A lower-resolved record from Lake Constance (Schwalb
2003) is in good agreement with the Ammersee record.
A low-resolution record also exists from Lake Geneva
covering the Holocene. More recently, the first results from a
long-term project aiming to produce a new high-resolution
record using a new sediment core from Mondsee (Austria)
show a remarkable synchrony for the negative excursion
during the 8.2 ky-event (Andersen et al. 2017) and over the
entire glacial-interglacial transition (Lauterbach et al. 2011)
Currently, a multidisciplinary team continues to work on the
complete high-resolution d
18 O P records from Ammersee
(Germany), Mondsee (Austria) and Lac d’Annecy (France)
for the last 15,000 years.
Hydro-Meteorological Effects
The Isotopic Composition of Atmospheric
Precipitation
The isotopic composition of atmospheric precipitation at a
given place and time is controlled by various fractionation
effects, including (1) the moisture formation in the (mainly
oceanic) source regions, (2) loss through precipitation along
the moisture transport pathways, (3) re-evaporation and
(4) the condensation and precipitation process at the sampling site. Thirty years of monthly observation at meteorological stations around the world show that, in high to mid
latitudes, losses of the original moisture content along horizontal temperature gradients seem to prevail, leading to the
existence of an overall positive correlation between d
18 O P
and local air temperatures on seasonal and inter-annual time
scales as well as geographically (Dansgaard 1964; Rozanski
et al. 1992). In low latitudes, a negative correlation between
d
18 O P and the amount of precipitation is observed and is
explained by the vertical temperature gradients. These
climate-d
18 O P relations are, however, overlain by noise due
to high short-term variability, reflecting the effects of the
different condensation processes and the admixture of vapor
from different marine or continental sources during a single
precipitation event, and through spatial variability, due to
relief effects. Evidently, the quantitative relation between
d
18 O P and climatic parameters, like temperature and precipitation, has to be checked for each basin before conclusions in terms of paleoclimate can be drawn from a record.
However, even if such a relation is weak or absent, the
reconstructed d
18 O P itself remains a valuable information
source for paleoclimate analyses and for the quantitative
interpretation of isotopic records with strong secondary
effects.
It is important to be aware of the seasonal variations of
d
18 O P , which are large even with respect to the greater shifts
recognised in paleo-records. In many cases, the retention
times of water within the catchment and in the lake are long
enough, so that the effective input into the lake carries an
average isotope signal over several years. This average is
often close to the average of d
18 O P , but might be offset by
preferential losses of a part of the precipitation by evaporation and transpiration (see below). In the special case of
short catchment retention times combined with a short residence time in the lake, the isotopic composition of the lake
water (d
18 O L ) can respond with oscillations either in or out
of phase with the seasonal d
18 O P cycle.
Catchment Effects
The term ‘catchment effect’ has been introduced and discussed in detail by Gat et al. (1995). It can be defined as the
sum of effects from all hydrological processes occurring in
the drainage basin that might alter the relation between long
term d
18 O P and d
18 O L . Besides the ‘catchment effect’ sensu
stricto, which is the measurable difference between the isotopic compositions of the drainage basin runoff d
18 O I and
d
18 O P (D P−I ), we also have to consider those processes that
control the amount of runoff (I) and the spatial and temporal
variability of both I and d
18 O I .
Potentially responsible for a significant D P−I are losses via
evaporation or sublimation, which can lead to an isotopic
enrichment of the remaining water. The importance of these
losses can be estimated by comparing the deuterium excess
of the runoff with that of the mean precipitation. Usually
they are small in humid climates and they depend on the
proportion of open-water surfaces within a drainage basin.
Alternatively, non-fractionating losses such as the uptake
and consequent transpiration and evaporation of water by
plants, or the recharge of deep groundwater, can be selective
against a portion of the annual precipitation and thus can
lead to a relative shift of d
18 O P . This effect theoretically
could be greater than 1‰, if the evapotranspiration is
restricted to the summer period and only summer precipitation is taken up. In practice, the largest part of water for
evapotranspiration comes from the soils, which retain significant amounts of water over the year and where the seasonal variation of d
18 O P and, in consequence, the effect on
D P−I is considerably reduced.
Stronger effects are to be expected in high-altitude drainage basins, including those in high alpine areas, where
build-up of glaciers can retain significant amounts of
isotopically-depleted winter precipitation and thus shift
d
18 O I to more positive values. Glacier growth is a relatively
long-term process, which takes place over tens to hundreds
of years during relatively cold periods, whereas ablation can
be almost spontaneous with a consequent discharge of
isotope-light water easily exceeding the amount of yearly
retention during glacier growth. The isotopic composition of
15 Air-Interface: d
18 O Records of Past Meteoric Water Using …
181
2003) is in good agreement with the Ammersee record.
A low-resolution record also exists from Lake Geneva
covering the Holocene. More recently, the first results from a
long-term project aiming to produce a new high-resolution
record using a new sediment core from Mondsee (Austria)
show a remarkable synchrony for the negative excursion
during the 8.2 ky-event (Andersen et al. 2017) and over the
entire glacial-interglacial transition (Lauterbach et al. 2011)
Currently, a multidisciplinary team continues to work on the
complete high-resolution d
18 O P records from Ammersee
(Germany), Mondsee (Austria) and Lac d’Annecy (France)
for the last 15,000 years.
Hydro-Meteorological Effects
The Isotopic Composition of Atmospheric
Precipitation
The isotopic composition of atmospheric precipitation at a
given place and time is controlled by various fractionation
effects, including (1) the moisture formation in the (mainly
oceanic) source regions, (2) loss through precipitation along
the moisture transport pathways, (3) re-evaporation and
(4) the condensation and precipitation process at the sampling site. Thirty years of monthly observation at meteorological stations around the world show that, in high to mid
latitudes, losses of the original moisture content along horizontal temperature gradients seem to prevail, leading to the
existence of an overall positive correlation between d
18 O P
and local air temperatures on seasonal and inter-annual time
scales as well as geographically (Dansgaard 1964; Rozanski
et al. 1992). In low latitudes, a negative correlation between
d
18 O P and the amount of precipitation is observed and is
explained by the vertical temperature gradients. These
climate-d
18 O P relations are, however, overlain by noise due
to high short-term variability, reflecting the effects of the
different condensation processes and the admixture of vapor
from different marine or continental sources during a single
precipitation event, and through spatial variability, due to
relief effects. Evidently, the quantitative relation between
d
18 O P and climatic parameters, like temperature and precipitation, has to be checked for each basin before conclusions in terms of paleoclimate can be drawn from a record.
However, even if such a relation is weak or absent, the
reconstructed d
18 O P itself remains a valuable information
source for paleoclimate analyses and for the quantitative
interpretation of isotopic records with strong secondary
effects.
It is important to be aware of the seasonal variations of
d
18 O P , which are large even with respect to the greater shifts
recognised in paleo-records. In many cases, the retention
times of water within the catchment and in the lake are long
enough, so that the effective input into the lake carries an
average isotope signal over several years. This average is
often close to the average of d
18 O P , but might be offset by
preferential losses of a part of the precipitation by evaporation and transpiration (see below). In the special case of
short catchment retention times combined with a short residence time in the lake, the isotopic composition of the lake
water (d
18 O L ) can respond with oscillations either in or out
of phase with the seasonal d
18 O P cycle.
Catchment Effects
The term ‘catchment effect’ has been introduced and discussed in detail by Gat et al. (1995). It can be defined as the
sum of effects from all hydrological processes occurring in
the drainage basin that might alter the relation between long
term d
18 O P and d
18 O L . Besides the ‘catchment effect’ sensu
stricto, which is the measurable difference between the isotopic compositions of the drainage basin runoff d
18 O I and
d
18 O P (D P−I ), we also have to consider those processes that
control the amount of runoff (I) and the spatial and temporal
variability of both I and d
18 O I .
Potentially responsible for a significant D P−I are losses via
evaporation or sublimation, which can lead to an isotopic
enrichment of the remaining water. The importance of these
losses can be estimated by comparing the deuterium excess
of the runoff with that of the mean precipitation. Usually
they are small in humid climates and they depend on the
proportion of open-water surfaces within a drainage basin.
Alternatively, non-fractionating losses such as the uptake
and consequent transpiration and evaporation of water by
plants, or the recharge of deep groundwater, can be selective
against a portion of the annual precipitation and thus can
lead to a relative shift of d
18 O P . This effect theoretically
could be greater than 1‰, if the evapotranspiration is
restricted to the summer period and only summer precipitation is taken up. In practice, the largest part of water for
evapotranspiration comes from the soils, which retain significant amounts of water over the year and where the seasonal variation of d
18 O P and, in consequence, the effect on
D P−I is considerably reduced.
Stronger effects are to be expected in high-altitude drainage basins, including those in high alpine areas, where
build-up of glaciers can retain significant amounts of
isotopically-depleted winter precipitation and thus shift
d
18 O I to more positive values. Glacier growth is a relatively
long-term process, which takes place over tens to hundreds
of years during relatively cold periods, whereas ablation can
be almost spontaneous with a consequent discharge of
isotope-light water easily exceeding the amount of yearly
retention during glacier growth. The isotopic composition of
15 Air-Interface: d
18 O Records of Past Meteoric Water Using …
181