[S-EM13] Geomagnetism, paleomagnetism and rock magnetism

[SEM13-02] Palaeomagnetism of ca. 3-4 Ma lavas from Borgarfjordur, Iceland - magnetostratigraphy and paleosecular variation of direction

Maxwell Christopher Brown^1,2, *Yuhji Yamamoto³, Hiroyuki Hoshi⁴, Masaru Kono⁵, Hidefumi Tanaka³, Takehiro Koyaguchi⁶, Brian Jicha⁷, Kohei Masaoka^3,8, Vivian Sinnen², Justin Tonti-Filippini^2,9, Hiroto Ishikawa³, Richard Bono¹⁰ (1.University of Minnesota, 2. University of Iceland, 3.Kochi University, 4.Aichi University of Education, 5.Tokyo Institute of Technology, 6.University of Tokyo, 7.University of Wisconsin-Madison, 8.Kyushu University, 9.Ludwig Maximilian University of Munich, 10.University of Liverpool)

Iceland is a unique location for palaeomagnetism. Glacially and fluvially eroded valleys expose sequences of hundreds of easily accessible lavas that dip gently towards rift zones: ideal candidates for reconstructing past variations in Earth’s magnetic field. In this study we present new palaeomagnetic directions from lavas spanning approximately 3.1-4.3 Ma from the Lundarreykjardalur valley in the Borgarfjörður region of western Iceland. In total ~1600 specimens from ~250 lavas spanning 14 sections along an 18 km transect of the Lundarháls ridge have been measured for palaeodirectional analysis. This provides one of the most detailed records of field changes recorded by lavas for this time.

This study is the culmination of two phases of field work and laboratory measurements. The first phase began in 1994 with the sampling and analysis of nine sections comprising the older half of the magnetostratigraphy. The second phase began in 2016, with an additional five sections expanding the magnetostratigraphy to younger ages. Through a combination of pilot K-Ar ages carried out in the 1990s and new ⁴⁰Ar/³⁹Ar ages we assigned a series of reversed and normal polarity lavas to chrons within the geomagnetic polarity time scale (GPTS) spanning approximately 4.3 to 3.1 Ma. ⁴⁰Ar/³⁹Ar ages confirm that the composite lava section records the reversal boundaries of C3n.1r/Cochiti (GPTS age, 4.300 Ma), Cochiti/C2Ar (4.187 Ma), C2Ar/C2An.3n (Gilbert/Gauss, 3.596 Ma) and C2An.2n/Kaena (3.116 Ma), but misses the Mammoth subchron.

Detailed variations in Earth's magnetic field can relate to changes in the geodynamo. We look at the distribution of virtual geomagnetic poles (VGPs), as well as their dispersion. Our analysis suggests the field recorded in the Lundarháls lavas is consistent with the geocentric axial dipole assumption, but that VGP dispersion may differ between chrons. This has implications for our understanding of the long-term variability of the field.