Quaternary TL Surveys

Guide to TL Date Measurement


The purpose of environmental dosimetry in TL dating is to measure the gamma and cosmic ray contribution to the total radiation dose received by the TL sample during its burial at the archaeological or geological site. In the case of flint and calcite, this contribution is commonly a major part of the total dose. Moreover, it is often the uncertainties in determining this contribution that govern the overall error limits of the TL date. The on-site environmental dosimetry is therefore an important part of the dating procedure.

This chapter set out the considerations involved in gamma ray dosimetry and some of the strategies employed in the measurements. In the first section, the two components of the total radiation dose absorbed by the TL sample are differentiated. The methods of on-site measurement of the environmental dose are then described. The final sections set out the difficulties that are encountered when interpreting the data collected on-site. These are associated with variabilities of the environmental dose rate, either of a spatial or of a temporal nature.

5.1. Internal and Environmental Doses

Before TL measurements are carried out upon a flint flake or stalagmite, the outer 3 mm of the sample are cut away. The purpose of this is two-fold. Firstly, the surface which has been exposed to light is removed, as a strong exposure to light can affect the TL. Secondly, since alpha particles have a range of 0.03 mm and betas a range of 3 mm, the volume which has been exposed to alpha and beta rays originating from outside the sample is also removed. The remaining portion has therefore received its alpha and beta dose entirely from within the volume of the flint or calcite. This dose, termed the internal dose, can be determined from an examination of the sample alone.

In contrast to alphas and betas, the range of gamma rays is approximately 0.3 m, and therefore much greater than the dimensions of most TL samples. This means that the contribution to the total radiation dose which comes from gammas is dependent mainly upon the radioactive content of the sediment surrounding the TL sample, and only slightly upon the radioactivity of the sample. Like the alpha and beta rays, gamma radiation derives from the decay of naturally occurring radionuclides present in the ground, such as potassium-40 and members of the uranium and thorium decay series. Together with the contribution from cosmic rays, which is a minor one, the gamma component of the total radiation dose is referred to as the environmental dose.

The purpose of environmental dosimetry is to estimate the mean rate at which the TL sample has received its environmental dose during burial. Ultimately this can only be an estimate, because measurements record only the present day radiation environment. One concern is therefore the temporal variability of the burial conditions which, as discussed below, is often associated with changes of ground water content. The other main concern is spatial variability, since it is rare to find sites that are homogeneous in their radioactivity. It is these spatial variations that are the principal subject for investigation in the field.

5.2. Methods of On-Site Dose Rate Measurement

The instrument most commonly used for environmental dosimetry is the portable gamma spectrometer. This usually consists of a crystal scintillation detector with electronics for sorting gamma rays according to the quantity of energy they deposit in the crystal. The detector is housed in a cylinder of 55 mm diameter, and is inserted to a depth of 0.3 m or more into the sediment under study using an auger. Counting times are generally between 15 mins and an hour, depending on the radioactivity of the environment. The resulting spectrum of gamma energies is analysed to determine the concentrations of uranium, thorium and potassium in the sediment. Figure 1 shows spectral data measured with a gamma spectrometer, and indicates the regions of the spectrum which are of particular value in determining radionuclide concentrations.

[ Figure 1 ]

Figure 1. Gamma ray spectrum measured with a portable spectrometer. Gamma ray energies are characteristic of the radionuclides which emit the radiation, and peaks in the energy spectrum therefore indicate the presence of particular radiation sources. Shaded areas show regions of the spectrum which are important for calculating the concentrations of potassium (K), uranium (U) and thorium (Th) radionuclides in the burial environment.

An alternative method is to use small copper capsules containing a sensitive TL phosphor such as calcium fluoride. These capsules are left buried in the sediment for a period of several months, and then returned to the laboratory where the TL acquired by the phosphor during its burial is measured. The disadvantage of this technique is the lack of an immediate feed-back of information which is available when using a gamma spectrometer. This feed-back is essential on sites of inhomogeneous gamma activity as it shows where the spatial variations are occurring, and allows the operator to plan a suitable strategy for measuring them.

In conjunction with gamma dosimetry, samples of the burial sediments are routinely collected and sealed inside two plastic bags. The water contents of these samples are measured on return to the laboratory. The dried samples may also be analysed to provide a cross-check to the on-site dosimetry However, laboratory analyses are only informative if the samples are from homogeneous deposits and therefore representative of the larger sediment body.

5.3. Spatial Variability of the Environmental Dose Rate

Only in rare circumstances can an environmental dose rate measurement be taken at the exact position of a TL sample. This is because the uncovering of the flint or stalagmitic floor involves the removal of at least part of the sediment body which was responsible for the environmental dose received by the TL sample. However, there are two cases in which this problem is overcome. The first arises if a burnt flint is fortuitously unearthed in the bottom of the auger hole cut to insert the gamma spectrometer probe. Such good fortune is only likely in dense artefact scatters. The second case sometimes occurs in cave sites when a large stalagmitic boss needs to be dated. If a corer can be used to penetrate 0.3 m into the stalagmite and extract the TL sample, a direct measurement of the gamma dose rate received by that sample can then be obtained.

Except in the two circumstances given above, the normal method of dosimetry is to extrapolate measurements taken at nearby positions where the sediments are intact. In this context, "intact" means that a body of undisturbed sediment remains which is at least 0.6 m across. If the volume is smaller than this, part of the gamma radiation recorded at its centre will originate from outside the undisturbed sediment. In practice, gamma measurements are often made by augering into a standing section of the stratigraphy.

In order to extrapolate the gamma measurements to the position of the TL sample, it is necessary to have an understanding of the spatial dose rate variations in neighbouring parts of the site. These variations can be divided into those which occur down the stratigraphy and those which occur horizontally within stratigraphic units.

One situation where marked vertical changes in gamma dose rates have been found is at the boundary of an ancient land surface and peat which has formed on top of it. The largely organic remains which compose the peat contain very much less radioactivity than the underlying deposits. This difference is accentuated by the higher water content of the peat compared to the sediment. The rate of change of the gamma dose with height is greatest at the old land surface, which is usually where datable flints are found. In these circumstances the vertical positions of flints need to be accurately recorded if dating precision is not to be lost.

Having noted the problems associated with peat covered sites, it should be pointed out that peat also has beneficial effects. For TL dating, desirable conditions include the rapid burial of the dated sample under 0.3 m of deposit which is subsequently undisturbed by erosion. The water content of the deposit should also remain stable. Peat forms an excellent protection against erosion, and helps stabilise the water content of the underlying sediments. Because its radioactivity is low, it preserves rather than alters the original environmental dose rate of the unburied artefacts, and this means that it is less critical to know how long the TL samples lay uncovered.

Another example where a stratigraphic boundary is accompanied by a change in radioactivity is found in limestone caves. Limestone and calcite often contain very much lower concentrations of radionuclides than deposits which enter the cave from outside. Gamma dose rates may therefore vary significantly in the vicinity of cave walls, fallen blocks, and stalagmitic formations that are more than a few centimetres across. Problems arise when a stalagmitic floor seals, or is covered by, a deposit containing limestone fragments, as the environmental dose rate will vary according to the proximity and sizes of the nearby fragments.

Even where a stalagmitic floor is situated between two homogeneous sediments, the attenuation of gamma rays penetrating the floor from outside needs to be taken into account. Since the size of this dose rate reduction increases with the thickness of the floor, thin stalagmitic floors (less than 50 mm across) are preferred as TL samples.

Spatial variations that occur horizontally within stratigraphic units are generally less marked than vertical dependences. Where it is necessary to extrapolate gamma measurements to the find location of a TL sample it is often assumed that environmental dose rates are uniform within a given unit. This assumption needs to be justified by observations taken over an area of the site which is greater than the extrapolation distance. The interval between measurement locations should take account of the nature of the deposits. Where the stratigraphy is locally variable, measurement intervals should be small enough to observe any correlated dose rate variations.

If the TL sample derives from a thin layer which is radioactively different from adjacent units, then dose rate variations can arise merely as a consequence of changing layer thickness. Because of the long range of gamma rays, such variations will also occur (to a lesser extent) in adjacent units. Waterlogged parts of the site will generally give lower dose rates than drier areas. These and other considerations mean that uniformity within horizontal strata cannot be assumed unless it is supported by evidence.

5.4. Time Variations in the Environmental Dose Rate

Undoubtedly the most intractable problems associated with TL dating are those concerned with dose rate variations over the burial period of the TL sample. Changes in dose rate over time can arise either by (i) variations in the amount of energy released by radioactive decays within a given volume, or by (ii) changes in the amount of matter absorbing that energy. Variations in the quantity of energy radiated are brought about by movements of radionuclides within the burial environment, by equilibriation of decay series, or, on agricultural land, by the recent addition of potassium fertilizer. Changes in the mass of the sediment are most commonly caused by fluctuating water retention.

The long-lived radioactive elements uranium and thorium each head a series of radionuclides which are formed by the decay of the preceding member of the chain. The series is in equilibrium when the decay frequencies of its members are all equal. Disequilibrium can arise if there is partial or complete removal of some types of radionuclide from the chain. Uranium is prone to such removal by virtue of its long half-life. Radon, although it is short-lived, has mobility as a gas and can also be lost from the system. However, in most situations the losses from a given volume of sediment will be balanced by gains from outside.

The environmental dosimetry is not sensitive to the presence of disequilibria. Most of the gamma radiation energy is emitted by sections of the uranium and thorium chains which are not prone to breakage, and the dosimetry examines these sections directly. However, overall activity is controlled by long-lived members at the top of the chain whose concentrations respond only slowly to conditions of disequilibrium.

An example of this is seen in stalagmitic calcite. The ground waters from which stalagmites are formed carry uranium but are virtually free of thorium. Over a period of some 350,000 years, the thorium activity recovers to equilibrium through decay of its parent uranium. The gamma activity of nearly all the uranium decay chain is tied to this recovery. Allowance must be made for this long-term trend in gamma dose rate when computing the dose accumulated by a stalagmitic TL sample which has been removed from a large formation.

Water in the burial environment can be thought of as additional shielding between the gamma emitters and the TL sample. The dose rate received by the sample decreases as the water content increases, and vice versa. Water contents vary on both seasonal and geological time scales. Seasonal fluctuations can mean that gamma measurements taken in the winter may differ from summer readings. Also, where excavation has been carried out much in advance of the dosimetry, readings may be affected by drying out of the sediments. For such reasons, the measured dose rates may not accurately reflect average present day conditions and allowance should be made for this.

Variations of sediment water content over geological time scales occur as the result of changes in climate and drainage. Clues about past conditions can be gleaned from a wide range of sources (e.g. pollen analyses, faunal and sedimentological studies). However, given the impossibility of producing reliable assessments of past water contents, the only safe response is to attach wide uncertainties to the estimates. Quite often it is this uncertainty which dominates the error limits of the final date.

Queries and comments on TL dating procedures and the above literature are welcome.

E-mail n.debenham@qtls.globalnet.co.uk

Return to Quaternary TL Surveys Home Page