Pierre Camps

Evidence for a 20° tilting of the Earth’s rotation axis 110 million years ago

True polar wander (TPW), the shift of the Earths rotation axis with respect to the entire globe, is most probably due to mass redistribution in the Earths mantle as a result of convection. Using a new rigorously selected palaeomagnetic database gathering only directions obtained from magmatic rocks, we find that TPW has been clearly intermittent over the last 200 Ma with two long periods of strict standstill from the present to 80 Ma and from approximately 150 to 200 Ma. A single period of shifting is observed, between 80 and about 150 Ma ago. This period culminates around 110 Ma ago in an 20° abrupt tilting during which an angular speed exceeding 5°/Ma (0.5m/yr) may have been reached. Assuming that the time-averaged geomagnetic field is axial, our results indicate that the changes in the position of the rotation axis, and therefore in the inertia tensor of the Earth are intermittent. We suggest that a major reorganization of the mass distribution in the Earths mantle occurred in the Lower Cretaceous. This event, concomitant with plume hyperactivity at the Earths surface and probable drastic changes at the core/mantle boundary attested by the inhibition of geomagnetic reversals, suggests unmixing of upper and lower mantle by avalanching of upper mantle material down to the core/mantle boundary. The astonishingly strict stability of the time-averaged position of the rotation axis before and after this episode of shifting implies the existence of some steady convection which does not modify the large scale distribution of mass within the mantle. Given the intermittence of mantle avalanching, we suggest that these long periods of stability correspond to the temporary reestablishment of a basically two-layered convection system within the mantle.

Increasing the efficiency of paleointensity analyses by selection of samples using first-order reversal curve diagrams

The global paleointensity database is restricted by the high failure rate of paleointensity analyses. Excluding thermal alteration, failure is usually caused by the presence of multidomain grains and interactions among grains, two properties that can be identified using first-order reversal curve (FORC) diagrams. We measured FORC diagrams on sister samples of about 200 samples that had been used for Thellier paleointensity determinations and determined criteria to discriminate samples that gave acceptable paleointensity results from those that did not. The three most discriminating criteria are the vertical spread of the FORC distribution (indicative of interactions), expressed as the full width at half maximum (FWHM), the spread of the FORC distribution along the Hc = 0 axis (width), and the bulk coercivity Hc (both indicative of domain state). Setting thresholds at 132 mT for the width of the distribution and 29 mT for the FWHM maximizes the number of unsuccessful rejected samples. Using an additional threshold of Hc = 5.4 mT results in rejection of 32% of unsuccessful samples. Seven samples that barely satisfy the paleointensity selection criteria would also be rejected using these selection criteria. Most of the samples that fail the paleointensity experiment without being detected by our selection criteria have ideal noninteracting single-domain magnetic properties but fail because of the thermal alteration that results from repeated heating. Being able to eliminate at least one third of unsuccessful samples using our FORC diagram-based prescreening procedure should provide a significant improvement in efficiency of paleointensity measurements.

Mono Lake or Laschamp geomagnetic event recorded from lava flows in Amsterdam Island (southeastern Indian Ocean)

SUMMARY We report a survey carried out on basalt flows from Amsterdam Island (Southeastern Indian Ocean) in order to check the presence of intermediate directions interpreted as belonging to a geomagnetic field excursion within the Brunhes epoch, completing this palaeomagnetic record with palaeointensity determinations and radiometric dating. Because the palaeomagnetic sampling was done over a few hours during the resupply of the French scientific base Martin du Viviers by the Marion Dufresne vessel, we could collect only 29 samples from four lava flows. The directional results corroborate the findings by Watkins & Nougier: normal polarity is found for two units and an intermediate direction, with associated virtual geomagnetic poles (VGPs) close to the equator, for the other two units. A notable result is that these volcanic rocks are well suited for absolute palaeointensity determinations. 50 per cent of the samples yields reliable intensity values with high-quality factors. An original element of this study is that we made use of the thermomagnetic criterion PTRM-tail test of Shcherbakova et al. to help in the interpretation of the palaeointensity measurements. Doing this, only the high-temperature intervals, beyond 400 ◦ C, were retained to obtain the most reliable estimate of the strength of the ancient magnetic field. However, not applying the PTRM-tail test does not change the flow-mean values significantly because the samples we selected by conventional criteria for estimating the palaeointensity carry only a small proportion of their remanence below 400 ◦ C. The normal units yield virtual dipole moments (VDM) of 6.2 and 7.7 (10 22 Am 2 ) and the excursional units yield values of 3.7 and 3.4 (10 22 Am 2 ). These results are quite consistent with the other Thellier determinations from Brunhes excursion records, all characterized by a decrease of the VDM as the VGP latitude decreases. 40 Ar/ 39 Ar isotopic age determinations provide an estimate of 26 ± 15 and 18 ± 9 kyr for the transitional lava flows, which could correspond to the Mono Lake excursion. However, the large error bars associated with these ages do not exclude the hypothesis that this event is the Laschamp.

A Statistical Model of the Fluctuations in the Geomagnetic Field from Paleosecular Variation to Reversal

The statistical characteristics of the local magnetic field of Earth during paleosecular variation, excursions, and reversals are described on the basis of a database that gathers the cleaned mean direction and average remanent intensity of 2741 lava flows that have erupted over the last 20 million years. A model consisting of a normally distributed axial dipole component plus an independent isotropic set of vectors with a Maxwellian distribution that simulates secular variation fits the range of geomagnetic fluctuations, in terms of both direction and intensity. This result suggests that the magnitude of secular variation vectors is independent of the magnitude of Earths axial dipole moment and that the amplitude of secular variation is unchanged during reversals.

Geomagnetic paleosecular variation recorded in Plio-Pleistocene volcanic rocks from Possession Island (Crozet Archipelago, southern Indian Ocean)

Possession Island, in the Crozet Archipelago, consists of volcanic units erupted mainly between ~5 and 0.5 Ma. A paleomagnetic sampling was carried out along several sections distributed near the northern, eastern, and southeastern coasts. A total of 45 independent flows were sampled (320 samples). For each flow a precisely defined characteristic remanence direction was usually isolated after a careful progressive cleaning in alternating fields. However, particularly complex remanence behavior is often observed. The magnetostratigraphy of the lava pile is quite simple, with reversed rocks in the lower part and normal units in the upper part of two sections. A third section is of normal polarity throughout its whole thickness, including three excursional directions. We did not find any intermediate directions between the normal and reverse magnetozones. Thus we have no evidence for the recording of the Matuyama-Brunhes transition expected from a previous study (Watkins, 1972). The amplitude of paleosecular variation, estimated from between-flow dispersion from the field of an axial dipole, is 11.8 with 95% confidence limits between 9.3 deg and 14.0 deg. This value is consistent with the general anisotropic statistical model for paleosecular variation of Constable and Johnson, 1999.

Transitional geomagnetic impulse hypothesis: Geomagnetic fact or rock-magnetic artifact?

A striking feature of the Steens Mountain (Oregon) geomagnetic polarity reversal is the two (maybe three) extremely rapid field directional changes (6 degrees per day) proposed to account for unusual behavior in direction of remanent magnetization in a single lava flow. Each of these very fast field changes, or impulses, is associated with a large directional gap (some 90°) in the record. In order to check the spatial reproducibility of the paleomagnetic signal over distances up to several kilometers, we have carried out a paleomagnetic investigation of two new sections (B and F) in the Steens summit region which cover the second and the third directional gap. The main result is the description of two new directions, which are located between the pre second and post second impulse directions. These findings weigh against the hypothesis that the geomagnetic field cause the unusual intraflow fluctuations, which now appears to be more ad hoc as an explanation of the paleomagnetic data. However, the alternative baking hypothesis remains also ad hoc since we have to assume variable rock magnetic properties that we have not yet been able to detect within the flows at the original section Steens A and D 1.5 km to the north. In addition, new results for 22 transitional and normal lava flows in section B are presented that correlate well with earlier results from section A.

Paleomagnetic and geochronological study of a geomagnetic field reversal or excursion recorded in pliocene volcanic rocks from Georgia (Lesser Caucasus)

A 250-m thick volcanic sequence of some 63 Pliocene lava flows from Georgia (Akhalkalaki plateau) which was previously thought to record the Gauss-Matuyama reversal was re-studied in detail. About 400 oriented cores were sampled along four sub-vertical sections partially overlapping each other. At the base of the section, 16 consecutive lava flows were found to record intermediate directions. The overlying flows are all reversely magnetized. With the exception of the upper part of the sequence, four groups of three to 11 consecutive lava flows with strictly the same paleodirection can be recognized. The location of the intermediate virtual geomagnetic poles (VGPs) does not correspond to that of the VGP clusters previously proposed as indicating a long lasting near dipole configuration during reversals. The field paleointensity is close to one fifth of the present field at the bottom of the section and tends to increase uphill. Age determination (40Ar39Ar) provides an estimate of 3.6 ± 0.06 Ma, which leaves us with two possible interpretations: the transitional vectors could correspond either to an excursion within chron 2Ar or to the upper (N-R) Cochiti-Gilbert reversal.

On the features of the geodynamo following reversals or excursions: by absolute geomagnetic paleointensity data

We carried out a Thellier paleointensity study of a ∼3.6 million year Pliocene geomagnetic excursion recorded in a lava flow succession from southern Georgia (lesser Caucasus). Previous paleomagnetic study [Phys. Earth Planet. Int. 96 (1996) 41] revealed that several consecutive lava flows record an intermediate polarity direction at the base of the section followed by a thick reverse polarity zone. Samples of 71 from 26 flows from both polarity zones were pre-selected for paleointensity experiments because of their low viscosity index, stable remanent magnetisation and close to reversible continuous thermomagnetic curves. Altogether 54 samples from 21 flows yielded reliable paleointensity estimates. The mean paleointensity of the intermediate field is 7.8 ± 2.4T (three flows). The stable polarity paleointensity is higher with a mean 24.2 ± 8.2T (15 flows), which corresponds to a mean virtual dipole moment (VDM) of 4.6 ± 1.8 × 10 22 Am 2 . This value is significantly lower than the average Pliocene geomagnetic dipole moment and post-intermediate dipole moments recorded in volcanic sequences at Hawaii (∼4 Ma) and Steens mountain (∼16.2 Ma). However, our results are quite similar to the post-intermediate field recorded in Iceland during the Gauss–Matuyama reversal. These results suggest that the regime of the geodynamo following reversals or excursions may vary significantly from one case to the next without any apparent systematic features.

Revisiting the Jurassic geomagnetic reversal recorded in the Lesotho Basalt (Southern Africa)

SUMMARY We carried out a detailed and continuous palaeomagnetic sampling of the reversed to normal geomagnetic transition recorded by some 60 consecutive flow units near the base of the Lesotho Basalt (183 ± 1 Ma). After alternating field or thermal cleaning the directions of remanence are generally well clustered within flow units. In contrast, the thermal instability of the samples did not allow us to obtain reliable palaeointensity determinations. The geomagnetic transition is incompletely recorded due to a gap in volcanic activity attested to both by eolian deposits and a large angular distance between the field directions of the flows underlying or overlying these deposits. The transition path is noticeably different from that reported in a pioneering work carried out in 1962. The most transitional virtual geomagnetic poles (VGPs) are observed after the volcanic hiatus. Once continents are replaced in their relative position 180 Myr, the post-hiatus VGPs cluster over Russia. However, two successive rebounds from that cluster are found, with VGPs reaching repeatedly the Eastern Asia coast. Thus, the VGP path is not narrowly constrained in palaeolongitude. The decrease in intensity of magnetization as the field deviates from the normal or reversed direction suggests that the decrease in field magnitude during the reversal reached 80‐90 per cent. We conclude that although the reversal is of a dipole of much weaker moment than that which existed on average during Cenozoic time, the characteristics of the reversing geodynamo seem to be basically similar.

Post-emplacement tilting of lava flows inferred from magnetic fabric study: the example of Oligocene lavas in the Jeanne d’Arc Peninsula (Kerguelen Islands)

Abstract Statistical analysis of the magnetic fabric of samples from several successive lava flows emplaced under similar conditions can allow determination of the mean flow direction when magnetic fabric data from individual flows do not lead to reliable results. A difference between the obtained flow direction and the present dip direction indicates that the flows were tilted after emplacement. For 2 successive series of flows on the Jeanne d’Arc Peninsula presently NNW dipping, this method shows lava emplacement along a SSW–NNE direction. This indicates a gentle tilting acquired during a period of weak deformation in the whole archipelago. Additionally, the magnetic fabric data allow the reconstruction of the different conditions of emplacement of these two series of lava flows and of formation of the local thick conglomerate interbedded between these series.