Holocene geomagnetic secular variation in the western United States: implications for the global geomagnetic field

Abstract

A record of Holocene geomagnetic secular variation in the western 14 United States has been measured from volcanic rocks dated by 14 C. Analysis of errors associated with the paleomagnetic techniques used suggests that all laboratory measurement errors are smaller than those introduced by deformation of the volcanic units and the errors of orientation of cores in the field. The dispersion in the calculated virtual geomagnetic poles introduced by the geographic spread of sample localities around the western United States is less than the average uncertainty from an individual locality, about 2 deg. The history of secular variation of field direction in the western United States has been reconstructed where sufficient time control exists. This history can be approximately reconstructed for the past 3000 years; in this time period, inclination varied from 44 to 72 deg and declination varied from 344 to 22 deg. These variations of the local field direction have periods between 500 and 1500 years. Prior to 3000 years B.P., the available observations allow reconstruction of the secular variation history only for short time intervals. The mean VGP obtained from paleofield directions measured from 77 volcanic units of Holocene and late Pleistocene age in the western United States does not coincide with the geographic pole of the Earth, but is displaced 2.5 deg from the geographic pole in a direction away from the sampled region. This displacement is just at the limit of significance at 95 percent confidence. Using a 10 -5 torr vacuum furnace, paleointensities have been obtained by the Thelliers' method from both high-temperature oxidized and unoxidized basic volcanic rocks. Specimens which have undergone incomplete high-temperature oxidation in nature commonly fail the Thellier experiment; the relationship between NRM lost and TRM gained in a known magnetic field is not linear. Despite the vacuum, oxidation can still occur above 300°C in the furnace ovens. The best determinations of paleointensity come from specimens in which the remanence resides primarily in single phase titanomagnetite or in intergrowths of Ti-poor magnetite in ilmenite which have been formed by natural high-temperature oxidation. Paleointensity determined from these specimens varies from 0.364 x 10 -4 T to 0.865 x 10 -4 T (1.0 x 10 -4 T = 1 0e); the period of oscillation of paleointensity is about 500 to 1000 years. Close agreement has been found between the paleointensity measurements from this study and those from archeomagnetic studies in the southwestern United States. The similarity of the periods of intensity variation to those of field direction changes may indicate that variations of the nondipole components of the geomagnetic field dominate both. By combining paleomagnetic and archeomagnetic data from globally distributed localities, the magnetic dipole can be shown to have wobbled with a period as short as 2000 years. Local variation of the field can be described as the consequence of fluctuation of the dipole moment, wobble of the dipole, and longitudinal drift of a nondipole field similar to that of the present. During the Holocene, the dipole moment of the Earth has fluctuated with a period of approximately 8000 years, though when viewed from a single region, the fluctuation of the dipole moment is obscured by variation of nondipole components of the field.