High-resolution record of geomagnetic secular variation from Late Pleistocene Lake Lisan sediments (paleo Dead Sea)

https://doi.org/10.1016/S0012-821X(98)00146-0Get rights and content

Abstract

We measured geomagnetic secular variation in Lake Lisan sediments (paleo Dead Sea). More than 1500 oriented samples were collected from a 27.3-m section of alternating aragonite and detritus laminae in the Dead Sea basin ranging in age from 67 to 32 ka. The natural remanent magnetization (NRM) is carried by titanomagnetite in the detrital laminae whereas the aragonite is diamagnetic. The NRM is very stable and was acquired several hundred years after deposition. The mean direction of 878 horizons is D = 005°, I = 45° (α95=1°; κ=22). We observed three modes of directional geomagnetic variation as a function of (and by inference, time): very rapid inter-sample changes, slow variation in mean direction, and inclination shallowing of about 1°/m. The overall rate of change in direction is 0.57±0.57°/year, not significantly different from zero. For about 83% of the record the rate of change is less than 1°/year and comparable to historical values. High rates of change are observed more frequently in the Lisan than in historical records, and peak rates are up to ten times faster. A smoothed curve resulting in a maximum rate of change of 0.66°/year and a mean 0.10±0.10°/year may be a more realistic representation of the field behavior. No reverse NRMs were observed, but geomagnetic field excursions may be present where the VGPs deviate by more than 40° from the geographic north at about 52 and 41 ka; the latter may represent the Laschamp event.

Introduction

Time variations of the Earth's magnetic field have a wide spectrum of about 20 orders of magnitude, from over 103 Hz to more than 100 million years [1]. The frequencies less than 1/year are termed geomagnetic secular variation (SV). SV is reflected in the field intensity and directional changes. SV is the combined effect of variation in the axial dipole field component and variation in the strength and location of the non-dipole field components. Changes in the dipole field are synchronous on a global scale, whereas the non-dipole component of the geomagnetic SV varies over smaller geographical distance. The record of geomagnetic SV is punctuated by geomagnetic polarity transitions which take perhaps 103–104 years to complete and by even shorter geomagnetic excursions or events which may represent aborted polarity transitions or more localized non-dipole effects and take 102–103 years to complete. When coupled with precise age measurements, paleomagnetic variations recorded in sedimentary sequences are an important source of information about the geomagnetic field, the timing and duration of polarity transitions and excursions [1]. The temporal resolution of paleomagnetic studies is generally higher where sedimentation rates are higher. In this paper we demonstrate that the remanent magnetization of the Late Pleistocene Lisan Formation sediments can provide a high-resolution paleomagnetic record. We examine the rock-magnetic properties of the Lisan sediments, in particular lock-in depth, inclination shallowing, and the effects of soft-sediment deformation. We next show that the Lisan is a faithful recorder of the geomagnetic secular variation and conclude with discussion of the observed directional changes.

The Lisan Formation is exposed along 220 km of the Dead Sea Transform (Fig. 1). In the study area in the south Dead Sea graben it is comprised mainly of alternating laminae of white aragonite and dark detritus, and a few thick clastic layers and gypsum. In addition, the section contains `mixed layers' of disturbed laminae which were interpreted as seismites 2, 3. The aragonite was precipitated chemically from the upper surface of Lake Lisan whereas the detritus which contains calcite, dolomite, aragonite, quartz and clay, was carried by annual floods [4]. The chemical preservation of aragonite in the Lisan is superb due to extreme dry conditions in the Dead Sea region and the retention within the aragonite of abundant interstitial chloride salt of a high Mg/Ca ratio, inherited from the original lake, which prevented its transformation to calcite 5, 6. Lake Lisan was too saline to support bottom burrowers, leaving the sediment structure intact down to its finest details.

The Lisan aragonite was dated by 14C and U-series. U–Th dating was done by the α counting technique 7, 8, and more recently the studied section was dated by thermal ionization mass spectrometry (TIMS) [9]. The average sedimentation rate is 0.86±0.02 mm/year [9]. Each 2-cm-thick paleomagnetic sample therefore provides, on average, a 23-year time-slice of the paleofield.

We sampled 27.3 m out of the 43-m PZ section of Lisan sediments in the least disturbed and best exposed outcrop found at Peratzim Valley, southwest of the Dead Sea (Fig. 1). The lower part of the section, 0–1400 cm, was sampled at site PZ3 where it is best accessible. The upper part, 1445–2730 cm was sampled about 500 m west, at site PZ1. Continuous exposure along the steep canyon walls enables a direct correlation. This section was selected for detailed geochronological, paleoseismic, geochemical and paleomagnetic studies 3, 6, 9, 10, 11.

Sampling was done by carving a 2 cm cube pedestal with a sharp knife, placing a clear plastic box over the pedestal, measuring its orientation, removing and sealing the box. A total of 878 horizons were sampled almost contiguously, on average every 3 cm. A few parts, e.g., the bottom of PZ1, were sampled more sparsely because of gypsum layers that obstruct sampling. At several horizons we collected as many as 36 samples to test reproducibility and scatter at a single stratigraphic position. Sample numbers indicate their elevation in the section.

Measurements of the magnetization vectors were done with a 2G cryogenic magnetometer at the Geophysical Institute of Israel. All the samples were treated by alternating field (AF) demagnetization in 5 to 10 mT steps up to 60 mT. Several samples were subjected to thermal demagnetization in quartz tubes in 50°C steps up to 650°C.

The characteristic direction of magnetization was defined by principal component analysis. The reference direction (the `expected direction') is the geocentric axial dipole field (GAD) direction in the study area, I = 51° and D = 0°.

Section snippets

Rock-magnetic tests

Natural remanent magnetization (NRM) intensities are commonly of the order of 10−2 A/m. Between 80% and 95% of the NRM is typically removed by AF demagnetization of 50 mT or 70 mT. The mean destructive field is between 15 and 30 mT (Fig. 2).

Horizontal and vertical variations

To understand how the NRM varies in space and in time we compared the scatter of fifteen horizontally and seventeen vertically sampled sites. At each site twelve samples were collected along about 0.5 m. Samples from a single horizontal level average about 23 years, and therefore record the same paleomagnetic direction. Scatter is probably introduced during sampling, handling, and variability of the recording process. Twelve specimens sampled as close as possible vertically span about 300

Lock-in depth

A series of tests was performed to investigate the temporal relations between the magnetization and syn-depositional deformations in the Lisan.

Timing of magnetization

The magnetization of all the deformed elements (folded layers, mixed layers, and an intraclast in the mixed layer) was acquired after deformation. This deformation is intraformational, and thus the magnetization does not postdate the entire Lisan Formation. The deformation features and the tests indicate that the Lisan sediments remained water-saturated for hundreds of years, enabling the magnetic particles to rotate in response to the magnetic torque; only after consolidation was the

Conclusions

(1) NRM in the Lisan Formation was acquired several hundreds of years after deposition, evident in post-deformation magnetization of a variety of syn-sedimentary features.

(2) The mean direction of the magnetic field is 45°/005° (α95=1°; κ=22) and the mean VGP is at 84°N, 171°E (dp = 1.7; dm = 1.3).

(3) Three types of secular variation are observed: high-frequency swings, low-frequency variation of the mean direction, and a trend of inclination steepening toward the top of the section.

(4) During

Acknowledgements

This project is part of a large-scale research on the geology of the Lisan Formation conducted by Abraham Starinsky (A.S.), the Institute of Earth Sciences, the Hebrew University of Jerusalem. It was funded by the US–Israel Binational Science Foundation grant #9200346 to A.S., M.O.M., H.R., and M.S. We are grateful to Amotz Agnon, Alexandra Schramm, and Steve Goldstein for constructive and fruitful discussions and help in fieldwork. Thanks to Ori Gonen and Revital Ken-Tor for assistance in

References (33)

  • R.T. Merrill et al.

    Geomagnetic field stability: reversal events and excursions

    Earth Planet. Sci. Lett.

    (1994)
  • N. Bonhommet et al.

    Paleomagnetism and potassium argon determinations of the Laschamp geomagnetic polarity event

    Earth Planet. Sci. Lett.

    (1969)
  • S. Levi et al.

    A sixty thousand year paleomagnetic record from Gulf of California sediments: secular variation, late Quaternary excursions and geomagnetic implications

    Earth Planet. Sci. Lett.

    (1989)
  • V. Courtillot et al.

    Time variations of the Earth's magnetic field: from daily to secular

    Annu. Rev. Earth Planet. Sci.

    (1988)
  • S. Marco et al.

    Prehistoric earthquake deformations near Masada, Dead Sea graben

    Geology

    (1995)
  • S. Marco et al.

    Long term earthquake clustering: a 50,000 year paleoseismic record in the Dead Sea Graben

    J. Geophys. Res.

    (1996)
  • Cited by (34)

    • Zones of inelastic deformation around surface ruptures detected by magnetic fabrics

      2020, Tectonophysics
      Citation Excerpt :

      The adjacent aragonite-rich and detritus-rich layers have similar magnetic fabrics as seen in the stereograms of Fig. 7. The thermomagnetic curves (Fig. 8a,b) of the aragonite-rich and detritus-rich layers show that the dominant magnetic carrier of all specimens is magnetite, characteristic of the Lisan Formation (Levi et al., 2006b; Marco et al., 1998). The heating and cooling curves of both materials decreases gently around 580 °C, attributed to the Curie temperature of magnetite (Butler, 1998).

    • Identifying soft-sediment deformation in rocks

      2019, Journal of Structural Geology
      Citation Excerpt :

      Although the upper part of the Lisan Formation has never had a significant (typically <10 m) overburden, this could still be sufficient to generate compaction-related fabrics in some cases. Indeed, the palaeomagnetic inclination record from the Lisan Formation is interpreted to show a shallowing effect possibly linked to compaction (Marco et al., 1998). Thus, a degree of ambiguity remains as to the amount of overburden required to generate compaction fabrics in unconsolidated sediment, and limited overburden cannot be used as a basis to discount such fabrics.

    • Beryllium isotopes as tracers of Lake Lisan (last Glacial Dead Sea) hydrology and the Laschamp geomagnetic excursion

      2014, Earth and Planetary Science Letters
      Citation Excerpt :

      The main goal of this study was to assess the feasibility of using the Lisan Formation as a high-resolution archive of 10Be atmospheric production. The Lisan Formation is an excellent site for such a study because it has a well-established high-resolution paleomagnetic record around the Laschamp geomagnetic event (Marco et al., 1998). In order to establish the use of Lisan Formation detrital laminae as faithful recorders of cosmogenic isotopes production rates it is crucial to demonstrate that atmospheric dust is rapidly washed from the drainage basin (short residence time) and deposited in the lake.

    • A high-resolution paleointensity stack of the past 14 to 68 ka from black sea sediments

      2013, Earth and Planetary Science Letters
      Citation Excerpt :

      Corresponding VGP latitudes are generally low for about three thousand years, some descend down to clearly less than 45°N (Fig. 5a). This corresponds to a low inclination interval from about 65 to 60 ka recorded in the Lisan Formation, Dead Sea, Israel (Marco et al., 1998; Fig. 4). Although the evidence is not very consistent in Black Sea sediments the VGP low is interpreted as an imprint of the Norwegian–Greenland-Sea (NGS) excursion (Bleil and Gard, 1989).

    • Problems with radiocarbon dating the Middle to Upper Palaeolithic transition in Italy

      2009, Quaternary Science Reviews
      Citation Excerpt :

      Giaccio et al. (2006) and Pyle et al. (2006) have suggested that the reason for these variable ages may be due to larger changes in the historic levels of 14C concentration in the Earth's atmosphere than previously attested. There is a demonstrable 14C production increase between ∼40–42 ka cal BP associated with the Laschamp Geomagnetic Excursion, associated with peaks in 10Be and 36Cl in lake and marine sediments in both hemispheres (see Robinson et al., 1995; Yiou et al., 1997; Marco et al., 1998; Laj et al., 2000; Lal and Charles, 2007). Giaccio et al. (2006) presented evidence from a marine core sequence in the Tyrrhenian Sea (core CT85-5) which they suggest shows a much more significant increase in the radiocarbon concentration in the atmosphere at ∼40,000 BP which may be related to the Laschamp event.

    View all citing articles on Scopus
    View full text