Journal of Conference Abstracts

Magnetic Self Reversals from Pyroclastic Flows on Lascar Volcano Chile: Possible Implications for the Cooling History of the Deposits

R. M. E. Thomas (Rick_Thomas@cartograph.co.uk)

CARtograph, Cambridge.

A palaeomagnetic study of two recent (<14000 yr old) pyroclastic flow deposits on Lascar volcano Chile found that they had distinct magnetic inclination directions, differing by 8.2°. This was attributed to secular variation, and may be used to correlate the flows with contemporary lava flows, or other pyroclastic events. Reversly magnetised samples were found in both flows, and their origin was shown to be self reversals rather than geomagnetic reversal. Most samples, however, were normally magnetised. The presence of these self reversed samples along with that of exsolution lamellae in titanomagnetite grains was used to constrain the cooling history of one of the flows. Samples can only self reverse if they have been fairly rapidly cooled from above a structural order disorder transition in titano-hematites, and if they have not subsequently been annealed for long times at lower temperatures. On the other hand a minimum time spent at intermediate temperatures (300-500°C) is required to generate exsolution in the titanomagnetites. The results indicate that parts of the flow were at between 400 and 500°C for 100's of hours, whilst the samples which self reversed cannot have spent much more than 10 hours over 400°C. Lithic samples from the flow indicated that no part of the flow was emplaced at less that 400°C, whilst the coarsest lamellae observed were consistent with a maximum emplacement temperature of 500°C. The dependence of self reversal on thermal history could lead to the flow bases and tops becoming reversed whilst the centres remain normally magnetised. This applies not only to pyroclastic flows, but also to lavas or hot fallout deposits.

U-Th Disequilibria and Dynamic Melting Models ­ An Example from Lanzarote, Canary Islands

L. E. Thomas (L.E.Thomas@open.ac.uk), C. J. Hawkesworth (C.J.Hawkesworth@open.ac.uk), P. V. Calsteren (P.V.Calsteren@open.ac.uk) & S. P. Turner (S.P.Turner@open.ac.uk)

Department of Earth Science, The Open University, Walton Hall, Milton Keynes, Bucks MK7 6AA.

Simple batch melting models (Shaw, 1970) can be used to model many minor and trace elements, however they cannot reproduce U-series disequilibria except at unrealistically small melt fractions. To successfully model U-Th disequilibria it is necessary to involve dynamic melting. Thus a number of authors (e.g. McKenzie, 1985) have presented detailed dynamic melting models to elucidate the controls and the processes involved in mantle melt generation and extraction from source regions.

The Canary Islands are underlain by a region of low buoyancy flux, it has been argued that the degree of U-Th disequilibria in OIB varies inversely with the buoyancy in the underlying mantle (Chabaux and Allegre, 1994) hence we would expect to find significant disequilibria. The historic lavas analysed from Lanzarote (samples from 1730-36, 1824 and Corona eruptions) are some of the most primitive rocks found in ocean intraplate settings with Mg numbers close to 70 and high Ni and Cr contents. This primitive nature allows comparison with primary mantle melts after only minor correction for olivine fractionation. The rocks range from silica saturated to undersaturated, and exhibit marked negative arrays between K₂O, Zr, Nb/Y and silica, which can be used to constrain the degrees and depth of melting. The source regions for Lanzarote are relatively deep within the garnet melting zone and the overall degrees of melting range from 1-10%. They exhibit significant U-Th isotopic disequilibrium with ²³⁰Th/²³⁸U varying from 1.06 to 1.81 and ²³⁰Th excesses of 6-81%: however the restricted ranges of Sr, Nd and Pb isotopes lie within the fields generally accepted for OIB. ²³⁰Th/²³⁸U in the silica saturated rocks are consistent with melt generation rates of approximately 10^-4kgm^-3yr^-1 but the silica undersaturated rocks require a more complex melting model, perhaps one with smaller degree melts being extracted more rapidly than the larger degree melts.