V13E-01:
Variations in Pyroclast Porosity: The 2010 Gunung Merapi Eruption
Monday, 15 December 2014: 1:40 PM
Kimberly D Genareau, University of Alabama, Tuscaloosa, AL, United States, Shane J Cronin, Massey University, Palmeston North, New Zealand and Gert Lube, Massey University, Institute of Environment and Agriculture, Palmeston North, New Zealand
Abstract:
The 2010 eruption of Gunung Merapi (Java, Indonesia) began with an explosion resulting from the rapid development of shallow conduit overpressure, which destabilized the pre-existing lava dome and generated large PDCs on 26 October. This was followed by rapid ascent and extrusion of a juvenile dome that also collapsed, generating the largest surges of the series on 5 November. Lava lapilli within the PDC deposits were analyzed using helium pycnometry to determine variations in pyroclast porosity during the transition from dome collapse to rapid dome regrowth and subsequent collapse. Lapilli were separated into pyroclasts derived from the earlier phase of dome growth from those derived from the later and more voluminous phase of dome growth. For all lapilli, regardless of eruptive phase, open porosity of the pyroclasts averages 0.19 (0.08). However, closed porosities of the lapilli are generally higher for the earlier phase, which were derived from deposits of the 26 October PDCs. Efficient degassing during both stages of dome growth produced similar open porosities in pyroclasts, but a higher closed porosity in the 26 October samples resulted from late-stage vesiculation of the residual melt during the initial explosion. This vesiculation event was triggered by the rapid development of shallow conduit overpressure, evidence of which is provided by secondary ion mass spectrometry (SIMS) depth profiling analyses of plagioclase phenocrysts from the ash-sized component of the PDC deposits. Phenocrysts from the earlier phase of dome growth show a build-up of Li in the groundmass glass (attached to crystal surfaces) due to a stage of protracted gas accumulation prior to the 26 October explosion. Conversely, phenocrysts from the later stage of dome growth reveal decreasing Li contents in the groundmass glass due to efficient volatile loss during magma ascent. Vesiculation of the residual melt during the initial explosion resulted in a higher closed porosity in pyroclasts derived from the 26 October PDCs. Thus, although efficient degassing during ascent may result in similar open porosities of lava lapilli, sudden variations in conduit conditions will alter the textural characteristics of pyroclasts immediately prior to their eruption and deposition, which are not revealed through a measurement of density alone.