ESR dating of sedimentary quartz: Possibilities and limitations of the single-grain approach

Usually, multiple-grain aliquots are used for electron spin resonance (ESR) dating of sediments. However, this approach excludes the ability of detecting insufficient bleaching, as would be the case when measuring several single aliquots or single grains. In this paper, we present preliminary results of single-grain ESR dating experiments on three different sedimentary deposits (fluvial, fluvio-aeolian and desert aeolian), ranging in age between 100 ka and 2 Ma. Titanium-related impurity centres (Ti–Li and Ti–H) were measured in a Q-band ESR spectrometer for estimating equivalent doses (D_e) of individual quartz grains (0.5–1 mm). Both additive and regenerative dose methods were used. The resulting D_e plots show a large scatter in D_e—from 100 Gy to more than 600 Gy—for the fluvial sample, probably reflecting different bleaching histories of the grains. On the contrary, a clear plateau can be observed in D_e for sands with a strong aeolian component. Preliminary single-grain ESR ages are in relatively good agreement with independent estimates, the best results being obtained for the desert aeolian sample (200 ka). It is concluded that ESR has a promising potential for estimating large naturally accumulated doses of well-bleached sediments. Nevertheless, the method is challenged by several experimental difficulties, such as long measurement times, excessive errors, and presumably a complex interrelationship between Ti–Li and Ti–H components.     

Resetting of sediments mobilised by the LGM ice-sheet in southern Norway

Former geological field investigations in the Rondane area, east-central southern Norway, have proposed that the maximum Fennoscandian ice-sheet coverage occurred during the Late Weichselian Glacial Maximum (LGM, ca. 20 ka) and that subsequent glaciofluvial sediments were first deposited in the early Holocene (after 10 ka). However, recent field investigations with ages from three internally consistent quartz optically stimulated luminescence (OSL) age series show an apparent deglaciation of northern Rondane in the period 20.0–13.8 ka. We examine here the possibility that these ages are too old because the sediment was not completely zeroed prior to deposition. Our investigations of incomplete bleaching use modern analogues, small aliquots, and single grains of quartz. First, the symmetric shape of small aliquot equivalent dose distributions suggests that the sediment was probably well bleached at deposition. This is supported by 5 modern analogue equivalent doses (D_e) of 0.6 Gy, 1.5% of the typical D_e from the deglaciation sediments. Finally, from single grain studies on three samples, we conclude that there is no evidence for poor bleaching in these samples; thus the weighted mean gives the best estimates of D_e, and these are completely consistent with both large and small aliquot estimates for these samples. These comparisons between large aliquots, modern analogues, small aliquots and single-grain analyses help to validate the OSL ages and confirm the complete resetting of these sediments prior to deposition.     

Detecting post-depositional sediment disturbance in sandy deposits using optical luminescence

Post-depositional mixing or exhumation is common in surficial sediments, yet may be unobservable from field evidence. However, any disturbance may have significant consquences in terms of establishing a reliable luminescence age determination. Optically stimulated luminescence (OSL) measurements, particularly measurements at the single grain level, can be used to gain an insight into both contemporary and past post-depositional processes.

This paper examines sites from Texas and Florida (USA) with independent chronological control to demonstrate the potential effects of varying degrees of bioturbation on OSL. Results show that contemporary soil forming processes clearly impact on the palaeodose (D_e) replicate distributions which are measured in order to derive an OSL age. Significant levels of scatter and apparently zero dose grains are observed in the upper-most sediments; declining with depth from the surface. D_e replicates from undisturbed and fully bleached sediments are unskewed, show low overdispersion (OD) and comparable single grain and single aliquot OSL ages. Bioturbated sediments, however, may show highly skewed multi-model D_e distributions with higher OD values, zero dose grains at depth, and significant diffences between single grain and single aliquot results. True burial ages may be derived from minimally bioturbated sediments through the application of statistical analysis such as finite mixture modelling to isolate D_e components. However, for significantly bioturbated sediments, the latter approach, even at the single grain level, produces inaccurate ages. In such cases we argue that additional evidence (both dating and contexual) may be required to identify with confidence the burial D_e population.     

Chronology of an oyster reef on the coast of Bohai Bay, China: Constraints from optical dating using different luminescence signals from fine quartz and polymineral fine grains of coastal sediments

Fossil oyster reefs are indicators of past sea levels, and their formation is usually dated by means of radiocarbon. However, radiocarbon dating of the shells from coastal areas may be complicated by the varying sources of carbon. Here we applied optical dating methods to date the samples from above and below a fossil oyster bed in a section on the coast of Bohai Bay, China. The optical ages of the sediments were used to constrain the oyster bed. Single-aliquot regenerative-dose procedures using the OSL signal from fine grain quartz, the IRSL and post-IR OSL signals from polymineral fine grains were employed to determine equivalent dose (D_e). The behaviors of the different luminescence signals from quartz and polymineral grains during D_e measurements were examined. The results showed that the quartz OSL signal is more reliable than the polymineral IRSL and post-IR OSL signals with respect to dating for these coastal samples. The optical ages indicated that the oyster reef formed between ca. 6.2 and 5.0 ka.     

Quartz and polymineral luminescence dating of Japanese loess over the last 0.6 Ma: Comparison with an independent chronology

There are only a few luminescence dating studies of loess sediments in Japan, but interleaved with these deposits are many well-described tephras of known age based on ¹⁴C and fission track analysis; these independent age controls provide an opportunity to test the reliability of loess luminescence ages. This study provides such a comparison at two sites in central Honshu, the largest island in the Japanese archipelago. Samples were collected from sequences of interleaved volcanic tephra and loess deposited on a Middle Pleistocene river terrace in the Niigata Prefecture and on an Early to Middle Pleistocene dissected fluvial surface at in the Tochigi Prefecture, Honshu. Equivalent doses (D_e) were estimated from fine grains (4–11 μm) using both polymineral IR-OSL and (post-IR) blue-OSL, and quartz blue-OSL. The blue-stimulated luminescence signals could be represented by up to three exponentially decaying components; only the most light sensitive of these components was used in the final D_e estimation. Almost all the estimates of D_e from polymineral IR-OSL are smaller than those from polymineral (post-IR) blue-OSL and quartz blue-OSL, whereas the latter two are in good agreement. The blue-light stimulated luminescence ages using the most light sensitive component are shown to be in good agreement with the independent control, up to 500 ka. Although the IR-OSL signals underestimate the known age, a simple laboratory fading test is found empirically to correct for this underestimation. We conclude that the most rapidly blue-stimulated luminescence signals from quartz extracted from our loess samples give reliable ages, and that future loess dating studies should concentrate on these signals.     

The Rhafas Cave (Morocco): Chronology of the mousterian and aterian archaeological occupations and their implications for Quaternary geochronology based on luminescence (TL/OSL) age determinations

The Middle Palaeolithic chronology of the Maghreb region of North Africa is poorly known because of the paucity of sites with a long stratigraphy and the limited number of available radiometric dates. In this paper, we report the age-estimates obtained by the TL and OSL methods on sediments and burnt lithic samples from the Rhafas Cave in Eastern Morocco. Our results indicate that the Mousterian is largely earlier than 100 ka, that the latest Mousterian dates to OIS 5 between 90–80 ka and the Mousterian to Aterian transition occurred about 70–80 ka, at the end of OIS 5 or during OIS 4. These dates will be most useful in the construction of a chronological framework of this unique sequence and for the interpretation of paleoenvironmental information.     

Problems associated with luminescence dating of Late Quaternary glacial sediments in the NW Scottish Highlands

Linking the timing of glacial episodes and behaviour to climatic shifts that are documented in ice and marine sedimentary archives is key to understanding ocean-land interactions. In the NW Scottish Highlands a large number of closely spaced (‘hummocky’) moraines formed at retreating glacier margins. Independent age control on one palaeo-glacier limit is consistent with the timing of Younger Dryas (YD) glaciation in the area, but adjacent glacier lobes have remained undated due to the lack of sites and material for ¹⁴C dating. Direct dating of ice-marginal moraines using optically stimulated luminescence (OSL) techniques has never been attempted before in Scotland, but if successful, they may be the most appropriate methods for constraining the age of sediment deposition in the absence of organic material. Coarse-grained quartz and K-feldspar minerals from supraglacial sheet flow deposits and glacilacustrine sediments within ice-marginal moraines were analysed using the single-aliquot regenerative-dose (SAR) protocol. Independent age control and clear geomorphological relationships indicate that all samples should yield YD or post-Last Glacial Maximum (LGM) ages. Quartz OSL shine down curves showed low luminescence sensitivity, significant medium-to-slow components, a weak fast component, and scattered SAR data; Linearly Modulated-OSL (LM-OSL) measurements confirmed that the fast component was weak or absent. In contrast, feldspar infrared stimulated luminescence (IRSL) was highly sensitive with excellent SAR data. However, SAR data from 3 mm diameter aliquots of feldspar (200 grains) give higher than expected equivalent doses (D_e) by an order of magnitude. SAR measurements of small clusters of feldspar grains (ranging from 1–8) considerably broaden the apparent D_e distribution, but even the lowest value is about 2–3 times the expected D_e. Two possibilities arise to explain the quartz and feldspar data: (1) that glacial sequences in the NW Highlands re-work inherited (older) glacial deposits and that some of the pre-Devensian existing glacial landforms have only been modified during Devensian glaciation; or (2) that the sedimentary processes operating in these ice-marginal environments are not conducive to adequate bleaching of quartz and feldspar grains. Our study implies that ice-proximal supraglacial sediments from this region in NW Scotland reflect older ages of deposition, but dating YD sediments has not been possible.     

Luminescence chronology of fluvial and aeolian deposits in the Russian Altai (Southern Siberia)

Geomorphological investigations in the Russian Altai Mountains provide evidence for the extent and timing of Late Quaternary glaciations and aeolian sediments. Infrared optically stimulated luminescence (IR-OSL) and thermoluminescence (TL) dating of aeolian sediments from the central part of the Russian Altai were carried out on silty and sandy sediments that cover fluvial and glacio-fluvial terraces. Most samples from loess, loess-like sediments and sandy loess taken from different terrace sequences within the Altai Mountains provide IRSL ages suggesting a main aeolian deposition period between 25 and 15 ka. These data are consistent with those from the adjacent Mongolian Altai. Sand and silt layers within moraines yielded IRSL age estimates between 22 and 19 ka and correlate to the ice margins of the Last Glacial Maximum. Aeolian dune sands overlying or neighbouring the loess-like sediments and the fluvial terraces at some places provide evidence for Late Glacial (around 15 ka) and Early Holocene (around 9 ka) aeolian activity. The youngest sand sheets gave deposition ages of about 1.5 ka. These sediments covered graves and are related to overgrazing. They provide evidence for the significant human impact on the environment. Fluvial sediments, including silt and sandy gravel intermingled with charcoal, yielded contradicting radiocarbon and luminescence ages. IRSL age estimates obtained for the silty layers range from 19 to 18 ka and are significantly older than the radiocarbon ages, which gave sub-recent ages between 1170 and 910 BP. These data indicate fluvial sedimentation of debris and mudflows in a period of deforestation and strong soil erosion related to mining activities during the 9th and 10th century.     

Was southern Sweden ice free at 19–25 ka, or were the post LGM glacifluvial sediments incompletely bleached?

Glacifluvial deposits along an ice-marginal zone in Småland, southern Sweden, have been dated using post-IR blue OSL. To test for incomplete bleaching, we adopted two strategies: analysis of modern analogues and small-aliquot dose distributions. Samples of modern fluvial sediments show no significant incomplete bleaching; they yield equivalent doses of only 0.5–2 Gy (0.25–4% of our glacifluvial sediment doses). Small-aliquot dose distributions do not provide any evidence for incomplete bleaching.

The sediments are believed to have been deposited during deglaciation and appear to fall into two age groups: 19–25 ka (mainly sandur sediments) and 33–73 ka (mainly deltaic sediments). Compared to the expected ages (13–15 ka), even the younger glacifluvial OSL ages appear up to 10 000 years (25 Gy) too old. The ages are nevertheless stratigraphically consistent and correspond between sites; we deduce that the 19–25 ka ages are true deposition ages. For glacifluvial sedimentation to take place on the South Swedish Upland at this time either a very early deglaciation is required, or alternatively ice-free conditions just prior to the LGM. The deltaic sediments (33–73 ka) were most likely not significantly bleached during deposition and thus they date events prior to the latest ice advance.     

Luminescence dating of old (>70 ka) Chinese loess: A comparison of single-aliquot OSL and IRSL techniques

The applicability of two different approaches in the luminescence dating of old (>70 ka) Chinese loess is investigated. Both SAR-OSL ages obtained on 63–90 μm quartz grains and SAR-IRSL ages obtained on 4–11 μm polymineral grains, for samples collected from two sites in the Chinese Loess Plateau (Luochuan and Dongchuan) are presented. The characteristics of the luminescence signals stimulated by blue and infrared light are investigated in terms of dose response and dose recovery, and as a function of age. Additionally, anomalous fading measurements from the 410 nm IRSL emission in polymineral fine-grains are reported. An average value of g_2days amounting to 3% per decade was measured and seems to be independent of site location and age. For the samples from Luochuan, independent age control (pedostratigraphy and palaeomagnetism) is available. At both sites, the SAR-OSL ages are always lower than the SAR-IRSL ages after they have been corrected for anomalous fading. It seems that the quartz-based SAR-OSL ages are accurate for the younger ages, but that they underestimate the true age of deposition for loess that was deposited about 60–70 ka ago. The fading-corrected SAR-IRSL ages are in better agreement with the pedostratigraphic age control (75 and 130 ka) and allow dating beyond the quartz OSL range. Based on our results, we suggest that conventional SAR-OSL and SAR-IRSL protocols at these sites should be restricted to samples of ages not exceeding 40–50  and 100–120 ka, respectively.     

Dating of lower terrace sediments from the Middle Rhine area, Germany

Modern and known-age Pleistocene fluvial sediments were investigated by optical dating of quartz to test the suitability of the approach for dating deposits from the deeply incised Middle Rhine Valley. Samples from modern flood sediments revealed skewed distributions indicating different residual levels of equivalent dose (D_e) within the different aliquots. Nevertheless, a substantial number of aliquots from the modern deposits reflect D_e values close to zero. For the Pleistocene samples, optical ages are in general consistent with age control given by the presence of the Laacher See Tephra and radiocarbon dating. However, some samples overestimate the known age by a few thousand years when using the arithmetic mean. This is apparently explained by including aliquots in the determination of mean D_e where the optical signal was incompletely bleached at deposition. The most difficult issue in this context is identifying a suitable approach that can distinguish between the variability of D_e due to partial bleaching and microdosimetry. However, even when considering these limitations it appears that optical dating will by a quite suitable method to date Pleistocene sediments from such a complex fluvial environment, especially when focusing on a precision scale beyond a few thousand years.     

Optical dating of Holocene sand dune activities in the Horqin sand-fields in inner Mongolia, China, using the SAR protocol

The Horqin sand-field in northeastern Inner Mongolia, China, had been the fertile grassland in North China, but desertification and sand-dust storm have increasingly occurred in the past decades [Zhu and Wang, 1992. Theory and practice of sandy desertification in China (in Chinese with English abstract). Quaternary Sciences 2, 97]. To understand the Holocene sand dune activities in this region, five sand dune profiles were investigated, and 32 coarse grain quartz samples were dated by OSL using the single-aliquot regenerative-dose (SAR) protocol [Murray and Wintle, 2000. Luminescence dating of quartz using an improved single-aliquot regenerative-dose protocol. Radiation Measurements 32, 57–73]. For cross-checking, six organic-rich samples from the paleosols and sandy peat/mud were dated by both ¹⁴C and quartz OSL. With one exception, ¹⁴C and quartz OSL dating results show good agreements. Based on the consistent results of five sand dune profiles, a chronology of Holocene dune activity in Horqin sand-field is established as follows: (i) active sand dunes built up widely before 10 ka; (ii) dunes semi-stabilized between 10 and 7.5 ka ago; (iii) the dunes solidify and chernozem soils developed between 7.5 and 2.0 ka ago; and (iv) partially re-mobilization of dunes occurred since about 2.0 ka ago.     

Temporal–compositional trends over short and long time-scales in basalts of the Big Pine Volcanic Field, California

Primitive basaltic single eruptions in the Big Pine Volcanic Field (BPVF) of Owens Valley, California show systematic temporal–compositional variation that cannot be described by simple models of fractional crystallization, partial melting of a single source, or crustal contamination. We targeted five monogenetic eruption sequences in the BPVF for detailed chemical and isotopic measurements and ⁴⁰Ar/³⁹Ar dating, focusing primarily on the Papoose Canyon sequence. The vent of the primitive (Mg# = 69) Papoose Canyon sequence (760.8 ± 22.8 ka) produced magmas with systematically decreasing (up to a factor of two) incompatible element concentrations, at roughly constant MgO (9.8 ± 0.3 (1σ) wt.%) and Na₂O. SiO₂ and compatible elements (Cr and Ni) show systematic increases, while ⁸⁷Sr/⁸⁶Sr systematically decreases (0.7063–0.7055) and ε_Nd increases (− 3.4 to − 1.1). ¹⁸⁷Os/¹⁸⁸Os is highly radiogenic (0.20–0.31), but variations among four samples do not correlate with other chemical or isotopic indices, are not systematic with respect to eruption order, and thus the Os system appears to be decoupled from the dominant trends. The single eruption trends likely result from coupled melting and mixing of two isotopically distinct sources, either through melt-rock interaction or melting of a lithologically heterogeneous source. The other four sequences, Jalopy Cone (469.4 ± 9.2 ka), Quarry Cone (90.5 ±17.6 ka), Volcanic Bomb Cone (61.6 ± 23.4 ka), and Goodale Bee Cone (31.8 ± 12.1 ka) show similar systematic temporal decreases in incompatible elements. Monogenetic volcanic fields are often used to decipher tectonic changes on the order of 10⁵–10⁶ yr through long-term changes in lava chemistry. However, the systematic variation found in Papoose Canyon (10⁰–10² yr) nearly spans that of the entire volcanic field, and straddles cutoffs for models of changing tectonic regime over much longer time-scales. Moreover, ten new ⁴⁰Ar/³⁹Ar ages combined with chemistry from all BPVF single eruption sequences show the long-term trend of BPVF evolution comprises the overlapping, temporal–compositional trends of the monogenetic vents. This suggests that the single eruption sequences contain the bulk of the systematic chemical variation, whereas their aggregate compositions define the long-term trend of volcanic field evolution.     

Study of the effect of a thermal treatment on the DE determination in ESR dating of speleothems

The choice of the ESR signal in the dating of speleothems is still debated. Among the peaks observed in speleothem ESR spectra, the most frequently signals used for the determination of the palaeodoses are the peaks at g=2.0007 (h3 line), g=2.0040 (Broad Line) and g=2.0057 (h1 line).

In order to determine the better analytical way to obtain reproducible D_E values from these three ESR signals, successive additional annealing treatments ranging from 60–300 °C were performed on several samples of a stalagmitic floor from the Mousterian site of the Caverna delle Fate (Liguria, Italy), also dated by U-series between 90 and130 ka. The results confirm that D_E values can be obtained from h3 and BL signals without annealing, while the h1 signal requires a thermal pre-treatment, to obtain the maximum intensity of this line. Then, on our samples, the D_Evalues derived for the three lines are equivalent. It gives, hence, the opportunity to obtain data when the g=2.0007 line cannot be used.     

Statistical treatment of fluvial dose distributions from southern Colorado arroyo deposits

Many of the small-scale arroyo systems found across southern Colorado contain well-preserved sedimentary records of prehistoric fluvial erosion and aggradation epicycles. In the following paper, we date a set of 50 ephemeral fluvial samples from four southern Colorado arroyo systems using a combination of single-grain and single-aliquot OSL techniques. Analysis of the sample D_e distribution characteristics reveals that these arroyo sediments were subjected to a diverse array of bleaching conditions prior to deposition. The use of appropriate burial dose estimation procedures is therefore deemed vital to ensuring that accurate age estimates are produced for each of these samples. In this study we apply the formal ‘age model’ decision procedures of Bailey and Arnold [Statistical modelling of single-grain quartz D_e distributions and an assessment of procedures for estimating burial dose. Quaternary Science Reviews, in press.] and Arnold [2006. Optical dating and computer modelling of arroyo epicycles in the American Southwest. D.phil. Thesis, University of Oxford, unpublished] to our fluvial sample dataset in order to enable a more objective selection of appropriate burial dose estimates. The resultant OSL chronostratigraphies are examined and discussed. These formal decision procedures yield sample ages that are stratigraphically consistent for 94% of the 50 fluvial samples examined. The resulting OSL ages also display a greater degree of stratigraphic consistency in comparison to those ages that would have been generated by simply applying a single type of age model to all samples.     

Aeolian landform processes since the last deglaciation revealed by OSL chronology and stratigraphy in the Hulunbuir dune field in NE China

The Hulunbuir dune field (HLB) is situated near the northern limit of the East Asian summer monsoon (EASM), and vulnerable to climate change. The aeolian sand–paleosol sequences of this region are crucial for understanding the past landform processes in response to climate change, but not yet understood well due to chronological controversies. Here, we presented 20 optically stimulated luminescence (OSL) ages from five aeolian sand–paleosol profiles in the HLB, and reconstructed the aeolian landform processes since 18 ka. The findings of this study suggested that the HLB was probably dominated by mobile dunes before 18 ka, as 10 out of 11 aeolian samples were dated to 18–12 ka. Two strong sandy paleosol layers were found and dated to ∼9 ka and 5–0.5 ka, indicating that strong in situ pedogenic process on the accumulative sand could occur during the Holocene. The OSL ages of samples near the top of three profiles were >9.5 ka, indicating two possible surface processes. First, the land surface was stable since 9.5 ka after stabilization, with no accumulation or erosion. Alternatively, the surface could have been erosive with the eroded sediments feeding downwind active dunes. The latter explanation is consistent with the current local landforms, which has widespread blowout pits, indicators of strong wind erosion. We emphasized that the OSL age of a sand layer sample in fossil dunes implied the onset of mobile dune stabilization, not the age of dune activity, as previously stated.     

Possible resetting of quartz OSL signals during earthquakes—Evidence from late Pleistocene injection dikes, Dead Sea basin, Israel

Clastic dikes are formed either by passive deposition of clastic material into pre-existing fissures (depositional dikes), or by fracturing and injection of clastics during earthquakes (injection dikes). We proposed to use optically stimulated luminescence (OSL) dating to distinguish between the two modes of formation and hypothesized that (1) depositional dikes filled from above show OSL ages younger than the host rock; and (2) injection dikes filled from below show the same OSL ages as that of the host rock. We studied the mechanisms of clastic-dike formation and their ages within the seismically active Dead Sea basin, where hundreds of dikes crosscut the late Pleistocene (70–15 ka) lacustrine sediments of the Lisan Formation. Field observations and analysis of magnetic tensors show unequivocally that most of these dikes were emplaced by injection, inferred to be due to seismically triggered fluidization–liquefaction during earthquakes. Twenty-eight samples were collected from the Lisan source material and dikes that, based on field observations, are unmistakably either depositional dikes or injection dikes.

Quartz single aliquot OSL ages of the source Lisan layers are between 43 and 34 ka, and are typical for the Lisan Formation. The ages of both depositional and injection dikes are between 15 and 17 ka, younger than the Lisan host rock. Depositional dikes show a highly scattered distribution of single grain ages, suggesting several episodes of infill. Single grain ages of injection dikes are of latest Pleistocene to Holocene, and do not contain recently bleached grains that infiltrated from above. These results imply that the OSL signals were reset at the time of fluidization–liquefaction and buildup of fluid pressure within the injection dikes. If this resetting mechanism has a physical ground, then OSL dating is an important tool for constraining the ages of earthquake-induced injection dikes and recovering paleoseismic data from them.     

Characteristics of the burial environment related to quartz SAR-OSL dating at St. Vincent Island, NW Florida, USA

St. Vincent Island, on the northwest Gulf coast of Florida, USA, preserves a well-developed beach ridge plain that began to form 4000 years ago based on inferred ages of pottery artefact assemblages. Seven vibra cores up to 2.88 m in length (uncompacted sediment) were retrieved from six ridges across the island. The single-aliquot regenerative-dose (SAR) procedure was used to obtain two optically stimulated luminescence (OSL) ages per core at different depths within each core.

The geographically youngest ridges gave ages of 370±50–1900±300 years, yielding an inter-ridge accumulation time of 150 years assuming uniform sediment accumulation. The oldest ridge sets yielded ages of 2700±400–2800±300 years. Dating results are evaluated in terms of equivalent dose (D_E) distributions and other characteristics in relation to aspects of the burial environment, including pedogenesis.     

Optically stimulated luminescence dating of aeolian sand in the Otindag dune field and Holocene climate change

The dune system in Otindag sand field of northern China is sensitive to climate change, where effective moisture and related vegetation cover play a controlling role for dune activity and stability. Therefore, aeolian deposits may be an archive of past environmental changes, possibly at the millennial scale, but previous studies on this topic have rarely been reported. In this study, thirty-five optically stimulated luminescence (OSL) ages of ten representative sand-paleosol profiles in Otindag sand field are ob-tained, and these ages provide a relatively complete and well-dated chronology for wet and dry varia-tions in Holocene. The results indicate that widespread dune mobilization occurred from 9.9 to 8.2 ka, suggesting a dry early Holocene climate. The dunes were mainly stabilized between 8.0 and 2.7 ka, implying a relatively wet climate, although there were short-term penetrations of dune activity during this wet period. After ~2.3 ka, the region became dry again, as inferred from widespread dune activity. The "8.2 ka" cold event and the Little Ice Age climatic deterioration are detected on the basis of the dune records and OSL ages. During the Medieval Warm Period and the Sui-Tang Warm Period (570-770 AD), climate in Otindag sand field was relatively humid and the vegetation was denser, and the sand dunes were stabilized again. These aeolian records may indicate climate changes at millennial time scale during Holocene, and these climatic changes may be the teleconnection to the climate changes elsewhere in the world.     

Devensian glacial events in Buchan (NE Scotland): A progress report on new OSL dates and their implications

Optically stimulated luminescence (OSL) dating of a series of glaciofluvial/glaciodeltaic sediments in central Buchan and Aberdeen has recently been undertaken. The aims of this project are to test the chronological model proposed by the most recent regional review and the suggestion that parts of Buchan may have remained ice-free during the last glacial maximum (late marine isotope stage (MIS) 3 and MIS 2 29–15 ka). The preliminary results indicate that during the Devensian (ca. 116–12 ka), extensive areas of Buchan may have been glaciated earlier than previously believed (possibly during MIS 4, 72–60 ka), but parts of the region show no depositional evidence of later glaciation. Some waterlain sediments from the Ugie Valley have yielded OSL ages indicating deposition during MIS 5d to MIS 5a (116–72 ka). We discuss whether the absence of overlying glaciogenic sediments at these locations raises questions about the reliability of the OSL ages or about existing models of the Devensian glaciation of Buchan.