Geochemical perspectives on large silicic magma systems: Role of crustal assimilation and magma supply

Abstract:

The formation of large (1000-5000 km³) silicic magma systems requires accumulation of eruptible magma in the upper crust for ≥10⁵ years. Critical information on how this condition can occur comes from isotope geochemical studies of magmatic systems, models for magma generation in the mantle, energy balance considerations for magma being transported through or stored in the crust, and the properties of wallrock surrounding a magma chamber. The single most important parameter is the rate and duration of magma supply to mid-crustal levels. Magma supply can be amplified by melting in the lower parts of thick crustal sections if the ambient lower crustal temperature is high enough; isotopic data suggest that this is critical for producing large silicic systems in subduction settings. Isotope studies show that in regions of thick, hot continental crust, the mantle magma supply can be enhanced by factors of 2 to 10 by melting in the lower crust. A high magma supply assures that magma can accumulate at mid- to shallow crustal levels; the continuing addition of new magma being sufficient to overcome losses due to crystallization and small eruptions. With a supply of 0.01 ckm/yr, which is high, it requires 100,000 years to accumulate 1000 ckm of magma, hence it is a requirement that magma be stored in the crust for a long time rather than erupted. Extended storage is a special circumstance; most magma supplied to the mid-crust is either rapidly erupted or crystallizes. Magma can accumulate in the middle and upper crust, even though it is buoyant and eruptible, if the wallrocks behave viscoelastically and have a sufficiently low yield stress (Jellinek and DePaolo, Bull. Volc., 2003; DeGruyter and Huber, EPSL, 2014), which suggests that wallrocks need to be pre-heated, perhaps by precursory magmatic activity. To erupt a large fraction (20-50%) of magma stored in a chamber requires that the roof detach during eruption; it is impossible to have a large eruption without forming a caldera. The general conditions for roof detachment can be inferred from observations of caldera dimensions and estimates of magma chamber depths.