Potentially explosive activity may accompany the extrusion of silicic lava domes. Explosive activity is manifest in the formation of explosion craters, the development of coarsely vesicular pumiceous textures and the collapse of lava flow fronts which can generate pyroclastic flows. Previously, modelling the emplacement of silicic domes has been hampered by a lack of accurate rheological parameters for input into eruptive models. Also, questions have been recently raised concerning the accuracy of models which predict the effect of water on the viscosity of rhyolitic magmas. Using the parallel-plate method, viscosity data have been experimentally determined on 5 natural rhyolitic samples from a stratigraphically well-constrained vertical section through a lava dome, the Ben Lomond rhyolite dome, Taupo Volcanic Centre, New Zealand. Rheological experiments were performed at volcanologically-relevant temperatures for the eruption and cooling of silicic lavas (780 - 950°C) and strain rates (10-5 - 10-7 s-1) appropriate to the emplacement of viscous flows. In addition, micropenetration viscometry was performed on an hydrothermally hydrated sample containing 0.5 wt% water. During viscometry, melt compositions, water contents and crystal contents were stable.
For samples containing <0.2 and 5 vol% crystals, the composition of the melt, rather than the physical effect of suspended crystals, influenced the resultant effective viscosity of the magma. Samples containing 10 vol% microlites, oriented parallel and perpendicular to microlite long axes, showed no significant difference in apparent viscosity. In addition, for a natural lava flow pumice with 50 vol% vesicles, an increase in effective viscosities with respect to the viscosity of the melt phase alone was recorded. This increase is likely to be due to bubble size, bubble shape, amount of elongation or interconnectivity and surface tension effects.
Cooling rates of these silica-rich volcanic glasses could not be reliably modelled from calorimetric heat capacity curves owing to the broadness of peaks in the calorimetric traces. Consequently, an existing numerical thermal cooling model was used in tandem with the viscosity data to construct temperature-time and viscosity-time profiles through the Ben Lomond flow. Input parameters include an estimated lava flow thickness of 90 m and initial emplacement temperature of 850°C. The time taken for the emplacement of the lava dome is calculated to be 1.5 years based on an average effusion rate of 10-4 m/s. The timescale for the formation of textural units within the lava flow can be constrained as the time taken for the whole unit to cool to the glass transition region (Tg = 700 - 730°C). For example, the vesicular upper 10 - 15 m of a rhyolite lava results from the diffusion during cooling of small amounts of water from the melt into growing bubbles after a certain lag-time for the nucleation of bubbles. According to the model, the time taken for the upper part of the flow to cool below Tg is predicted to be < 2 years. Bubble growth can only occur within this cooling-limited time as vesicles cannot form below Tg.
The formation of spherulites in the upper obsidian layer occurs after a lag-time of 5 - 25 years, as spherulites are interpreted to form just above Tg, and crystallisation of spherulites in the flow centre commenced at least 20 years after the flow stopped moving. Explosion pits can form by the intersection of tension cracks into volatile-rich pockets within the lava flow. Modelling results suggest that explosion pits can form > 2 years after emergence of the lava from the vent and can be a significant volcanic hazard for these intrusions.