Paleointensity and 40Ar/39Ar geochronology of basalts at Gamarri, Ethiopia: Correlation with the Réunion subchron and Huckleberry Ridge excursion?

Characterization of the geomagnetic field during subchrons, reversals, and excursions is vital to understanding geodynamo processes and interactions across the core-mantle boundary. Moreover, an accurate timescale for geomagnetic field instabilities is critical to global high resolution stratigraphy. The Réunion subchron and Huckleberry Ridge excursion are ideal candidates for study due to globally distributed recordings in both sedimentary and igneous rocks. We present new full-vector paleomagnetic data for 30 basaltic flows from the Gamarri volcanic section in the Afar region of Ethiopia and 11 ⁴⁰Ar/³⁹Ar ages. Paleointensities were calculated using the LTD-DHT Shaw technique and results generally agree with those of Carlut et al. (1999). Two geomagnetic instabilities are recorded, an older excursion and a younger period of normal polarity within the reversed Matuyama chron. Our results show a longer duration of low (<20 μT) paleointensity in the oldest flows and more variable low paleointensity values in the younger flows, and are generally lower than Thellier-style values of Carlut et al. (1999). Relative to 28.201 Ma Fish Canyon sanidine, plateau ⁴⁰Ar/³⁹Ar ages of the youngest (GB21) and oldest (GA02) flows are 2.029 ± 0.041 (2σ) and 2.410 ± 0.130 Ma, respectively. This eruptive duration is longer than that reported by Kidane et al. (1999), where the unspiked K-Ar method yields ages for GB23 (2 flows overlying GB21) and GA02 of 2.02 ± 0.08 (2σ) and 2.14 ± 0.12 Ma, respectively. ⁴⁰Ar/³⁹Ar ages of 4 lavas within the normal polarity zone in the upper section are between 2.063 ± 0.044 and 2.118 ± 0.057 Ma, but are indistinguishable at 2σ. These flows may sample the Huckleberry Ridge excursion (2.086 ± 0.016 Ma, Singer et al. 2004), the Réunion subchron (2.153-2.115 Ma, Channell et al. 2003), or both. Given several ⁴⁰Ar/³⁹Ar ages >2.2 Ma, the older excursion in the Gamarri section is not consistent with the Réunion subchron, and can be linked to any of several excursions occurring between ~2.2 and ~2.5 Ma. These excursions have been observed within records from ODP 982 (Channell & Guyodo, 2004) and IODP U1314 (Ohno et al., 2012), as well as within the GPTS as cryptochron C2r.2r-1 (originally dated as 2.420 to 2.441 Ma by Cande & Kent, 1995). Thus, we no longer interpret the excursion recorded in the lower portion of the Gamarri section to be part of the Réunion subchron and recommend that it be omitted from efforts to construct integrated global field models across the Huckleberry Ridge excursion and Réunion subchron.