from the analytical error is indicated on the figure 4, uncertainty of the correction method is not indicated, but it is bigger), the variation of δ18O values by time showed on Fig. 4 clearly demonstrates that the temperature changed in that period of time from less cool to cooler and back to less cool. We must know that the mean temperature was not the same during the Ice Age. There were cooler and less cool periods. The last cooler period was the Würm III between about 29000 and 18000 years BP (before present). The three points, which lie in the Würm III period have more negative δ18O values than those, which lie outside of this period. One point on the left side and one on the right side are outside the Würm III. From the δ18O-T relation [3] we calculated the mean annual air temperature of the infiltration area (Fig. 4). If the δ18O-T relation determined in the 20th century was valid in the Ice Age as well, then we can conclude, that the mean annual temperature in a part of the Würm III was lower than zero °C.
Fig. 4 The stable oxygen isotope composition and calculated air temperature vs. calculated, 13C corrected radiocarbon age of some old deeper groundwater on the Csepel Island (south of Budapest).
The stable isotope composition of river water reflects the climate of its catchment area. Although the river water flows away, it does not remain in time, shells of molluscs living in the river are in somehow isotopic equilibrium with the water, so the isotopic composition of the remaining carbonate shells in the river sediments can be a good proxy for the climate of the catchment area.
Since the catchment area of the Tisza River is lower than that of the Danube, the annual mean δ18O value of the Tisza is less negative than that of the Danube. Molluscs grow their carbonate shells from about April to October. If we consider the difference between the δ18O values of the two rivers in this period, it is 1.5‰ [8]. This difference is well reflected by the stable oxygen isotope composition of carbonate shells of Unionidae living in these rivers (Fig. 5). The difference between the mean δ18O values of shells of the Danube and Tisza Rivers is also 1.5‰. This nowadays difference between the two rivers is good basis for studying whether the climate on the catchment areas changed in the same way or not. We can study this problem by measuring the δ18O values of the shells in the Holocene and Ice Age sediments of the two rivers.
Figure 5 The stable carbon isotope composition vs. stable oxygen isotope composition of carbonate shells of Unionidae molluscs living in the Danube and Tisza rivers. Samples were collected in the year 2001. The big open circles indicate the average values for the Danube and Tisza groups.