Extended cooling and viscous flow of large, hot rhyolite lavas: implications of numerical modeling results

Abstract

Abstract A common misconception about rhyolite lava flows is that they cannot advance far from their vents and are constrained to be small due to their high viscosities. Although a high-viscosity lava will indeed flow quite slowly, it may advance a great distance if the lava efficiently retains heat and thus mobility, and if the available magma supply is large. To evaluate this possibility, a finite-difference numerical model was used to determine the rate of heat loss from units interpreted to be large rhyolite lavas. Cooling of such 100–300 m-thick units is very slow and is further retarded by a vesicular carapace and by evolution of latent heat. Thermal history alone suggests that large, thick lava flows could remain active for several decades. To constrain the duration of flow advance of these units, their inferred apparent viscosities are compared with those of active lava flows. Mean flow velocity of some large rhyolites in southwestern Idaho, USA, was estimated based on the observation that the units must still have been advancing by viscous flow at the start of spherulite crystallization, indicating perhaps a 10- to 18-year flow duration, based on studies of Obsidian Dome, California. Mean flow velocities would have been roughly 0.59 to 2.5 km/yr, and calculated average apparent viscosities of the flows ranged from 1.8 to 3.6 orders of magnitude greater than the viscosity of the lava itself. This relationship also holds for active andesite and basaltic andesite block lava flows, which have apparent viscosities 2.7 to 4 orders of magnitude greater than lava viscosities. This implies that large rhyolites may indeed have been emplaced in the same manner as less-silicic block lavas, and that long distance viscous flow is not unrealistic for rhyolite lavas. A good example of a large-volume lava is the 15-km 3 Rhyolite of the Badlands (SW Idaho), where vent relations show that lava effused from a fissure and advanced up to 9 km. More extensive rhyolites could also be emplaced solely by effusion. Extended flow of these lavas may lead to the remelting and removal of basal crumble breccias, and to the formation of welded shard textures in their margins, both processes tending to obscure the lava flow origin of the units. The great dimensions of lava-like rhyolitic units with volumes of 10 to at least 200 km 3 are not a sufficient reason to assume they erupted as ignimbrites.