The Pleistocene overburden deposits in the St. Lawrence Lowlands were dated by Th/U to test the Rosholt U-trend method which, in principle, is applicable to the age determinations in open systems. The A horizon of a Holocene soil developed on the Gentilly Till (Upper Wisconsinian) is characterized by a minor positive flow of uranium (230Th/234U > 1). The Gentilly Till as such is marked by an uranium loss, the maximum loss being between 1.5 and 3 m below the soil surface (1.2< 230Th/234U < 1.8). Also, the Bécancour Till (Lower Wisconsinian or Illinoian) lower in the profile differs from the former by a minor secondary uranium fixation (230Th/234U ≤ 1). The U-trend model does not appear to be applicable for the dating of these units, either because of the large uncertainty on the slope of the isochron lines (Gentilly Till), or because the age obtained (~260 ka for the Bécancour Till) does not agree with the regional chronostratigraphy. This last anomaly is due probably to a reversal of the uranium flow in the till with time. Assuming an opening of the Th–U system at the onset of the postglacial emergence in the area, at about 10 ka, rates of ablation of uranium of 10−5 to 10−6 a−1 and uranium flow of ~30 pg∙g−1∙a−1 can be calculated for the leached horizons (soil and Gentilly Till). The uranium loss affects the calculation of ages by thermoluminescence; it leads to an underestimation by 10% of the total radioactivity of the 238U family when this radioactivity is evaluated on the basis of the uranium concentrations. Though this study does not confirm the validity of the Rosholt U-trend model, it does support the conclusion that an analysis of the disequilibrium of Th/U in overburden deposits is of intrinsic interest in matters concerning the calculation of ambient radiation to measure ages by indirect physical methods. [Journal Translation]

You do not currently have access to this article.