GP11B-0827 0800h
Paleomagnetism of Ar-Ar Dated Lava Flows From the Ceboruco-San Pedro Volcanic Field (Western Mexico): Evidences for the Matuyama-Brunhes Transition Precursor and a Fully Reversed Geomagnetic Event in the Brunhes Chron
We report a detailed paleomagnetic and rock-magnetic study of seventeen independent lava flows belonging to the Trans-Mexican Volcanic Belt. 175 oriented samples were collected in the Ceboruco-San Pedro volcanic field. These sites were previously dated by means of the state-of-the-art 40Ar-39Ar geochronological method (Frey et al., GSA Bulletin, 116, 259-276, 2004) and span from 819 to 2 Ky. Rock-magnetic experiments which included continuous susceptibility and hysteresis measurements point to simple magnetic mineralogy. In most of cases, the remanence is carried by Ti-poor titanomagnetite of pseudo-single-domain magnetic structure. Fourteen flows give normal magnetic polarities while 2 are reversely magnetized and only one cooling unit yields intermediate paleodirections. The paleodirections of the flow dated as 819±25 ka correspond to a VGP latitude of 18°N. This anomalous field behavior apparently recorded prior to the Matuyama-Brunhes reversal may coincide with the transitionally magnetized lavas on La Palma, Canary Islands (40Ar-39Ar age - 822.2±8.7 ka, Singer et al., 2002) and with an event featured in several marine sediment records. Thus, this geomagnetic event, defined as M-B precursor, is probably global in extent. Two independent lava flows, dated as 623±91 and 614±16 ka respectively, yield reverse paleodirections. Age uncertainties make it extremely difficult to claim the discovery of a new geomagnetic event. It is possible that these lavas erupted during the worldwide observable Big Lost event (40Ar-39Ar age - 580.2±7.8 ka, Singer et al., 2002) which probably has been longer and more complex than it is generally believed for geomagnetic excursions. This study underlines the possible existence of a new, fully reversed geomagnetic event, but it is naturally premature to claim its presence in the Brunhes chron from this study alone.
GP11B-0828 0800h
Paleomagnetic Directions and 40Ar/39Ar Ages From Tatara-San Pedro Volcanic Complex, Chilean Andes: Lava Record of a Matuyama-Brunhes Precursor?
Lava flows in Quebrada Turbia, Tatara San Pedro Volcanic Complex, central Chile (36S, 289E), preserve a detailed record of what was presumed to be the Matuyama-Brunhes geomagnetic reversal. Sampling in the west wall of the valley yields reverse and transitional polarity flows; two km to the north, another section yields 17 transitional flows from the same sequence overlain by flows with normal polarity. 40Ar/39Ar incremental heating experiments on lavas of the two sections provide 9 independent age determinations and yield a weighted mean of 792 ka for the magnetic transition. Alternating field and thermal demagnetization, rock magnetic analyses, and petrographic observations indicate primary magnetization carried by titanomagnetite. The polarity change is characterized by a jump from reverse poles to poles clustered over Australia, followed by a jump to normal polarity latitudes. Magnetization of these lavas is consistent with either a brief period when the field was dominated by a subequatorial dipole, or a more complex non-dipolar field that may reflect the influence of a long-lived regional lower mantle control over a weakened dynamo. The Quebrada Turbia lavas are ca. 16 kyrs older than those recording the reversal at Haleakala volcano, Maui. Moreover, the 792 ka age of the Chilean lavas is older than most astronomical estimates for the Matuyama-Brunhes reversal suggesting that this section may record a precursor to the actual field reversal, seen in low paleointensities from more than a dozen marine sediment cores.
GP11B-0829 INVITED 0800h
Structural and Temporal Requirements for Geomagnetic Field Reversal Deduced From 40Ar/39Ar Dated Lava Flows
40Ar/39Ar dating of lavas on Tahiti, long thought to record the primary part of the Matuyama-Bruhnes (M-B) reversa1, gives an age of 795+/- 7 ka, indistinguishable from that of transitional lavas in Chile and La Palma, but older than the accepted age for the reversal. Only the transitional lavas on Maui and one from La Palma (dated at 776 +/- 2 ka), agree with the astronomical age for the M-B reversal. Virtual geomagnetic poles (VGPs) associated with the Tahitian and Chilean lavas cluster near Australia, as do VGPs recorded on Tahiti during the Big Lost and Punaruu events, two apparently unsuccessful reversals. These findings, suggestive of a recurring, mantle-held flux pattern at the outer core surface during reversal attempts, are also theoretically equivalent to the situation that would arise today if the axial dipole were to continue to weaken and vanish. Hence, we propose that the 795 ka lavas record the onset of a dynamo process--one which only on occasion would result in polarity change. This initial instability, associated with the first of two decreases in field intensity, began 18 kyrs prior to the actual polarity switch. These data may provide the first observational support to the claim that complete reversals require a significant interval of time for magnetic flux to escape from the solid inner core and sufficiently weaken its stabilizing effect.
GP11B-0830 0800h
Calibration of the Cenozoic Polarity Time Scale Based on Intervals of Constant Sea Floor Spreading Rates
Astronomical calibration of the polarity time scale is now nearly complete for 0-12 Ma and recent work is making rapid progress on calibrating 12-40 Ma. Integration of radiometric dating and sea floor spreading histories allows calibration of the Middle and Early Miocene to an accuracy arguably 0.1 Myr, supporting an Oligocene-Miocene boundary near 23.0 Ma. Spreading rate histories for the Neogene show that all plate pairs have sudden rate changes, that sudden ($<$0.5 Myr) rate changes are more common than gradual ($>$2 Myr) rate changes, and that rate changes usually correlate with recognizable changes in fracture zone trends and location of the pole of relative motion. These relations suggest a calibration strategy of identifying intervals of stable relative plate motion and assuming constant, rather than smoothly changing, spreading rates in these intervals. For Oligocene and Late Eocene (C6C-C17), Australia-Antarctica and Pacific-Nazca (Farallon) spreading distances provide excellent constraints when calibrated against Southwest U.S. ignimbrite polarity stratigraphy [McIntosh et al, 1992] with revisions for the FC 40/39 monitor at 28.21 Ma [Kuiper et al 2004]. Calibration of this interval differs subtly from published alternatives, with the Late-Early Oligocene boundary in C10N at about 28.2 Ma, the Eocene-Oligocene boundary in C13R at about 33.7 Ma, and the Middle-Late Eocene boundary in C17N at about 36.9 Ma, all with accuracy arguably 0.2 Myr or better. Calibration of Paleocene through Middle Eocene is much more speculative due to lack of ideal spreading records, lack of consensus calibration points, and apparently global plate-motion reorganization at or near C21. A strategy assuming constant rate for Pacific-Juan de Fuca (Vancouver) for C13-C21 and minimizing rate changes for Pacific-Farallon for C21-C32 appears to be at least as viable as other, published time scales. This alternative strategy brings ages for much of the Middle Eocene younger than on all commonly used time scales, with C21N spanning about 44-45 Ma and the Early-Middle Eocene boundary at about 47-48 Ma. Large differences between this work and the calibrations of Cande and Kent result from this work allowing an interval of slow South Atlantic spreading about C21-C31 to start and end suddenly whereas their cubic-spline technique forces spreading rates to change gradually.
GP11B-0831 INVITED 0800h
Age Offsets of the Matuyama-Brunhes Polarity Transition in Records From the Atlantic: Lock-in Depth Variations or Site Dependent Field Behavior?
A number of high -resolution Matuyama-Brunhes transitions have recently been obtained from deep-sea sediments in the Atlantic Ocean. Three of these transition records were obtained from sites on the Blake-Bahama Outer Ridge and the Bermuda Rise as part of depth transect drilled during ODP Leg 172. The polarity transition records from these sites are remarkably similar, both during the reversal and in excursions that occurred prior to the reversal. Previous work has shown that in this region initial magnetic susceptibility records provide a useful proxy for marine isotopic stages. However, using the shipboard susceptibility records as a correlation tool, it becomes apparent that while the transition records exhibit very similar features, these features are offset relative to the susceptibility records at ODP Sites 1060, 1061 and 1063. The distance between Sites 1060 and 1061 is insignificant on a geomagnetic scale, implying that the differences are not likely due to time transgressive field behavior. Instead, the offset of the magnetization record relative to the susceptibility record may result from a difference in the lock-in depths at these sites. A possible explanation is that the different water depths at the sites affect the amount of organic matter that is oxidized before reaching the sea-floor. This in turn would lead to different depths of the redox boundaries within the sediment column. If the remanence lock-in zone is related to processes associated with the redox boundary, the difference in water depths may explain the offsets. To test this hypothesis, we compare the other Matuyama-Brunhes records from the Atlantic with the positions of the reversal relative to the marine isotopic stages, including the record from Site 1083, which was recorded in anoxic sediments.
GP11B-0832 0800h
Local Reversal Durations and Inverse Magnetic Field Patches in Geodynamo Simulations
Paleomagnetists commonly infer reversal properties based on records from one site only. In an attempt to gain more global information about geomagnetic field reversals Bradford Clement (Nature, 2004) analyses the latitude dependence of reversal durations in several paleomagentic records. He finds that the field transitions takes signifincantly longer close to the poles than near the equator. Such a piece of information is important for testing numerical geodynamo simulations. Many simulated reversals show a comparable latitudinal dependence. These models offer the opportunity to identify the cause of such an effect. I have tried to link the reversal duration to the transitional magnetic field at the core-mantle boundary. Unfortunately, the connection is complex. Reversals tend to last longer then average where the first inverse patches appear at the core-mantel boundary. They are shorter than average where secondary inverse patches tilt over an already weak core-mantel boundary field. Any equatorial and large scale azimuthal transitional field asymmetries can be identified in a global map of reversal duration. It has been speculated that the intensifying inverse patches at the Earth's core-mantel boundary may constitute the start of a field reversal. So far, they are too weak to lead to significant changes in field inclination. This, however, may change if the inverse patches close to the equator move to higher latitudes. Such a behavior is frequently found in dynamo simulations and indeed can mark the start of a field reversal. However, more often the inverse patches disappear again causing only slight oszillations in inclination.
http://www.linmpi.mpg/~Wicht/Reversals
GP11B-0833 0800h
Geomagnetic Dipole Strength and Reversal Rate
We present a first 2 million years long composite curve obtained after stacking records of relative paleointensity from a selection of sedimentary sequences. The composite Sint-2000 record was calibrated using the virtual dipole moments (VDMs) of the 2004 updated volcanic database of absolute paleointensity over 0.1 Myr long intervals. The value of the time-averaged VDM (7.46+/-1.16 x 1022 Am2) for the past 0.8Myr was used for calibration and the mean values of the successive 0.1 Myr intervals were found in very good agreement with the relative paleointensity for the same periods. A striking characteristic of this Sint-2000 curve is the succession of periods with different mean values of paleointensity. During the Brunhes chron the dipole oscillated around a value of 7.51+/-1.66 x 1022 Am2, which was significantly larger than during the previous 400 kyrs (5.3 +/- 1.5 x 1022 Am2). To provide a more quantitative picture of field strength as a function of reversal frequency, we calculated successive running averages of the field intensity over 100 kyrs long intervals and found that the time-averaged field was higher during periods without reversals. We also observe that the amplitude of the short-term oscillations remained the same. As a consequence, less intervals of very low intensity are expected during periods associated with a strong average dipole moment, whereas more excursions or other instabilities are produced during periods of weak field intensity. The relation between the mean dipole strength and the frequency of reversals suggests also the existence of a large field during long periods without reversals, under the assumption that they would be governed by the same processes. Prior to reversals, the axial dipole decays during 60 to 80 kyrs, but rebuilds itself in the opposite direction much more rapidly, in a few thousand years at most. These time constants suggest that the decay phase is caused by diffusion while advection would dominate the dipole recovery.
GP11B-0834 INVITED 0800h
Microwave Palaeointensity Study of Jorullo Volcano (Central Mexico)
A single lava flow from the earliest eruption period (ca.1759-1774) of Jorullo Volcano, Michoacan-Guanajuato volcanic field, Mexico has been sampled to determine the palaeointensity of the magnetic field at the time of extrusion. Both conventional Thellier-Thellier experiments and two variants of the microwave palaeointensity technique (Coe variant of the Thellier-Thellier method (Coe, 1967) and the perpendicular applied field method (Kono & Ueno, 1977)) have been carried out. High quality palaeointensity results have been obtained, with both methods mostly in good agreement. This study presents further evidence of the equivalence of both microwave and thermal methods and adds to the database of palaeomagnetic data from the monogenetic volcanoes of Central America.