Palaeomagnetic Secular Variation of Early Miocene Volcanics from the Rhön (Germany)

Summary Approximated by a geocentric axial dipole oscillating between two states of polarity, the main geomagnetic field experiences directional fluctuations during stable periods termed palaeosecular variation (PSV). How the changes in reversal frequency correlate with the vigour of PSV over geological timescales remains an open question. To better constrain this relationship during the Cainozoic, we sampled 107 palaeomagnetic sites (1,206 oriented cores) from the Rhön volcanic field emplaced ∼18–20 Ma in Central Germany. We successfully interpreted 97 site-mean palaeodirections, leading to a collection of 80 independent geomagnetic snapshots after combining directionally correlated sites. This collection passes the reversal test but exhibits a large proportion of outlying or extreme directions, the origin of which remains uncertain and cannot be explained by the magnetic mineralogy of the remanence carriers. In particular, 25 of the 80 records yield virtual geomagnetic poles (VGPs) outside a fixed 40○ cutoff, yielding a VGP scatter SB=23.3∘\, |^26.2∘_20.6∘ (N = 55). By contrast, the adaptive VGP cutoff with A = 71.1○ is ineffective to remove extreme directions, yielding SB=36.7∘\, |^41.1∘_32.6∘ (N = 79) statistically indistinguishable from SB=37.5∘\, |^42.1∘_33.3∘ (N = 80) without VGP cutoff. When the collections from the Rhön and the neighbouring ∼15–18 Ma Vogelsberg volcanic field are combined, the fixed and adaptive cutoffs yield more consistent estimates of SB, ranging from 20.2∘\, |^21.7∘_18.8∘ (N = 169) for a fixed 40○ cutoff to 20.9∘\, |^22.5∘_19.4∘ (N = 173) for an adaptive VGP cutoff. This suggests that datasets of 100–200 independent directions should be favoured for PSV studies. Finally, focussing on a compilation of SB estimates at 40○–50○ magnetic palaeolatitude over the past 150 Myr, we confirm that no robust relationship can be established between SB and reversal frequency, irrespective of the employed filters.