EAE03-A-14209
A NOBLE GAS PALAEOTEMPERATURE RECORD FROM
THE LEDO-PANISELIAN AQUIFER IN BELGIUM
Institute of Environmental Physics
University of Heidelberg
University of Bern
W. Aeschbach-Hertig1, 2, R. Kipfer 2, P. C. Blaser3, K. Walraevens4
Environment al Isot opes Group
Dept . of Wat er Resources
and Drinking Wat er
1Institute
of Environmental Physics, University of Heidelberg, D-69120 Heidelberg, Germany
Isotope Group, W+T, EAWAG and Isotope geology, ETHZ, CH-8600 Dübendorf, Switzerland
3Climate and Environmental Physics, Physics Institute, University of Bern, CH-3012 Bern, Switzerland
4Applied Geology and Hydrogeology, Ghent University, B-9000 Gent, Belgium
Contact: [email protected]
2Environmental
1. Study Area
H
RT
NO
51°13’N
Western
A
SE
The LPA forms a model example of a freshening aquifer, where the initial marine porewater is flushed by meteoric recharge.
Scheldt
Ledo-Paniselian Aquifer
A’
THE NETHERLANDS
(Zealand)
BELGIUM
(Flanders)
0
3.
5.0
10.0
Eeklo
Bruges
Ursel
51°02’N
A
Ghent
10 km
44 samples from 39 wells taken between 1997 and
2001 were analysed for hydrochemistry, radiocarbon, stable isotopes, and noble gases.
3°42’E
Dutch - Belgian frontier
Profile A-A’
10.0
calculated hydraulic head [m]
in natural state (before 1920)
A
SSW
A’
NNE
Boom Clay
Oligocene Sands
Bartonian Clay
-100
recharge
area
Ledo-Paniselian Sands
-200
TAW: Belgian reference level
(low spring tide ~
~ -2.3 m asl)
Ca2+
0.10
0.01
0
5
1.10-10 cm3STP g-1 yr-1
8
6
4
normal
2
degassed
high Cl, TDS
0
0
10000
20000
14C
30000
40000
Cl- [meq l-1]
5.10-12 cm3STP g-1 yr-1
10000
-5.5
δ18O normal
δ18O degassed
-6
0.8
0.6
-6.5
0.4
model age [yr]
-7
excess air
A = 0.1, F = 0.9
A = 0.001, F = 0
0
A = 0.1, F = 1.8
0.6
0.2
A = 0.1, F = 3.5
Ne
The same model can also describe the phenomenon of degassing,
as shown here for some samples.
In the model equation
Ci = Ci* (T , S , P ) +
A = 0.1, F = 1.2
degassing
concentration relative to equilibrium
symbols: data
lines: model
Ar
Kr
Noble Gas
Xe
30000
40000
C model age [yr]
(1 − F ) Azi
1 + FAzi Ci*
values of the fractionation parameter F < 1 describe excess air,
whereas F > 1 corresponds to
degassing by equilibration with a
secondary gas phase (e.g. CH4).
The parameter A describes the
initial concentration of trapped air.
0
2
4
6
8
10
12
-7.5
NGT [°C]
δ18O [‰]
Estimated in situ accum. rate:
0
20000
14
Cl- degassed
1
1.4
0.4
30
age gap?
-5
2 10-6
4. Interpretation of noble gas concentrations
0.8
25
10
Cl- normal
1.2
Inverse modeling of the observed noble gas concentrations was used to interpret the
data in terms of recharge temperature and excess air. The closed-system equilibration
(CE) model was used to describe the excess air component (Aeschbach-Hertig et al.,
2000). The model assumes equilibration between groundwater and gas bubbles.
1
20
12
1.4
3 10-6
1 10-6
A = 0.04, F = 0.7
15
The absence of samples with ages between 23 and 29 kyr may be due to inhibition of
groundwater recharge by permafrost conditions during the Last Glacial Maximum.
Observed He accumulation rate:
4 10-6
The noble gas temperature (NGT)
record from the LPA indicates
strong glacial cooling. Holocene
samples yield NGTs of 8 to 10 °C,
in agreement with modern air
temperatures. At 14C-ages of
about 20 kyr, NGTs near the
freezing point (< 2 °C) occur. Clrich samples older than 30 kyr
show intermediate NGTs. A
certain gap exists between the
latter two groups of samples.
0.2
1.2
10
5. Paleoclimate indicators
5 10-6
Herad [cm3STP g-1]
K+
Mg2+
1.00
Ypresian Clay
Spreadsheet, NETPATH and PHREEQC calculations were performed to account for
chemical reactions and isotope exchange. The initial δ13C and 14C values of soil CO2
and dissolved CO2 at the time of groundwater recharge were calculated for distinct
climatic scenarios, according to van der Kemp et al. (2000), but additionally taking
temperature effects into account. Log pCO2 was varied from -1.5 (Holocene) to -3.5
(glacial), temperature from 10 °C (Holocene) to 2 °C (glacial), and the atmospheric CO2
and δ13C history was taken from Leuenberger et al. (1992) and Marino et al. (1992).
These adapted input values were then inserted in the usual 14C-correction models (e.g.,
Fontes & Garnier). Details are described in Blaser (2003).
0
SO42-
Paniselian Clay
Ypresian Sands
10 km
3. Radiocarbon dating and He accumulation
The concentrations of radiogenic 4He correlate well with
the 14C model ages, supporting the corrections applied
for dating. In or near the
recharge area, the He
accumulation rate is low and
close to the expected rate
due to in situ He production.
Further downstream, He
accumulation is about an
order of magnitude stronger.
10.00
The marine influence is still visible in the downstream
part of the investigated flow line, characterised by
high concentrations of Cl-, Na+, and TDS. The noble
gases also reveal a zonation of the aquifer. Most of
the wells in the recharge area show low He accumulation. Degassed samples occur mainly in a narrow
band further downstream and in the southern outcrop.
N
Quaternary
HCO3-
Distance from recharge [km]
m TAW
0
Na+
Cl-
main flow line
normal
low He
degassed
high Cl
recharge area
100.00
∆T = 8.5 °C
The LPA is part of an alternating sequence of marine Tertiary clay and sand deposits. Recharge takes place in topographically higher regions through the semiconfining cover of Bartonian Clay. Some
discharge occurs in the southern outcrop
area. In the confined part of the aquifer the
main flow direction is northwards. Upward
outflow through the Bartonian Clay causes
flow velocities to diminish gradually.
2°59’E
The marine cations Na+, K+ and Mg2+
were replaced by Ca2+ in the order of
increasing affinity to the clay. This
cation exchange induced a pronounced
chromatographic pattern upflow of the
fresh/salt-water interface, resulting in a
typical sequence of NaCl, NaHCO3,
MgHCO3 and CaHCO3 watertypes.
Concentration [meq/L]
The Eocene Ledo-Paniselian Aquifer (LPA) is a
sandy, confined coastal aquifer, extending from
Flanders in north-western Belgium to the Dutch
province of Zealand. The study area is located in
Flanders NW of Ghent. The natural hydraulic
gradient follows dip from south to north. Flow
modelling shows substantial changes in piezometric
levels due to exploitation (Walraevens, 1988).
2. Hydrochemical evolution
NGT [°C]
Rare Gas Group
Inst . of Isot ope Geology and
Mineral Resources
The stable isotope ratios
show little climate sensitivity in the LPA. δ18O correlates only weakly with NGTs.
In contrast, Cl- concentrations exhibit a significant
correlation with NGTs
(excluding the downstream
region with elevated Cl-). In
coastal aquifers, Cl- records
changes in sea-salt input
due to sea-level variations.
6. Conclusions
• The NGT-record indicates a glacial cooling of at least 8.5 °C for Belgium
• Detailed hydrochemical and isotopic modeling yields reliable 14C ages
• Radiogenic 4He, mostly non-locally produced, correlates with 14C ages
• Chloride concentrations record a climatic signal (sea-level changes)
• The CE-model describes all noble gas data, even degassed samples
References
Aeschbach-Hertig et al. (2000). Palaeotemperature reconstruction from noble gases in ground water taking into account equilibration
with entrapped air. Nature 405, 1040-1044.
Blaser (2003). Tracermethoden in der Hydrologie – Kombination verschiedener Methoden und Anwendungen in Aquifersystemen in
Belgien und Estland. Diss. Univ. Bern.
Leuenberger et al. (1992). Carbon Isotope Composition of Atmospheric CO2 during the Last Ice Age from an Antarctic Ice Core. Nature
357, 488-490.
Marino et al. (1992). Glacial-to-interglacial variations in carbon isotopic composition of atmospheric CO2. Nature 357, 461-466.
Van der Kemp et al. (2000). Inverse chemical modeling and radiocarbon dating of palaeogroundwaters: The Tertiary Ledo-Paniselian
aquifer in Flanders, Belgium. Water Resour. Res. 36(5), 1277-1287.
Walraevens (1998). Natural isotopes and noble gases in groundwater of the Tertiary Ledo-Paniselian aquifer in East and West
Flanders. Natuurwet. Tijdschr. 78, 246-260.