No evidence for strong fields during the R3–N3 Icelandic geomagnetic reversal. Earth Planet

A previous study of the R3–N3 Icelandic reversal, which corresponds probably to the Matuyama–Réunion reversal, provided relatively large virtual dipole moments when the pole is close to equator. To check these former results, obtained using the Shaw paleointensity method, we resampled the very same sections and flows and reanalysed the samples using the Thellier method. 38 lava units (and one baked horizon) from 4 sections provide intermediate R3–N3 paleofield directions. Most of them, possibly related to a single eruption, display similar directions corresponding to the near equator VGP cluster. Unfortunately, most of the intermediate lava flows are thermally unstable, as attested by the irreversibility of the susceptibility versus temperature (k–T) curves, which precludes their use for paleointensity purpose. The rest of the units provide fairly reversible k–T curves and a near magnetite Tc due to Fe-rich titanomagnetite resulting from oxyexsolution at high temperature. 25 baked units or lava flows of this kind yield reliable Thellier paleointensity data. Average paleostrengths of 44:7 14:9 mT and 32:5 22:9 mT were found for the reversed R3 and the normal N3 polarity zones, respectively. Combining these non-transitional paleostrengths with other Thellier data from over the world indicates that the Pliocene dipole moment was close to 8 1022 A m2. The average paleostrength is reduced to 11:3 3:0 mT during the R3–N3 transition. We find no indication for the ‘strong fields ’ (20.4 to 38.6 mT) previously obtained by the Shaw method. Thus our results do not confirm the existence of large virtual dipole moments when the pole approaches th