Andreas Nilsson

Accurate and precise chronologies are needed to evaluate the existence and effect of a post-depositional remanent magnetization lock-in process on sedimentary palaeomagnetic records. Here we present lock-in modelling results of two palaeomagnetic records from varved lake sediments (lakes Kalksjon and Gyltigesjon) in Sweden by using model predictions based on archaeomagnetic data. We used the C-14 wiggle-match dating technique to improve the precision of the Kalksjon varve chronology in the period between 3000 and 2000 cal BP, which is characterized by pronounced palaeomagnetic secular variation. This method allowed us to infer an age model with uncertainties of +/- 20 years (95.4% probability range). Furthermore, we compared the palaeomagnetic record of Kalksjon to Gyltigesjon, which has a corresponding 14C wiggle-matched chronology. The ages of palaeomagnetic features derived from the wiggle-matched varve chronologies are older than those predicted by the archaeomagnetic models. Lock-in modelling was performed with different filter functions to explain the temporal offset and the amplitude of the lake sediment palaeomagnetic data. The analyses suggest that a linear lock-in function with lock-in depths (the depth below which no more natural magnetic remanence is acquired) that range between 30 and 80 cm in Kalksjon and 50 and 160 cm in Gyltigesjon are most appropriate to explain the data. These relatively deep lock-in depths in sediments without a bioturbated 'mixed-zone' can be attributed to the relatively high organic contents and low density of the lake sediments, which contribute to a thick unconsolidated upper zone of the sediment sequence in which re-alignment of magnetic particles can take place. (C) 2014 Elsevier B.V. All rights reserved.