Geomagnetism and Paleomagnetism [GP]

GP51B MCW:3007 Friday 0800h

Earth's Magnetic Field Variability, Links With Orbital/Rotational Motions, and Paleoclimate: Data and Models I

Presiding:G Acton, University of California, Davis; C Constable, IGPP, Scripps Institution of Oceanography, University of California, San Diego

GP51B-01 INVITED

Millennial- to Centennial-Scale Changes in Earth's Magnetic Field Intensity and Direction From High Resolution Marine and Lacustrine Sedimentary Records: Implications for a Geomagnetism-Climate Relationship?

* St-Onge, G (guillaume_st-onge@uqar.qc.ca) , ISMER and GEOTOP, 310 allee des Ursulines, Rimouski, QC G5L 3A1 Canada
Stoner, J (jstoner@coas.oregonstate.edu) , Oregon State University, College of Oceanic and Atmospheric Sciences (COAS) 104 COAS Administration Building, Corvallis, OR 97331-5503, United States

It is well established that the geomagnetic field shields the Earth from cosmic rays, and that variation in its strength controls the long-term (>10$^{3}$ yr) production rate of cosmogenic isotopes such as $^{10}$Be, $^{14}$C and $^{36}$Cl. Based on relative paleointensity estimates from well-dated and high-sedimentation rate marine and varved lake sediment cores, it was also recently suggested that millennial-scale variations in Earth's magnetic field intensity could have modulated the production of cosmogenic isotopes through the Holocene and into the Late Pleistocene. Here, we will reinforce that notion by showing that high-resolution Holocene/Late Pleistocene marine paleointensity records from Eastern Canada, Greenland and as far afield as the Chilean Margin depict millennial- to centennial-scale variability that can be correlated to changes in the flux of $^{10}$Be from the GRIP/GISP2 ice core and to the $\Delta$$^{14}$C residual record derived from tree-ring data. This implies that the paleoclimatic inferences drawn from cosmogenic isotope studies alone could be misleading. In addition, new paleomagnetic secular variation studies from the vicinity of Iceland suggest that geomagnetic field morphology may also play a significant role. A more complete description of the space/time variations of the geomagnetic field is therefore required to assess the possible relationships between geomagnetism and climate.

GP51B-02

Does the Earth's Magnetic Field Influence Climate?

Fluteau, F (fluteau@ipgp.jussieu.fr) , Institut de Physique du Globe, Place Jussieu, Paris, 75005 France
* Courtillot, V (courtil@ipgp.jussieu.fr) , Institut de Physique du Globe, Place Jussieu, Paris, 75005 France
Gallet, Y (gallet@ipgp.jussieu.fr) , Institut de Physique du Globe, Place Jussieu, Paris, 75005 France
Le Mouel, J (lemouel@ipgp.jussieu.fr) , Institut de Physique du Globe, Place Jussieu, Paris, 75005 France
Genevey, A (genevey@ipgp.jussieu.fr) , Centre de Recherche et de Restauration des Musees de France, Palais du Louvre, Paris, 75001 France

Much of the observed increase in global surface temperature over the past 150 years occurred prior to the 1940's and after the 1980's. The main agents which are invoked are solar variability, changes in atmospheric greenhouse gas content or sulfur, due to natural or anthropogenic action, or internal variability of the coupled ocean-atmosphere system. Magnetism has seldom been invoked, and evidence for connections between climate and magnetic field variations have received little attention. We review evidence for such connections, starting with suggested correlations, on three time scales: recent secular variation (10-100 years), historical and archeomagnetic change (100-5000 years) and excursions and reversals (1000-1 million years). We attempt to suggest which mechanisms could account for observed correlations. Evidence for correlations in field intensity changes, excursions and reversals, which invoke Milankovic forcing in the core, either directly or through changes in ice distribution and moments of inertia of the Earth, is still tenuous. Correlation between decadal changes in amplitude of geomagnetic variations of external origin, solar irradiance and global temperature is stronger. The correlation applies until the 1980's, suggesting that solar irradiance is the prime forcing function of climate until then, when the correlation breaks and anomalous warming may emerge from the signal. Indeed, only solar flux of energy and particles can jointly explain parallel variations in temperature and external magnetic field. The most intriguing feature may be recently proposed archeomagnetic jerks (see abstract by Gallet et al). These seem to correlate with significant climatic events. A proposed mechanism involves tilt of the dipole to low latitudes, resulting in enhanced cosmic-ray induced nucleation of clouds. Intense data acquisition over a broad range of durations is required to further probe these indications that the Earth's and Sun's magnetic fields may have significant bearing on climate change at various time scales.

GP51B-03

Eastward and westward drift of the Earth's magnetic field for the last three Millennia

* Dumberry, M (m.dumberry@earth.leeds.ac.uk) , School of Earth and Environement, University of Leeds, Leeds, LS2 9JT United Kingdom
Finlay, C C (cfinlay@erdw.ethz.ch) , Institut fur geophysik, ETH, Zurich, CH-8093 Switzerland

We analyse the secular variation captured by the archaeomagnetic field model CALS7K.2 in an effort to determine episodes of eastward and westward drift of Earth's magnetic field over the past 3000 years. The direction, amplitude and geographical distribution of azimuthal motion of field features at the core-mantle boundary are described. We find that the clearest azimuthal motions are observed at mid to high latitudes in the Northern hemisphere, where both eastward and westward motion occur. These azimuthal motions correspond to displacements and distortions of the two main, quasi-stationary, high latitude magnetic flux patches. Similar motions are not observed in the Southern hemisphere, although this may be a consequence of the poorer data coverage there. The globally averaged drift has been westward since 1400 AD, was eastward between 800 AD and 1400 AD, and generally westward prior to 800 AD except perhaps for a short episode of eastward motion near 100 AD. Under Europe, the times of transition in the direction of the drift coincide with the times at which ``archaeomagnetic jerks'' have been reported. Our results suggest that these are caused by a relatively rapid ( < 100 yr) change in the direction of the underlying azimuthal flow near the core surface. The amplitude and time-dependency of the mean azimuthal drift recovered by our method are generally consistent with the changes in core angular momentum required to explain the 1500-yr variation in the length of day (LOD) reconstructed from ancient records of eclipses. Though a detailed match in not obtained, it suggests nevertheless that core-mantle coupling is responsible for the 1500-yr changes in LOD. The equatorial westward drift observed during the past 4 centuries appears to have been generally present over the past 3000 years. When it is observed, it is restricted to the Atlantic hemisphere, as is the case for the present-day westward drift, which we interpret as a signature of core-mantle thermal coupling.

GP51B-04

The evolution of the core-surface flow and the length of day variation over the last seven thousands years

Korte, M (monika@gfz-potsdam.de) , GFZ Potsdam, Section 2.3, GFZ Potsdam, Telegrafenberg, Potsdam, 14473 Germany
* Wardinski, I (ingo@gfz-potsdam.de) , GFZ Potsdam, Section 2.3, GFZ Potsdam, Telegrafenberg, Potsdam, 14473 Germany

We present the results of our studies about the core-surface flow over the last seven thousands years. Our analysis is based on the CALS7k.2, a time--dependent model of the Earth's magnetic field and its secular variation for the period 5000 B.C to 1950. This model relies on indirect observation of the Earth's magnetic field taken from natural archives (lake sediments) and archaeological artifacts. In order to compute the core--surface motion we invert the magnetic induction equation, which relates the magnetic field at the CMB, its temporal field variation and the fluid motion atop the core. We seek instantaneous flow solutions without considering the effect of magnetic diffusion. Invoking magnetic diffusion in the inversion scheme to gain a time dependent fluid flow requires information about the morphology and the state of the toroidal field at the CMB. These information are unavailable. However, instantaneous flow solutions may also be appropriate to evaluate processes at the CMB. These flow solutions suggest that the core-surface flow undergoes different regimes of zonal flow direction. Epochs of a mainly westward flow alternate with epochs of eastward flow. Also, the mean velocity of the flow seems to fluctuate. These changes in the flow direction seem to recur periodically, with a period of about 500 -- 800 years. While the mechanism causing these variations is unassigned, they should also reflect in the length of day variation, where due to core--mantle coupling angular momentum is transfered from the core to the mantle. Much about the exchange of angular momentum on millennial timescales is yet unknown. A comparison between the deduced length of day variation from historical eclipses and the prediction of our flow model shows a significant mismatch. We argue that this mismatch can be indirectly attributed either to millennial scale tidal forcing of the moon, or to solar radiation variability. Both mechanisms have been suggested to influence the Earth's climate with periods of 1500 to 1800 years. The long term climate variations cause fluctuations in the mass distributions at the Earth's surface, which in turn alter the length of day.

GP51B-05

Geomagnetic 100-kyr Variation Extracted From Sedimentary Paleointensity Records in the Pacific Ocean

Yokoyama, Y (yokoyama@big.ous.ac.jp) , Department of Biosphere-Geosphere System Science, Faculty of Informatics, Okayama University of Science, Ridai-cho 1-1, Okayama, 700-0005 Japan
* Yamazaki, T (toshi-yamazaki@aist.go.jp) , Institute of Geology and Geoinformation, Geological Survey of Japan, AIST, 1-1-1 Higashi, Tsukuba, 305-8567 Japan
Oda, H (hirokuni-oda@aist.go.jp) , Institute of Geology and Geoinformation, Geological Survey of Japan, AIST, 1-1-1 Higashi, Tsukuba, 305-8567 Japan

Recent paleomagnetic studies suggest timescales of geomagnetic variation close to the Earth�fs orbital elements (e.g., Channel et al., 1998; Yamazaki, 1999). This suggests a possible energy source of geomagnetic field from outside of the Earth's core. To study this problem in detail, Yokoyama and Yamazaki (2000) analyzed paleointensity series from 100 to 600 ka obtained from 5 sediment cores in a wide area of the Pacific Ocean, and found a paleosecular variation with 100-kyr time scale. We here analyzed another three sets of paleointensity series in a longer duration, from 300 to 1600 ka, to confirm an existence of the variation. The data series are from the equatorial Pacific sediment cores MD982185 (3N, 135E) (Oda and Yamazaki, 2002) and MD982187 (4N, 135E) (Oda and Yamazaki, 2005), and from the north Pacific sediment core KR0310-PC1 (35N, 175E) (Yamazaki et al., in preparation). We interpolated the series into 2-kyr intervals, and transformed the series into a wavelet space in a similar way to Yokoyama and Yamazaki (2000). We first checked an influence of rock-magnetic properties on the relative intensity. We calculated wavelet correlation between the series of the paleointensity and the rock-magnetic parameters of each core. The wavelet correlation coefficients of the three cores are small so that the change of relative intensity in each core is independent from that of magnetic properties. In order to confirm the independence of the relative intensity variation, we secondly calculated wavelet correlation among the three cores. As the results, the relative intensity variations of the three cores have correlation, and rock-magnetic parameters do not have correlation. This is because that phases of the rock-magnetic variations are different by cores though there are common scale variations. On the other hand, relative intensity variations are in-phase, and these must not be originated in the changes of rock-magnetic properties. Because the extracted variation of the paleointensity series is not rock-magnetically contaminated and because the variation is observed in a wide area, we conclude that the variation is of global geomagnetic field.

GP51B-06

Geomagnetic Dipole Lows and Excursions of the last 800 ka : connection with Interglacials and/or low Obliquity Times?

* THOUVENY, N (thouveny@cerege.fr) , CEREGE-CNRS and Aix-Marseille University, Europole de l'ARBOIS BP80, AIX EN PROVENCE, 13545 France
CARCAILLET, J (Julien.Carcaillet@ujf-grenoble.fr)
SARACCO, G (saracco@cerege.fr)
BOURLES, D L (bourles@cerege.fr)

Paleomagnetic directions, relative paleointensities (RPI) and authigenic 10Be/9Be ratio were measured along sedimentary clayey-carbonate sequences in high accumulation rate sites of the Portuguese margin (0-400 ka BP) and West-Equatorial Pacific (600-1300 ka BP. Series of high and low RPI features are placed on the chronological scale using C-14 ages, using correlations with of delta O-18 records with the Greenland ice cores and SPECMAP records, and using the ages of polarity reversals. During the time intervals of dramatically low RPI anomalous paleodirections document excursions or polarity reversals. Significant peaks of the authigenic 10Be/9Be ratio point in stratigraphic layers recording all low RPI phases. Plotted against RPI data the 10Be/9Be ratios statistically follow the expected power law (Elsasser et al. ,1958 and Lal, 1988), which strongly establishes the unique and direct link between the recorded cosmogenic enhancement and dipole moment loss, allowing us to univocally interpret our 10Be/9Be ratio and RPI records in terms of geomagnetic dipole moment lows and highs (DML and DMH) alternation. delta O-18 records of the same cores (e.g. Abreu et al. 2004), provide the frame to interpret dipole moment variations in a paleoclimatic context within strict stratigraphic terms. We note that most DML of the last 400 ka fall in the end of interglacial stages, while DMH are rather related with full glacials. We then confirm this coincidence, though whithout strict stratigraphic control, by comparing the SINT-800 curve (Guyodo and Valet, 1999) and the S. E. Pacific near Sea Floor mag. record (Gee et al., 2000), with the highest resolution ÃïÂ18O record yet available (Bassinot et al. 1994). Complex Wavelet analyses using modulus and phase reveal that the geomagnetic moment proxy records contain a maximum power for periods between 30 and 100 ka. DML do not occur in any fixed eccentricity context, but several DML fall at the time of obliquity minima. The comparison of phases of the SINT-800 and Indian Ocean ÃïÂ18O records, at the 100 ka period, indicates that the ice volume fluctuations lead the geomagnetic moment fluctuations. These observations shed the light on mechanical couplings via Earth's rotation rate and/or orbital changes : 1) Alternations of ice accretion/melting at high and mid-latitude may have resulted in acceleration/deceleration of the Earth's rotation (Doake, 1977), although the mass of the 130 m equivalent sea level only represents 7 ppm of the Earth mass. However, acceleration/braking effects generated by tides friction on large/reduced continental platform at low/high sea level conditions (in glacial/interglacial conditions resp.)may have contribute. A coupling between Earth's rotation and geomagnetic regime was suggested for the 20th century (Courtillot and Le Mouel, 1984; Jing, 1992). An alternative Ãâ‚“ or complementary- explanation can be envisaged: lots of DML occured at the time of low obliquity, i.e. at the time of low angle precession, (rejoining Fuller's (2006)observation and re- introducing the long standing debate about possible forcing by astronomical precession on the geodynamo (Malkus, 1963,1968, Rochester, 1976). It has to be emphasized that because deglaciation are explained by maximum summer insolation of the Northern hemisphere (i.e. at high obliquity), dipole lows occurring at low obliquity, would logically lagg the deglaciation by  obliquity period (20.5 ka), and thus occur near the end of interglacials.

GP51B-07

Geomagnetic Links to Climate Change and Orbital Cycles

* Acton, G (acton@geology.ucdavis.edu) , University of California, Davis, Department of Geology, Davis, CA 95616, United States

Years of speculation, newly recognized mechanisms for interactions, and a sparse but expanding number of observations support some form of link between geomagnetic field variability and climate change and/or orbital cycles. Early paleomagnetic observations only hinted at the links and failed to withstand scrutiny for a number of reasons including poor data quality, poor age control, poor resolution of short-term geomagnetic directional variability over sufficiently long time periods, and a reliance on relative paleointensity records. Even though Milankovitch periodicities have been observed in the latter, proving that these are not influenced by climatically induced lithologic changes rather than by geomagnetic field variability is difficult. At this point, the speculation has been more interesting that the evidence has been convincing. New long continuous records of short-term paleomagnetic directional variability that span the past 1 m.y., however, show intriguing correlations of geomagnetic excursions with precession cycles and with deglacials. The changes in directions for these excursions are too large to be attributed to lithologic variations nor can they be attributed to local sedimentary or tectonic processes as the excursion are observed regionally or globally. Although such correlations might have been regarded as fortuitous in the past, age constraints have improved significantly by obtaining stable isotope records or other climate proxies directly from the same stratigraphic sections as the geomagnetic records. Furthermore, speculation about mechanisms for geomagnetic links to climate and orbital cycles have been succeeded by climate studies that have found that cloud formation is associated with the amount of cosmogenic radiation, which is largely controlled by the geomagnetic field. Similarly, precession had been disregarded as a driving force for the geodynamo, but recent modeling shows that such conclusions were premature. Thus, causal relationships between geomagnetic field variability, climate change, and orbital cycles are not unexpected nor are they unobserved.

GP51B-08

Testing the Influence of Orbital Cycles on Paleointensity Records and Timing of Reversals and Excursions

* Xuan, C (xuan2005@ufl.edu) , Department of Geological Sciences, University of Florida, 241 Williamson Hall, P.O. Box 112120, Gainesville, FL 32611, United States
Channell, J E (jetc@geology.ufl.edu) , Department of Geological Sciences, University of Florida, 241 Williamson Hall, P.O. Box 112120, Gainesville, FL 32611, United States

Orbital cycles with 100 kyr and 41 kyr periods, detected in sedimentary normalized remanence (relative paleointensity) records by power spectral analysis or wavelet analysis, have been attributed either to orbital control on the geodynamo, or to lithologic contamination and hence ineffectiveness of the normalization procedure. A test is designed to determine the significance level for peaks in the power spectra by modeling the "noise" using an auto-regressive (AR) function derived from individual paleointensity records after extraction (filtering) of the dominant (orbital) power peaks. A Monte Carlo method is then used to determine the 90%, 95%, and 99% confidence level. The results, for a number of North Atlantic paleointensity records, indicate that the 100 kyr and 41 kyr power is often significant at the 99% confidence level. The significance tests are then extended to the wavelet analysis of the ~2 Myr paleointensity records from ODP Sites 983 and 984. Wavelet power spectra of these two records indicate the presence of significant power at 100 kyr and 41 kyr. An apparently significant transfer of power from a dominant 41 kyr period to a 100 kyr period occurs at about 800 ka. As this is a characteristic of many Pleistocene climate records, and is also exhibited in wavelet power spectra of the normalizers such as anhysteretic remanent magnetization (ARM), it implies that the orbital power in the normalized remanence records is a lithologic contamination due to incomplete normalization of the natural remanent magnetization (NRM). Following Fuller (2006), we test the relationship between the phase of orbital obliquity and eccentricity, and reversal/excursion ages. For the last 5 Myr, a Rayleigh test of circular uniformity suggests that the phases at which reversals and excursions occur are uniformly distributed at 95% confidence level, however this analysis is obviously influenced by reversal/excursion age uncertainties. Over the last 20 Myr, a similar result is obtained. A test for the phase difference between the maximum obliquity envelope and reversal age for the last 5 Myr indicates that the phases are, however, not uniformly distributed but follow a Von Mises distribution with a mean value of 103.7\deg, and 95% confidence interval of (53.16\deg, 154.15\deg), implying that reversals preferentially occurred during decrease of the maximum obliquity envelope in the last 5 Myr. For the last 20 Myr, two recent timescales imply a uniform distribution (at the 95% confidence level) for the phases of reversal ages relative to the maximum obliquity envelope.