GP11C-01
A Paleomagnetic and Paleointensity Study on Late Pliocene Volcanic Rocks From Southern Georgia (Caucasus)
A paleomagnetic, rock-magnetic and paleointensity study was carried out on 21 basaltic lava flows belonging to four different sequences of late Pliocene age from southern Georgia (Caucasus): Diliska (5 flows), Kvemo Orozmani (5 flows), Dmanisi (11 flows) and Zemo Karabulaki (3 flows). Paleomagnetic analysis generally showed the presence of a single component (mainly in the Dmanisi sequence) but also two more or less superimposed components in several other cases. All sites except one clearly displayed a normal-polarity characteristic component. Susceptibility-versus-temperature curves measured in argon atmosphere on whole- rock powdered samples yielded low-Ti titanomagnetite as main carrier of remanence, although a lower Tc- component (300-400ºC) was also observed in several cases. Both reversible and non-reversible k-T curves were measured. A pilot paleointensity study was performed with the Coe method on two samples of each of those sites considered suitable after interpretation of rock-magnetic and paleomagnetic results. The pilot study showed that reliable paleointensity results were mainly obtained from sites of the Dmanisi sequence. This thick sequence of basaltic lava flows records the upper end of the normal-polarity Olduvai subchron, a fact confirmed by 40Ar/39Ar dating of the uppermost lava flow and overlying volcanogenic ashes, which yields ages of 1.8 to 1.85 My. A new paleointensity experiment was carried out only on samples belonging to the Dmanisi sequence. Although this work is still in progress, first results show that paleointensities are low, their values lying between 10 and 20 µT in many cases, and not being higher than 30 µT. For comparison, present day field is 47 µT.
GP11C-02
The First Radiocarbon-Constrained Full-Vector Holocene Paleomagnetic Secular Variation Reconstruction for Eastern Canada
Here we present the first high-resolution Holocene paleomagnetic secular variation (PSV) master curve and relative paleointensity stack constrained by radiocarbon dates for Eastern Canada. This reconstruction is based on seven sedimentary sequences collected from the St. Lawrence Estuary and Gulf from the head to the mouth of the Laurentian Channel. The natural remanent magnetization (NRM), the anhysteretic remanent magnetization (ARM) and the isothermal remanent magnetization (IRM) were measured on u-channel samples at the Paleomagnetism Laboratory of ISMER using a 2G Enterprises cryogenic magnetometer. Magnetic mineralogy was additionally assessed using a recently installed alternative gradient force magnetometer (AGM). The magnetic mineralogy of the Holocene postglacial sediments is mainly carried by low-coercivity ferrimagnetic minerals (most likely magnetite) in the pseudo-single domain size range. The characteristic remanent magnetization (ChRM), assessed by principal component analysis, reveals the presence of a stable and well-defined magnetization characterized by maximum angular deviation (MAD) values generally lower than 5°. Furthermore, ChRM inclinations fluctuate around the expected geocentric axial dipole (GAD) magnetic inclination for the latitude of the coring sites (from 63° to 66°). The similarity of these records on their own independent timescales implies that all of the individual sedimentary sequences record a reliable Holocene PSV record for Eastern Canada. Relative paleointensity (RPI) was estimated by normalizing the NRM by the ARM which provided the best coercivity match. Lastly, the paleomagnetic directional and RPI records were stacked on a common time scale spanning the last ~10 000 cal BP. The smoothed PSV stack reveal centennial- to millennial-scale geomagnetic features concordant with the CALS7K.2 time-varying spherical harmonic model, as well as with the US eastern stack (King and Peck, 2001). Comparisons further a field with the Fish Lake record from Oregon (Verosub et al., 1996), the Icelandic and E. Greenland continental margin records (Stoner et al., 2007) and the Fennoscandia stack (Snowball et al., 2007) suggest large scale coherence of geomagnetic features. These and other comparisons will be made. In conclusion, the full-vector Holocene PSV and RPI master curves from Eastern Canada provide a new high-resolution regional chronostratigraphic tool as well as new insights into the Holocene geomagnetic field behaviour.
GP11C-03
Assessment of Magnetostatic Interaction Effects on Thellier Paleointensity Determination by Experimental Simulations
The ability to control magnetic interactions between grains is of fundamental importance in paleointensity studies. We continued to perform experimental simulations to help understand the effect of magnetostatic interaction on Thellier type paleointensity experiments, using artificial synthesized magnetite grains mixed with both pseudo-single domain (PSD) and multidomain (MD) particles. Magnetite powders were mixed either with an Aron ceramic or were dispersed in matrix of Seto porcelain clay. The effects of interaction between grains can be observed from the magnetic behavior of specimens with different inter-grain distances. The maximum effect of domain's interaction can be estimated by comparing the behavior of specimens with large inter-grain distance (i.e., mostly dispersed-grains) with that of ideal non-interacting SD grains. Our results clearly show that (1) the interaction between grains (rather than domain's interaction) has particular disastrous effects on the Thellier-Coe paleointensity experiment; (2) interaction of large inter- grain distance samples adds an almost negative constant value to the applied external field (i.e., acting as an internal demagnetizing field); (3) interaction in shorter inter-grain distance samples mainly generates the difference in blocking and unblocking temperatures of the sample. Detailed results will be presented and discussed at the meeting.
GP11C-04
The Influence of Cooling Rates on the Paleointensities of Volcanic Glasses Tested on Synthetic and Natural Glass
For a better understanding of the ancient magnetic field of the earth, paleointensities are essential. Pristine
glass is often dominated by single domain magnetic remanence and alteration can be excluded to the most
extent. However a major biasing factor is the magnetic cooling rate dependency. In this study, we validate the
method used to correct this effect. A natural phonolitic glass sample from Tenerife, Spain was homogenized at
1600 ° C in order to let the volatiles escape. It was then tempered at 900 ° C for 10 h to grow a
sufficient amount of magnetic remanence carriers. Six samples of this glass were heated above the glass
transition temperature and cooled down with constant rates of 0.1 to 15 Kmin-1 in a SiC furnace with
bifilar winding in the present ambient magnetic field of Munich (∼48 μT). Rock magnetic
measurements suggest SD titanomagnetites with different Ti impurities but did not show trends related to the
cooling rates. Modified Thellier experiments (MT4) were performed, showing a mean intensity of
56.3 ±0.9 μT, which is considerably too high. However, after performing the cooling rate correction
reasonable intensities of about 46 μT were obtained.
To apply the correction method to natural samples, a vertical profile through a Lava flow in Tenerife was drilled
and the natural cooling rates of the samples were estimated by relaxation geospeedometry. The structure of
glass depends strongly on its thermal history. Structural relaxation processes are expressed in properties,
such as the heat capacity (Cp) which was measured in several heating/cooling cycles across the glass
transition interval. By modelling the Cp-T curves, the natural cooling rates were obtained. They show an
unsymmetrical gradient of 10 Kmin-1 at the top and the bottom and 0.05 Kmin-1 inside the flow. The
cooling rates were used to correct the paleointensity data of the flow. The mean uncorrected paleointensities
are about 67 μT, with anisotropy correction about 69 μT and after the cooling rate correction about
57 μT for the fourteen samples from this site.
GP11C-05
Paleointensity of the Hawaii 1960 Lava Flow
Samples from the 1960 lava flow on Hawaii Island were studied using the multi-specimen parallel difference pTRM method to validate the ancient field strength of about 36 T. Several sets of specimens were treated at different temperatures between 400 and 550C. Only the lowest and highest temperatures yielded significant overestimates, while between 440 and 500C paleointensities between 32.2 and 34.2 T were obtained. At 550C the result was clearly affected by thermal alteration processes, while at 400C this possibly was caused by a cooling rate effect. Experiments with the same, thermally stabilized specimens and a laboratory TRM provided PI within the error limits, when same cooling rates were used for the full TRM and for the pTRM acquisition. Using a slower cooling rate for the TRM acquisition lead to an overestimate of the PI. Systematic variations of the unblocking temperature spectra of different specimens were observed on the between drill-core level, while such variability was much smaller between specimens of one drill-core. Based on these observations a new protocol of the multi-specimen PI method presented, that is much less affected by such rock magnetic variability. The protocol is tested on synthetic TRM and natural TRM, using specimens from the 1960 lava flow.
GP11C-06
Relative Paleointensity: a Proxy for Fluctuating Fluid Velocity in Earth's Core
GP11C-07
First Paleomagnetic Results from Holocene Alluvial Sediments in the Iberian Peninsula
We present the first paleomagnetic results from Holocene alluvial sediments that crop out in the western sector of the Ebro basin at the Bardenas Reales de Navarra natural park (N Spain). The Holocene alluvial record in this area is characterized by a sequence of four morphopedosedimentary units arranged in a cut-and- fill and nested fill terrace system. We have sampled the lowermost two units in order to test the reliability of these sediments for secular variation studies. Unit 1, which is 17 meters thick and ranges between 7150-5300 BC in age, has been sampled with a mean resolution of 25 cm. The lowermost 6 meters of Unit 2, with ages between 1700-1000 BC, have been sampled with a mean resolution of 50 cm. Thermal demagnetization of the natural remanent magnetization has enabled identification of a low-temperature component carried by magnetite. In most of the samples, an additional high-temperature component carried by hematite has been identified. Although both components are affected by inclination shallowing, they give similar secular variation trends. A comparison of these results with secular variation data derived from European lacustrine sediments and with available global geomagnetic models is presented. Our results suggest that the studied alluvial sediments might provide secular variation data of comparable quality to those of lacustrine sediments. Although these results are preliminary, they constitute the first step towards the extension of the Holocene secular variation curve for the Iberian Peninsula beyond archeomagnetic data.