Kristian Vasskog

A Holocene record of snow-avalanche and flood activity reconstructed from a lacustrine sedimentary sequence in Oldevatnet, western Norway

Two lacustrine sediment cores from Oldevatnet in western Norway have been studied in order to produce a record of floods, mass-wasting events and glacier fluctuations during the last 7300 years. River floods, density currents and snow-avalanches have deposited distinct ‘event layers’ at the lake floor throughout this time interval. In this study, a novel approach has been applied to distinguish event layers from the continuous background sedimentation, using Rb/Sr-ratios from X-Ray Fluorescence data. Grain-size distribution and the sedimentological parameters ‘mean’ and ‘sorting’ were used to further infer the depositional processes behind each layer. Our data suggest a record dominated by snow-avalanches, with the largest activity occurring during the ‘Little Ice Age’ (LIA). This increase in snow-avalanche activity observed during the LIA was probably caused by a combination of generally increasing winter precipitation and the advance of local glaciers towards the steep valley sides. Several fluctuations in snow-avalanche activity are also recognized prior to the LIA. Proxies of glacial activity from the background sediments indicate a similar development as earlier palaeoclimatic reconstructions from the area. It differs from previous reconstructions, however, by suggesting a lower glacial activity in the period from 2200 to 1000 cal. yr BP.

Effects of hydrogen peroxide treatment on measurements of lake sediment grain-size distribution

Grain-size analysis is routinely performed on soft sediments, and has been applied in numerous paleoenvironmental studies using lake sediment archives. Despite the frequent use of this method, no common protocol exists for the treatment of lake sediment samples prior to analysis. In this study, differences in grain-size distribution before and after treatment with hydrogen peroxide (H2O2) were evaluated for sediment sequences from four lakes in Norway characterized by different geological and environmental settings. We found that removal of organic matter had a profound effect on the grain-size distribution of the sediment samples, which might have important bearing on how such data are interpreted. One of the analysed sediment sequences showed a systematic shift toward more fines after treatment, whereas no clear systematic effects were observed for the remaining three. The observed differences might derive from (1) variable concentrations of biogenic silica, which could cause shifts in the Sedigraph readings as a consequence of the porosity and fragility of diatom frustules, (2) clogging of organic matter during initial mechanical sieving, and (3) disturbance of the internal flow regime in the Sedigraph as a result of the dispersive agent having a viscosity that is too low. These possible explanations require further testing. Nevertheless, we propose that the paleolimnological community should develop a common protocol for pre-treatment of sediments from lacustrine settings that both enables reproducibility of results and reduces the risk of misinterpretation. Our observations suggest that sediments from boreal lakes should be treated with hydrogen peroxide prior to grain-size analysis, although treatment should be relatively mild to avoid possible mineral degradation.