Correlation of the fluvial magnetic susceptibility (MS) record of borehole Devavanya‐1 in the Körös Basin (eastern Hungary) with Chinese aeolian MS records (Jingbian, Lingtai) and the marine δ18O record from the Equatorial Pacific (V28‐239) is established here based on cross‐correlations and singular spectral analysis. A basin‐scale well‐to‐well correlation based on magnetic susceptibility records was also performed involving unpublished cores. To refine the age model, a Monte Carlo simulation was conducted using the Chinese Jingbian section as a tuning target. Spectral analysis of the tuned record revealed c. 400, c. 100 and c. 41 ka cycles over the 2.5 million years of the Quaternary fluvial succession. To ensure a complementary palaeoclimate proxy, the full width at half maximum of smectites was measured as a facies‐independent indicator of weathering intensity. This investigation was carried out on a subset of samples involved in MS measurements representing a c. 400 ka time interval across the top of the Olduvai subchron. A phase‐shift between MS and weathering intensity recorded in the clay mineralogy indicates different response times of the considered proxies. The fluvial MS record is determined by the climatic control on delivery and preservation of magnetic minerals, mainly of magnetite. Under cold‐and‐dry climate these minerals were released owing to frost shattering in the adjacent hinterlands and were transported to alluvial plains in the early postglacial periods thanks to the increasing discharge of rivers. With further warming the weathering‐sensitive magnetic minerals soon disappeared from the soils of the catchment area and thus from the fluvial load. As a result, in fluvial successions early postglacial warmings are expressed by the occurrences of MS maxima (magnetic episodes), while the palaeotemperature maximum and the subsequent cooling remain concealed within the tract of low MS values. The early postglacial magnetic episodes may serve as ideal stratigraphical markers in regional and global correlations.
Read full abstract