Deuterium and oxygen-18 contents as an index of the properties
of snow covers
H. Moser and W. Stichler
Abstract.
On the basis of the results of observation on natural snow covers supplemented by
laboratory tests in cold chambers it is shown and explained in part how and whereby the original
deuterium and oxygen-18 contents of snow undergo a change in the snow cover in the course of
time. Variations in isotope content thus are produced among other factors by evaporation of the
snow, by condensation of the air humidity into the snow cover, and by melt, sublimation and
«crystallization (metamorphism) phenomena. Since all of these effects occur in layers close to
the surface, the isotope contents of the deeper strata of a snow profile remain unaltered. Also a
possible percolation of melt water through these layers is practically of no influence. A comparison
of the variation of the deuterium content with that of the oxygen-18 content provides information
on whether the changes in isotope content take place in a thermodynamical state of equilibrium
of the phases involved.
Résumé.
Au moyen des résultats d'observation des couvertures de neige naturelles complétés
par des expériences de laboratoire en chambres froides, on montrera et expliquera, comment et
pourquoi le contenu de deuterium et d'oxygène-18 change au cours du temps dans la couverture
de neige. Les variations du contenu d'isotopes se produisent donc dans la couverture de neige par
la vaporisation de la neige, la condensation de l'humidité de l'air dans la couverture de neige et
par les phénomènes de fonte des neiges, de sublimation et de changement de cristallisation.
La teneur en isotopes des couches plus profondes reste intact, car tous ces événements se déroulent
dans des couches plus proches de la surface. Une éventuelle percolation de l'eau ne change
pratiquement rien. La comparaison de la variation de la teneur en deuterium avec celle de
la teneur en oxygène-18 nous donne des informations quant aux variations de la teneur en
isotopes se déroulant dans un état d'équilibre thermodynamique des phases participantes.
INTRODUCTION
The amounts of stable isotopes of hydrogen and oxygen* contained in natural water
are not constant. Thus, for example, the concentrations of HDO vary between 220
and 340ppm,f and those of H 2 18 0 between 1920 and 2010 ppm. Using sensitive
mass spectrometric measuring methods with corresponding preceding preparation of
the water samples it becomes possible to determine concentrations of HDO and
H 2 18 0 to about 0.3 ppm (see Moser and Stichler, 1971, for example), and thereby
distinguish separate waters by their isotope contents. Usually, the isotope contents
are indicated, according to Craig (1961), in terms of relative deviations SD and S180,
respectively, from an ocean water standard 'SMOW. These 'S-values' are defined by the
relation
6D or S180 =
i?sample i?standard
'
"standard
x 1000 per thousand
-^sample a n d ^standard being the isotopic ratios of hydrogen D/H and oxygen 1 8 0/ 1 6 0,
respectively, in the sample and the standard water measured under equal conditions.
* In nature, hydrogen occurs principally in the isotopic forms of 'H = H (hydrogen of relative
mass 1) and 2H = D (deuterium); oxygen, in the isotopic forms of 160 and 18 0.
f ppm = parts per million (1CT6).
Deuterium and oxygen-18 contents
123
The measuring accuracy in this case is ±1 per thousand for 5D and ±0.2 per thousand
for ô 18 0.
The changes in isotope content taking place in nature are the result mainly of
isotope fractionations during phase transitions, the isotopic ratios for a transition
from Phase I to Phase II being given by the separation factor a = RPhase i/#PhaSe nIn addition, the separation factors differ according to whether a thermodynamical
equilibrium between the two phases exists or not (compare Dansgaard, 1964).
Considering the isotope fractionations of significance to snow hydrology, it is
possible to conclude for the transition from the gaseous to the solid phase that the
contents of heavy isotopes in snow precipitation (as with rain) rise with the condensation temperature and are dependent therefore not only on the season, but on the
orographic altitude as well.
Isotope fractionations in the transition from the solid phase to the liquid phase
were verified for the first time by Friedman et al. (1964) through measurements on a
melting icicle. Employing the separation factor, it was possible from the difference in
isotope contents between the point and the base of the icicle to determine the
amount of water frozen fast. Further theoretical and experimental studies on isotope
fractionation of water when changing from solid to liquid have been described, for
example, by Arnason (1967, 1970) and Merlivat et al. (1965).
The phase transition solid—gaseous has not been treated in detail so far because in
general the model proceeded from was that of a layerwise evaporated ice surface
which, due to the slow diffusion within the solid state, does not permit an isotope
fractionation to be expected (see Ambach et al. 1968). These conditions, however,
are not present in a snow cover, since through convection and diffusion phenomena in
the air spaces of the porous medium the isotope fractionations occurring at the
boundary layer solid—gaseous become an effect of volume. Thus, the authors (Moser
and Stichler, 1970) were able to report on measurements on natural snow covers at
temperatures below freezing, which showed an enrichment decreasing with the
distance from the snow surface in the heavy isotopes D and 18 0 in the snow cover a
few days after the snowfall. It was concluded therefrom that the frequently observed
enrichment in heavy isotopes in the course of the ageing of the snow cover may also
be caused by evaporation of the snow.
OBSERVATIONS ON NATURAL SNOW COVERS
Isotope content as a function of altitude
The 'isotopic altitude effect' of precipitation, explained in the Introduction, could
only be found in fresh snow. Figure 1 gives an example from which can be seen that
the deuterium content of the samples of fresh snow declines linearly with increasing
altitude. Other measurements have been reported by the authors (Moser and Stichler,
1970). Overall, the decrease of the deuterium content A5D and of the oxygen-18
content AS180, depending on the region under observation, falls into the range of
A5D = —4 ±2 per thousand and A5180 = - 0 . 5 ±0.25 per thousand, respectively, per
100 m increase in elevation. This range covers the isotopic altitude effects mentioned
in the literature; for example, the 5180 values given by Gonfiantini (1970) for snow
samples from Mount Kilimanjaro (Tanzania) (A§180 « 0.3 per cent/100 m), and the
dependence of the deuterium content of snow samples from the western side of the
Sierra Nevada (California, U.S.A.) (AôD = - 4 per thousand/100m), published by
Friedman and Smith (1970).
Isotope content and ageing of a snow layer
In contrast to the results obtained on samples of fresh snow, the behaviour of samples
of pack snow is highly complex as regards their isotope contents. The recrystallizations
caused by solar radiation and temperature variations, as well as melt and sublimation
124
H. Moser and W. Stichler
SD
(•ho)
I.
-60-70-80--90-
-100--110--
2
(j,
-120
1
-130
(m) A.S.L
2000
3000
4000
5000
FIGURE 1. 6D values of samples of fresh snow from the Mont Blanc region. The
measurement values denoted by 1 and 2 belong to two sample series taken at the same
altitude at a horizontal distance between each other of about 30 m. x = mean value.
(X) = samples with a high degree of deviation that were not considered in the evaluation.
The straight line is determined by adjustment calculation.
processes, may substantially alter the original isotope content of the fresh snow. It is
possible in this way, for example, for an inverse altitude effect to occur, i.e. an
increase of the SD or ô180 value with increasing altitude, as demonstrated by pack
snow samples from the Kitzsteinhorn (Salzburg) in Fig. 2. Other examples of this
effect are furnished by Moser and Stichler (1970) and by Moser et al. (1972), and
explained in part by the respective local meteorological conditions. It became
apparent in these cases that in general the S values of the isotope contents rise with
increasing metamorphism, whereas dry and fine grained powdery snow retains its
5 value at the surface even after longer exposure. To confirm these findings, snow
samples were gathered in the region of the Vernagtferner (Ôtztal Alps, Austria) at the
end of September 1969 from a heavily aged summer snow cover dating from the second
half of August, the samples being taken from subsurface layers (at a depth of from 1 to
5 cm). Figure 3 contains a representation of the 5 18 0 values of the snow samples as a
function of the altitude of the sampling site with a notation at the test points indicating
the state of the snow at the time of sampling. It is evident that in the samples showing
no or only a slight metamorphism (above 3200 m) the normal isotopic altitude effect
6D
(*.)|
-80 t-100
-120
2000
3000
(m) AS.L
FIGURE 2.
5D values of samples of pack snow from the Kitzsteinhorn (near Kaprun/
Austria) in relation to the altitude of the sampling sites.
Deuterium and oxygen-18 contents
125
6 *0 (%.)
» v«ry coarse, wet
•, coarse, wet
^firned
-13
^slightly timed
•*. 5 dry, consolidated
-u
"•--. dry, consolidated
dry, powdery
(m) a.s.l.
—i—*
3000
—I
3500
•
(000
FIGURE 3.
5 , 8 0 values of snow samples from the Vernagtferner region (Ôtztal Alps,
Austria) in relation to the altitude of the sampling sites. The state of the snow is indicated
at each test point.
of the fresh snow of 5 18 0 » —0.3 per thousand/lOOm has been conserved. (The
relatively low inconsistent 5180 value of the sample taken at 3445 m can be explained
by drift snow from greater altitudes; the sampling point was located in a hollow below
a steep slope rising to a ridge.) On the other hand, the samples gathered in sampling
areas at lower altitudes, which were heavily exposed to the sun, particularly in the
lower part, evidence an enrichment in heavy isotopes with increasing metamorphism
of the snow exceeding the isotopic altitude effect.
The following observation proves that it is not alone the exposure to radiation
and the associated snow metamorphosis that affects the isotope content of a snow
cover: at the Weissfluhjoch (2540m, Switzerland), after a snowfall, the following fair
weather period was taken advantage of to collect two snow samples per day from the
surface layer (at a depth of from 1 to 2 cm) in the course of an 8-day test programme.
Figure 4 shows the 5D values of these snow samples for the respective sampling times.
These values reveal, to begin with, that during the daytime (between 8.00 and 16.00
hours) a general enrichment in deuterium occurs due to snow evaporation. The 5D
values grow at an average of about 6 per thousand. In the first nights of the observation period (21—25 March 1973), this enrichment is annulled for the greater part
probably by condensation of the air moisture into the snow cover so that about equal
5D values result for the same periods of the day. A change takes place in the night
from 25 to 26 March, the trend becoming more pronounced from 26 to 27 March;
during these nights an increase of the 5D value occurs also at night so that, overall, a
substantial deuterium enrichment of the snow blanket is observed between 25 and 27
March. Towards the end of the test the alternation between an increase of the SD
value on account of evaporation during daytime hours and a decrease of the 5 D value
through condensation at night appears to become effective again, although at a
higher deuterium level.
Of the meteorological data additionally entered on Fig. 4, only air humidity
undergoes a significant change from 25TVlarch 1973. We must therefore assume that
the increase of the 5D value was caused by condensation of a fresh supply of air
moisture that had a greater deuterium content than in the preceding nights. Regrettably,
it was not possible to supplement each sampling by a sample of air moisture for
isotopic measurements to verify this finding.
126
H. Moser and W. Stichler
*6D(%o)
A H (•/.)
march 1973
FIGURE 4.
SD values of snow samples from the surface, air temperature (T) and air
humidity {H) in the course of a test series at Weissfluhjoch/Davos (Switzerland).
In Fig. 5, the ÔD and ô180 values of the separate snow samples have been plotted
against each other (5D—<5180 line), and the test points relating to the same day have
been connected. In addition, the SD-6 18 0 line for precipitations in Central Europe
(compare Dansgaard, 1964), that represents the relation 5D = 8ÔO + 10, has been drawn. It
can be recognized that the measurement values of the samples taken in the afternoon
(16.00 hours) contain relatively less deuterium than expected from the relation given
above, the samples collected in the morning (8.00 hours), on the other hand, being
closer to the 'precipitation line'. In consequence, the evaporation processes by day do
not take place in a state of equilibrium between solid and gaseous phases (see Ehhalt
and Knott, 1965), whereas at night this equilibrium is evidently restored by condensation. In the next main section of this paper, first laboratory experiments are
Deuterium and oxygen-18 contents
-12
-15
127
6 D (%.)
^6 ,f t>(%<,)
• 8:00
x 16:00
FIGURE 5.
SD-5 , 8 0 relation of the same snow samples as plotted on Fig. 4. The test
points relating to the same day have been connected.
Isotope contents of snow profiles
Isotope contents have been measured on snow profiles both to determine the annual
accumulations of glaciers (comp. Ambach et al, 1968) and study the water balance
and water movement in snow covers (comp. Arnason et al, 1972). It was possible
for Ambach etal. (1972) to show on the basis of measurements on a snow profile
near Innsbruck that the isotope contents of the individual snow layers of the profile
remain essentially unchanged during the ablation period despite melt water and rain
infiltration, thus labelling the several layers.
Figure 6 gives the deuterium values of snow profile surveys made under the auspices
of the Amtlicher Bayerischer Lawinenwarndienst (Official Bavarian Avalanche
Warning Service) during the first semester of 1972 on the Zugspitzplatt near GarmischPartenkirchen (Upper Bavaria) in a pegged-out snow field.* The first profile recorded
on 2 January 1972 showed pronounced differences in deuterium content (more than
100 per thousand) which can be correlated separately with the 5D values of the
snowfalls in late autumn (Fig. 7). Slight enrichments in deuterium content result in
this regard between precipitation and snow layer, reflecting the changes of the isotope
content of the snow surface mentioned before. Tracing the isotope content profiles
up to end of March, it is noted that the stratification, save for small variations due
to sampling, is maintained. (The dropping of the layer boundary from an original
70 cm on 2 January 1972 to about 50 cm on 15 March 1972 is the result probably
of the advancing sampling points in the test area; on 31 March 1972 samples once
more were taken in the vicinity of the sampling point of 2 January 1972, the layer
boundary therefore being higher again.) From beginning of April, heavy snowfalls
with intermediate thaw periods considerably increased the snow depth. The isotope
contents of precipitations, due to climatic conditions, in these cases differed so
widely from one precipitation event to the next that it was not possible to reproduce
the layer boundary throughout. In contrast, the deeper layers, down to about
50—60 cm above the glacier surface, retained their 5D values over the whole observation period of some 7 months, although, for instance, the driving resistance approxi-
i
r"
128
<=—
H. Moser and W. Stichler
O —;
— OJ
i
J
"i
S
o —
-m
o
- O CM
!
o —
I
i
o—
I
-in
i
f-9
Deuterium and oxygen-18 contents
fH(cm)
" N(mm)
129
b)
150"
. J 20/21 dec
10/11 dec.
] 9/10 dec.
200
J 8 79"dec.
,0(
*
19/2* nov. 1971
'I.
8/10 nov.
M2/15oi
127/29 sept. &D(%o)
H
-150
1
1
1
1
1
-50
01
6D(7«
-150
-100
-50
FIGURE 7.
(a) 5D values of separate successive snowfalls on the Zugspitzplatt
(Bavarian Alps) with the respective summed water equivalents TV and precipitation dates.
(b) 6D values of a snow profile sampled on 2 January 1972; H = snow height.
mately doubled between 2 January 1972 and 15 March 1972, and the entire snow
cover was thoroughly wet on 13 July 1973. Measurements of the SD value of the
snow surface furnished SD values higher by about 15 per thousand in comparison
to the SD value of the sample from the uppermost layer, agreeing with the observations made in the second subsection on the isotope content and ageing of a snow
layer. From the evolution pattern of the profiles in Fig. 6 can be seen that the
reduction of the snow blanket in this case occurs from the top. In agreement with
Ambach etal. (1972), no mentionable influence was noted on the SD values of the
deeper layers by melt water percolation.
VARIATIONS IN, ISOTOPE CONTENT IN THE CASE OF PURE
EVAPORATION OF SNOW
Experimental ptocedure and preliminary tests
In a cold chamber with a temperature of about — 10°C snow samples varying in amount
were set out in bowls with different diameters. It had been ascertained in preliminary
tests that in the weight range examined the net loss in weight per day depends
primarily on the diameter of the bowl, thus on the snow area, and practically not on
the snow amount exposed (Fig. 8). In the main experiment the snow samples remaining
after varying losses in weight were then withdrawn for a mass spectrometric determination of their isotope contents, temperature and humidity in the cold chamber having
been maintained fairly constant during the evaporation process.
Results
Figure 9 shows the variations in deuterium content SD and oxygen-18 content S180
as a function of the per cent weight loss Ag. Following an initial content of SD =
— 143.7 per thousand and of 5180 = —18.77 per thousand in the snow introduced
into the chamber (average values for five snow samples), the snow suffers a linear
enrichment in D content by about 1 per thousand and in 18 0 content by about
0.2 per thousand per 1 per cent of weight loss.
130
H. Moser and W. Stichler
n weight loss /day
u • ( g/t )
snow a m o u n t
• 40-60 g
o 80-150g
10 -
snow area >
(cm2)
—,
o
100
200
300
400
600
700
800
FIGURE 8. Weight loss per day of snow samples in relation to the area of the snow
surface for different amounts of snow samples. Measurements taken in a cold chamber
with a temperature of about — 10°C.
A & D (%. )
-100
-U0--
&"b(%.)
-10
-12-K-16'
Ag(%)
50
20
60
70
ls
FIGURE 9.
6D and 6 O values of snow samples exposed in a cold chamber of about
-10°C in relation to the sample weight loss Ag.
In Fig. 10 these §D and ô180 values have been plotted against each other. The test
points lie significantly below the graph of the function typical for precipitation water
samples and form a straight line with a slope of 5.7. These measurement results allow
the conclusion that the snow evaporation produced an isotope fractionation that did
not take place in a state of equilibrium between solid and gaseous phases.
Deuterium and oxygen-18 contents
131
- -150
FIGURE 10. SD -6 I8 0 relation of the samples of Fig. 9. m = slope of graph. The
dashed curve corresponds to the 'precipitation line' (Dansgaard, 1964). X = snow
originally introduced.
ISOTOPE VARIATIONS IN A SNOW COVER WITH A TEMPERATURE
GRADIENT
In a snow layer with a temperature gradient, water vapour develops at the warmer
side, which condenses again on colder snow crystals (comp. de Quervain, 1972;
Yosida et al, 1955). It is difficult on account of the small magnitudes involved to
detect this mass transport experimentally. On the other hand, as demonstrated below,
measurements of the isotope content are well suited to follow this phenomenon.
Experimental procedure
Figure 11 contains a schematic of the test set-up: into thermally insulated plexiglass
cylinders (inside diameter, 10 cm; length 10 and 20 cm, respectively) set out in a cold
chamber with a temperature of about —20°C snow samples are introduced. A heater
is used to maintain the bottoms of the cylinders at a temperature of circa — 10°C; in
total, a temperature differential resulted over the length of the snow samples in both
columns of 7.5°C. Temperature measurements by six thermocouples along the
cylinder lengths showed that a temperature differential of 3°C had developed in the
undermost 1.5 cm of the two cylinders, the zone above this section, on the other hand
having a linear temperature gradient (Fig. 12).* To determine the initial isotope
content, two snow samples were taken prior to the test. After completion of the
experiment in which the snow had been exposed to the temperature gradient for
11 days the core of the snow column was extracted by a hollow cylinder (inside
diameter, 5 cm) and separated into snow discs with a thickness of 1 cm. The snow
crystals that had grown on the cold lid were also removed.
* The daily defrosting of the cold chamber raised the temperatures for a short time to -6°C at
the bottom and to —14°C on the cold side of the columns. This signifies that the temperature
gradient remained more or less the same. It is not very probable therefore that the disturbance
amounting to about 8 per cent of the test period substantially distorted the results.
132
H. Moserand W. Stichler
sN=fe
E^rfq
IM
"^tm
M
-zr
1cm
FIGURE 11. Experimental device for investigations on snow layers with a temperature
gradient, a = Styropor insulation, b = temperature measuring points.
Results
The snow filled into the columns had a deuterium content of SD = -89.8 per
thousand. Figure 12 shows the deuterium content in the snow layers of the two
columns after the exposure in the cold chamber described in the previous subsection.
In both columns there first appears, in correspondence with the steep temperature
gradient, a substantial enrichment in heavy isotopes (to about —85 per thousand)
towards the bottom. In the 20-cm long column the SD value in the zone extending
upwards from a level of 3 cm above the column bottom remains constant within the
limits of the SD value of the snow introduced, making allowance for the given
measuring accuracy. Only the ice crystals removed from the column lid show a slightly
depleted heavy isotope value. In the 10-cm long column likewise a value approximate
to the SD value of the snow introduced results in the section between 3 and 6 cm from
Deuterium and oxygen-18 contents
133
si
S2 '
H (cm)
-10
tCC)
-90
-85
i5D C/..)
FIGURE 12.
6D values and temperature distribution in relation to the distance H from
the bottom of the cylindrical experimental device of Fig. 11 after a test period of 14 days
for two different snow column heights (SI: 10 cm, S2: 20 cm). *Sk = 6D value of snow
crystals sublimated on the lids of the columns.
the bottom; above this level, however, the ÔD value drops to about —92 per thousand,
and the snow crystals from the lid have a 5D value even of only —99.3 per thousand.
It must be concluded that the marked increase of the SD value was caused in the
first instance by a considerable mass transport from the deepest layers to the top as a
result of the steep temperature gradient. The isotopically light water vapour condenses
in the superjacent layers which in turn, however, evaporate water molecules towards
the top and thus keep their isotope contents constant. Solely from the surface of the
snow column a few snow crystals completely sublimate on the column lid and therefore hardly alter their isotope contents. In the shorter column with its steeper temperature gradient a predominant condensation becomes evident in the upper layers,
leading to a §D value less than that of the snow introduced. The crystals in this case
are formed of water vapour which has suffered a continuous depletion in heavy
isotopes in the column and therefore has a very low SD value. It evidently originated
from the deeper layers of the snow column. In general, the mass transport processes
in the columns about correspond to the combined lamella-pore model of a snow cover
proposed by de Quervain (1972). Moreover, the isotope fractionation phenomena in
this test also in no way operate in a thermodynamical equilibrium, as shown by the
SD-S 18 0 diagram of Fig. 13 for four selected snow samples (1,2,5 and 9 cm above
the bottom) of the 10-cm column.
134
H. Moser and W. Stichler
A 6D (•/..)
-13
-12
-11
-10
I
1
1
«
-9
B"0{'/..)
1
>
• -80
6D= 86 ,a O -10 /
/
/
j
/
/
--85
/ ^
/
,/
/
/ x
/
,/
/^°
s'
._
9 0
FIGURE 13. 6D-S180 relation of some snow samples from the experiment described
in Fig. 12 (= = •>). The test point (X) relates to the original snow, the dashed line corresponds
to the 'precipitation Une' (Dansgaard, 1964).
CONCLUDING REMARKS
The still qualitative results from laboratory and field tests indicate that measurements
of the stable isotope content are capable of providing information on origin and ageing
of a snow cover as well as insights into internal mass transport processes. A quantitative
analysis will be attempted after further measurement series have been evaluated.
Acknowledgements. The authors wish to thank the Swiss Federal Institute for Snow and
Avalanche Research, Weissfluhjoch/Davos, for having put the cold chambers at our disposal for
the experiments. We are grateful to the Official Bavarian Avalanche Warning Service for data on
the snow surveys on the Zugspitzplatt. We are indebted to Dipl. Met. O. Reimvarth for the
collection of samples. Mr P. Trimborn, Mr D. Scharf and Mrs I. Hirschmann actively assisted in
the mass spectrometric measurements.
REFERENCES
Ambach, W. et al. (1968) The altitude effect on the isotopic composition of precipitation and
glacier ice in the Alps. Tellus, 20,595-600.
Ambach, W. et al. (1972) Isotopic oxygen composition of firn, old snow and precipitation in
Alpine regions. Zeit. Gletscherk. Glazialgeol. 8, 125-135.
Arnason, B. (1967) The exchange of hydrogen isotopes between ice and water in temperate
glaciers. Earth Planet. Set Let. 6,423-430.
Arnason, B. (1970) The exchange of deuterium between ice and water applied to glaciological
studies in Iceland. Isotope Hydrology, IAEA, pp. 59-71, Vienna 1970.
Arnason, B. et al. (1972) Movement of water through snow pack traced by deuterium and tritium.
Proceedings of the Symposia on the Role of Snow and Ice in Hydrology, Banff.
Craig, H. (1961) Standard for reporting concentrations of deuterium and oxygen-18 in natural
waters. Science, 133, 1833-34.
Dansgaard, W. (1964) Stable isotopes in precipitation. Tellus, 16, 436-468.
Ehhalt, D. and Knott, K. (1965) Kinetische Isotopentrennung bei der Verdampfung von Wasser.
Tellus, 17,389-397.
Friedman, I. et al. (1964) The variation of the deuterium content of natural waters in the
hydrologie cycle. Rev.. Geophys. 2, 177-224.
Deuterium and oxygen-18 contents
135
Friedman, I. and Smith, G. (1970) Deuterium content of snow cores from Sierra Nevada area.
Science, 1 6 9 , 4 6 7 - 7 0 .
Gonfiantini, R. (1970) Discussion. Isotope Hydrology, IAEA, p. 56, Vienna 1970.
Merlivat, L. et al. (1965) Formation de la grêle et fraktionnement du deuterium. Abhand. Deut.
Akad. Wiss. Berl. 7, 839.
Moser, H. and Stichler, W. (1970) Deuterium measurements on snow samples from the Alps.
Isotope Hydrology, IAEA, pp. 4 3 - 5 7 , Vienna 1970.
Moser, H. and Stichler, W. (1971) Die Verwendung des Deuterium- und Sauerstoff-18-Gehalts
bei hydrologischen Untersuchungen. Geologica Bavarica, 64, 7 - 3 5 .
Moser, H. et al. (1972) Messungen des Deuterium- und Tritiumgehaltes von Schnee-, Eis- und
Schmelzwasserproben des Hintereisferners. Zeit. Gletscherk. Glazialgeol. 8 , 2 7 5 - 2 8 1 .
de Quervain, M. R. (1972) Snow structure, heat and mass flux through snow. Proceedings on the
Symposia on the Role of Snow and Ice in Hydrology, Banff.
Yosida, Z. et al. (1955) Physical studies on deposited snow. I. Thermal Properties. Contributions
from the Institute of Low Temperature Science, No. 7, pp. 1 9 - 7 4 .
DISCUSSION
J. Martinec:
Referring to Fig. 4: Is the enrichment by evaporation likely to affect the isotopic
balance of the whole snowpack? This would necessitate a revision of models
describing the isotopic exchange between the percolating water and the snow
which do not take this effect into account.
H. Moser:
According to our experience of studying snow profiles, we found that only near-tosurface layers will be isotopically enriched by evaporation. Deeper layers are not so
much affected. But this depends on the porosity of the snow blanket and on the
temperature gradient, as I showed before.
J. Martinec:
Could the laboratory experiment demonstrating the effect of evaporation be arranged
to show the enrichment of the remaining snow by the meltwater leaving the snow
cylinder?
H. Moser:
We performed the described experiments for studies of evaporation and sublimation.
It should be possible to change the arrangement for melting processes.
L. Gold:
Would you please comment on the possible implication of your findings on the use
of the isotope method for drawing conclusions concerning past climates (for example,
as done from studies of cores from Greenland and the Antarctic).
H. Moser:
At the moment, I cannot see any implications of our findings for paleoclimatical
studies. We tried to explain that the isotope content of snow can be used as an
additional tool for characterizing the snow properties, especially its origin and
metamorphism.