Vagn Mejdahl

FLUCTUATIONS OF THE SVALBARD–BARENTS SEA ICE SHEET DURING THE LAST 150 000 YEARS

Abstract On Spitsbergen, western Svalbard, three major glacial advances have been identified during the Weichselian. All three reached the continental shelf west of the Svalbard archipelago. Radiocarbon, luminescence and amino acid dating of interbedded interstadial and interglacial sediments indicate that these glacial advances have Early (Isotope Stage 5d), Middle (Stage 4), and Late Weichselian ages (Stage 2). An additional, more local, advance has been dated to Isotope Stage 5b. The Late Weichselian ice sheet expanded across the entire Barents Sea. However, in the south-western Barents Sea, the Late Weichselian till is the only till above Eemian sediments, indicating that the Early- and Middle Weichselian ice advances were restricted to the Svalbard archipelago and the northern Barents Sea. A major problem with the onshore sites is the dating of events beyond the range of the radiocarbon method. To overcome this, the onshore record has been correlated with marine cores from the continental slope and the deep-sea west of Svalbard, where a chronology has been established by oxygen isotope stratigraphy. Ice rafted detritus (IRD) was used as the main monitor of glaciation. The IRD record closely mirrors the glaciation history as interpreted from the onshore sections. During the Late Weichselian, the largest IRD peak occurred during deglaciation, a pattern also postulated for the earlier events. Given this, the results from the marine cores indicate that the ages for the first glacial advances during the Weichselian were a few thousand years older than interpreted from the onshore stratigraphy.

New light on OSL

Abstract We have tested recently released green and blue light emitting diodes (LEDs) for the optical stimulation of quartz, with a view to using these in routine optically stimulated luminescence (OSL) dating and retrospective dosimetry. We compare results obtained with blue LEDs with those obtained using broad-band blue/green light filtered from a halogen lamp. We describe two practical blue LED OSL configurations and conclude that a small number of blue LEDs provide the same stimulation efficiency for quartz as can be obtained from filtered halogen lamps and other commonly used sources. From the direct comparison of OSL decay curves produced by the two light sources, and by examination of the dependence of blue LED OSL on preheat temperature, we conclude that there is no evidence that the blue LED light stimulates deep traps in a different manner from broad-band halogen light. Preliminary intercomparisons, using 34 quartz samples taken from both heated and unheated materials, between equivalent doses obtained using blue LEDs and using broad-band blue/green light from a halogen lamp show that, on average, the two data sets are indistinguishable (blue/halogen=0.98±0.02). We conclude that blue LEDs offer a practical alternative to existing stimulation sources. They have the significant advantage that the output can be controlled electronically; thus the power can be readily controlled by software.

Luminescence dating of archaeological materials using a new technique based on single aliquot measurements

Abstract A dating method based on single aliquot luminescence dose determination was applied to quartz and feldspars from heated archaeological materials. The problem of changes of sensitivity to radiation resulting from signal erasure in regeneration measurements was studied in some detail and a new method based on single aliquot regeneration measurements within an added-dose framework which appears to eliminate the problem of sensitivity change is proposed. Preliminary dating results, mainly using quartz, obtained with the new method are presented.

Procedures used for luminescence dating of sediments

Abstract An account of the procedures for thermoluminescence (TL) and optically stimulated luminescence (OSL) dating of sediments used at the Nordic Laboratory for Luminescence Dating is given. The topics discussed include sampling, sample preparation, separation of quartz and feldspars using heavy liquids, dose rate estimation and methods for estimating the equivalent dose. Special attention is paid to the partial bleach TL technique of Mejdahl (1988), which appears to give reliable results for partially bleached sediments. Optically stimulated luminescence dating using stimulation by infra-red (IR) and green light has been taken up and especially added-dose methods using stimulation by IR or green light have been tried with some success. Some sources of error are discussed, including the adverse effect of shallow traps.

Combined OSL (infrared) and TL studies of feldspars

An infrared (i.r.) diode array unit for optically stimulated luminescence (OSL) was developed which can be fitted directly onto the automated Riso TL dating apparatus, thus permitting combined TL and OSL measurements using the same PM tube. Thirty-two diodes with an i.r. emission of 880 Δ 80 nm are used in the array and the power delivered at the sample is about 40 mW cm-2. A specially designed feed-back system serves as control for the light output from the diodes. The i.r. unit and preliminary OSL studies of feldspar samples are described.

The plateau method for dating partially bleached sediments by thermoluminescence

Abstract The plateau method for dating sediments that have been incompletely bleached prior to deposition is described. By exposing aliquots of a sample to sunlight for different periods of time, a range of residual values is obtained and combined with the additive dose or the regeneration technique to yield a series of paleodoses. The correct paleodose is taken as that for which the longest plateau in a plot of paleodose vs. temperature is obtained.

Long-term stability of the TL signal in alkali feldspars

Abstract The long-term stability of the TL signal in alkali feldspars was investigated by studying infinitely old samples extracted from Tertiary sands. Mean lives in the range 800–5400 ka were obtained, indicating that long-term fading has to be taken into account. Fading corrections were calculated for a range of TL ages. The study indicated that the stability of the TL signal in alkali feldspars is less than that predicted from kinetic studies, and suggested a practical dating limit of 500 ka.

Determination of potassium in feldspars by beta counting using a GM multicounter system

Abstract A gas flow multicounter system developed at Riso National Laboratory for low-level beta counting applications was adopted for measuring potassium in feldspars extracted from archaeological and geological materials, in connection with TL dating work. A guard counter reduces the background by using an anticoincidence technique. Calibration was accomplished by measuring KCl and a feldspar standard. A potassium content of 1% yields a net beta count rate (background subtracted) of about 40 c.p.h.; the background is around 15 c.p.h. The precision in repeated measurements is better than 2% and the estimated accuracy is about 3%.

Thermoluminescence dating of late-glacial sand sediments

Abstract Ten Late-Glacial sediments were dated by the regeneration method combined with the plateau criteria. The thermoluminescence (TL) dates were compared with independent dates assessed indirectly by radiocarbon dating or varve chronology. For comparison, the 14 C dates were calibrated by adding 1000 yr to the conventional results. Eight TL dates agreed with the 14 C dates within the uncertainty of the TL dates (±1 ka), one was overestimated and one underestimated.

Determination of archaeological doses for TL dating using an automated TL apparatus

Abstract An automated reader was designed with special regard to TL dating. The reader is a microprocessor-controlled sample changer system that can construct the primary and secondaty response curves required for determination of the archaeological dose of quartz and feldspar extracted from pottery. Results are presented to demonstrate the performance.