Using OTOR(X) fit functions to improve estimation of high natural luminescence doses

While highly desired, it remains a challenge for luminescence dating to determine high doses, hence high ages (e.g., >300 ka). The challenge is to project a natural dose close to saturation to a dose-response curve generated with high laboratory doses. The single saturating exponential (SSE) function mostly delivers poor fits to this type of dose responses. Other functions, e.g., the single saturating exponential plus linear function, are then often employed, but these include constants that have no direct physical meaning. Such an approach is inconsistent with the OSL/IRSL measurement parameters (e.g. detection wavelength) by which the signal from a dosimeter’s specific trap-hole pair is targeted out of a broad light spectrum. It is therefore beneficial to employ a physically based model that allows to interpret observations obtained from high laboratory dose responses.

Here we employ the analytical expression, Lambert W, developed by Pagonis et al. (2020) which is an exact solution of the well-studied OTOR (one trap one recombination centre) model, and extended by Lawless and Timar-Gabor (2024) to the OTORX model. We compare results obtained from SSE fits, in particular the characteristic saturation dose (“D₀”) parameter, with those obtained from the OTOR(X) functions. Well-bleached fine-grained polymineral samples irradiated up to ~5000 Gy were used and measured using the pIRIR₂₂₅ protocol.

For the SSE function the results point to the 80% rule of thumb: at ca 80% of the saturation dose the SSE-fitted dose response tend to underestimate the natural dose. The OTOR(X) functions reveal that this is due to the ratio of trapping rate versus recombination rate of free electrons which changes as the regenerated dose response approaches saturation. Consequently, the shape of the dose response curve flattens out in a way that the SSE function is unable to predict. We show here how the change of shape affects the dose interpolation point and how the accuracy of dose estimate is tested using the 63% (D₀) and 80% dose values. We conclude that the OTOR(X) functions provide accurate estimates of natural doses close to saturation.