Forward and Inverse Modeling of Magma Chamber Dynamics and Crystal Zoning Provides Insights into Rates and Processes Leading to Eruption

Abstract:

Chemical and textural features of crystals provide unique constrains on the processes and rates that occur during the growth, replenishment and eruption of magma reservoirs. Recent studies that use different methods have highlighted that many eruptions are issued from reservoirs that may grow for up to hundreds of thousands years (e.g, U-Th series of mineral separates, zircon dating). In contrast, the effect of new magma injections (recharges) can influence the thermodynamic state of reservoirs on a much shorter scale. In extreme cases it can lead to remobilization of magmas over months (for small reservoirs) to thousands of years (for larger systems). Modeling of mineral zoning provides a lower bound estimate for the timescales associated with magma recharge because (a) they do not record dissolution/resorbtion of phases and (b) they only record the time at which a crystal reacts to a thermodynamic change, which does not necessarily equate with the timing of an intrusion. In open system reservoirs, magmas may experience significant thermal and compositional fluctuations over time, and so far, only simple cooling scenarios [1] have been considered. In particular, under what conditions does a recharge lead to an eruption? Can large recharge events be mostly recycled and partially involved in later eruptions? Here we combine new forward models of magma chamber replenishment and evolution [2] that track the thermodynamic evolution of shallow resevoir with crystal zoning models. We explore several scenarios of magma reservoir growth and evolution and compare them to the time scales obtained from the crystal zoning (e.g. thermal history and boundary condition changes with time). Inverse modeling of natural crystal zoning profiles from the silicic eruptions of Rabaul caldera in 1994 and 2006 allows us to distinguish between scenarios where inputs of fresh magma leads to eruption and recharges that are accommodated by the reservoir over longer time scales and are expressed over many subsequent eruptions [3]. Differentiating between these scenarios is crucial for proper interpretation of monitoring signals and anticipation of the eruptions. [1] Costa et al (2008), Rev. Min. Geochem. 69: 545 [2] Degruyter and Huber (2014) EPSL, 403:117; [3] Bouvet de Maisonneuve et al (2014) Geol. Soc. London Special Pub.