Date of Award
Doctor of Philosophy (PhD)
Professor Henry P. Schwarcz
Professor Derek C. Ford
The magnetic field of the Earth undergoes slow changes with time known as the secular variation (SV). The SV apparently arises from changes in both the central main dipole and in the non-dipole sources of the core dynamo. The study of the past field and its SV is contingent upon the reliability of the various paleomagnetic records. Considering SV mainly within the Brunhes chron, however, the available recorders are known to be generally deficient, as follows:
1) The recorded signal is not always directly dateable. The assigned ages often lack accuracy and resolution.
2) The intensity and direction record from lavas and baked materials is fragmentary.
3) The sediment records are often either imperfect or have unknown reliability. Several core records are required in order to establish a dated SV master curve for a given region.
In summary there are very few age-dated combined intensity and direction data so necessary for a full vectorial description of the field.
The calcium carbonate cave deposits known as speleothems are shown here to provide dateable continuous records going some way towards satisfying these problems. Stalagmites and flowstones may be reliably dated by the uranium-thorium disequilibrium method. After allowance has been made for any initial contamination of the sample and precautions taken against analytical chemical errors, growth rate curves may be fitted by least squares methods.
Some speleothems have too low a magnetic content to have a measurable NRM. Others have shown themselves to possess a measurable and stable NRM which is inferred to be primary. These latter speleothems have yielded reliable paleofield directions and this claim is based upon:
1) In one sample, TS, the directions from known modern deposition was in good agreement with modern field values at the site.
2) Signals were internally consistent. Precision of the direction data was often good, and the directions were serially correlated across the growth layers.
3) In six samples the NRMs were shown to be detectably free of depositional errors.
4) In one sample no magnetic anisotropy was apparent within the experimental error.
The speleothem deposit thus presents a major discovery in the search for a reliable dateable and continuous paleomagnetic recorder.
It was thus found possible to safely make geo-physical interpretations from these records as follows:
1) Sample VCCL (Vancouver island) clearly showed clockwise looping of the field from about 6.4 to 2.2 Ka BP. This is interpreted as evidence for westward drift of the field. The virtual geomagnetic poles (VGPs) were mostly far-sided.
2) TS (England) showed clockwise looping though this was not as clear as in the VCCL record. The VGPs were biassed far-sided and right-handed for the period 11.3 to 7.5 Ka BP. It was found possible to correlate the declination features with those of a Swiss lake sediment record. The lack of a correlation of the coeval inclination features appears to suggest that the sources were assymetrical.
3) A stalagmite SJLS, from Mexico, showed far-sided VGPs but no readily discernable cyclic features, for the period 9.0 to 4.5 Ka BP.
4) Stalagmite SJHS, from Mexico, was found to have ages straddling the Blake event. It showed no evidence for such anomalous field behaviour.
5) By comparing the declinations at latitude 22°N around the world with another Mexican record (DAS2), westward drift of the field was shown to have occurred for the last 1.3 Ka, at these latitudes. The record almost covers one cycle of the westward drift of declination features. At the start of this cycle inclinations were shallower than they are today.
These findings suggest that the dated SV records from speleothems may provide excellent magnetostratigraphic master curves for age-dating purposes back to about 350 Ka BP - the limit of the U-Th method.
An attempt was made to recover paleointensity using anhysteretic remanent magnetization (ARM) as an analogue of the NRM. ARM was found to be approximately linear with weak field and was apparently successful in normalising for the variable magnetic content across the growth layers. Although it was not found possible to make a full assessment of the ARM analogue, the results are encouraging enough to warrant further study.
Many of the more strongly magnetized samples owe their NRM to magnetite-bearing detritus from cave floods. If the bulk of such an NRM is a detrital remanent magnetization (DRM) then it is free from depositional errors. Weaker remanences may have arisen from chemical precipitation, also of magnetite. The evidence for a chemical remanent magnetization (CRM) is largely circumstantial. It may be related to the ubiquitous presence of soluble or insoluble organic material in speleothems. It is suggested that the NRM of most speleothems may be a combined DRM and CRM.
Magnetite (or less likely, maghemite) was shown to be the remanence carrier in most samples. A few titano-magnetite grains were observed in the thin sections from two stalagmites. Oxyhydroxides of iron were shown to be present in samples associated with terra rosa soils. The remanence carrier from a Greek sample is hematite though this finding awaits confirmation.
Latham, Alfred G., "Paleomagnetism, rock magnetism and U-Th dating of speleothem deposits" (1981). Open Access Dissertations and Theses. Paper 569.