Reply to Preusser et al. on Frechen et al. “Late Pleistocene fluvial dynamics in the Hochrhein Valley in the upper Rhine Graben: chronological frame”

The River Rhine provides one of the most detailed late Cenozoic terrestrial records of fluvial activity in Europe triggered by climatic and tectonic activities (Boenigk and Frechen 2006). Frechen et al. (2008, 2010) provided OSLbased chronologies for the late Pleistocene and Holocene fluvial dynamics in the Hochrhein Valley and in the Upper Rhine Graben (URG). The OSL age estimates are in excellent agreement with the stratigraphical concept of sedimentary dynamics in the southern URG (see Frechen et al. 2008, 2010; Lammermann-Barthel et al. 2009). All samples taken from below an event layer (discontinuity) yielded OSL age estimates older than 20 ka and those from above the event layer gave estimates younger than 20 ka. The results are all in stratigraphic order and for the Holocene fluvial sediments in agreement with three radiocarbon ages. The validity of this approach was checked against known age samples from Bremgarten. A comparison between radiocarbon ages of wood and four OSL ages suggested that the OSL ages have a residual (due to insufficient bleaching) of maximum 1–2 ka. This residual value would be insignificant for sediments older than 10 ka. This information made us to decide the application of large aliquots for the fluvial sediments from the Hochrhein and from the URG. The problem of OSL age overestimation in fluvial and glaciofluvial environments is well-known, and we are of course aware of the great potential to overcome the problems of insufficient bleaching by using small aliquots or single grains and by applying statistical approaches. These latter approaches can indeed help improve the numerical dating, as demonstrated for fluvial sediments in the northern URG and the Lower Rhine area by our own group (e.g. Lauer et al. 2010, in press). However, even after applying statistical methods for luminescence dating to glaciofluvial and fluvial sediments, the luminescence age estimates may still be problematic. Overestimates up to 50% and greater were reported by DeLong and Arnold (2007) of single-aliquot measurements when compared to single-grain measurements from samples under 1,000 years old. Negligible differences were found for late Pleistocene samples from various regions and environments (DeLong and Arnold 2007; Porat et al. 2008; Thomas et al. 2005). Furthermore, Mayya et al. (2006) suggested that even for well-bleached samples, a heterogenous distribution of K-feldspar grains in a sample can cause a wide dose scatter. In the comment of Preusser et al., (this issue) the experimental design of Frechen et al. (2010) is questioned owing to the application of large aliquots, which are not sufficient to detect insufficiently bleached samples and thus result in age overestimation, a statement, which is already given in the paper under discussion by Frechen et al. (2010): ‘‘... De values ... from large aliquots ... are not suitable for detecting poor bleaching’’. Preusser et al. (this issue) claim that the application of small aliquots including statistical approaches result in ‘‘better’’ deposition ages for glacial and glaciofluvial sediments and that the OSL age M. Frechen (&) A. Techmer Leibniz Institute for Applied Geophysics (LIAG), Stilleweg 2, 30655 Hannover, Germany e-mail: Manfred.Frechen@liag-hannover.de