Environmental

Isotope Hydrology
Environmental isotope hydrology is a relatively new field of
investigation based on isotopic variations observed in natural waters.
These isotopic characteristics have been established over a broad
space and time scale. They cannot be controlled by man,
but can be observed and interpreted to gain valuable regional
information on the origin, turnover and transit time of water in the
system which often cannot be obtained by other techniques. The cost
of such investigations is usually relatively small in comparison with
the cost of classical hydrological studies.
The main environmental isotopes of hydrological interest are the
stable isotopes deuterium (hydrogen-2), carbon-13, oxygen-18, and the
radioactive isotopes tritium (hydrogen-3) and carbon-14. Isotopes
of hydrogen and oxygen are ideal geochemical tracers of water
because their concentrations are usually not subject to change by
interaction with the aquifer material. On the other hand,
carbon compounds in groundwater may interact with the aquifer
material, complicating the interpretation of carbon-14 data.

A few other environmental isotopes such as 32Si and 2 3 8 1 / / 2 3 4 U have

been proposed recently for hydrological purposes but their use
has been quite limited until now and they will not be discussed here.

Stable Isotopes of Hydrogen and Oxygen in the Hydrological Cycle

The variations of the isotopic ratios D/H and 18 O/ 16 O in water samples are expressed in
terms of per mille deviation (6%o) from the isotope ratios of mean ocean water,
which constitutes the reference standard SMOW:

5%o= (^
R
SMOW

The isotope ratio, R, is measured using a special mass spectrometer.

The principal value of stable isotope data of a hydrological system is to identify its recharge
area on the basis of general knowledge of the isotopic variations of precipitation and
surface waters. The most important natural processes that cause variations of the
stable isotopic composition of natural waters are evaporation and condensation. The light
molecules of water, H 2 1 6 0, are more volatile than the heavier ones, so on cooling of
atmospheric moisture the heavy molecules condense preferentially, leaving a

Tracing water sources in the volcanic island of Cheju in the Republic of Korea . . . the Chungji Fall, ^
Sampling Point No. 5.
42
residual vapour more and more depleted in heavy isotopes. Such sequential condensation
occurs as air masses move inland from the sea or rise to higher altitudes. Because the
residual vapour requires a lower and lower temperature for further condensation,
a relationship between temperature of condensation and isotopic composition of
precipitation is observed; as the condensation temperature decreases, the heavy isotope
content of the precipitation is lower.
This dependency on temperature produces (i) seasonal isotope variations of precipitation
(winter precipitation is depleted in heavy isotopes with respect to summer precipitation),
(ii) latitude variations (high latitude precipitation is depleted with respect to low
latitude precipitation) and (iii) altitude variations (the heavy isotope content
of precipitation decreases with increasing altitude). The last effect is especially important in
regional hydrological studies, where for instance groundwaters deriving from recharge
areas at different elevations may be differentiated.
When precipitation infiltrates to feed groundwater, mixing in the unsaturated zone smoothes
the seasonal isotopic variations so that water in the saturated zone has a composition
corresponding to the mean isotopic composition of infiltration in the area. Thus,
the isotopic composition of groundwater is related to that of precipitation in the recharge
region of the aquifer at the time of recharge and may serve to identify the area of
recharge on the basis of altitude or distance from the sea. One must consider
any climatic differences of the time of recharge that could have produced a different
isotopic composition of the original infiltration from that occurring at the present time.
Groundwater recharge by lateral seepage from rivers and lakes or by vertical infiltration
from ponded waters may be identified since such waters often have a different
isotopic composition than local precipitation. Rivers may bring in water
precipitated at high altitude, hence depleted in heavy isotopes. In ponds or lakes, the water
may be considerably enriched in heavy isotopes through evaporation and they will
deviate from the D-18O relationship usual for precipitation.

Tritium in the Hydrological Cycle

Tritium is produced in the atmosphere by cosmic radiation and has been released in
thermonuclear tests. Most of it is rapidly oxidized to HTO and incorporated in the hydro-
logical cycle where it constitutes a very useful marker for water that has been in the
atmosphere within the past 20 years. The great dilution by H2O results in very
low concentrations which can be measured only by means of tritium's radioactivity, usually
after an isotopic enrichment treatment (IAEA Bulletin Vol. 15, No. 1 1973). The
tritium content of natural waters is expressed in Tritium Units (TU). One
Tritium Unit corresponds to a concentration of 1 tritium atom per 10 !8 hydrogen atoms.
Cosmic radiation establishes a concentration of about 10 TU in temperate zone continental
precipitation. Large amounts of man-made tritium have been released to the atmosphere
by thermonuclear tests since 1953. Concentrations up to 10,000 TU were reached
in northern hemisphere precipitation in 1963. Since then concentrations have decreased as
a consequence of limitations on the explosion of thermonuclear devices in the
atmosphere. Considerable seasonal and geographical variations of the
tritium content of precipitation result from the location of the major releases (the northern
stratosphere) and the mechanisms of transfer from the stratosphere to the troposphere
44
to surface waters. The IAEA publishes data on the concentration of tritium in
precipitation collected at stations around the globe, from which tritium deposition at most
places may be estimated.
In hydrological studies, tritium measurements give information on the transit or turnover
time of water within a system. In a confined groundwater system, tritium concentrations
may be correlated with the known tritium deposition modified by considerations of
seasonally-selected recharge and dispersive mixing effects within a granular
aquifer. Qualitatively, concentrations of less than 2 or 3 TU at a northern continental site
indicate that the water was recharged before the release of thermonuclear tritium, i.e.
before 1953, whereas concentrations above 20 TU indicate a much more rapid
flow of water through the system. In an open system, surface or subsurface, mixing of
waters of different ages predominates and tritium concentrations are interpreted in
terms of a Mean Residence Time in a mathematical model consisting of a mixing cell or
combination of cells that are considered reasonable from the geohydrological
standpoint.

Carbon Isotopes in the Hydrological Cycle

Like tritium, carbon-14 occurs in the environment as a result of cosmic ray irradiation of the
atmosphere and release in nuclear tests. It is oxidized to carbon dioxide, mixes with the
carbon dioxide of the atmosphere and enters the global carbon cycle. Its longer
half-life (5730 y) makes it of interest in studying groundwater systems where transit times
may be very long. Bomb carbon-14 is not normally of interest for hydrological
studies since tritium is a much better marker for recent waters.
The use of 14C for dating groundwater is based on the fact that soil zone carbon dioxide is
of biogenic origin, resulting from the respiration of plant roots and plant decay and
hence contains 14C derived by plants from the atmosphere. This biogenic carbon
dioxide dissolves in infiltrating water and is carried down to the groundwater reservoir. Its
14
C content decreases by radioactive decay and the fraction of the original remaining
is a measure of the time since it was removed from the soil zone, i.e. the time since
infiltration of the associated water.
14
C is measured relative to the total carbon content of the sample, so one must consider
the origin of both the 14C and the stable carbon of the sample. Not all of the stable
carbon of groundwater carbonate is of the same origin as the 14C. Infiltrating
water, containing carbon dioxide dissolved from the soil zone, dissolves carbonate minerals
in the soil. However, the carbon originating from limestone in general contains no 14C,
so that the water reaching the water table contains dissolved carbon (in the chemical
forms of H2CO3, HCOJ" and COj") with a 14C content lower than that present in the soil
biogenic CO2. The evaluation of the dilution of soil CO2 originally containing 100% of
modern 14C with 14C-free carbonate to estimate the initial 14C concentration in recharge
water reaching the water table constitutes the most difficult problem in the 14C
age determination of water.
Experience has shown that 85% of modern 14C is a good average value for the content of
recent, prethermonuclear infiltrating water in central Europe. In an attempt to obtain
a better estimate of this value at specific sites, the stable 13C content of the
groundwater carbonate sample may be studied. Since biogenic carbon dioxide has a
significantly lower 13C content than limestone, one could hope to estimate the proportions
45
of biogenic and limestone carbon from the 13C content of the groundwater carbonate.
However, knowledge of the 13C content of plants in the place and time (thousands
of years ago) of recharge is usually very limited. Furthermore, it is recognized that the
carbon isotope ratios can be affected by much more complicated groundwater
chemistry than the simple carbonate system considered above.

The 14C method can be used for waters younger than 30,000 years. In general, it is applied
to study the movement of water in confined aquifers. Where recharge occurs only in
the outcrop area and the water chemistry and isotopic composition of dissolved
carbon species are relatively uniform, the age differences in space are not affected by the
uncertainties which affect the absolute age determination of water. Thus, it is
possible to determine the flow velocity of water, by determining the age
differences between two sampling points at known distance. This proves the hydrologist
with an estimate of the mean regional permeability. 14C measurements, especially in
association with tritium measurements, may also give information on mixing
processes of waters of different ages within a given aquifer.

Examples of Application
These general principles have been put to practical use in many parts of the world during the
past few years in projects supported by the IAEA at the request of the Member States
in whose regions hydrological problems occured.
For example, a project in the Hodna region of Algeria concerns a study of the groundwater
in two aquifers in the region of a dry salt lake, Chott-el-Hodna, where flood waters from
the nearby mountains evaporate completely. The heavy isotope content of water in
the deeper aquifer is lower than that in the shallow aquifer, despite the fact that the two-
aquifers are presumed to have been recharged from the same general area. However,
carbon-14 measurements giving ages of the order of some 30,000 years for the
deep aquifer, indicate that the difference in stable isotopic composition is most probably due
to an age effect (climatic change since the time of recharge). The deep aquifer is
connected with the shallow one through buried alluvial fan (revealed by geophysics). In
this area, the 14C content of water jn the deep aquifer increases together with the
18
O and C content. This is a clear indication of slow recharge from the shallow
aquifer. In this study tritium measurements also demonstrated the occurrence of recharge
of the shallow aquifer from wadis.
Stable isotope and tritium analyses were used in the volcanic island of Cheju, Republic of
Korea, to characterize groundwaters with respect to time and place of recharge and
to determine the nature of mixing of the different groundwater sources and
estimate their residence times. From a preliminary sampling of 14 springs, 8 wells and
2 streams from different parts of the island, nine points were chosen for periodic sampling.
All the waters sampled contained appreciable amounts of thermonuclear tritium, thus
indicating rapid circulation in the various groundwater systems. Upon the basis
of their tritium content and stable isotopic composition, the waters were classified into a
number of flow regimes.

This small well in Whasoon-ri was Sampling Point No. 7 in the Cheju isotope hydrology study. In the
lower partitioned well, water is used for drinking, while the washing is done and stock is allowed ^
to drink outside the enclosed area.
46
47
Springs at medium and high altitude exhibited a relatively high tritium concentration and a
wide spread of stable isotopic composition. This was interpreted as the water having a
short transit time with poor mixing, in general agreement with the hydrogeology of
the area. On the other hand, large coastal springs had a much lower tritium concentration,
but with a similar spread in stable isotopic composition. This suggested a similar
source as for the high altitude springs, but involving a somewhat longer transit
time. Small springs and wells near the east coast of the island had a similar tritium
concentration but, at the same time, somewhat higher heavy isotope content. The latter
appeared to reflect the somewhat lower terrain in the eastern part of the island.
Although the period covered by the tritium analyses was relatively short, an estimate was
made of the mean transit time of the waters. A well-mixed reservoir model was adopted
assuming that recharge was only effective for monthly precipitation values in excess
of 100mm. Values of from 2 to 8 1/2 years were estimated, the longer times
being associated with waters discharged from an extensive well-mixed fresh water lens.
A project in the coastal plain of Nicaragua concerns the study of recharge mechanism to a
phreatic aquifer. The project area is sub-rectangular in shape, having approximate
dimensions of 30 X 55 km, oriented to NW-SE. The SWboundary is the
Pacific Ocean, and the drainage divide of the volcanic chain, La Cordillera de Marrabios,
constitutes the NE limit of the area. Across the short dimension, elevations rise
uniformly to about 200 m above sea-level at about 20 - 25 km from the
ocean (Chinandega Plain), and abruptly thereafter to the Cordilleran crest which has an
average elevation of about 1000 m and a maximum of 1745 m. The plain is covered
by Quarternary "tobas", ash deposits and the stratigraphic sequence of the Tertiary. Stable
isotopes and tritium have been employed to study the problems associated with the
recharge mechanism to the groundwater system in the area with particular
emphasis on the importance of low plain recharge versus the elevated slopes of the
Cordillera. Periodic sampling at 5 precipitation stations and two springs, all located at
different elevations enabled experts to establish the relation between the elevation and
stable isotopic composition of recharging water. Both shallow and deep wells in the plain
are sampled periodically to determine the isotopic composition of local recharge as
well as that of the deeper groundwater. The rather shallow wells on the plain
with most enriched stable isotopic composition and highest tritium content were good
indices for the isotopic composition of local recharge. The rather uniform stable
isotopic composition of drilled wells, together with the known isotopic composition of
local recharge on one hand and the stable isotopic composition of the water recharging from
elevated slopes, enabled the study of the relative importance of plain recharge versus
the recharge from elevated slopes. It was found that a much higher proportion
of recharge to the groundwater system occurs from the elevated slopes with a mean altitude
of about 500 m. The tritium content of wells in the plain enabled the study of the
transit time involved in the recharge to groundwater and it was concluded that
the water in the shallow wells had been recharged locally no more than 3 years ago and the
water tapped by the drilled wells had a mean transit time in the order of 20 years.
Similar hydrological studies are at present being continued in Austria, Brazil, Bolivia,
Canary Islands, Colombia, Crete, Jamaica, Katar, Lebanon, Mexico, Senegal,
Turkey and Yugoslavia.

48