THE ROLE OF SHEAR HEATING IN THE FORMATION OF OBSIDIAN

With ~74-78 wt% SiO₂ and <1 vol% crystals, obsidian is the end-member of high-silica, crystal-poor volcanic rocks, making it ideal for evaluating current hypotheses of high-silica magmatism. In traditional models, high-silica melts are generated through liquid separation from a crystal mush in compositionally, thermally, and mineralogically zoned magma bodies. Such liquids should be saturated in quartz and one or two alkali feldspars at separation. However, due to its high viscosity (10⁶-10¹² Pa s), high-silica magma is difficult to separate from a crystal mush. An alternative method for producing crystal-poor liquids is via shear heating and resorption of crystals, because high-viscosity magmas are susceptible to shear heating. This effect has only recently been accounted for in thermal models of silicic magma flow, with possible temperature increases of 150 K. This study focused on modal analyses and crystal morphology of obsidian samples from the Long Valley and Coso volcanic areas east of the Sierra Nevada, California. To understand the potential for shear heating to create superliquidus conditions in silicic melts, petrographic analysis is combined with thermal models of conduit flow using composition to calculate temperature-dependent viscosity. Rounded crystals suggestive of heating dominate in obsidian from Coso, Mono Craters, the Long Valley resurgent dome, and Glass Mountain; crystal faces are rare. Samples have an average of 0.35 vol% phenocrysts and contain dominantly plagioclase (An_14-48) or anorthoclase (An_9-12, Or_9-12) with minor quartz or orthopyroxene (En_54-57). The high-silica content (74-78 wt%) of these rocks predicts saturation in quartz and sanidine, yet potassic feldspar (Or_30-34) is only present in trace amounts in one sample. These observations contradict mush-model predictions of silicic mineral assemblages and suggest crystal resorption via shear heating during ascent as a mechanism for the petrogenesis of obsidian.