Ar/Ar ages from transitionally magnetized lavas on La Palma, Canary Islands, and the geomagnetic instability timescale

Abstract

A detailed study of 43 lava flows comprising two stratigraphic sequences exposed along the north and south walls of Barranco de los Tilos on the island of La Palma, Canary Islands, reveals a complex, temporally segmented record of geodynamo behavior that contains no less than three distinct geomagnetic events. The Matuyama‐Brunhes (M‐B) reversal is recorded in five transitionally magnetized lava flows from the north (TN) section. The isochrons obtained from three of the lower four M‐B lavas are defined by 14 incremental heating experiments that, together with a previous age determination, yielded a weighted mean of 798.4 ± 6.2 ka (all uncertainties ±2σ). In addition, a 780.3 ± 10.3 ka isochron was determined for the overlying transitionally magnetized flow, indicating that it was erupted during a distinctly younger portion of the transition. Near the base of the south (TS) section one finds a sequence of weakly magnetized flows associated with virtual geomagnetic pole (VGP) positions in the southwest Indian Ocean between latitudes 56°S and 65°S, suggesting instability of the geomagnetic field beyond that of typical secular variation. 40Ar/39Ar isochrons from three of these flows, defined by 11 separate incremental heating experiments, gave a weighted mean of 822.2 ± 8.7 ka. This anomalous field behavior recorded 24 ± 11 kyr prior to the M‐B reversal may coincide with an event featured in several marine sediment records. Directly above two normal polarity flows (40Ar/39Ar isochrons of 751.9 ± 18.1 ka and 675.0 ± 15.7 ka) are nine transitionally magnetized lavas having magnetization directions associated with low to midlatitude VGPs spanning 23°–60°N. These flows are then capped by a single flow possessing normal polarity. Based on 12 incremental heating experiments, 40Ar/39Ar isochrons of five of these nine lavas, along with the uppermost flow, gave a weighted mean age of 580.2 ± 7.8 ka for this period of transitional to normal field behavior. From these same transitional lavas, Quidelleur et al. [1999] reported three unspiked K‐Ar ages with a weighted mean of 602 ± 24 ka and proposed a new event called the “La Palma” excursion. However, the 40Ar/39Ar age presented here is three times more precise than the K‐Ar age and is indistinguishable at the 95% confidence level from the 40Ar/39Ar age of a lava from the Snake River Plain, Idaho, that originally defined the Big Lost event. Transitional field behavior of similar age observed in astronomically dated marine cores further establishes that the Big Lost event recorded at La Palma was indeed global in extent. Rigorous temporal and geomagnetic constraints for several additional periods of geomagnetic field instability during the last several million years will comprise a geomagnetic instability timescale that can be factored confidently into models of the dynamo process.