過去230万年間の地磁気伏角永年変化に見られる10万年周期

*山崎 俊嗣[1], 小田 啓邦[1]

産業技術総合研究所　地質調査総合センター[1]

Long-term secular variation of geomagnetic field inclination with 100 kyr periodicity during the last 2.3 Myrs

*Toshitsugu Yamazaki[1] ,Hirokuni Oda [1]
Geological Survey of Japan, AIST[1]

A continuous record of secular changes in inclination during the last ca. 2.3 Myrs has been obtained from a marine sediment core of 42m long. We found long-term cyclic changes in inclination with ca. 100 kyr periodicity. The period of 100 kyr is significantly longer than electrical diffusion time of the core, and thus condisered to be excited from outside the core. The periodicity suggests a possible connection between the geodynamo and the orbital eccentricity.

Recent progress in relative paleointensity estimation during the Brunhes Chron revealed the presence of long-term variations (order of 10 to 100 kyrs). Some studies have found the Milankovitch orbital frequencies in the paleointensity records, but others result in contradicting conclusions: no stable frequency, or artefact caused by paleoclimatically induced rock-magnetic changes. Long-term changes are expected to exist also in directions if outer forces energize the geodynamo. A few studies have indeed suggested possible occurrence of such changes, but they were suspicious because of the lack of continuous and long-enough data with good age-control. We will present for the first time a continuous record of secular changes in inclination during the last ca. 2.3 Myrs.
A giant piston core of 42m long (MD982183) was taken from the West Caroline Basin at 3 deg 05 min N, 135 deg 00 min E. The sediments have rock-magnetically favorable nature, which enabled recovery of geomagnetic filed variations both intensity and direction. A magnetic polarity sequence down to the Reunion I subchron was identified clearly from declinations. Average inclination is approximately zero and thus the polarity sequence cannot be recognized in inclination. Relative paleointensity was estimated from NRM intensities normalized by ARM after AF demagnetization of 30mT. The variations during the Brunhes closely resemble to the standard curve Sint 800, from which the depths of the core were converted to ages. The age control before the B/M boundary is based on the magnetostratigraphy.
The characteristic feature of the inclination fluctuation is the occurrence of long-term cyclic changes throughout the record. Frequency analysis indicates the presence of ca. 100 kyr period. Cyclic lithological changes may induce cyclic changes in the magnitude of the inclination error, and then inclination. However, this cannot explain the variation observed here. In equatorial regions like this site, inclinations are close to zero, and thus the inclination error should be negligibly small. Furthermore, a shift from 40 to 100 kyr periodicity at about 800 ka occurs only in magnetic susceptibility reflecting the paleoclimatic change known as the mid-Pleistocene revolution, but not in inclination. This fact strongly suggests that the 100-kyr periodicity in inclination would reflect geomagnetic field behavior.
The period of 100 kyr is significantly longer than electrical diffusion time of the core, and thus condisered to be excited from outside the core. The 100-kyr periodicity suggests a possible connection between the geodynamo and the orbital eccentricity. The cycles of the 100-kyr changes seem to be not interrupted by polarity changes. The secular variations in directions would be maintained flows near the surface of the core, which may be independent to the main flows deeper in the core which sustain the axial dipole field.