Chemical differentiation, cold storage and remobilization of magma in the Earth’s crust

Abstract

The formation, storage and chemical differentiation of magma in the Earth's crust is of fundamental importance in igneous geology and volcanology. Recent data are challenging the high melt fraction 'magma chamber' paradigm that has underpinned models of crustal magmatism for over a century, suggesting instead that magma is normally stored in low melt fraction 'mush reservoirs'. 1-9 A mush reservoir comprises a porous and permeable framework of closely packed crystals with melt present in the pore space. 1,10 However, many common features of crustal magmatism have not yet been explained by either the 'chamber' or 'mush' reservoir concepts. 1,11 Here we show that reactive melt flow is a critical, but hitherto neglected, process in crustal mush reservoirs, occurring because buoyant melt percolates upwards through, and reacts with, the crystals. 10 Reactive melt flow in mush reservoirs produces the low crystallinity, chemically differentiated (silicic) magmas which ascend to form shallower intrusions or erupt to the surface. 11-13 The magmas can host much older crystals, stored at low and even sub-solidus temperatures, consistent with crystal chemistry data. 6-9 Changes in local bulk composition caused by reactive melt flow, rather than significant increases in temperature, produce the rapid increase in melt fraction that remobilizes these cool-or cold-stored crystals. Reactive flow can also produce bimodality in magma compositions sourced from mid-to lower-crustal reservoirs. 14,15 Trace element profiles generated by reactive flow are similar to those observed in a well-studied reservoir now exposed at the surface. 16 We propose that magma storage and differentiation primarily occurs by reactive melt flow in long-lived mush reservoirs, rather than by the commonly invoked process of fractional crystallisation in magma chambers. 14 Magma reservoirs occur at several depths within the crust and typically grow incrementally through the intrusion of dykes or sills. 1,11,13,16,17 High melt fractions must sometimes be present in these reservoirs to produce eruptible, low-crystallinity magmas. 1,7,8,9,13 However, geophysical data suggest that reservoirs have low melt fraction even beneath active volcanoes 2-5 and crystal chemistry data indicate that long-term magma storage occurs at low or even sub-solidus temperature. 6-9 High melt fractions are therefore ephemeral, yet geochemical models typically assume differentiation occurs by crystal fractionation from low-crystallinity magmas; 11,14 moreover, geochronological data demonstrate that crustal magma reservoirs can be long-lived, spanning hundreds of thousands to millions of years. 17-21 Existing models of crustal magma storage and differentiation cannot reconcile these conflicting observations. We use numerical modelling to investigate the storage and chemical differentiation of magma in crustal reservoirs. The model describes repeated intrusion of mafic to intermediate sills into the mid-to lower crust, 12,13,16,21-23 the associated transport of heat via conduction and ad...

Magma Reservoir Below Laguna del Maule Volcanic Field, Chile, Imaged With Surface‐Wave Tomography

“…Moreover, these eruptible magma chambers and the upper reaches of their surrounding crystal-rich reservoirs reflect transcrustal magma systems that extend vertically into the lower crust (Cashman et al, 2017). Yet the architecture of such transcrustal systems (Marsh, 2015), the mechanisms and rates of melt transport through their mostly cool crystalline mush columns (Jackson et al, 2018), and the thermochemical conditions required to generate and store significant volumes of eruptible magma in shallow chambers that form within them (Barboni et al, 2016;Cooper & Kent, 2014) all remain elusive.…”

Crystal Fractionation by Crystal‐Driven Convection

“…The worldwide compositional abundances of igneous rocks, including both volcanic and plutonic settings, show bimodal peaks at basaltic and rhyolitic compositions, referred to as the Daly gap (Daly, 1925). One of the key processes contributing to the formation of compositional gaps is fractional crystallization (Clague, 1978;Dufek & Bachmann, 2010;Jackson et al, 2018), a reactive transport process that requires crystals to precipitate from the melt and segregate from their residual melt. While some authors (e.g., Bonnefoi et al, 1995) have hypothesized that fractional crystallization leads to even distributions of composition, studies of compaction show variable efficiency of fractional crystallization (Dufek & Bachmann, 2010;Jackson et al, 2018).…”

Magma diversity reflects recharge regime and thermal structure of the crust

“…We performed thermal modelling of pulsed magma injection into the mid to lower crust coupled to experimental phase relations 32 , which allows us to track the temporal evolution of eruptible magma chemistry. Our model design builds on the theory that silicic magmas originate at mid to deep crustal levels 14,24,25,26,27,52 and that magmatic bodies are built over protracted periods of time by incremental assembly 30,31 .…”

“…To establish the link between magma recharge rates and the evolution of eruptible magma chemistries in mid crustal reservoirs, we developed a numerical model that couples heat transfer and phase petrology. The design of our model is motivated by geophysical 21,22,23 and petrological 14,24,25,26,27,28,32,52 data, which show that magmas acquire their chemistry mostly at mid to lower crustal depths, before being erupted. We do not simulate magma extraction and eruption, but we trace the evolution in the chemistry of potentially eruptible magma over time, using this as a proxy for what is likely be sampled in eruptions.…”