Paper No. 4
Presentation Time: 8:45 AM
CRYSTAL SIZE DISTRIBUTIONS IN EVOLVING SILICIC MAGMA CHAMBERS: SLOW GROWTH AND EPISODIC FRAGMENTATION
Measured CSDs of quartz and zircon from individual pumice clasts provide a quenched snapshot view of growth and fragmentation conditions in preclimactic magma chambers. I describe an acid (HF and HBF4) technique to extract phenocrysts from these clasts, coupled with camera- and computer-assisted measurements of phenocryst length, width, 3D shape, and volume abundance. CSDs of quartz and zircon are presented for several well-known voluminous ash-flow tuffs and small-volume lavas: Bishop, Lava Creek, Lower Bandelier, Toba, Katmai, and Timber Mt. A common feature of CSDs of unfragmented phenocrysts is a concave-down, lognormal shape, in contrast to the reported linear CSDs in more mafic systems. This feature in silicic magmas is interpreted to be a general result of surface-controlled, size-dependent crystal growth by layer nucleation at small supersaturation. CSD slopes on log-linear frequency-size graphs in large volume tuffs, and smaller volume intracaldera lavas are similar, and do not simply correlate to the eruptive volume, or SHRIMP-determined zircon ages. CSDs of quartz in clasts with known stratigraphic positions document single evolving reservoir, fingerprint different magma batches (Lower Bandelier and Lava Creek), and overgrowth and gravitational redistribution (Bishop). Fragment size distributions (FSDs) in the same clasts document fragmentation due to decrepitation of melt inclusions, and syneruptive breakage. FSDs are treated with lognormal, Weibull, and fractal distributions. Among studied clasts, asymptotic and fractal FSDs are found to be more common. However, the genesis mechanisms (e.g. fractal, scale-invariant vs. size-dependent lognormal) inferred from CSD or FSD should be treated with caution. I present relevant examples from experimental and theoretical physics on fragmentation, published examples of breakage from volcanic and metamorphic environments, and my experiments on crystal shapes, and FSDs in experimental breakage due to crushing and melt inclusion decrepitation in quartz. Decrepitation results in a smaller number of fragments (2-6) than crushing and in shapes that can be distinguished on perimeter/area vs. length diagrams. CSD and FSD have potential implications as important mechanisms of trace elemental and isotopic exchange in magma chambers.