runoff from partially glaciated areas may therefore be biased
to more positive values compared to d
18
O P during cold
phases with glacier growth, and marked by strong negative
spikes during subsequent warm periods.
Environmental changes including climate shifts can
increase or decrease the runoff without significantly affecting
D P−I . The most evident is a change in precipitation, but
evapotranspiration and groundwater recharge and discharge
can also change either in concert with precipitation, or
independently, as a response to climate change, vegetation
change and human activity. Such variations of the runoff will
alter the lake water balance and thus the link between d
18 O P
and the isotopic composition of lake water d
18 O L by an
amount dependent on the hydrological sensitivity of the lake
(see below). Changes in the seasonal runoff characteristics
can affect the link between d
18 O P and d
18 O L . These may
occur without any measurable changes in the annual runoff
and mean annual d
18 O I , by moderation of the short-term
reservoirs (soil, snow cover). Such effects are important in
lakes with water residence times of close to or less than one
year, but may also be visible in lakes with longer response
times.
Probably the most efficient way to characterize the
catchment effects in a given basin is through a survey of the
isotopic composition of both the precipitation and the river
runoff. Figure 15.1 gives an example for the Ammer River
(southern Germany), the main contributor to Lake Ammersee. River water was sampled just upstream of the lake in
two-weekly intervals for more than two years and on a daily
base for the second of these two years (von Grafenstein et al.
1996). Atmospheric precipitation is collected routinely at
one meteorological station (Hohenpeissenberg) within the
drainage basin and at two additional stations (Neuherberg
and Garmisch), about 50 km to the north and south,
respectively, providing monthly values for d
18 O P since 1970
AD. The oxygen isotope composition of the total two-years
of runoff is −10.3‰, with a tendency to increase from
−10.5‰ to −10.1‰ during this period. Interestingly, the
two-year precipitation mean for the same period is −10.1‰,
indicating that at least a part of the runoff is from longer term
(groundwater) reservoirs. Indeed, the daily record of the
second year allows the spontaneous, soil water and
groundwater components to be separated out with a mean
retention time of 9 months for the soil water component and
of at least several years for the groundwater component. In
the long term, the total contribution of soil water and
spontaneous runoff is about 30%. It is therefore impossible
to define an exact estimate of D P−I without considering the
history of d
18 O P for the past few decades. The best explanation for the tendency of d
18 O I over more than 4 years
(including lower frequency measurements from 1989 on) is a
mean age of three years for the groundwater component.
However, even if we consider a much longer retention in the
groundwater, D P−I is less than +0.3‰ for the present-day
conditions in the Ammersee drainage basin. A small
decrease in the deuterium excess of the mean runoff (8‰)
compared to that of the precipitation over the last decade
(10‰) might indicate that this probable catchment effect in
the case of the Ammer river is due to surface evaporation
(from some smaller lakes in the basin).
In summary, catchment effects, even if they are seemingly
small as in our example, have to be at least considered as a
possible reason for changes of d
18 O L in the past. Quantification is however rather complicated and has to be based on
a proper description of the modern geomorphologic conditions and the vegetation cover of the drainage basin.
Lake Water Balance Effects
The isotopic composition of lake water is controlled by the
amounts and the isotopic compositions of the inflowing
water (I, I ) and loss via evaporation (E, E ). In steady state, i.e.
if the environmental conditions are considered as stable for
some time, d L approaches a constant value representing a
mixture between d I (weighted by I) and d* (weighted by the
evaporation and the atmospheric water vapor deficit):
d L ¼ d P I þ d
Ã
Eh= l À h
ð
Þ
ð
Þ = I þ Eh= l À h
ð
Þ
ð
Þ ð 1Þ
d* is the isotopic composition of water being in isotopic
equilibrium with the atmospheric water vapor and thus the
maximum that can be reached by evaporation:
d
Ã
¼ h d A þ e
ð
Þ = h À e
ð
Þ
ð2Þ
where h is the relative humidity, d A the isotopic composition
of the atmospheric water vapour and e the sum of the equilibrium and kinetic fractionation between water and vapour
(Gat et al. 1994; Gibson et al. 1993) Several important facts
with respect to the reconstruction of d P can be derived from
these relations: (1) the isotopic enrichment due to evaporation
is independent of the residence time of water (I/V), but
depends, in addition to the atmospheric moisture conditions,
on I (which is the product of P-ET, PrecipitationEvapoTranspiration, and the surface of the catchment basin)
and E (which is the product of evaporation per surface unit
and the lake’s surface). Lakes with differing ratios between
the lake surface A L and drainage basin area A C will therefore
have significantly differing d L , even if the atmospheric conditions including d P , d A and the evaporative flux are equal.
(2) Any change of P-ET or E will be weighted by this factor
A L /A C , i.e. will induce a stronger reaction of d L for lakes with
a larger A L /A C . The smaller the lake surface compared to the
drainage basin, the smaller will be the influence of changing
hydrology on the quantitative link between d I and d L (and
consequently the link between d P and d L ). (3) Comparison of
182
U. von Grafenstein and I. Labuhn
to more positive values compared to d
18
O P during cold
phases with glacier growth, and marked by strong negative
spikes during subsequent warm periods.
Environmental changes including climate shifts can
increase or decrease the runoff without significantly affecting
D P−I . The most evident is a change in precipitation, but
evapotranspiration and groundwater recharge and discharge
can also change either in concert with precipitation, or
independently, as a response to climate change, vegetation
change and human activity. Such variations of the runoff will
alter the lake water balance and thus the link between d
18 O P
and the isotopic composition of lake water d
18 O L by an
amount dependent on the hydrological sensitivity of the lake
(see below). Changes in the seasonal runoff characteristics
can affect the link between d
18 O P and d
18 O L . These may
occur without any measurable changes in the annual runoff
and mean annual d
18 O I , by moderation of the short-term
reservoirs (soil, snow cover). Such effects are important in
lakes with water residence times of close to or less than one
year, but may also be visible in lakes with longer response
times.
Probably the most efficient way to characterize the
catchment effects in a given basin is through a survey of the
isotopic composition of both the precipitation and the river
runoff. Figure 15.1 gives an example for the Ammer River
(southern Germany), the main contributor to Lake Ammersee. River water was sampled just upstream of the lake in
two-weekly intervals for more than two years and on a daily
base for the second of these two years (von Grafenstein et al.
1996). Atmospheric precipitation is collected routinely at
one meteorological station (Hohenpeissenberg) within the
drainage basin and at two additional stations (Neuherberg
and Garmisch), about 50 km to the north and south,
respectively, providing monthly values for d
18 O P since 1970
AD. The oxygen isotope composition of the total two-years
of runoff is −10.3‰, with a tendency to increase from
−10.5‰ to −10.1‰ during this period. Interestingly, the
two-year precipitation mean for the same period is −10.1‰,
indicating that at least a part of the runoff is from longer term
(groundwater) reservoirs. Indeed, the daily record of the
second year allows the spontaneous, soil water and
groundwater components to be separated out with a mean
retention time of 9 months for the soil water component and
of at least several years for the groundwater component. In
the long term, the total contribution of soil water and
spontaneous runoff is about 30%. It is therefore impossible
to define an exact estimate of D P−I without considering the
history of d
18 O P for the past few decades. The best explanation for the tendency of d
18 O I over more than 4 years
(including lower frequency measurements from 1989 on) is a
mean age of three years for the groundwater component.
However, even if we consider a much longer retention in the
groundwater, D P−I is less than +0.3‰ for the present-day
conditions in the Ammersee drainage basin. A small
decrease in the deuterium excess of the mean runoff (8‰)
compared to that of the precipitation over the last decade
(10‰) might indicate that this probable catchment effect in
the case of the Ammer river is due to surface evaporation
(from some smaller lakes in the basin).
In summary, catchment effects, even if they are seemingly
small as in our example, have to be at least considered as a
possible reason for changes of d
18 O L in the past. Quantification is however rather complicated and has to be based on
a proper description of the modern geomorphologic conditions and the vegetation cover of the drainage basin.
Lake Water Balance Effects
The isotopic composition of lake water is controlled by the
amounts and the isotopic compositions of the inflowing
water (I, I ) and loss via evaporation (E, E ). In steady state, i.e.
if the environmental conditions are considered as stable for
some time, d L approaches a constant value representing a
mixture between d I (weighted by I) and d* (weighted by the
evaporation and the atmospheric water vapor deficit):
d L ¼ d P I þ d
Ã
Eh= l À h
ð
Þ
ð
Þ = I þ Eh= l À h
ð
Þ
ð
Þ ð 1Þ
d* is the isotopic composition of water being in isotopic
equilibrium with the atmospheric water vapor and thus the
maximum that can be reached by evaporation:
d
Ã
¼ h d A þ e
ð
Þ = h À e
ð
Þ
ð2Þ
where h is the relative humidity, d A the isotopic composition
of the atmospheric water vapour and e the sum of the equilibrium and kinetic fractionation between water and vapour
(Gat et al. 1994; Gibson et al. 1993) Several important facts
with respect to the reconstruction of d P can be derived from
these relations: (1) the isotopic enrichment due to evaporation
is independent of the residence time of water (I/V), but
depends, in addition to the atmospheric moisture conditions,
on I (which is the product of P-ET, PrecipitationEvapoTranspiration, and the surface of the catchment basin)
and E (which is the product of evaporation per surface unit
and the lake’s surface). Lakes with differing ratios between
the lake surface A L and drainage basin area A C will therefore
have significantly differing d L , even if the atmospheric conditions including d P , d A and the evaporative flux are equal.
(2) Any change of P-ET or E will be weighted by this factor
A L /A C , i.e. will induce a stronger reaction of d L for lakes with
a larger A L /A C . The smaller the lake surface compared to the
drainage basin, the smaller will be the influence of changing
hydrology on the quantitative link between d I and d L (and
consequently the link between d P and d L ). (3) Comparison of
182
U. von Grafenstein and I. Labuhn