@phdthesis{01H4FYKPNBG95NEDNDX9YD54PQ,
  abstract     = {{Optically stimulated luminescence (OSL) dating of sedimentary quartz and feldspar is now widely used in a range of Quaternary research topics. It is advantageous as it uses the main mineral constituents of the sediments (e.g. quartz and feldspar), covers a wide age range (typically 100 - 105 years in case of quartz), and provides numerical age information (in calendar years), without the need for calibration through independent age information. The event being dated is the last exposure of the minerals to light (e.g. during erosion and transport) or heat (e.g. during anthropogenic firing or natural heating events). At present, quartz-based OSL dating is mainly used for establishing deposition chronologies for unheated sediments. Nevertheless, the origins of luminescence dating lie in the application in archaeological context, and the dating of heated sedimentary materials (such as ceramics) in particular. 

This dissertation investigates the potential of quartz-based OSL signals for dating anthropogenically heated sediments associated with two particular features – relic charcoal kilns and cooking hearths – over the particular timespan of the last few centuries. Both types of features are receiving increasing interest by the scientific community but are notoriously difficult to date with sufficient accuracy and/or precision. Radiocarbon (14C) dating is commonly applied, for instance, but essentially provides no meaningful time-resolution from around 1650 CE onwards. Archaeomagnetic dating has great potential but requires well-preserved, unweathered in-situ/oriented fragments, relatively high firing temperatures, and an appropriate standard calibration curve of the archaeomagnetic field; very often these conditions are lacking or contentious, at least for the types of features investigated here. 

The first part of this dissertation focusses on the optical dating of relic charcoal kilns as preserved in two different lithological substrates, a loamy one (Sonian Forest, Belgium) and a sandy one (Zoersel Forest, Belgium and the Veluwe, The Netherlands). The majority of the relic charcoal kilns found in Europe date to between the 17th and the 19th century CE, but cannot be constrained in time more precisely using 14C dating. The emphasis of the research was on the testing of the accuracy and precision that can be attained using OSL dating. 
For each of the investigated features and samples, the quartz OSL characteristics are documented in terms of the various procedural tests that are commonly used to assess the appropriateness of the single-aliquot regenerative-dose (SAR) protocol. The results suggest that, in all cases, the measurement procedure is reproducible and is able to accurately measure known laboratory doses. The SAR protocol was therefore used to determine equivalent doses in both large aliquots (composed of thousands of grains) and small aliquots (composed of 100-200 grains); for a few samples, single grains were analysed as well. The dose rate was derived from radionuclide analysis using high-resolution gamma-ray spectrometry, and our best possible estimates of the various parameters involved in the calculations (such as time-averaged moisture content, burial depth and others). As such, a more or less widely adopted mainstream OSL-dating methodology was deployed. The accuracy of the resulting OSL-dates is evaluated through comparison with independent age control (14C and archaeomagnetic dating and historical sources); precision is evaluated in comparison to that of 14C-dating. 
The study at Sonian forest (Central Belgium; loamy substrate) used a set of 14 samples collected from seven relic charcoal kilns and underlying parent material. The kilns were independently dated using 14C of carefully selected charcoal fragments; the results show that five of the investigated kilns pre-date 1650 CE, while the other two are more recent. The OSL ages are entirely consistent with the 14C-ages, indicating that the approach is accurate. The precision of OSL dating is comparable to that of 14C-dating for pre-1650 CE features, and can be significantly better for post-1650 CE ones. Moreover, in case of complete resetting, it is suggested that OSL-dating might allow distinguishing between comparable features with a relative time-resolution of 2-4%. Finally, it is illustrated that the linear part of the dose response curve can be exploited to obtain large datasets, without compromising accuracy or precision of the results. 
The exercise was extended to relic charcoal kilns preserved in sandy substrates. In the Veluwe (Central Netherlands), seven samples from five relic charcoal kilns and 11 complementary samples from the underlying sandy substrate were investigated. The research adopted the same approach as for the kilns in Sonian Forest. For four of the five investigated kilns, OSL ages are consistent with independent age information (14C-dating and written sources). Also here, it is observed that the precision can be significantly higher than that of 14C-dating for post-1650 CE features, with potential for establishing relative age relationships on decadal to even multi-annual timescales. For one of the kilns, both 14C and OSL dating yield ages that are consistent with (poorly constrained) historical age information, but with the 14C date being significantly older. The cause for this discrepancy remains to be understood; at least from a methodological perspective (including examinations of the distributions of equivalent dose), there is at present no evidence for the OSL-date for this particular feature to be less accurate than those obtained for the other kilns. The same exercise was performed in Zoersel Forest (Northern Belgium), where a suite of 32 samples from 11 relic charcoal kilns were investigated. In contrast to the two aforementioned studies, the OSL dates overestimate the 14C-dates for most of the features. For some of the samples, this could be anticipated based on the spread observed in the distribution of equivalent dose in small aliquots of quartz. For four of the investigated features (8 samples), OSL ages were also obtained using single grains of quartz, which are either consistent or overestimate the 14C-dates to the same extent as the ages obtained using small aliquots. A potential and partial explanation for this discrepancy is that heating is a homogenous process, but may not have been sufficiently stringent to fully reset the quartz-OSL signal. The character of charcoal production at Zoersel Forest differs from that in other regions in Europe, however, implying that other factors (related to e.g. dose rate variations through time and/or independent age determination) might be partially responsible as well. Apart from this, it is hypothesized that distributions of signal brightness might give information about heating regimes experienced by the grains; this remains to be explored. 

The second part of this dissertation examines the same optical dating methodology for application to relic (cooking) hearths associated with early-Modern (late 15th – late 18th CE) and Modern (late 18th – present) military encampments in Ninove (Central-East Belgium). A vast amount of such traces were discovered at this site and are allocated to 1692 CE and/or 1693 CE, and 1745 CE, essentially on the basis of more or less circumstantial evidence. As for the relic charcoal hearths, 14C-dating was not expected to help distinguishing between features from the end of the 17th and those from the mid-18th century CE. Optical ages were obtained for six samples from three relic hearths – the ages do not differ significantly, indicating that the sampled features are coeval. The average optical age of 1748 ± 59 CE (~ 95 % probability) practically coincides with insights subsequently gained from archaeological research (1745 CE). The results illustrate how OSL dating cannot only provide information for archaeomagnetic dating, but reinforces the potential for distinguishing – relatively - between comparable features from this period of time at a decadal time scale with 95% confidence. The time-resolution that can be achieved is thus unrivalled compared to that of other radiometric techniques. 

The overall conclusion of this study is that quartz-based SAR-OSL dating is a powerful tool for dating relic (charcoal) hearths, regardless of whether they pre- or postdate 1650 CE. Given that 14C-dating is well embedded in archeological and young-Quaternary research, it is unlikely that OSL-dating will be readily considered for dating pre-1650 CE features. This is unfortunate given that the method is not material specific, directly targets the event of interest (no association), does not rely on independent age information for calibration purposes, and offers a potential relative time-resolution of 2-4%. For post-1650 CE features, OSL dating offers the most obvious and significant advantages over other radiometric techniques such as 14C-dating. Some issues remain, however. OSL analyses of multiple samples from a single feature, as part of a multiple-method-dating strategy and in close collaboration with all researchers involved, is recommended.}},
  author       = {{Karimi Moayed, Nasrin}},
  language     = {{eng}},
  pages        = {{XXII, 173}},
  publisher    = {{Ghent University. Faculty of Sciences}},
  school       = {{Ghent University}},
  title        = {{Testing the potential of OSL dating for application to heated anthropogenic sediments (hearth and charcoal kiln remains)}},
  year         = {{2023}},
}