Late Quaternary Geomagnetic Excursions at High Northen Latitudes in Marine Sediments: Reproducing Results for the Wrong Reasons?

Paleomagnetic records based on multiple sediment cores from the Arctic Ocean and Greenland Seas (e.g. Nowaczyk et al. 2003 Geophys. J. Int. 155) have suggested that several Late Quaternary geomagnetic excursions took place in this region, with the possibility of two excursions with calibrated radiocarbon ages younger than Laschamp (~40ka). During a study of the effect of the Laschamp geomagnetic excursion and the associated decrease in dipole moment on cosmogenic radionuclide production in the atmosphere, we have measured the palaeomagnetic properties of three sediment cores: JM05-31GC1 (from the Yermak Plateau), JM04-025PC2 (western Svalbard shelf) and LINK15 (Faroe-Shetland Channel). Magnetic susceptibility logs of the three cores show millennial-scale trends that can be correlated to the Dansgaard-Oeschger stadial/interstadial cycles recorded in the Greenland ice cores. Laschamp is the only geomagnetic excursion in marine isotope stage 3 that can be securely identified in the LINK15 core and it took place during D-O cycles 9 and 10, as determined from the record of magnetic susceptibility. This core was recovered from a sediment drift receiving detrital minerals from a mainly basaltic province and, therefore, the sediments contain relatively high concentrations of magnetic minerals and are characterized by stable natural remanent magnetizations. On the other hand, the two other cores were recovered from areas with a detrital source dominated by sedimentary bedrock. Thus, the sediments have a relatively low magnetic mineral concentration and relatively weak natural remanent magnetizations. In the two cores from the Arctic, a few apparent excursions were revealed by the application of successively stronger alternating field demagnetization steps. Thus, we were able to reproduce the results of Nowaczyk et al. (2003). However, we discovered that many of the apparent excursions in our cores, which are also characterized by low relative paleointensity estimates, were artifacts caused by the acquisition of anhysteretic remanent magnetizations perpendicular to the final demagnetization (vertical geographic) axis by samples with weak natural remanent magnetizations. We cannot claim that the earlier published reports of excursions were due to a similar systematic laboratory magnetization; however, we do not find conclusive evidence of an excursion younger than Laschamp (e.g. Mono) in these relatively high-resolution sediment records.