Magmatic differentiation in a chaotic background: A comparison of multiphase simulations and magmatic chronometers

Abstract:

The location and timescales of silicic magma production has received much recent attention and these questions are at the forefront of understanding incubation times for eruptive episodes and ultimately, the production of continental crust. While idealizations of differentiation in simplified magma reservoir geometries have been useful to frame end-member behavior, most magmatic systems, and particular large magma reservoirs, are open systems with time varying geometries. Evidence of open systems and assembly of magmatic systems incrementally are present in a range of plutonic and volcanic rocks. To evaluate the timescales of silicic magma production, a 3D multiphase dynamics model was implemented that includes heat transfer, phase change and magma dynamics. The size of the magmatic systems under consideration are not prescribed, but rather grow or shrink in response to the flux of heat and intrusions. To compare simulations to a range of data, major element chemistry, phase assemblage, and tracking of representative crystals are made through time.

A particular focus of this presentation is a comparison of dynamic processes to proxies used as chronometers. This includes recording the timescale of appearance of different phases that can be compared to timescales inferred from diffusion profiles and monitoring zircon saturation and dispersal. Both the differentiation timescale and timescales of the major growth of zircons are a relatively small fraction of the melt-present lifetime of magma reservoirs, and in particular, typically represent relatively smaller fractions for larger magmatic systems. Melt can exist at low melt fraction (<0.2) for timescales of 100s kyr for the largest systems, while spending only a small amount of time at high melt fraction. Nevertheless, these reservoirs can be assembled incrementally with magma fluxes in the ranges estimated for arcs. A mid-upper crust location is important to have phase assemblages with sufficient leverage to produce the most silicic magmas. A generalized view from these simulations is one in which magmas reside for long periods at low melt fraction, but where many chronometers and dynamics operate on much shorter timescales, reconciling some apparently inconsistent observations on the duration of magmatism.