Earth and Planetary Science Letters 193 (2001) 617^630
www.elsevier.com/locate/epsl
Testing optically stimulated luminescence dating of
sand-sized quartz and feldspar from £uvial deposits
J. Wallinga a;b;c; *, A.S. Murray b , G.A.T. Duller c , T.E. To«rnqvist d
a
The Netherlands Centre for Geo-ecological Research (ICG), Faculty of Geographical Sciences, Utrecht University, P.O. Box 80115,
NL-3508 TC Utrecht, The Netherlands
b
Nordic Laboratory for Luminescence Dating, Department of Earth Sciences, Aarhus University, RisÖ National Laboratory,
DK-4000 Roskilde, Denmark
c
Institute of Geography and Earth Sciences, University of Wales, Aberystwyth SY23 3DB, UK
d
Department of Earth and Environmental Sciences, University of Illinois at Chicago, 845 West Taylor Street, Chicago,
IL 60607-7059, USA
Received 17 May 2001; received in revised form 22 August 2001; accepted 30 September 2001
Abstract
We apply single-aliquot optically stimulated luminescence (OSL) dating to quartz- and feldspar-rich extracts from
fluvial channel deposits of the Rhine^Meuse system in The Netherlands. The time of deposition of these deposits is
tightly constrained by radiocarbon dating or historical sources. This allows us to compare OSL ages obtained on quartz
and infrared OSL (IR-OSL) ages obtained on potassium-rich feldspar with independent ages over the range of 0.3^13
ka. We show that the quartz OSL ages are in good agreement with the expected age. Using IR-OSL dating of feldspar,
we find a slight age overestimate for the youngest sample, whereas for older samples the age is significantly
underestimated. We also apply OSL dating to older fluvial and estuarine channel deposits with limited independent
chronological constraints. Comparison of feldspar IR-OSL ages with the quartz OSL ages up to V200 ka shows a clear
trend, where the former severely underestimates the latter. This trend is similar to that found for the samples with
independent age control, indicating that the feldspar IR-OSL ages are erroneously young for the entire age range. In the
youngest samples, incomplete resetting of the IR-OSL signal prior to deposition probably masks the age
underestimation. We show that the IR-OSL age underestimation is partly caused by changes in trapping probability
due to preheating. Correction for this phenomenon improves the IR-OSL ages slightly, but does not provide a complete
solution to the discrepancy. We suggest that, in the light of the problems encountered in the IR-OSL dating of feldspar,
quartz is the mineral of choice for OSL dating of these deposits. However, feldspar dating should continue to be
investigated, because it has potential application to longer time scales. ß 2001 Elsevier Science B.V. All rights
reserved.
Keywords: optically stimulated luminescence dating; quartz; feldspar group; stream sediments; age
1. Introduction
* Corresponding author. Tel.: +31-30-253-2749;
Fax: +31-30-253-1145.
E-mail address: j.wallinga@geog.uu.nl (J. Wallinga).
Optically stimulated luminescence (OSL) dating
is a rapidly developing technique that provides
absolute chronologies for late Quaternary clastic
0012-821X / 01 / $ ^ see front matter ß 2001 Elsevier Science B.V. All rights reserved.
PII: S 0 0 1 2 - 8 2 1 X ( 0 1 ) 0 0 5 2 6 - X
618
J. Wallinga et al. / Earth and Planetary Science Letters 193 (2001) 617^630
sedimentary records [1]. OSL dating of the sandy
fraction can be applied to quartz up to an age of
100^200 ka, or to feldspar, which o¡ers the potential to extend the age range to 1 Ma. Fluvial
records now represent one of the most common
sedimentary environments dated by OSL methods. However, a potential problem in £uvial systems is that light exposure of sand during transport may not have been su¤cient to completely
reset the OSL signal of all grains prior to deposition. Incomplete resetting of the OSL signal will
lead to an overestimation of the age. Most studies
applying OSL dating to £uvial deposits have focused on chronologically unconstrained, or poorly
constrained, Pleistocene successions. With the exception of a few studies (notably [2^5]), little effort has been made to provide a sound justi¢cation for the use of luminescence dating in £uvial
settings.
Rigorous comparison of OSL ages with independent ages is only possible if the latter are
highly accurate and if the stratigraphic relationship between OSL samples and independent ages
is indisputable. Studies must also provide enough
background information to assess the quality of
luminescence and independent ages. For coarsegrain quartz OSL dating, several comparative
studies that meet these guidelines have been conducted in aeolian environments [6,7]. For coarsegrain feldspar infrared OSL (IR-OSL) dating, on
the other hand, validation of the technique is surprisingly limited ; few studies ¢t the criteria for
rigorous comparison outlined above. Good agreement with independent age control was found for
samples younger than 30 ka for aeolian dune
sands from New Zealand [8]. On the other
hand, IR-OSL ages on samples from the same
type of deposit from Germany showed an underestimation of age [9].
In the present study, we apply OSL dating to
both quartz and feldspar extracts from sandy
channel deposits from the Rhine^Meuse system
in The Netherlands. We sampled four channel
belts for which the period of activity is accurately
known from historical sources or radiocarbon
dating. Comparison of OSL dating results from
these samples with unusually tight independent
age constraints presents an unparalleled opportu-
nity to test the accuracy of OSL dating results
from submodern (0.3 ka) to Late Weichselian
(V13 ka) £uvial sediments. Additionally, OSL
dating of samples from chronologically unconstrained £uvial and estuarine channel deposits
from the same area allows comparison of quartz
and feldspar OSL ages up to an age of V200 ka.
2. Study area and independent age constraints
The Rhine^Meuse Delta (Figs. 1 and 2) is located in the southeastern part of the North Sea
Basin. The Holocene delta is underlain by Weichselian (oxygen-isotope stages (OIS) 2^4), sandy to
gravelly £uvial channel deposits. The Holocene
Rhine^Meuse Delta was formed in response to
relative sea-level rise [10,11] and contains wide-
Fig. 1. The Rhine^Meuse Delta in The Netherlands and location of sampling sites.
619
J. Wallinga et al. / Earth and Planetary Science Letters 193 (2001) 617^630
Fig. 2. Schematic east^west cross section showing relative positioning of the samples. The Winssen, Rumpt and Schelluinen samples were taken from sandy £uvial channel deposits
within the Holocene Rhine^Meuse Delta. The Elden and
Leidschendam samples were taken from Pleistocene, predominantly £uvial channel deposits underlying the Holocene delta. Sample location (local coordinates) and depth (meters below the surface): Winssen (168.485/435.475, 32.75 m),
Rumpt I-3 (139.355/433.880, 34.35 m), Rumpt IV-2
(139.920/433.330, 32.95 m), Schelluinen II-3 (123.290/
429.875, 33.15 m), Schelluinen II-6 (123.290/429.875, 35.55
m), Elden (187.985/440.000, 36.25 m), Leidschendam
(87.540/454.380, for sample depths see [20]).
spread in situ organics interbedded with £uvial
clastics, thus enabling highly accurate 14 C chronologies to be established. We collected OSL samples from three Holocene channel belts (Waal,
Linge, and Schaik systems) and one Late Weich-
selian channel belt (Elden core), and additionally
from Weichselian to Saalian (OIS 2 and older),
£uvial and estuarine channel deposits (Leidschendam core) (Figs. 1 and 2).
According to historical maps, the deposits at
the sampling site of the Waal (Winssen core)
were formed between AD 1688 and 1723 [12],
which corresponds to an age of V300 yr (note
that we use AD 2000 as a reference for calendar
ages). The Winssen sample was collected to assess
the potential e¡ects of poor bleaching. We selected a 300-yr-old sample rather than a modern
one to avoid the in£uence of locks and waterworks that have altered sediment transport in
the present-day river, and thus the bleaching environment.
The ages of the other two Holocene channel
belts are well constrained by 14 C dating (Table
1) with periods of activity no longer than 1700
yr (Linge system) and 800 yr (Schaik system).
Radiocarbon dating [13] employed accelerator
mass spectrometry (AMS) of terrestrial botanical
macrofossils [14]. For both the Linge and the
Schaik systems, a larger number of 14 C ages are
available; we used those that are located closest
to the sites of our OSL samples. The beginning of
activity of both channel belts was determined by
14
C dating the top of peat directly underlying
Table 1
Independent age constraints of OSL-dated £uvial channel deposits
Fluvial system
OSL samples
Waal
Linge
Schaik
Late Weichselian
Winssen
Rumpt I-3, Rumpt IV-2
Schelluinen II-3, Schelluinen II-6
Elden
14
Calendar ageb
(yr before AD 2000)
(median and 2c con¢dence interval)
C age
(yr BP)
Beginning
End
2 235 þ 35a1
5 050 þ 85a2 4 605 þ 45a3
11 063 þ 12a4
Beginning
End
312c
277c
2 300d1 (2 190^2 370)
V700e
5 850d1 (5 670^6 030)
5 360d2 (5 210^5 470)
13 155f (12 990^13 320)
a
Laboratory numbers: (1) UtC-1481/1482; (2) UtC-1144/1300; (3)UtC-1128/1129/1130/1141/1142 (all based on AMS 14 C measurements of terrestrial plant macrofossils, except for samples UtC-1128, 1141 and 1481 [13]); (4) Hd-19607, Hd-18648, Hd19098, Hd-19092, Hd-18622, Hd-19037, Hd-18438 (AMS 14 C measurements of decadal samples of poplar buried by the Laacher
See Tephra [16]).
b
As derived from historical sources or calibrated 14 C ages. For the latter 50 yr was added to each age (initially determined in yr
BP = AD 1950) to enable direct comparison with OSL ages.
c
Obtained from historical maps [12].
d
Calibration of 14 C ages according to the Groningen CAL25 program [18] using smoothed curves [19]; smoothing parameters
used: (1) 100; (2) 200.
e
Historical age of damming of the river at its upstream bifurcation [15].
f
Following calibration by Friedrich et al. [16], including a systematic uncertainty of 70 yr.
620
J. Wallinga et al. / Earth and Planetary Science Letters 193 (2001) 617^630
overbank deposits (sampling strategy is extensively discussed by To«rnqvist and Van Dijk
[13]). The end of activity of the Linge system
was a result of arti¢cial damming at its upstream
bifurcation around AD 1300 [15]. The end of activity of the Schaik system was derived from the
age of the base of peat overlying overbank deposits and basal peat in the residual channel [13].
We also sampled channel deposits of Late
Weichselian age, which underlie the Holocene deltaic deposits at Elden (Figs. 1 and 2). The £uvial
sands at this location are extremely rich in pumice
that originated from the Laacher See volcanic
eruption in the Eifel (Germany), 14 C dated at
11 063 þ 12 yr BP [16]. Pumice from the Laacher
See eruption blocked the River Rhine for several
days [17], and was transported downstream when
the dam collapsed. The pumice has been found
throughout Late Weichselian £uvial deposits in
The Netherlands [15], and the high concentration
(up to 25% by volume) at our sampling site makes
it highly likely that these sands were deposited
shortly after the volcanic eruption.
In addition to the sites with independent age
control (summarized in Table 1), samples were
taken from a core retrieved from Saalian (OIS
6^8) to Weichselian (OIS 2^4) £uvial and estuarine channel deposits near Leidschendam [20]
(Figs. 1 and 2). OSL dating of the quartz and
feldspar separates from this core allows us to
compare the results obtained on both mineral
fractions up to an age of V200 ka.
3. Methods
Most samples were taken using a simple handoperated suction corer, enabling us to obtain 30cm-long samples in opaque PVC tubes [21]. The
Leidschendam core was obtained using a mechanized bailer drilling unit [22] of the Netherlands
Institute of Applied Geoscience TNO, yielding an
undisturbed core with 10 cm diameter. The cores
were opened in subdued red light, after which
samples for OSL dating were taken. All samples
were water-washed and treated with 10% HCl and
30% H2 O2 to remove carbonates and organic material. After drying, the samples were sieved and
subsequently density-separated using an aqueous
solution of sodium polytungstate to extract the
potassium-rich feldspar fraction lighter than 2.58
g/cm3 . The denser fraction was treated with concentrated hydro£uoric acid for 40 min to obtain a
clean quartz sample and to etch away the outer 10
Wm of the quartz grains. No hydro£uoric acid
etching was used for the feldspar.
Measurements were made on an automated
RisÖ TL/OSL reader, using an internal 90 Sr/90 Y
L-source [23]. The sample grains were mounted
on aluminum or stainless steel discs using silicone
spray. Blue light emitting diodes were used for
stimulation of quartz (at 125³C) and the resulting
luminescence signal was detected through 9 mm
of Schott U-340 ¢lters (detection window 250^390
nm). The single-aliquot regenerative dose (SAR)
protocol [24] was used for estimation of the equivalent dose. A relatively low preheat (200³C for 10
s) was used to avoid thermal transfer e¡ects [1,25]
that were shown to a¡ect the equivalent dose of
the Winssen sample when more stringent preheats
were used (Fig. 3a). For the Leidschendam samples a more usual 10 s preheat at 260³C was used
because here thermal e¡ects are expected to be
negligible. This was con¢rmed by the preheat plateau obtained for sample Leidschendam I (Fig.
3b). The test-dose response was measured after
heating to 160³C for all samples.
For the feldspar separates, we used Schott BG39 and Corning 7-59 ¢lters, giving a transmission
window between 320 and 480 nm. The single-aliquot additive dose (SAAD) procedure [26,27],
with a 10 min preheat at 220³C, was used for
estimation of the equivalent dose. Optical stimulation was provided by infrared diodes (emitting
round 880 nm). The equivalent dose was also estimated using the SAR protocol for feldspar [28],
using a 10 s preheat at 290³C for natural and
regenerative doses, and heating to 210³C for the
test doses. An infrared laser diode (emitting at
830 nm) was used for stimulation.
The natural dose rate was estimated in the laboratory using high-resolution Q-spectrometry [29]
on bulk samples (results in Table 2), that were
taken from around the sample used for equivalent-dose determination. It is fair to assume that
the samples have been saturated with water
J. Wallinga et al. / Earth and Planetary Science Letters 193 (2001) 617^630
Fig. 3. Equivalent doses obtained for a range of preheat temperatures (each held for 10 s) for aliquots of quartz separate
from the Winssen (a) and Leidschendam I (b) samples. Each
estimate is the average from three aliquots, and error bars
indicate the standard error on the mean. The equivalent dose
(circles), and the recycling ratio (triangles) [24] are both
shown. The equivalent-dose estimate used for age determination (Table 2; Winssen preheat 200³C, Leidschendam I preheat 260³C) is indicated by the dotted line.
throughout their lifetimes, which diminishes the
dose rate [38].
4. OSL dating results and discussion
4.1. Quartz OSL
The quartz OSL dating results and independent
ages are presented in Table 2 and Fig. 4. For the
youngest sample (Winssen), the OSL age slightly
overestimates the known historical age of the
sample. For the other samples with independent
age control all OSL ages are in excellent agree-
621
ment with the radiocarbon-dated periods of activity. The slight o¡set (600 yr) found for the Winssen sample is likely caused by incomplete
bleaching, and is comparable to that found for
modern channel deposits in other parts of the
world [2,4,5]. We observed that application of
higher preheat temperatures (e.g. 260³C) resulted
in a greater o¡set for the Winssen sample (Fig.
3a), and also in a slight overestimation of age for
the other samples with independent age control
(not shown). These results are in accordance
with those of Rhodes [25] and indicate that stringent preheats should be avoided when dating
young samples that might not have been thoroughly exposed to light prior to deposition.
In addition to our test of the validity of quartz
OSL dating, we applied OSL dating to older deposits with limited age constraints to allow comparison of quartz OSL and feldspar IR-OSL dating results for a wider age range. The geological
context and the quartz OSL dating results on
these samples (Leidschendam core) are discussed
by To«rnqvist et al. [20]; a brief summary is given
here. The OSL data suggest that the base of the
succession (samples VIII^X; Table 2) was deposited during the Saalian glaciation (OIS 6). The
£attening of the OSL dose^response curve at the
relatively high doses these samples have absorbed
ampli¢es the scatter in the OSL measurements,
and this is re£ected in the relatively large uncertainties in the age estimates. The OSL age of sample VII suggests an Eemian or Early Weichselian
(OIS 5) origin, which is supported by the ¢rst
occurrence of transgressive marine shell remains
at this level [20].
OSL ages for the upper part of the succession
(samples I^VI) point to deposition around OIS 4,
i.e. during a period with relatively low sea level.
However, mud drapes containing warm pollen
and estuarine diatoms were encountered near the
base of this unit (sample VI, see also [20]), suggesting that these sediments must have been deposited during a period of relatively high sea level.
Hence, we cannot preclude the possibility that
these deposits were formed during the preceding
sea-level highstand, i.e. OIS 5a (V80 ka), and
that they are possibly older than the quartz OSL
ages suggest. The slight age reversal for samples
622
J. Wallinga et al. / Earth and Planetary Science Letters 193 (2001) 617^630
Table 2
Quartz OSL dating results
Sample
Grain
size
(Wm)
Radionuclide concentrationa
(Bq/kg)
238
Winssen
Rumpt I-3
Rumpt IV-2
Schelluinen II-3
Schelluinen II-6
Elden
Leidschendam I
Leidschendam II
Leidschendam III
Leidschendam IV
Leidschendam V
Leidschendam VI
Leidschendam VII
Leidschendam VIII
Leidschendam IX
Leidschendam X
180^212
180^212
180^212
180^212
180^212
180^212
180^212
180^212
180^212
180^212
180^212
180^212
180^250
180^250
180^212
180^212
U
12 þ 2
15 þ 2
35 þ 5
12 þ 3
11 þ 2
13 þ 2
10 þ 3
6þ3
6þ2
12 þ 3
6þ3
8þ3
12 þ 3
10 þ 2
11 þ 2
7þ3
226
Ra
10.8 þ 0.2
10.1 þ 0.4
31.7 þ 0.5
9.9 þ 0.3
9.3 þ 0.2
14.3 þ 0.4
9.5 þ 0.5
9.5 þ 0.3
9.5 þ 0.4
12.8 þ 0.3
6.9 þ 0.5
9.8 þ 0.2
6.8 þ 0.3
6.5 þ 0.4
5.5 þ 0.3
6.9 þ 0.3
232
Th
10.9 þ 0.2
12.3 þ 0.4
33.0 þ 0.5
10.7 þ 0.3
9.9 þ 0.2
16.4 þ 0.4
9.6 þ 0.4
10.8 þ 0.3
7.5 þ 0.3
14.1 þ 0.3
8.2 þ 0.4
10.7 þ 0.2
6.7 þ 0.2
8.2 þ 0.4
7.4 þ 0.2
8.2 þ 0.2
40
Dose
ratea;b
(Gy/ka)
Equivalent
dose
(Gy)
OSL
age
(ka)
Independent
age
(ka)
1.27 þ 0.05
1.36 þ 0.05d
2.10 þ 0.14
1.33 þ 0.05
1.22 þ 0.09d
1.51 þ 0.05
1.11 þ 0.06
0.99 þ 0.05
0.78 þ 0.05
1.29 þ 0.06
0.92 þ 0.05
1.13 þ 0.05
0.86 þ 0.04
0.79 þ 0.05
0.74 þ 0.04
0.87 þ 0.04
1.17 þ 0.12c
1.67 þ 0.13d
3.7 þ 0.2
6.9 þ 0.5
7.5 þ 0.2d
20.0 þ 1.0
54 þ 3
54 þ 5
64 þ 6
92 þ 7
56 þ 3
66 þ 3
104 þ 6
124 þ 7
107 þ 10
156 þ 23
0.92 þ 0.10
1.23 þ 0.10d
1.75 þ 0.10
5.1 þ 0.4
6.1 þ 0.5d
13.3 þ 0.8
48 þ 4
55 þ 6
82 þ 9
71 þ 6
61 þ 5
58 þ 4
120 þ 9
158 þ 13
145 þ 16
180 þ 28
0.3
0.7^2.4
0.7^2.4
5.2^6.0
5.2^6.0
13.0^13.3
^
^
^
^
^
^
^
^
^
^
K
362 þ 7
404 þ 13
482 þ 10
398 þ 8
363 þ 7
415 þ 12
297 þ 13
249 þ 7
200 þ 10
341 þ 7
253 þ 13
314 þ 7
237 þ 6
201 þ 10
190 þ 7
240 þ 6
a
Spectral data from high-resolution Q-spectroscopy converted to activity concentrations and in¢nite matrix dose rates using the
conversion data given by Olley et al. [30].
b
The natural dose rate was calculated from the in¢nite matrix dose rate using attenuation factors given by Mejdahl [31], and includes a contribution from cosmic rays [32]. A contribution from internal K dose was calculated based on U and Th contents reported by Mejdahl [33], using an a-value of 0.04 þ 0.01 [34], which resulted in an internal dose rate of 0.028 þ 0.013 Gy/ka. All
dose rates calculated for a water content of 20 þ 2% (based on a porosity of 34 þ 3% [35]) using attenuation factors given by Zimmerman [36].
c
Four outliers (De s 6 Gy) from a total of 34 aliquots were not incorporated.
d
Values di¡er slightly from those reported by Wallinga and Duller [37] due to a small shift in the source calibration, and an improved water content estimation.
Leidschendam III^VI might also indicate that the
OSL ages obtained on samples V and VI (V60
ka) could be too young. It is interesting to note
that slight age underestimation was also reported
for quartz OSL ages of Eemian (OIS 5e) deposits
from Denmark [39].
This study for the ¢rst time rigorously compares OSL ages and independent age control for
£uvial deposits over a relatively wide age range
(up to 13 ka), and the results underline the validity of quartz OSL dating for establishing absolute chronologies for £uvial deposits. However,
the controversy over the age of the Weichselian
deposits in the Leidschendam core stresses the
need for comparisons of OSL ages with independent age control for pre-Holocene deposits. Future
research should focus on such comparisons to further increase the con¢dence in quartz OSL ages.
4.2. Feldspar IR-OSL
The feldspar IR-OSL dating results obtained by
the SAAD protocol [26,27] are presented in Table
3 and Fig. 4. For the Winssen sample the IR-OSL
age overestimates the independent age and is also
slightly higher than that obtained by OSL dating
of the quartz separate. For the Rumpt samples
IR-OSL ages are found in agreement with the
independent age range and with the quartz OSL
dating results. However, for the Schelluinen and
Elden samples an IR-OSL age underestimation is
found, both compared to independent ages and
compared to the quartz OSL ages. An increasing
deviation with increasing age (Fig. 4b) suggests
that the underestimation is a relative e¡ect, possibly masked in the younger samples by poor
bleaching. This underestimation of IR-OSL age
J. Wallinga et al. / Earth and Planetary Science Letters 193 (2001) 617^630
Fig. 4. (a) OSL ages obtained on quartz separates using the
SAR protocol [24] with a 10 s 200³C preheat (¢lled circles),
and IR-OSL ages obtained on potassium-rich feldspar separates using the SAAD procedure [26,27] (open circles) plotted
against the independent age of the samples. All errors indicate 2c con¢dence intervals. (b) Di¡erence between the luminescence ages and the independent age control. Errors include uncertainties in both the independent ages and the
OSL ages.
is unlikely to be caused by errors in determination
of the external dose rate; this would also a¡ect
the quartz ages. It is improbable that the di¡erences between OSL ages on quartz and IR-OSL
ages on feldspar are caused by di¡erences in
bleaching, as the quartz signal is more readily
reset by sunlight [1], and because the quartz
ages agree with the independent chronology.
To investigate whether the age underestimation
could be caused by sensitivity changes during the
measurement procedure, a SAR protocol for feldspar [28] was employed. Using this protocol, ages
similar to those obtained using the SAAD procedure were obtained (Table 3, see also [28]). An
additional advantage of the SAR procedure is
that it is straightforward to test whether preheating removes all unstable trapped charge caused by
laboratory irradiation. Using the SAR protocol,
we found equivalent doses to be independent of
preheat temperature for 10 s preheats above
200³C (Fig. 5).
623
The IR-OSL age of feldspar separates from the
Leidschendam core was measured using the SAR
protocol. The average equivalent dose of three to
six aliquots of each sample is shown in Table 3. In
Fig. 6 the IR-OSL ages obtained on the samples
are plotted as a function of the quartz OSL age.
All but two samples follow a trend where the
feldspar IR-OSL age is only half that of the
quartz OSL age from the same sample. For the
two samples that do not follow this trend (Leidschendam VII and VIII), an atypically large scatter was observed between equivalent doses obtained on di¡erent feldspar aliquots, suggesting
that the IR-OSL signal was not completely reset
for all grains at the time of deposition [41].
From the results for the samples with independent age control, we deduced that the accuracy of
the quartz OSL ages is superior to that of the
feldspar IR-OSL ages. Combining the results obtained on the samples with independent age control, and that from the Leidschendam samples, we
conclude that the IR-OSL ages obtained on the
potassium-rich feldspar separates severely underestimate the age of our samples, but that in some
Fig. 5. Equivalent doses obtained for a range of preheat temperatures (each held for 10 s) for aliquots of feldspar separate from sample Schelluinen II-3. The sample was heated to
150³C after the test dose was administered. Each estimate is
the average of three aliquots. The equivalent dose obtained
by the SAR procedure (¢lled circles) and the recycling ratio
(¢lled triangles) are indicated, as are the correction factor for
di¡erent preheat temperatures (open triangles) and the equivalent dose after correction with this factor (open circles). The
derivation of the correction procedure is described in the
main text.
624
J. Wallinga et al. / Earth and Planetary Science Letters 193 (2001) 617^630
Table 3
Feldspar IR-OSL dating results
Sample
Internal Dose
rateb
Ka
(%)
(Gy/ka)
Winssen
8.5
Rumpt I-3
7.4
Rumpt IV-2
9.5
Schelluinen II-3
7.9
Schelluinen II-6
9.8
Elden
8.0
Leidschendam I
9.2
Leidschendam II
8.2
Leidschendam III
9.2
Leidschendam IV
9.1
Leidschendam V
9.2
Leidschendam VI
9.0
Leidschendam VII
8.9
Leidschendam VIII 8.5
Leidschendam IX
10.1
Leidschendam X
9.1
1.88 þ 0.08
1.92 þ 0.08e
2.87 þ 0.13
1.93 þ 0.07
1.91 þ 0.11e
2.12 þ 0.08
1.68 þ 0.09
1.51 þ 0.09
1.35 þ 0.08
1.87 þ 0.09
1.49 þ 0.08
1.69 þ 0.08
1.47 þ 0.12
1.36 þ 0.12
1.36 þ 0.08
1.43 þ 0.07
Equivalent dose
IR-OSL
age
(ka)
(Gy)
Independent
age
(ka)
SAADc
SARd
SAADc
SARd
3.0 þ 0.2
2.38 þ 0.05e
4.53 þ 0.13
7.7 þ 0.2
7.9 þ 0.2e
19.6 þ 0.5
^
^
^
^
^
^
^
^
^
^
2.9 þ 0.2
2.8 þ 0.4
4.2 þ 0.3
7.4 þ 0.5
7.6 þ 0.4
16.8 þ 1.2
46.9 þ 1.8
47.6 þ 1.5
59 þ 4
71 þ 5
52 þ 3
65 þ 3
146 þ 17
191 þ 13
96 þ 6
112 þ 4
1.58 þ 0.14
1.24 þ 0.05e
1.58 þ 0.08
4.0 þ 0.2
4.2 þ 0.3e
9.2 þ 0.4
^
^
^
^
^
^
^
^
^
^
1.52 þ 0.14
1.5 þ 0.2
1.47 þ 0.13
3.8 þ 0.3
4.0 þ 0.3
7.9 þ 0.6
28 þ 2
32 þ 2
44 þ 4
38 þ 3
35 þ 3
38 þ 3
100 þ 15
140 þ 16
71 þ 6
78 þ 5
0.3
0.7^2.4
0.7^2.4
5.2^6.0
5.2^6.0
13.0^13.3
^
^
^
^
^
^
^
^
^
^
a
The potassium content of the potassium-feldspar separates was determined by L-counting in a GM multicounter system [40]. A
calibration uncertainty of 5% was assumed.
b
Calculations similar to quartz, apart from a contribution from internal potassium [33]. An a-value of 0.08 þ 0.02 [34] was used
to calculate the contribution from internal K-radiation (0.06 þ 0.03 Gy/ka). As the separates were not HF-etched a small contribution from external K-radiation was also included (typically V0.02 Gy/ka).
c
SAAD protocol [26,27] used for equivalent-dose estimation.
d
SAR protocol [28] used for equivalent-dose estimation.
e
Values di¡er slightly from those reported by Wallinga and Duller [37] due to a small shift in the source calibration, and an improved water content estimation.
cases the underestimation is masked by incomplete bleaching. The quartz OSL ages of the samples from the Leidschendam core might slightly
underestimate the true age, as was discussed in
Section 4.1. If this is the case, the feldspar IROSL age underestimation is even more severe
than indicated by the comparison of quartz OSL
and feldspar IR-OSL results. In the following sections, the causes for the feldspar IR-OSL age
underestimation are considered.
5. Possible reasons for age underestimation in
feldspar IR-OSL
5.1. Anomalous fading
Anomalous fading is the loss of electrons from
traps on a time scale that is short compared with
the lifetime predicted on the basis of their trap
depth. This phenomenon is only known to a¡ect
feldspars and gives rise to an age underestimation
[42^44]. We carried out fading tests in three di¡erent ways.
Firstly, the decay in IR-OSL was monitored
following laboratory irradiation of previously unmeasured natural samples. Doses similar to the
natural dose of the sample were used for this experiment. After irradiation, the samples were preheated to 220³C for 10 min and measurements
were made using short exposure to infrared light.
To correct for the decay due to this measurement,
the same measurements were made on natural
samples that did not receive a laboratory dose
but were otherwise treated identically. The fading
ratio is given by the ratio of irradiated to natural
J. Wallinga et al. / Earth and Planetary Science Letters 193 (2001) 617^630
Fig. 6. Feldspar IR-OSL ages as a function of the quartz
OSL ages of the same samples. The SAR procedure was
used for equivalent-dose determination in both (see Tables 2
and 3 for sample names, equivalent doses and ages). Nearly
all samples follow a trend where the feldspar IR-OSL age is
only half the quartz OSL age (¢lled circles, dash^dot trend
line). The two samples that are indicated by triangles show
unusually wide scatter in the IR-OSL equivalent-dose determinations and were not incorporated in the regression. Also
shown are the feldspar IR-OSL ages after application of the
correction factor as determined for each sample (open symbols, dotted trend line; Table 5).
IR-OSL response before storage, divided by the
same ratio after storage at ambient temperature
for 6 months.
For the second and third tests, fading was
checked using the SAR procedure [28]. We compared the sensitivity-corrected OSL signal after
irradiation (V50 Gy) and storage with that measured immediately after irradiation (time between
irradiation and measurement was a maximum of
4 h). These fading tests were performed on aliquots that had previously been used for equivalent-dose determination using the feldspar SAR
procedure [28]. The samples were preheated to
290³C for 10 s prior to measurement of the IROSL signal. Samples were stored either for
4 months at ambient temperature (second test),
or for 10 days at 100³C (third test). Due to a
lack of material, it was not possible to apply all
three methods to each sample.
Some fading was detected in the Elden sample
using the third test (Table 4); unfortunately there
was not enough material to check this fading with
the other two methods. The fading in this sample
might be associated with the recent volcanic origin of some of the feldspar, as the material was
deposited shortly after the Laacher See volcanic
Table 4
IR-OSL fading tests on the feldspar separates
Sample
625
Fading ratio
Conventional
6 months ambient
SAR
4 months ambient
SAR
10 days 100³C
Winssen
Rumpt I-3
Rumpt IV-2
Schelluinen II-3
Schelluinen II-6
Elden
Leidschendam I
Leidschendam II
Leidschendam III
Leidschendam IV
Leidschendam V
Leidschendam VI
Leidschendam VII
Leidschendam VIII
Leidschendam IX
Leidschendam X
^
0.95 þ 0.04
1.01 þ 0.08
0.96 þ 0.02
0.87 þ 0.03
^
^
^
^
^
^
^
^
^
^
^
^
^
^
^
^
^
1.003 þ 0.005
0.98 þ 0.02
1.00 þ 0.02
0.934 þ 0.012
1.00 þ 0.03
0.957 þ 0.006
0.972 þ 0.008
0.98 þ 0.02
0.978 þ 0.006
0.996 þ 0.007
1.00 þ 0.05
1.003 þ 0.013
^
1.06 þ 0.02
1.00 þ 0.02
0.907 þ 0.010
0.940 þ 0.007
1.034 þ 0.002
0.91 þ 0.03
0.970 þ 0.011
0.962 þ 0.011
0.965 þ 0.007
0.962 þ 0.014
0.97 þ 0.02
0.944 þ 0.009
0.936 þ 0.008
Group average
Overall average
0.95 þ 0.03
0.980 þ 0.007
0.970 þ 0.007
0.970 þ 0.011
626
J. Wallinga et al. / Earth and Planetary Science Letters 193 (2001) 617^630
Fig. 7. Equivalent-dose determination on feldspar separates
from sample Schelluinen II-6 using the SARA protocol [48].
The equivalent dose obtained by this procedure is not affected by changes in the trapping probability due to preheating, as the added doses are administered prior to any heating
of the aliquots. In the absence of trapping sensitivity
changes, the regression should follow the dotted line of unit
slope.
eruption [16]. Based on the results of the fading
tests (Table 4), we cannot completely rule out the
existence of anomalous fading in our samples
(average fading ratio 0.970 þ 0.007). In interpreting this fading ratio, one should also keep in mind
that the laboratory time scale for fading tests is
very short compared to the geological time scale.
However, tests on samples that show an IR-OSL
age underestimate as a result of anomalous fading
tend to give an unambiguous indication of the
presence of fading [45]. On balance we therefore
think it is unlikely that anomalous fading is the
main cause for the severe age underestimation
found for our feldspar samples.
5.2. Sensitivity change
The SAR procedure was developed to overcome problems with sensitivity change during
measurement [24,28,46]. However, Murray and
Wintle [24] pointed out that if sensitivity changes
occur between measurement of the natural OSL
signal and measurement of the OSL from the test
dose related to the natural, the SAR procedure
will not detect or correct these. Wallinga et al.
[47] have shown that heating of feldspar grains
to temperatures higher than 200³C for 10 s can
cause changes in the charge trapping probability,
and thus can change the overall sensitivity. In
both SAR and SAAD measurement procedures,
such a change in trapping probability would occur during preheating of the aliquot (prior to the
¢rst measurement of the IR-OSL). Using samples
from the same area as discussed here, Wallinga et
al. [47] showed that the change in trapping probability resulted in underestimation of a known
laboratory dose administered prior to any heating
of the sample.
Changes in trapping probability because of
heating can be avoided if laboratory doses are
given prior to heating of the sample, as is the
case in multiple-aliquot methods. To test for
this, we applied the single-aliquot regeneration
and added dose (SARA) procedure [48], which
is (despite its name) a multiple-aliquot procedure.
We used the SAR protocol for determination of
the dose in those aliquots that just retained their
natural signal and in those where a laboratory
dose had been added to the natural dose, thereby
slightly modifying the standard SARA procedure
[48]. Using this protocol, higher equivalent doses
are obtained than with direct SAR measurements
(Fig. 7, Table 5). Although clearly an improvement for feldspar separates, the SARA procedure
might not be the preferred protocol. Firstly, it is
extremely time consuming, as the method needs
equivalent-dose determinations to be carried out
on a large number of aliquots. Secondly, the
equivalent dose is obtained by extrapolation,
which is not desirable, especially for older samples. Finally, a linear extrapolation might not be
justi¢ed [3].
As an alternative to the use of a SARA procedure to circumvent the problem, a sample-dependent correction factor can be determined by measuring the extent of the change in trapping
probability. For this purpose, three aliquots
from each sample were bleached for at least 2 h
in a Ho«nle SOL2 solar simulator, and subsequently given a dose similar to their natural
dose. As this ¢rst dose is administered prior to
any heating, it is expected to have the same trapping sensitivity as the natural dose. Hence, a correction factor for the natural equivalent dose can
be derived by dividing the known laboratory dose
627
J. Wallinga et al. / Earth and Planetary Science Letters 193 (2001) 617^630
Table 5
Feldspar IR-OSL results corrected for sensitivity changes
Sample
Correction factor
SARa
Feldspar IR-OSL equivalent dose
Feldspar IR-OSL age
(Gy)
(ka)
b
Winssen
Rumpt I-3
Rumpt IV-2
Schelluinen II-3
Schelluinen II-6
Elden
Leidschendam I
Leidschendam II
Leidschendam III
Leidschendam IV
Leidschendam V
Leidschendam VI
Leidschendam VII
Leidschendam VIII
Leidschendam IX
Leidschendam X
a
b
c
d
0.94 þ 0.04
0.97 þ 0.02
1.217 þ 0.014
1.16 þ 0.02
1.13 þ 0.08
1.20 þ 0.06
1.22 þ 0.08
1.34 þ 0.08
1.32 þ 0.09
1.41 þ 0.13
1.21 þ 0.13
1.43 þ 0.04
1.324 þ 0.007
1.34 þ 0.07
1.36 þ 0.06
1.40 þ 0.05
c
Quartz OSL
age
(ka)
SAR
SARA
Corrected
SARd
SARb
SARAc
Corrected
SARd
2.9 þ 0.2
2.8 þ 0.4
4.2 þ 0.3
7.4 þ 0.5
7.6 þ 0.4
16.8 þ 1.2
46.9 þ 1.8
47.6 þ 1.5
59 þ 4
71 þ 5
52 þ 3
65 þ 3
146 þ 17
191 þ 13
96 þ 6
112 þ 4
4.2 þ 2.1
3.6 þ 0.6
6.9 þ 0.6
11.2 þ 2.2
9.5 þ 0.3
^
52 þ 6
^
^
^
^
^
^
^
137 þ 18
^
2.7 þ 0.2
2.7 þ 0.4
5.1 þ 0.4
8.5 þ 0.6
8.6 þ 0.8
20.2 þ 1.7
57 þ 4
64 þ 4
78 þ 7
99 þ 11
63 þ 8
93 þ 5
193 þ 23
256 þ 22
130 þ 10
156 þ 8
1.52 þ 0.14
1.5 þ 0.2
1.47 þ 0.13
3.8 þ 0.3
4.0 þ 0.3
7.9 þ 0.6
28 þ 2
32 þ 2
44 þ 4
38 þ 3
35 þ 3
38 þ 3
100 þ 15
140 þ 16
71 þ 6
78 þ 5
2.3 þ 1.1
1.9 þ 0.3
2.4 þ 0.2
5.8 þ 1.2
5.0 þ 0.3
^
31 þ 4
^
^
^
^
^
^
^
105 þ 8
^
1.44 þ 0.14
1.4 þ 0.2
1.8 þ 0.2
4.4 þ 0.4
4.5 þ 0.5
9.5 þ 0.9
34 þ 3
43 þ 4
58 þ 6
53 þ 7
42 þ 6
55 þ 4
136 þ 18
194 þ 20
96 þ 9
109 þ 8
0.92 þ 0.10
1.23 þ 0.10d
1.75 þ 0.10
5.1 þ 0.4
6.1 þ 0.5
13.3 þ 0.8
48 þ 4
55 þ 6
82 þ 9
71 þ 6
61 þ 5
58 þ 4
120 þ 9
158 þ 13
145 þ 16
180 þ 28
Correction needed to account for sensitivity changes due to preheating. Details are given in the main text.
The SAR protocol [28] was used for equivalent-dose estimation.
The SARA protocol [48] was used for equivalent-dose estimation.
As derived using the correction factors given in the ¢rst column.
by the dose estimated by the SAR procedure. Using SAR and these correction factors (Table 5),
similar ages were obtained as by the SARA protocol (Table 5). An apparently more elegant approach would be to preheat only to temperatures
below 200³C for 10 s, but this is outside the temperature range needed to remove unstable trapped
charge after laboratory irradiation (indicated by
the rising plateau in this region in Fig. 5).
Either the use of the correction factors or the
application of the SARA procedure improves the
feldspar IR-OSL ages, and brings them closer to
the independent ages and the quartz OSL ages
(Table 5). However, even after correction for
trapping probability there is still a clear age
underestimation (Fig. 6). It must be recognized
that the correction factor may be underestimated
as a consequence of accidental heating of the extracts prior to measurement. During sample preparation the extracts are exposed to temperatures
above ambient during drying at 60³C, during
treatment with H2 O2 (some samples were heated
signi¢cantly by exothermic reactions), and during
bleaching in the solar simulator. It is possible that
even at these relatively low temperatures the trapping probability changes. In radioluminescence
studies of potassium feldspar, sensitivity changes
up to 50% have been shown to occur at temperatures below 100³C [49]. We strongly recommend
that temperatures above ambient should be
avoided at any stage during sample preparation
for IR-OSL dating of feldspar separates.
It is not yet known whether the change in trapping probability in feldspar as a consequence of
heating is common. In a preliminary investigation, the correction factor was determined for
three feldspar separates from Denmark (samples
989201^989203 [50]) and one feldspar separate
from New Zealand (sample GDNZ6 [8]). The correction factors obtained for a 10 s, 290³C preheat
were all less than 1.1, indicating that the e¡ect
does not produce an error greater than 10% for
these samples. For the samples from Denmark,
identical results were obtained from the OSL dating of the quartz fraction and IR-OSL dating of
the feldspar separates (both used SAR procedures
628
J. Wallinga et al. / Earth and Planetary Science Letters 193 (2001) 617^630
for equivalent-dose determination [50]). Although
it is not clear how widespread this problem is, we
recommend that testing for the phenomenon
should be routine practice when single-aliquot
IR-OSL dating is applied to feldspar separates.
It should be stressed that underestimation in
the luminescence age obtained on feldspar separates has also been found when multiple-aliquot
procedures were used [9]. Moreover, correcting
for changes in trapping probability or application
of multiple-aliquot techniques does not eliminate
the age underestimation for our samples. This indicates that there must be additional reasons for
the underestimation. A possible candidate would
be incomplete removal of unstable charge from
laboratory irradiation. Incorporation of the correction factor for di¡erent preheats, as shown in
Fig. 4, suggests that a true preheat plateau does
not exist for this sample ; after correction for
changes in trapping probability the `plateau' rises
continuously. This indicates that unstable charge
is present up to temperatures where almost all
trapped charge is removed (10 s preheat
s 325³C). The apparent plateau shown for the
uncorrected data (Fig. 4) appears to be a consequence of two competing phenomena: incomplete
removal of unstable charge giving a rising trend
with temperature, and changes in trapping probability causing a decreasing trend.
6. Conclusions
OSL ages, obtained using the SAR protocol on
quartz from £uvial channel deposits in the Rhine^
Meuse Delta, are in excellent agreement with tight
independent age control for the range of 1^13 ka.
Thermal transfer was shown to result in a small
overestimation of age, but this unwanted e¡ect
was largely avoided by using a less stringent preheat regime. Our results con¢rm the applicability
of quartz OSL dating to establish absolute chronologies for late Quaternary sedimentary records
in general, and £uvial records in particular.
Single-aliquot IR-OSL dating of feldspar separates proved to be less successful. The IR-OSL
age of the feldspar samples is underestimated by
up to 50% in comparison with independent age
control (up to 13 ka), and quartz OSL dating
results (up to 200 ka). We show that part of this
age underestimation is caused by changes in the
charge trapping probability as a consequence of
heating of the sample during single-aliquot procedures. This problem can be circumvented by using
the SARA protocol, or by determining a sampledependent correction factor. Both procedures produce results that are in better agreement with the
independent age control, but they only partly
solve the underestimation problem. Clearly, our
results indicate that previously established luminescence chronologies based on coarse-grain feldspar may need re-evaluation.
Considering the problems encountered in the
IR-OSL dating of feldspar, we suggest that quartz
is the mineral of choice for OSL dating of these
deposits, and probably of late Quaternary sediments in general. Nevertheless, it is important
that the problems with coarse-grain feldspar dating continue to be investigated, in view of the
potential of feldspars to extend luminescence dating to much longer time scales than quartz.
Acknowledgements
This is a contribution to the NEESDI (Netherlands Environmental Earth System Dynamics Initiative) program. J.W. is grateful for additional
funding received from the Netherlands Organization for Scienti¢c Research (NWO), which allowed him to visit the Luminescence Laboratory
of the University of Wales in Aberystwyth, UK,
and the Nordic Laboratory for Luminescence
Dating, Aarhus University, Denmark, and to
carry out the OSL measurements at those facilities. Thanks are due to the Netherlands Institute
of Applied Geoscience (NITG-TNO) for drilling
and processing of the Leidschendam core. Lorraine Morrison (University of Wales Aberystwyth) kindly etched the quartz grains of the Holocene samples ; Mette Adrian and Anne SÖrensen
(Nordic Laboratory for Luminescence Dating)
did the same for the other samples and helped
with the Q-spectroscopy. The manuscript was
greatly improved following thorough reviews by
Ann Wintle, Michel Lamothe and an anonymous
J. Wallinga et al. / Earth and Planetary Science Letters 193 (2001) 617^630
reviewer. We greatly appreciate their interest and
constructive comments.[AH]
[15]
References
[1] M.J. Aitken, An Introduction to Optical Dating, Oxford
University Press, Oxford, 1998, 267 pp.
[2] A.S. Murray, J.M. Olley, G.G. Caitcheon, Measurement
of equivalent doses in quartz from contemporary waterlain sediments using optically stimulated luminescence,
Quat. Sci. Rev. 14 (1995) 365^371.
[3] A.S. Murray, Developments in optically stimulated luminescence and photo-transferred thermoluminescence dating of young sediments: application to a 2000-year sequence of £ood deposits, Geochim. Cosmochim. Acta
60 (1996) 565^576.
[4] J.M. Olley, G.G. Caitcheon, A.S. Murray, The distribution of apparent dose as determined by optically stimulated luminescence in small aliquots of £uvial quartz: implications for dating young sediments, Quat. Sci. Rev. 17
(1998) 1033^1040.
[5] S. Stokes, H.E. Bray, M.D. Blum, Optical resetting in
large drainage basins; tests of zeroing assumptions using
single-aliquot procedures, Quat. Sci. Rev. 20 (2001) 879^
885.
[6] S. Stokes, D.R. Gaylord, Optical dating of Holocene dune
sands in the Ferris Dune Field, Wyoming, Quat. Res. 39
(1993) 274^281.
[7] A.S. Murray, L.B. Clemmensen, Luminescence dating of
Holocene aeolian sand movement, Thy, Denmark, Quat.
Sci. Rev. 20 (2001) 751^754.
[8] G.A.T. Duller, Luminescence dating using feldspars: A
test case from southern North Island, New Zealand,
Quat. Sci. Rev. 13 (1994) 423^427.
[9] U. Radtke, A. Janotta, A. Hilgers, A.S. Murray, The
potential of OSL and TL for dating Lateglacial and Holocene dune sands tested with independent age control of
the Laacher See tephra (12 880 a) at the Section `MainzGonsenheim', Quat. Sci. Rev. 20 (2001) 719^724.
[10] T.E. To«rnqvist, Holocene alternation of meandering and
anastomosing £uvial systems in the Rhine^Meuse Delta
(central Netherlands) controlled by sea-level rise and subsoil erodibility, J. Sediment. Petrol. 63 (1993) 683^693.
[11] H.J.A. Berendsen, E. Stouthamer, Late Weichselian and
Holocene palaeogeography of the Rhine^Meuse delta,
The Netherlands, Palaeogeogr. Palaeoclimatol. Palaeolecol. 161 (2000) 311^335.
[12] H. Middelkoop, Embanked £oodplains in the Netherlands; Geomorphological evolution over various time
scales, Neth. Geogr. Stud. 224 (1997) 1^341.
[13] T.E. To«rnqvist, G.J. Van Dijk, Optimizing sampling strategy for radiocarbon dating of Holocene £uvial systems in
a vertically aggrading setting, Boreas 22 (1993) 129^
145.
[14] T.E. To«rnqvist, A.F.M. de Jong, W.A. Oosterbaan, K.
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
629
van der Borg, Accurate dating of organic deposits by
AMS 14 C measurement of macrofossils, Radiocarbon 34
(1992) 566^577.
A. Verbraeck, Toelichtingen bij de Geologische Kaart van
Nederland 1:50.000. Blad Tiel West (39W) en Blad Tiel
Oost (39O), Rijks Geologische Dienst, Haarlem, 1984, 355
pp.
M. Friedrich, B. Kromer, M. Spurk, J. Hofmann, K.L.
Kaiser, Paleo-environment and radiocarbon calibration as
derived from Lateglacial/Early Holocene tree-ring chronologies, Quat. Int. 61 (1999) 27^39.
H.U. Schmincke, C. Park, E. Harms, Evolution and environmental impacts of the eruption of Laacher See Volcano (Germany) 12.900 a BP, Quat. Int. 61 (1999) 61^72.
J. van der Plicht, The Groningen radiocarbon calibration
program, Radiocarbon 35 (1993) 231^237.
T.E. To«rnqvist, M.F.P. Bierkens, How smooth should
curves be for calibrating radiocarbon ages?, Radiocarbon
36 (1994) 11^26.
T.E. To«rnqvist, J. Wallinga, A.S. Murray, H. de Wolf, P.
Cleveringa, W. de Gans, Response of the Rhine^Meuse
system (west^central Netherlands) to the last Quaternary
glacio-eustatic cycles: a ¢rst assessment, Glob. Planet.
Change 27 (2000) 89^111.
J. Wallinga, J. van der Staay, Sampling in waterlogged
sands with a simple hand-operated corer, Ancient TL 17
(1999) 59^61.
E. Oele, W. Apon, M.M. Fischer, R. Hoogendoorn, C.S.
Mesdag, E.F.J. de Mulder, B. Overzee, A. Seso«ren, W.E.
Westerho¡, Surveying The Netherlands: sampling techniques, maps and their applications, in: M.W. van den
Berg, R. Felix (Eds.), Special Issue in the Honour of
J.D. de Jong, Geol. Mijnb. 62 (1983) 355^372.
L. BÖtter-Jensen, E. Bulur, G.A.T. Duller, A.S. Murray,
Advances in luminescence instrument systems, Radiat.
Meas. 32 (2000) 523^528.
A.S. Murray, A.G. Wintle, Luminescence dating of quartz
using an improved single-aliquot regenerative-dose protocol, Radiat. Meas. 32 (2000) 57^73.
E.J. Rhodes, Observations of thermal transfer OSL signals in glacigenic quartz, Radiat. Meas. 32 (2000) 595^
602.
G.A.T. Duller, Equivalent dose determination using single
aliquots, Nuclear Tracks Radiat. Meas. 18 (1991) 371^
378.
G.A.T. Duller, Luminescence dating using single aliquots
methods and applications, Radiat. Meas. 24 (1995) 217^
226.
J. Wallinga, A.S. Murray, A.G. Wintle, The single-aliquot
regenerative-dose (SAR) protocol applied to coarse-grain
feldspar, Radiat. Meas. 32 (2000) 529^533.
A.S. Murray, R. Marten, A. Johnston, P. Marten, Analysis for naturally occurring radionuclides at environmental concentrations by gamma spectrometry, J. Radioanal.
Nuclear Chem. 115 (1987) 263^288.
J.M. Olley, A.S. Murray, R.G. Roberts, The e¡ects of
disequilibria in the uranium and thorium decay chains
630
[31]
[32]
[33]
[34]
[35]
[36]
[37]
[38]
[39]
[40]
J. Wallinga et al. / Earth and Planetary Science Letters 193 (2001) 617^630
on burial dose rates in £uvial sediments, Quat. Sci. Rev.
15 (1996) 751^760.
V. Mejdahl, Thermoluminescence dating beta dose attenuation in quartz grains, Archaeometry 21 (1979) 61^72.
J.R. Prescott, J.T. Hutton, Cosmic ray contributions to
dose rates for luminescence and ESR dating large depths
and long-term time variations, Radiation Measurements
23 (1994) 497^500.
V. Mejdahl, Internal radioactivity in quartz and feldspar
grains, Ancient TL 5 (1987) 10^17.
J. Rees-Jones, M.S. Tite, Optical dating results for British
archaeological sediments, Archaeometry 39 (1997) 177^
188.
H.J.T. Weerts, Complex con¢ning layers; architecture and
hydraulic properties of Holocene and Late Weichselian
deposits in the £uvial Rhine^Meuse delta, The Netherlands, Neth. Geogr. Stud. 213 (1996) 1^187.
D.W. Zimmerman, Thermoluminescent dating using ¢ne
grains from pottery, Archaeometry 13 (1971) 29^52.
J. Wallinga, G.A.T. Duller, The e¡ect of optical absorption on the infrared stimulated luminescence age obtained
on coarse-grain feldspar, Quat. Sci. Rev. 19 (2000) 1035^
1042.
M.J. Aitken, Thermoluminescence Dating, Academic
Press, London, 1985, 359 pp.
A.S. Murray, J.M. Olley, Precision and accuracy in the
optically stimulated luminescence dating of sedimentary
quartz: a status review, Geochronometria, submitted.
L. BÖtter-Jensen, V. Mejdahl, Determination of potassium in feldspars by beta counting using a GM multicounter system, Nuclear Tracks Radiat. Meas. 10 (1985)
663^666.
[41] S.H. Li, Optical dating insu¤ciently bleached sediments,
Radiat. Meas. 23 (1994) 563^567.
[42] A.G. Wintle, Anomalous fading of thermoluminescence in
mineral samples, Nature 245 (1973) 143^144.
[43] N.A. Spooner, The anomalous fading of infrared-stimulated luminescence from feldspars, Radiat. Meas. 23
(1994) 625^632.
[44] M. Lamothe, M. Auclair, A solution to anomalous fading
and age shortfalls in optical dating of feldspar minerals,
Earth Planet. Sci. Lett. 171 (1999) 319^323.
[45] D.J. Huntley, M. Lamothe, Ubiquity of anomalous fading
in K-feldspars, and the measurement and correction for it
in optical dating, Can. J. Earth Sci. 38 (2001) 1093^1106.
[46] A.S. Murray, R.G. Roberts, Measurement of equivalent
dose in quartz using a regenerative-dose single-aliquot
protocol, Radiat. Meas. 29 (1998) 503^515.
[47] J. Wallinga, A.S. Murray, G.A.T. Duller, Underestimation of equivalent dose in single-aliquot optical dating of
feldspar caused by pre-heating, Radiat. Meas. 32 (2000)
691^695.
[48] V. Mejdahl, L. BÖtter-Jensen, Luminescence dating of archaeological materials using a new technique based on
single aliquot measurements, Quat. Sci. Rev. 7 (1994)
551^554.
[49] T. Trautmann, M.R. Krbetschek, A. Dietrich, W. Stolz,
The basic principle of radioluminescence dating and a
localized transition model, Radiat. Meas. 32 (2000) 487^
492.
[50] K. Strickertsson, A.S. Murray, H. Lykke-Andersen, Optically stimulated luminescence dates for Late Pleistocene
sediments from Stensn×s, Northern Jutland, Denmark,
Quat. Sci. Rev. 20 (2001) 755^759.