Calculating or simulating the dose rate? A comparison

Accurate determination of the environmental radiation dose-rate is crucial for dosimetric dating methods. Two approaches for dose-rate estimation are currently employed: (1) calculating and summing up individual doserate components, including propagation of uncertainties and (2) simulating the radiation field and calculating the absorbed dose using a Monte Carlo approach. Here we compare dose-rate estimation using the two aforementioned approaches, applied to six sedimentary quartz samples with differing radionuclide concentrations, chemical composition, grain-size distribution and water content and to one heterogeneous archaeological sample. For approach (1) two web-based, freely accessible calculators, “Dose-Rate calculator” (DRc), “Dose Rate and Age Calculator” (DRAC) and a bespoke spreadsheet were used. For approach (2) DosiVox, a Geant4 Monte Carlo simulation toolkit was employed which allows the definition of material properties such as chemical composition, density and porosity in addition to radionuclide concentrations and water content in a three dimensional geometry. The sensitivity of dose rate to material properties, usually assumed to be constant over the burial time scale, was also tested. The results obtained using the three dose-rate calculators for sedimentary quartz samples are consistent within uncertainties, mainly because the conversion and correction factors are all taken from the same references. Comparing these results with those obtained from DosiVox simulations show grain-size dependent doserate differences, caused by the charge build-up and attenuation of the irradiated grain-size fraction. Moreover, it reveals that the effect of porosity and moisture on dose rate is grain-size dependent. For the heterogeneous archaeological sample the total dose rate determined by a calculator is inaccurate for the alpha- and too precise for the beta dose-rate. We conclude that the standard approach to dose-rate estimation is robust and that DosiVox is an important diagnostic tool for samples originating from complex environments.