2003 Seattle Annual Meeting (November 2–5, 2003)

Paper No. 30
Presentation Time: 8:00 AM-12:00 PM

DYNAMICS OF THE HISTORICAL MT. ETNA MAGMA PLUMBING SYSTEM: EVIDENCE FROM TH DISEQUILIBRIA, SR ISOTOPE MICROSAMPLING, AND CRYSTAL SIZE DISTRIBUTION DATA


BOHRSON, Wendy A., Dept. Geol. Sci, Central Washington Univ, Ellensburg, WA 98926, SPERA, Frank J., Earth Science, University of California, Santa Barbara, Santa Barbara, CA 93106, GRAHAM, David W., Oceanic & Atmospheric Sciences, Corvallis, OR 97331, HERRERA, Tracey, Dept. Earth & Space Sci, UCLA, Los Angeles, CA 90095, RAMOS, Frank C., Department of Geology, Washington State Univ, Pullman, WA 99164 and TEPLEY, Frank, Dept. Earth Sci, UCSC, Santa Cruz, CA 95064, bohrson@geology.cwu.edu

Th disequilibria, 87Sr/86Sr microsampling, and crystal size distribution (CSD) data on Mt. Etna lavas erupted between 1329-1991 constrain timescales of magmagenesis and the size(s) and connectivity of magma bodies. (230Th/232Th) and (230Th/238U) (n=14) range from 0.95—1.00 and 0.99—1.14, respectively, and describe 2 horizontal arrays, suggesting the time between melt extraction and eruption was of order 103 yr. Such timescales are short with respect to the post-tholeiitic history of Etna (~2x105 yr). Generally, (230Th/238U) vs. eruptive volume describes a linear array, where higher ratios correlate with smaller volumes. If (230Th/238U) is a proxy for degree of partial melting, then batches of magma generated in distinct 'melting events' do not undergo homogenization. For 5 lavas (1329, 1669, 1950, 1983, 1991), microsampling of plagioclase and clinopyroxene phenocrysts demonstrates that 87Sr/86Sr ranges for crystals from different flows do not overlap: 87Sr/86Sr of crystals in the 1329 eruption are distinct from those in the 1950-91 eruptions. Similarly, 87Sr/86Sr of crystals from 1983 do not overlap those from 1991. Despite this, core to rim analyses of single crystals are typically within analytical uncertainty. These isotopic results suggest that crystals have relatively short residence times (~10-102 yr) and each individual crystal reflects growth in a simple magmatic environment. Plagioclase CSD data describe 2 linear arrays in ln (N) vs. L space that are interpreted to reflect crystallization in 2 different environments (e.g., shallow-magma reservoir and eruptive conduit). For reasonable growth rates (10-9—10-10 mm/sec) crystal residence times range from ~10-102 yr. Significantly, this interpretation is consistent with Sr microsampling conclusions. Collectively, the data suggest that the timescales of magma genesis at Etna are short: total processing time from source to eruption is less than a few thousand years and crystal residence times range from decades to centuries. The shallow-level magma storage and transport system is best envisioned as a series of small volume, poorly-connected chambers or a plexus of dikes and sills, in which melt and crystals reside for a short time before eruption.