[en] These proceedings contain the papers and poster presentations from the Symposium on Isotope Hydrology held 12-16 September 1983 in Vienna, Austria. The topics of the sessions were as follows: Thermal water studies, groundwater dating, hydrology of arid and semi-arid areas, field studies with environmental isotopes, precipitation-surface-groundwater relationships, pollution, artificial tracers and sediment transport. Twenty poster presentations in English have been indexed here separately. All other articles from this Proceedings Series are available under ISBN 92-0-040087-6

[en] Full text: Detailed studies of the isotopic variations in rainfall and unsaturated-zone water have been carried out at a research site on arable land in eastern England in order to investigate the mechanisms of water movement. The results also have implications for the interpretation of thermonuclear tritium profiles and the transport of pollutants such as nitrate towards the water table. The isotopic composition of rainfall has been monitored over several years during which four cored profiles of the soil and unsaturated chalk have been taken. Water has been extracted from this drill core at 20 cm intervals by centrifugation. An undisturbed lysimeter on grassland, 5 m deep, has also permitted natural drainage water to be sampled periodically. Two further profiles on permanent grassland permit comparisons for the effects of land use and lithology. Duplicated profiles generally show impressive similarity. The stable-isotope composition of unsaturated-zone interstitial water is virtually uniform below about 4 m depth. Drainage from the lysimeter is mostly homogeneous isotopically and chemically, although there is a rapid hydraulic response to rainfall; infrequent departures from uniformity may represent unusual infiltration events. Above 4 m, the profiles show light isotopic compositions at intermediate depth and marked enrichment of heavy isotopes in the top few tens of centimetres, except in one case in which the opposite occurs. The isotopically heavy water in the top zone shows very little evidence of evaporative modification and is predominantly from unmodified rainfall. The scale of variations and the isotopic balance at this site suggest that the penetration depth by infiltrating rainfall varies greatly with time and season. The presence of a zone of mixing, extending several metres down from the surface, does not permit a 'piston-displacement' mechanism to be applied. Below this zone of mixing, however, the diffusive interchange of water held in fissures and pore spaces with different sizes and hydraulic conductivities ensures an approach to isotopic homogeneity. The results of this study suggest that infiltration below permanent grassland at this site is about 5 per mille δ²H lighter than infiltration below adjacent arable land. There are insufficient data to give a valid comparison between these compositions and the average for rainfall, though it is observed tentatively that both infiltration below grassland and saturated-zone water have isotopic compositions closer to the composition of average rainfall (1979-82) than to that of infiltration below arable land. (author)

[en] Full text: Techniques in use at Chalk River for studying the hydrology of existing and potential waste management sites include: (a) devices to scan radioactivity in situ and to obtain samples or to inject tracers into groundwater systems; and (b) isotope measurements to follow dispersion, to differentiate between surface water and groundwater, and to obtain information on the age of the ground water. Groundwater discharge into a lake bed is measured by a 'seepage meter' consisting of the end of a steel drum which is embedded in a lake bed and vented to a plastic bag which inflates with the incoming water. Groundwater flow near the shoreline of a lake has been studied using several arrangements for injecting tracers in a uniform plane and collecting samples in the flow path. For sampling in a shallow borehole in rock, a string of simple inflatable packers with sampling intervals is constructed from readily available materials. Dispersion in a sand aquifer has been studied using ¹³¹I tracer; this has been monitored over 20 m of travel by gamma-scanning an array of dry PVC pipes in the flow system. This enabled detailed observation of tracer flow without the labour of discrete sample analysis and with no disturbance due to withdrawal of water from the flow system. Results from grain-size analysis and hydraulic conductivity measurements indicated the aquifer to be homogeneous, but the in situ scanning technique revealed three layers of different velocity. Within discrete layers, observed and calculated breakthrough curves yielded surprisingly low dispersivities (∼1 cm), in agreement with laboratory measurements. Measurements of environmental ³H, ¹⁴C, uranium isotopes and deuterium are being used to characterize groundwater discharging from bedrock and to estimate its age. Detectable tritium is taken as an indication of the presence of recent surface-water input, either residual drill fluid or leakage from the surface. Carbon-14 is measured by CO₂ absorption in an amine-type liquid scintillation cocktail when sufficient sample is available. The accelerator mass-spectrometry technique has been used for measuring samples of a few milligrams, e.g. from groundwater carbonate of 0.5 L samples taken in down-hole samplers or from fracture-coating calcites. (author)

[en] Full text: In many densely populated regions the groundwater resources can no longer meet the demand for drinking water. As a result, many water catchment areas using bank filtration water are being installed near rivers. In special cases the groundwater supplies from such catchment areas can be increased by artificially induced infiltration of river water. In three case studies the infiltration rates and flow parameters of the bank storage in relation to the extracted groundwater are determined using environmental isotopes in combination with the usual hydrological methods. (1) The water supply area of about 2 km² is situated in the confluence area of two rivers. The aquifer consists of sandy gravel layers of a thickness of 6-8 m. The ¹⁸O data from pumping and observation wells enabled the quantitative calculation of the relative proportions of infiltrated river water and their spatial distribution. The proportions of water in the pumping wells from each of the rivers were determined as a function of time, using the isotope curves for the pumping wells and for both rivers. Proportions of between 50 and 100% of specific river water have been found in the different pumping wells, with an estimated accuracy of ±5%. Correlation with results of pumping tests led to a localization of the infiltration areas. (2) The second case concerns a water catchment area on an approximately 0.3 km² large river island. The changing ¹⁸O content of the river water is well suited for the determination of the water infiltrated into the tapping wells and its flow velocities. Assisted by the large number of observation wells, it was possible to draw a detailed map of the proportions of river water in the groundwater (40-100%) and to determine flow velocities of between 2 and 15 metres per day. The results obtained by isotope measurements were supported and confirmed by dye tracer tests and measurements of electrical conductivity. (3) In the case of an artificial groundwater enrichment by injected river. water the amount and transit time of infiltrated groundwater were estimated by means of ²H data. Using a simple mixing equation, a proportion of 90 ± 5% of artificially infiltrated water in relation to the autochthonous groundwater was calculated for a pumping well. Utilizing hydrological models together with the annual isotope variation, the mean transit time was determined to be between two and four months. (author)

[en] Full text: The variations of oxygen-18 in four Swedish rivers are presented as monthly values for the period 1978-1982 and as daily values during the snow-melt period of 1980. Also included are the variation of electrical conductivity for the latter period as well as river discharges. There are obvious seasonal patterns in the oxygen-18 variations. The northern rivers show fairly constant oxygen-18 values, except for a marked decrease at snow melt in May. The southern rivers display more sinusoidal variations, with a minimum in spring and a maximum in late summer and autumn. This difference is partly due to different runoff regimes. In the north, all winter precipitation is accumulated as snow and does not affect runoff until snow melt. In the south, there are quite substantial autumn and winter rains in addition to a less pronounced spring flood. Other factors influencing the oxygen-18 variations are the catchment area and its physiographical characteristics, the number of lakes and the climate. These govern the ratio of groundwater contribution to direct runoff, response time, damping and heavy-isotope enrichment by evaporation. The short-time variations show that the decrease of oxygen-18 in spring is an immediate response to the increasing discharge. The electrical conductivity varies in a similar way, with low values during the spring flood and high values when the discharge is smaller. (author)

[en] Full text: In the United Kingdom, hydrological studies have in general neglected the use of stable isotopes as natural tracers, possibly because without the extremes of climate and/or relief found elsewhere, successful application is less certain. Such assumptions, however, have apparently not been supported by systematic investigations, and the purpose of the work reported here is to attempt to obtain methodical information by studying ¹⁸O/¹⁶O isotope relationships between natural waters in a selected area. Malham in North Yorkshire, an area of horizontal karstified limestone, was chosen because of its short throughput times, the presence of a significant altitudinal range (185-645 m O.D.) and the availability of information from previous conventional hydrological studies. The nature of the δ¹⁸O_SMOW input signal and its correlation with surface waters have been examined over short and extended time-scales (up to 2 1/2 years). On a monthly basis, the values are seen to follow a definable structure with some seasonal trends, normally within the range -4 per mille to -10 per mille. Examinations of weekly and within-storm sample measurements, however, reveal a wide range of values (at least 12 per mille), the details of which cannot be observed in a single month's rainfall collection. An average isotope depletion rate of 0.11 per mille per 100 m is discernible in summer months, although at other times the altitude effect appears erratic. On recharge, a rapid smoothing of the δ¹⁸O_SMOW input signal appears to occur, shown by a subdued and variable response in surface waters, which exhibit an average base-flow value of approximately -7.8 per mille. Unexpectedly, the most well-defined seasonal signal is found in the lake (Malham Tarn) which is fed by streams from the upper catchment, although only a weak correlation is found between the δ¹⁸O_SMOW values in these waters and in the previous month's rainfall (r=0.36). The naturally labelled lake water is clearly identified in lower-altitude springs, to which it is known to contribute. The definition of the signal is assumed to be due to evaporation, despite the lack of seasonal temperature extremes in the area. It is concluded that, although for many applications the use of stable isotopes is inevitably limited in the United Kingdom, the nature of the seasonal signal in the lake suggests an important application for studies of leakage from, or interconnections with, surface storage systems. (author)

[en] Full text: Significant variations in pH, molality of HCO₃^- and ¹³C/¹²C ratios of total dissolved inorganic carbon have been observed in samples taken at 24 points in the period January to May 1977. The sampling area is part of the southwestern Lublin Upland in which Cretaceous and Tertiary bedrock limestone is overlain by Pleistocene aeolian rock and/or Holocene alluvia. The observed values of mHCO₃^- and ¹³C/¹²C plotted versus pH show small and multi-directional variations in winter/early spring and a generally increasing trend in late spring. The observed variations are for a period in which the limestone aquifer was fed by water which penetrated the soil under different conditions. In late spring the feeding of the aquifer came to an end and the trends observed approached the evolution lines theoretically predicted by Deines et al. (Deines, P., Langmuir, D., Harmon, R.S., Geochim. Cosmochim. Acta 38 (1974) 1147). Applying a theoretical model, the δ¹³C of the soil-generated CO₂ is estimated as -21 to -25 per mille and the partial pressure of CO₂ as 10^-1.0 to 10^-1.6 atm. (author)

[en] Full text: Environmental isotope techniques (³H, ²H or ¹⁸O) can contribute to making the runoff and storage behaviour of Alpine catchment areas more transparent by means of runoff component separations and mean residence time calculations. Runoff components were separated for the following single events: snow melt, snow melt plus rain, and storm. The direct component was defined as having the isotope content of the actual input. Snow-melt events were found to yield the most reliable results, provided snow lysimeters are available (Herrmann, A., Martinec, J., Stichler, W., in Proc. Mtg. Modelling of Snow Cover Runoff, Crel, Hannover (1979) 288). This is mainly due to the isotopically fluctuating but always light inputs from snow-cover outflows compared with the stable-isotope contents of the pre-event waters or the indirect component. Direct runoff proportions can reach 30 % during snowmelt events and 50% during storms. Subsurface water is, therefore, considered to play a dominant role in peak discharge formation. During the main ablation season in glaciated areas the dominant role can be attributed to the ice meltwater component, which is free of tritium, In this special case, more than two components were separated. Additionally, a quick and a delayed snow/firn melt-water component were discriminated by stable isotopes, and the mineralized groundwater fraction was determined by measurements of the electrical conductivity (Behrens, H., et al., in Isotope Hydrology 1978 (Proc. Symp. Neuherberg, 1978), Vol.2, IAEA, Vienna (1979) 829. Annual data of the stable-isotope output allow distinct runoff seasons to be distinguished which have positive or negative deviations from the isotopic mean and are of different length. A comparison with the isotopical input curves of the respective time intervals permits the calculation of seasonal direct runoff proportions of between 30 and 50%. This confirms the results from the single-event analyses. The mean residence times T_r of the subsurface reservoirs were derived from comparisons of the ³H and ²H or ¹⁸O input and output functions, the specific hydrological models and the model age distributions applied. T_r varies between 1.5 and 4 years in the studied areas, the results from stable-isotope and ³H data being in good agreement (Maloszewski, P., Rauert, W. Stichler, W., Herrmann, A., J. Hydrol. 66 (1983)). By introducing hydrological data, such as mean and minimum stream flow, it was possible to discriminate between two separate sub-reservoirs with different T_r values, i.e. 0.1-1 year and 4-7 years. The combination of hydrological and isotopical investigation methods enables a new insight into the water balances of the studied catchment areas. For example, the infiltration rates amount to about 50% of the precipitation depths, allowing for an evaporation loss of 25%. The subsurface storage volumes are of the order of 10⁸ times the discharge, with T_r of 2.5 years. Assuming effective porosities of 20%, the calculated mean aquifer thickness is of the order of 10-15 times the precipitation depths. (author)

[en] Full text: The Crocodile River in the Transvaal, South Africa, forms part of an extensive irrigation scheme which serves farmers along more than 100 km of the river course. Occasional accidental releases of crop-damaging pollutants from an industrial area situated upstream of the river have resulted in the establishment of a monitoring station to warn farmers against the presence of excessive levels of pollutants in the river. Effective implementation of the system required information about transit times and dilution characteristics. These were consequently determined by means of a radioactive tracer study. The experimental work involved the release of 20-50 GBq of ⁸²Br into the river at two positions and the monitoring of downstream irrigation channels. Samples were collected and counted on a NaI(Tl) detector, and the required information was obtained from the analysis of the resulting tracer response curves. Mean transit times were calculated directly from the weighted means of the response curves. Estimates of the degree of mixing were made in terms of a dimensionless Peclet number, using a mathematical flow model and a non-linear regression technique. The results showed that a single value of the Peclet number could be used to predict downstream concentrations by means of the flow model to an accuracy of 25% in widely varying situations. This accuracy was considered sufficient for the intended purpose of these predictions. Finally, a semi-empirical relation between transit time and river flow rate was established, which made the prediction of downstream pollutant behaviour possible in terms of the transit time, the time taken by a pollutant to pass a specific point and the expected maximum concentration. (author)

[en] Full text: Observations of δ³⁴S and δ¹⁸O of the sulphate dissolved in rain and snow were carried out in 1979. These data, combined with the δ¹⁸O and pH values of water, and the SO₄^2- and Cl^- concentrations, provided information on the way of SO₄^2- formation in the atmosphere: (1) High and variable sulphate concentrations (from 1 to 19.3 mg/L) and low Cl^- content (below 3 mg/L in most cases) point to continental derivation of the sulphate. (2) Nearly constant values of δ³⁴S (+3.7 per mille) represent an average isotopic composition of numerous SO₂ emitters distributed over a large area. (3) The δ¹⁸O value of the sulphate is significantly correlated with that of water (r = 0.62, n = 32). The following regression line has been derived from the collected data: δ¹⁸O_sulphate = 16.0 ± 0.9 + (0.38 ± 0.07) δ¹⁸O_water. The slope and intercept of this line depend on the season. In winter, the values approach 1/4 and 14, respectively, for slope and intercept, which confirms the industrial origin of SO₂, with an initial δ¹⁸O value of +23 per mille and the incorporation of one atom of oxygen from water into SO₄^2-. Higher values of the slope and intercept must be due to partial oxygen isotope exchange between water and SO₂/SO₃. (author)