Spectral analysis of Pennocarboniferous geomagnetic variation data from glacial rhythmites

Summary. Palaeomagnetic results for a sequence of Permocarboniferous rhythmites presented in the previous paper have been submitted to maximum entropy spectral analysis to test whether these palaeomagnetic data could supply information on geomagnetic variations. There is a good correlation between the thickness of the rhythmites and sunspot spectra, suggesting that these sediments are really seasonal. The palaeomagnetic spectra are compared with those of observatory records. Periods of approximately 24.4, 12.4, 8.6, 6.7 and 5.5 found for palaeomagnetic data have corresponding values in the geomagnetic spectrum. Most of these periods, however, are the same as those found in the thickness data, implying that magnetization can be influenced by the sedimentation process as suggested by other investigators. On the other hand, both geomagnetic and climatic (thickness) variations seem to be related to solar activity. Therefore, at least indirectly, palaeomagnetic data may reflect geomagnetic variations.     

An investigation of the magnetic variation within two recent lava flows

Samples collected in vertical sections from two recent (1169 AD and 1971 AD) lava flows show significant variations in both the palaeomagnetic vector and the rock magnetic characteristics. The age and the short cooling time of the flows (both are less than 3 m thick) are such that the observed range of palaeomagnetic variation is more than can reasonably be explained by secular variation. The variations in palaeomagnetic and rock magnetic characteristics appear in part to be a response to the relative proportions of two coexisting ferrimagnetic phases, one that displays a primary coarsegrained and Ti-rich composition and a second that has finer-grained, Ti-poor characteristics. The use of AF demagnetization in this study means that the latter phase will have a disproportionate influence on the palaeomagnetic results. It has been suggested elsewhere that in such a magnetic system the Ti-poor phase may represent a CRM formed at some temperature below its final Curie temperature; if so, the results of this study imply that not only does it affect the palaeointensity estimate (which requires a pure TRM) but it appears also to influence the palaeomagnetic direction to an extent that far exceeds the normal secular change that might be expected to occur during delayed acquisition. Such a result has important implications for palaeomagnetic studies that use lavas, both in the selection of a suitable sampling strategy and in the geomagnetic interpretation of palaeomagnetic variation within single flow units.     

The geomagnetic field over the past 5 million years

The modern geomagnetic field is usually expressed as a spherical harmonic expansion. Although the palaeomagnetic record is very incomplete in both space and time, sufficient data are available from a span of ages to generate time-averaged spherical harmonic field models with many degrees of freedom. Here three data sets are considered: directional measurements from lavas, inclination measurements from ocean sediments, and intensity measurements from lavas. Individual data are analysed, as well as site-averages, using the same methods that have been developed for the modern field, to give models for the past 5 Myr. The normal-polarity field model has an axial-dipole intensity similar to that of the modern-day field, whilst the equatorial-dipole component is very much smaller. The field is not axisymmetric, but shows flux concentrations at the core's surface under Canada and Siberia similar to those observed in the field over historical timescales. Tests on synthetic data show that it is unlikely that these similarities result from the overprinting of the palaeomagnetic field due to inadequate cleaning of the samples. The reverse-polarity field model does not show such obvious features, but this may be due to the sparsity of the data.
The patterns observed in the normal-polarity field, with persistent features in the northern hemisphere and a smooth southern hemisphere, could be explained if the present pattern of secular variation is typical of the past several million years. This would reveal itself as large variations over time in the direction of the magnetic vector in regions of high secular variation, with relatively little change over quieter regions. However, we have been unable to find any evidence for a geographical pattern of secular variation in the data.     

Geomagnetic variations recorded in Older (≫ 23 000 B.P.) and Younger Yoldia Clay (∼ 14 000 B.P.) at Nørre Lyngby, Denmark

Summary. Detailed palaeomagnetic results from a rapidly deposited 12.5-m Younger Yoldia Clay sequence of age around 14 000 B.P. at the coast cliff at Nørre Lyngby (northern Jutland, Denmark) and a 2-m Older Yoldia Clay sequence of age somewhere between 23 000 and 40 000 B.P. are presented and discussed. The Younger Yoldia Clay spans some 1000–1500 yr and shows swings in inclination and declination of about that period, and also more rapid oscillations which are particularly marked in inclination, showing that rapid secular variations as have occurred during historic times were indeed also present back in time. There exist easterly declinations of 80° to 90° in the upper half of the Younger Yoldia Clay which cause the virtual geomagnetic pole to migrate clockwise to around 50° away from the rotation pole. This we name the 'Nørre Lyngby declination excursion'.
In the Older Yoldia Clay, as well as secular variations in both declination and inclination, significant low inclination values are found, confirming the existence of the recently named 'Rubjerg low inclination excursion', with the virtual geomagnetic pole moving first in a clockwise then in an anticlockwise sense at 40° to 60° away from the rotation pole.
It is therefore inferred that models for the Earth's geomagnetic field should involve at least local rapid eastward as well as westward 'drift' of the non-dipole field components at various times in the past.     

Palaeomagnetic record in Late Pleistocene and Holocene dry lake deposits at Tlapacoya, Mexico

Summary. A palaeomagnetic investigation of Carbon-14-dated marsh and near-shore lacustrine sediments deposited between about 25000–5000 yr bp at Tlapacoya, Mexico, reveals normal polarity of the geomagnetic field in all samples measured. At one site, anomalous palaeomagnetic directions in a mud unit dated about 14500 yr bp raised the possibility of a geomagnetic excursion, but subsequent work at six additional sites in the unit revealed no abnormal directions. Thus the anomalous directions are most likely not a true reflection of geomagnetic field behaviour, although no specific alternative explanation is entirely convincing. The preliminary Tlapacoya data of anomalous directions have been cited by others as positive evidence for an excursion. We strongly recommend it no longer be considered as such.     

Preliminary palaeomagnetic results of an Archaean dolerite dyke of west Greenland: geomagnetic field intensity at 2.8 Ga

The geomagnetic field intensity during Archaean times is evaluated from a palaeomagnetic and chronological study of a dolerite dyke intruded into the 3000 Ma Nuuk Gneisses at Nuuk (64.2°N, 51.7°W), west Greenland. Plagioclase from the dolerite dyke yields a mean K-Ar age of 2752 Ma. Palaeomagnetic directions after thermal demagnetization of the dyke and the gneiss reveal a positive baked-contact test, indicating that the high-temperature-component magnetization of the dyke is primary. Thellier experiments on 12 dyke specimens yield a palaeointensity value of 13.5±4.4 μT. The virtual dipole moment at ca. 2.8 Ga is 1.9±0.6 × 10²² Am², which is about one-quarter of the present value. The present study and other available data imply that the Earth's magnetic field at 2.7 ∼ 2.8 Ga was characterized by a weak dipole moment and that a fairly strong geomagnetic field similar to the present intensity followed the weak field after ca. 2.6 Ga.     

Long-term features of drifting and standing non-dipole fields as determined from Holocene palaeomagnetic secular variations

Geomagnetic field motions of Holocene secular variations are investigated using a separation method. The palaeomagnetic secular variations from Britain, North America and Australia have been subjected to maximum-entropy method analyses. Based on the results of spectral analyses, the secular variations are separated by band-pass filters into low-frequency components, generally including the period band 1800-3600 yr, and high-frequency components, generally including the period band 1000-1200 yr. There is an interval, from 4200 to 1700 yr BP, which shows clockwise rotational motions in the low-frequency components of all three sites. Westward drifting of geomagnetic fields may be globally dominant. Swinging or elliptical looping motions constrained to a certain direction were observed in the low-frequency components of the British data. The time duration for the persistence of the swinging motion constrained to a certain direction was 3500 years or so, which could be the lifetime of an oscillating stationary field. The duration of the transitional motion was 1000-1300 years, which may indicate the recurrence time of a stationary field.     

Palaeosecular variation over the past 5 Myr based on a new generalized database

A generalized database of global palaeomagnetic data from 3719 lava flows and thin dykes of age 0–5 Ma has been constructed for use with a relational database management system. The database includes all data whose virtual geomagnetic poles (VGP) lie within 45 of the spin axis and can be used for studies of palaeosecular variation and for geomagnetic field modelling. Because many of these data were collected and processed more than 15–20 years ago, each result has been characterized according to the demagnetization procedures carried out. Analysis of these data in terms of the latitude variation of the angular dispersion of VGPs (palaeosecular variation from lavas) strongly suggests that careful data selection is required and that many of the older studies may need to be redone using more modern methods. Differences between the angular dispersions for separate normal- and reverse-polarity data sets confirm that many older studies have not been adequately cleaned magnetically. Therefore, the use of the database for geomagnetic field modelling should be carried out with some caution. Using a VGP cut-off angle that varies with latitude, the best data set consists of 2636 results that show a smooth increase of VGP angular dispersion with latitude. Model G for palaeosecular variation, which is based on modelling of the antisymmetric (dipole) and symmetric (quadrupole) dynamo families, provides a good fit to these results.     

A physical model for palaeosecular variation

Summary. A new model to describe the latitude dependence of the angular dispersion of the palaeomagnetic field (palaeosecular variation) is developed following previous models, but with crucial differences. It is shown that if the probability distribution of virtual geomagnetic poles (VGPs) is circularly symmétric about the rotation axis then the geométry of the distribution of field directions is latitude dependent. This has a significant effect on the latitude dependence of dispersion and is accounted for in the model. The dipole and non-dipole parts of the field are not artificially separated but are intimately linked through an observationally determined relation that the time averaged intensity of the non-dipole field is dependent upon the intensity of the dipole field. It is shown that a consequence of this relation is that no knowledge of the probability distribution of the geomagnetic dipole moment is required. This is a fundamental improvement over previous models.
The model provides excellent fits to the palaeodata and, unlike previous models, is not inconsistent with the latitude variation of the non-dipole field dispersion determined from the present field. For the past 5 Ma the point estimate of the VGP dispersion due to dipole wobble is 7.2° and of the VGP dispersion at the equator due to variation in the non-dipole field is 10.6°. This estimate of the dispersion due to variation in the non-dipole field is in excellent agreement with the value predicted from an analysis of the variation in field intensities over the same period. Fits of the model to data from earlier periods indicate that dispersion due to variation in the non-dipole field is essentially independent of the geomagnetic reversal rate while dipole wobble is positively correlated with reversal rate.     

Palaeosecular variation in Central Mexico over the last 30 000 years: the record from lavas

13 lava flows of known age (ages from ¹⁴C dating), which have been erupted in the last 30 000 years, have been studied to determine the palaeosecular variation of the geomagnetic field in Central Mexico. Samples were taken from two different monogenetic volcanic fields: the Michoacan-Guanajuato volcanic field (six sites) and the Chichinautzin Formation (seven sites), both part of the Transmexican Volcanic Belt. The lavas were studied in detail using rock magnetic methods (magnetic susceptibility at room temperature, low-temperature susceptibility behaviour, hysteresis loops, Curie temperatures), combined with reflected light microscopy, in order to deduce their magnetic mineralogy and the domain states of the magnetic minerals. The magnetic carriers are titanomagnetites, which show differing degrees of high-temperature deuteric oxidation, and seem to be predominantly pseudo-single domain (PSD), though in many cases are probably a mixture of domain states. Mean palaeomagnetic directions and palaeointensity values using Shaw and Thellier techniques were obtained using several specimens from each flow. Our data seem to indicate a sharp easterly swing in declination about 5000 years ago, which is also observed in lake sediments from Central Mexico. The calculated values of the virtual dipole moment (VDM) range from 3.1 to 14.9 × 10²² A m². Our data indicate that the virtual dipole moment seems to have increased gradually in magnitude over the last 30 kyr, with a peak at about 9000 years BP. These are features that have been observed in other parts of the globe and are probably caused by variations in the dipole part of the geomagnetic field.     

Magnetostratigraphy of Palaeocene basalts from the Vaigat Formation of West Greenland

A palaeomagnetic study comprising the directional results from 289 individual lava flows, sampled along eight sections in the Palaeocene basalts of West Greenland, is reported. The eight individual sections are correlated using lithostratigraphical marker horizons to form a single composite profile. Generally, the lithological correlation is in good agreement with the record of geomagnetic secular variation.
  The total composite palaeomagnetic profile represents a stratigraphic thickness of 1.6  km through the Vaigat Formation, which is the lowermost of the two volcanic formations formed during the main stage of plateau volcanism. Only two polarity zones are found in the composite profile, suggesting a very short duration for the West Greenland main plateau-building volcanism. ⁴⁰Ar/³⁹Ar dates support a high extrusion rate and also indicate that the lower normal polarity zone is Chron C27n and that the upper reverse polarity zone is Chron C26r.
  The C27n–C26r transition is fully recorded along one of the sections (Nuusap Qaqqarsua), with intermediate directions covering a 200  m thick succession of lavas. A combined palaeomagnetic, field and geochemical study along this profile showed good agreement; that is, geochemically and geologically derived single magmatic events show groupings of the palaeomagnetic directions. Supposing a duration for the geomagnetic transition of 5000 years, the eruption frequency during this period was as high as one flow every 80 years.     

Authors' reply to a comment on 'An objective method for smoothing palaeomagnetic data'

Summary. King & Rees (1979), in commenting on our paper (Clark & Thompson 1978), have raised a number of subjects beyond the main theme of our paper. We described a statistical method of smoothing noisy data and illustrated its application using a set of previously published palaeomagnetic declination data. King & Rees acknowledge our success in achieving the main aim of our paper, namely the construction of objective confidence limits for the smooth curve presumed to underlie the data. However, they seem to question the value of our general technique of data-analysis when used to estimate past variations in the geomagnetic field. As they interpose in their comments on our statistical analyses references to 'techniques of sedimentary analysis', 'inclination errors', 'mode of acquisition of remanence' and 'stability tests', we take the opportunity to discuss these additional topics as well as those relating to our original paper.     

Palaeomagnetic results from two basic volcanic formations in the western Himalayas and a Phanerozoic polar wandering curve for India

Summary. We present palaeomagnetic results from the Durgapipal and Rudraprayag formations, which are basic volcanic formations in the Lesser Himalayas of Uttar Pradesh State. NRM measurements and AF demagnetization stability tests were made on specimens cored from oriented block samples collected at representative sites. Mean stable remanent magnetic directions were used for calculating the Virtual Geomagnetic Pole (VGP) positions; where necessary tectonic corrections were applied.
The virtual geomagnetic north poles were found to be located at:

• (a). 

  Durgapipal (Permian): λ _p = 10° S, L_p = 42° W;
• (b). 

  Rudraprayag (Silurian-Devonian): λ _p = 30° S, L_p = 12° W.

A new, continuous Phanerozoic apparent polar wandering curve for the Indian subcontinent has been plotted from the available palaeomagnetic data and the VGP positions reported in this paper. As a result, the gap in the Indian palaeomagnetic data from the Lower Carboniferous to the Cambrian has been partially filled. The locations of the pole positions for the two formations on the Phanerozoic polar wandering curve for the Indian subcontinent, have been found to coincide with the stratigraphic ages assigned to them on the basis of rather limited geological and palaeontological evidence.
The Cambrian and Permian poles for the Salt Range in the NW Himalayas and the Permian pole for the Kumaon Himalayas are grouped along with the pole positions of contemporaneous formations of the Peninsular Shield. The palaeomagnetic data thus suggests that the two formations are autochthonous in nature.     

On the interpretation of frequency response functions for oceanic gravity and bathymetry

Summary. In 1972–78, the late N. D. Watkins and others carried out ajoint field programme of geological mapping in the Mio-Pliocene flood basalts of Iceland, including sampling for K-Ar dating and palaeomagnetic research. The major part of the palaeomagnetic sampling is represented by 2462 lavas in five long composite sections through the lava pile. This paper deals with various statistical properties of this data set.
It is concluded that geomagnetic reversals occur more frequently than is assumed in the current ocean-floor polarity time-scale. There is no evidence for significant asymmetries between normal and reverse polarity states of the field, neither as regards chron lengths, secular variation, or virtual dipole moment magnitude. Intensities of remanence in these lavas are shown to be well approximated in terms of a hyperbolic distribution. The latitude distribution of virtual magnetic poles can be fitted with a Bingham function having k' ∼ 4.5, and low-latitude poles do not occur preferentially in any particular longitude interval.     

Palaeomagnetism of 935 Ma mafic dykes in southern Sweden and implications for the Sveconorwegian Loop

New palaeomagnetic results for the 935 Ma Göteborg-Slussen mafic dykes in southern Sweden provide a well-dated high-quality palaeomagnetic pole for Early Neoproterozoic Baltica. New U-Pb geochronological data for several palaeomagnetically studied mafic intrusions yield three additional well-dated palaeopoles and one virtual geomagnetic pole. This set of dated poles suggests minimal drift of Baltica in moderate latitudes between ∼965 and 915 Ma. They also support the hypothesis of a post-900 Ma regional remagnetization event in SW Sweden and SW Norway. The positions of three distinct clusters of ∼1100 to 850 Ma palaeopoles suggest a clockwise time progression of the Baltica apparent polar wander path (the Sveconorwegian Loop) during this time interval. New well-dated palaeomagnetic poles for ∼970 to 900 Ma from Laurentia are required to verify the palaeogeographic reconstructions of Baltica and Laurentia.     

Palaeomagnetism of coast-parallel alkaline dykes from western Norway; ages of magmatism and evidence for crustal uplift and collapse

Summary. Palaeomagnetic investigation of 18 coast-parallel, alkaline dykes from the Sunnhordland region, western Norway, reveal the presence of multi-component magnetization in half of the dykes. The distribution of reversely magnetized directional end-points corresponds to a Kiaman (Permian) geomagnetic field direction, but the intersection of great circle trends coincide with the well-grouped and stable normal magnetization present in the remaining nine dykes. The stable remanence deviates significantly from any known palaeomagnetic field direction for western Europe and, although a non-dipole field/spot-reading origin cannot be ruled out, it is tentatively suggested that the magnetically stable dykes were intruded when the Sunnhordland area was affected by a temporary eastward tilt associated with a phase of crustal uplift over the adjacent North Sea area. A tilt-connection of 15° will bring the stable remanence directions in close agreement with the Triassic/Jurassic pole position for western Europe, implying a Kimmerian age of these intrusions. The range of reported K-Ar ages of some of the dykes is not in accordance with the palaeomagnetic results and is tentatively attributed to: (a) partial argon loss in the reversely magnetized dykes (Kiaman) caused by reheating during the intrusion of the assumed younger sequence of the magnetically stable and normally magnetized dykes, (b) excess argon in the assumed younger dykes, inherited from xenoliths present. A Kiaman intrusive age of all the dykes cannot be excluded, however, in which case a complete remagnetization of some of the dykes, occurring possibly during the proposed event of crustal uplift, was associated with partial loss of argon.     

Magnetic stratigraphy observed in ferromanganese crust

Summary. A sample of ferromanganese crust deposit (SCHW-1D) was cut into two sets of 1 mm slices for palaeomagnetic study. Alternating field and thermal demagnetization, and isothermal remanent magnetization analyses were performed on each thin slice. The results reveal the presence of a stable magnetism and both normal and reversed polarity intervals in the specimens. The observed polarity intervals cannot be confidently correlated with the geomagnetic polarity time-scale of the last 10 Myr due to the polarity overlap inherent in the sampling technique. But the results confirm the slow accretion rate of ferromanganese deposit determined by ¹⁰Be method and suggest potential use of ferromanganese deposits in palaeomagnetic and tectonic studies.     

Palaeomagnetism of alkaline lava flows from El Salto-Almafuerte, Córdoba Province, Argentina

Summary. Stable components of magnetization have been isolated in 15 lava flows (mean K-Ar age 123 ± 4 Myr) from the alkaline sequence outcropping at El Salto-Almafuerte, Province of Cordoba, Argentina. Magnetic and geologic stratigraphy, as well as K-Ar ages indicate that this sequence was probably extruded in the Lower Cretaceous during the first volcanic cycle of the Sierra de los Cóndores Group (Vulcanitas Cerro Colorado Formation).
The palaeomagnetic pole-position for El Salto-Almafuerte lava flows, computed from the mean of 15 virtual geomagnetic poles and denoted SAK₇, is: 25° E, 72° S ( k = 35, α₉₅= 6.5°); it is fairly close to other Lower Cretaceous palaeomagnetic poles for South America. The elongated distribution of Cretaceous palaeomagnetic poles suggest recurrent drift for South America in early Cretaceous time.
The palaeomagnetic and radiometric data for the igneous rocks from El Salto-Almafuerte support the magnetic reversal time-scale for the early Cretaceous suggested by oceanic magnetic lineations.     

The magnetic polarity stratigraphy of a Hauterivian/Barremian carbonate sequence from southern Italy

Remanence directions, measured at 2  cm intervals along a composite 88  m bore-core, enable mean palaeomagnetic poles to be defined at 13.6°S, 25.2°W and 13.6°N, 154.8°E. The directions of remanence vary very smoothly away from each palaeomagnetic pole, extending more than 90° from them. This raises doubts about the physical meaning of polarity definitions based on the distance between virtual and mean palaeomagnetic poles. For practical purposes, intermediate polarity is defined as directions whose virtual poles lie more than 25° from the mean pole, enabling at least five normal subchrons to be specified within the upper predominately reversed quarter of the core and 11 reversed subchrons within the lower predominantly normal three-quarters of the core. The stratigraphic thickness between these subchrons shows a very high linear correlation ( r >0.99) with the stratigraphic thickness of other terrestrial sequences and the distances between marine polarity sequences of comparable age. The analysed sequence contains wavelength spectra which, when transformed to the temporal realm, match periodicities determined for three marine magnetic anomaly profiles of similar age. These also match planetary orbital periodicities for the Cretaceous. These observations suggest that secular variations and polarity transitions are driven by common core processes whose surface expression is influenced by changes in the planetary orbits. Such detailed geomagnetic features enable far greater reliability in establishing magnetostratigraphic correlations and also enable them to be dated astronomically.     

The British Tertiary igneous province: palaeomagnetism of the dyke swarm along the Sleat coast of Skye

Summary. A palaeomagnetic investigation has been made of a swarm of more than 400 dykes along the south coast of Skye, Scotland, by the Sound of Sleat.
Seven red lamprophyre dykes have palaeomagnetic directions inconsistent with Tertiary age, and not inconsistent with their previously held Caledonian age. The remaining 409 dykes have palaeomagnetic directions that are consistent with a Lower Tertiary age. We present evidence suggesting that the Tertiary dykes might have been emplaced during a short time, over which the geomagnetic polarity occupied as few as three polarity intervals (NRN or RNR).
Certain 'intermediate' directions of magnetization have also been found, and are presented here.