GP34A-01
Direct measurement of hematite individual particle anisotropy: implications for inclination shallowing in red bed DRMs.
Methods to correct for the observed inclination shallowing in sedimentary rocks have been proposed that are based on either models of the geomagnetic field and the resulting directional distribution of paleomagnetic vectors or the magnetic anisotropy of the magnetic minerals carrying the remanence. One limitation of the anisotropy method for hematite-bearing red beds has been the isolation and determination of a rock's detrital hematite individual particle anisotropy. Up to now, our red bed inclination shallowing corrections have been dependent on estimates of hematite individual particle anisotropy using data fit to theoretical correction curves. We have developed a technique for preferentially extracting the detrital hematite particles in a sample in order to directly measure their individual particle anisotropy. The method involves crushing of the sample followed by ball milling and sieving to ensure that the rock particles are smaller than 4Φ. The resulting slurry was then placed in an ultrasonic cleaner for at least 24 hours and finally centrifuged at 1000 rpm for 20 minutes in order to separate the dense, gray iron oxide particles from the red pigmentary grains. The gray, iron oxide-rich slurry was collected by hand and circulated in a magnetic extraction apparatus. The magnetic separate was then collected over a period of two to three weeks. Small amounts of the magnetic separates where mixed in a slow-drying epoxy resin for 24 hours and placed in a DC magnetic field (100 mT to 180 mT) in order to align the grains. The bulk IRM anisotropy of the epoxy samples provides an average individual particle anisotropy for the magnetic grains. Separates were collected from samples of the Mauch Chunk Fm. of Pennsylvania, the Maringouin and the Shepody Fms of New Brunswick/ Nova Scotia and the Kapusaliang Fm. of northwestern China. IRM acquisitions experiments were performed in fields of up to 1.2 T in order to identify the magnetic mineralogies present. Remanence appears to be carried by a low coercivity phase (~$50 mT) interpreted to be secondary magnetite and a higher coercivity phase (~$350 mT) interpreted to be primary hematite for the Shepody and Maringouin Fms or just one high coercivity component (200- 250 mT) interpreted as primary hematite for the Mauch Chunk and Kapusaliang Fms. Hematite individual particle anisotropy was measured by imparting a 1.2 T IRM to the specimens in 9 different orientations followed by AF demagnetization at 100 mT. Calculated individual particle anisotropy values ranged between 1.28 and 1.45 with bulk anisotropies of ~$40%. Inclination corrections using the directly measured individual particle anisotropies indicate significant inclination shallowing for the Mauch Chunk and Kapusaliang Fms, while more moderate shallowing for the Maringouin and Shepody Fms. Curve fitting techniques with added constraints give a good first order approximation of the individual particle anisotropy, however direct measurement is preferable. The measured particle anisotropies for hematite are low and suggest that there is the potential for significant amounts of shallowing for a hematite DRM. This observation is consistent with redeposition experiments performed by Tauxe and Kent [1984] and the notion that depositional inclination of hematite may suffer from more shallowing than magnetite because of its lower spontaneous magnetization making it more affected by gravitational forces.
GP34A-02
Determining the Anisotropy of Remanence Tensor Using a Conventional Electromagnet: Overcoming the Hematite Problem and Correcting for Inclination Error in Redbeds
The origin of the natural remanent magnetization of sediments is often investigated by studying their anisotropy, either of magnetic susceptibility or some form of remanence. An oblate magnitude ellipsoid with a minimum principal axis aligned with the bedding pole is necessary (almost, though not sufficient) evidence for a primary depositional fabric and therefore a primary characteristic remanence. The anisotropy of isothermal remanent magnetization (IRM) can also be used to correct for inclination error (Jackson et al., GJI, 104, 95-103, 1991). While this is straightforward for magnetite bearing sediments, the experimental procedure to determine the anisotropy of remanence for hematite bearing sediments is problematical (Tauxe et al., JGR, 95, 4391-4404, 1990). The problem arises from ‘magnetic memory', which for magnetite can be easily erased (reset) by alternating field (AF) demagnetization. The coercivity of hematite is prohibitively high for AF demagnetization and therefore only a fraction of the coercivity spectrum is accessible leaving a significant memory from previous exposures to high magnetic fields. By using a 14 T superconducting magnet Kodama and Dekkers (Stud. Geophys. Geod., 48, 747-766, 2004) have demonstrated that it is possible to derive the full anisotropy tensor for hematite bearing samples by imparting saturation IRMs, which activate almost the whole coercivity spectrum. While this is undoubtedly the ultimate solution to the hematite problem, such machines are not commonly available and are very expensive to operate. An alternative procedure that allows the use of more common and cheaper ~1 T electromagnets that obviates the AF demagnetization stage is to rotate the sample in the field in a manner that erases, or resets, the magnetic history so each IRM is equivalent. Admittedly, not all the coercivity spectrum is accessed, but typically something >50 percent is available compared to perhaps <10 percent for anhysteretic remanence. There is no unique set of rotations that resets the magnetic history but the least requirement, not unlike that for AF demagnetization, is that the sample needs to be exposed to a steady magnetic field along many axes. Instead of slowly reducing the field to zero as for AF demagnetization, the field is held constant with the rotation action converging on the desired axis in a spiral fashion. When the axis and field are aligned the field can be removed and the remanence measured before repeating the exercise for another axis. The effects of magnetite and goethite were eliminated by routine AF and thermal demagnetization to 150 mT and 120°C respectively after imparting the IRMs. Results from ‘rotating in the field' to date suggest they are comparable to the expensive superconducting magnetic results. Results from the simple oblique single IRM method will also be compared.
GP34A-03
Inclination Error Correction In Red Beds: Is It Possible ?
Highly detailed records - including processing of hundreds of samples - have been obtained from red beds in southern France, in the Dôme de Barrot and Lodève basins. The main purpose was to sample and determine paleosecular variation (PSV) - over sufficient time - during the Permian Superchron. We compared our records with older and earlier published literature data, and generally find good agreement. Since (hematite bearing) red beds are famous for their inclination error, we tried to correct our distributions using two independent methods. One method is using a PSV model (TK03.GAD; Tauxe and Kent, 2004) which - not surprisingly for such low latitudes - gives no significant correction on the distributions that sufficiently sample PSV. In addition, our data are in very good agreement with published APWP data, giving confidence in the recording qualities of these red beds, at least at paleoequatorial latitudes. Another method is to correct the inclination via an approach (the "a- factor"of Tan and Kodama, 2002) that uses the anisotropies of the magnetic susceptibility and of laboratory acquired (anhysteretic or isothermal) remanence. To this end, we sampled single layers, that we assumed to record - approximately - one single occurrence of PSV. A model approach was used to estimate the a-factor, rather than determining this parameter from laboratory experiments. We also used TK03.GAD on a large distribution (N=~200) of these single layer samples. This yielded interesting results. In one case a positive inclination was corrected - via the a- factor model - to a negative inclination, and in an another case the Permian red beds were corrected - using the TK03.GAD model - to a position at the latitude of the Netherlands, in contrast to their assumed paleo-equatorial position. We discuss the various merits of these different and independent methods for inclination error correction in these (and other) red bed sequences.
GP34A-04
Paleomagnetism of ~635 Ma "Elatina Rhythmites," Reconsidered
Extraordinary laminated sand-siltstone of South Australia's synglacial/deglacial Elatina Formation-equivalent at Pichi Richi Pass is an iconic outcrop for paleomagnetists, paleoclimatologists, and astrophysicists. Cyclic- bedded stacks of coarse/fine-grained doublets have been interpreted as rhythmites of semi-diurnal tidal origin in a deep-water delta front environment. Cycle analysis has yielded estimates of Earth's paleorotation and revolution periods and lunar distance. The reported 10 m (interpreted 60 year) record disagrees somewhat with coeval and younger biological "sediment-clocks," and Williams (2000) argues that the sedimentological context and extraordinary preservation of Elatina rhythmites makes that record preferable. A paleomagnetic "synsedimentary fold" test on <8 cm amplitude antiforms from the Pichi Richi Pass rhythmites confirmed low-latitude Neoproterozoic glaciation (Sumner et al., 1987) and inspired the formal "Snowball Earth" hypothesis (Kirschvink, 1989). Repeated study of the Elatina rhythmites (Schmidt and Williams, 1995 and citations therein) has documented NRM essentially equal to ChRM carried by detrital hematite, "locked-in" synchronously with "deformation" of "antiforms" (i.e., within hours to days of deposition). "Synfolding" remanence has been ascribed to superposition of slump-shear and DRM. Anomalously shallow inclination of rhythmites (Sohl et al., 1999) could represent inadequate averaging of ca. 635 Ma geomagnetic secular variation. New measurement of ten stratigraphic sections through rhythmite-bearing facies outside of Pichi Richi Pass supports a distributary environment since rhythmites are expressed in multiple lithologies closely beneath ultimate deglacial, Nuccaleena cap dolostone. At Pichi Richi Pass, a full section of rhythmites approx. 18 meters thick is now documented, supporting a novel magnetostratigraphic collection of 130 samples. It is necessary to re-examine the DRM interpretation of Elatina rhythmite magnetization, however, in light of new observations: The structures previously considered "slump folds" are in fact sedimentary ripple crests. At Pichi Richi Pass, rhythmite facies grade upward from undisturbed, planar lamination into disorganized dunes, then low- wavelength, low-amplitude, linear-crested ripples, increasing steadily in wavelength and amplitude and climbing upward in >50 cm packages to culminate in unambiguous ladderback ripples. This shallowing-upward sequence reflects high sediment supply over 60 years and/or else significant sea level regression during (possibly longer) rhythmite deposition. The exceptional character of climbing-ripple rhythmites can be explained by pervasive microbial binding of sediment, in the anomalous mode characterizing many Ediacaran siliciclastic environments. Detrital magnetization seemingly should not around ripple crests. Either the interval of Snowball Earth deglaciation was characterized by unusually low geomagnetic field strength, biasing magnetization fidelity, or else Elatina rhythmites do not record DRM over a continuous 60-year timescale. This latter alternative may find support in the original studies, which document considerable dispersion, unexpected for a VGP. Kirschvink, 1992. in Schopf, J.W. and Klein, C., eds. The Proterozoic Biosphere, Cambridge University Press, Cambridge, pp. 51-52. Schmidt and Williams, 1995. Earth and Planetary Science Letters, v. 134, pp. 107-124. Sohl, Christie-Blick, and Kent, 1999. Geological Society of America Bulletin, v. 111, pp. 1120-1139. Sumner, Kirschvink, and Runnegar, 1987. EOS, v. 68, pg. 1251. Williams, 2000. Reviews of Geophysics, v. 38, pp. 37-59.