Student : AMESTOY Julien
Advisor : MESLIN Pierre-Yves (IRAP), BARATOUX David (GET/IRD)
Start : 2017
Group : PEPS
Airborne γ-ray spectrometry is an efficient technique to carry out the radiological assessment of a site of large spatial extent. Its application to environmental and natural radioactivity monitoring requires temporal repeatability of radionuclide concentration maps in the soil. The innovative approach of this study is based on the coupled acquisition of γ spectra from a simulated airborne hover at 50 meters and measurements of environmental factors over a continuous 14 month time-series, between April 2019 and June 2020, which allows the characterization of environmental factors affecting γ-ray spectrometry. The study site (P2OA-CRA at Lannemezan) was selected for the presence of a 60 meter-tall meteorological mast capable of hosting a NaI(Tl) spectrometer, RSX-5, of the same kind as the one used during CEA γ-ray surveys. The main objective of the thesis is characterize and correct the different measurement biases to isolate the γ signal emitted from the ground. To this purpose, temporal chronicles were used to observe and quantify the environmental effects associated with the variation of soil moisture, atmospheric pressure, cosmic rays, precipitation, and atmospheric radon-222. Each parameter was the subject of a dedicated study, which led to the implementation of a correction on the measured γ signal. Theoretical fit curves of gamma-ray attenuation, coupled with Geant-4 simulations, validated the removal of soil moisture effects. These models are used to express the radionuclides’ concentration on a per dry soil mass basis, which is therefore no longer a function of soil moisture at the time of the γ-ray survey, an important source of variability. This corrective procedure was first validated for 40K, 232Th, and 137Cs, and then for 238U after the influence of atmospheric 222Rn was removed from the signal. This correction is also applicable using soil moisture acquired by remote sensing, in particular using SMAP L4 products. A complementary study on the monitoring of soil moisture profile by γ radiation of the same radionuclide at different energies has opened up prospects for applications in precision agriculture. The characterization of an energy window centered on cosmic rays has made it possible to eliminate their influence on the measured g-ray signal. If the effect of the semi-diurnal cycle of atmospheric pressure on the g-ray signal was evidenced for all radionuclides, its limited influence was not subject to a dedicated correction. The 238U-centric precipitation effect was studied on single rain event profiles of variable intensities. The increase in the 238U count rate associated with these events shows a linear relationship with the intensity of rainy episodes. The coupled 222Rn and 238U time-series were used to characterize and quantify the influence of atmospheric radon on the uranium γ signal. A spectral component of atmospheric radon was extracted to that purpose. Following this new quantification of the influence of atmospheric 222Rn on the signal of 238U, a novel method for estimating the atmospheric 222Rn volume activity from airborne γ spectrometry was developed. The 222Rn estimated this way can both be used in the correction protocol for the 238U signal and be the subject if dedicated studies. Measurements of its temporal evolution from γ spectrometry thus can thus be obtained. A global correction procedure was then implemented, PASTHEL, and applied to the spectra acquired during the airborne γ-ray spectrometry campaign carried out on the study site. Maps produced with this protocol show a very good diurnal and semi-diurnal repeatability, which is not the case of the uncorrected maps. This validates the protocol developed in this study.