Volatile Abundances in Basaltic Magmas and Their Degassing Paths Tracked by Melt Inclusions

Abstract

The abundances of CO2, H2O, S and halogens dissolved in basaltic magmas are strongly variable because their solubilities and ability to be fractionated in the vapor phase depend on several parameters such as pressure, temperature, melt composition and redox state. Experimental and analytical studies show that CO2 is much less soluble in silicate melts compared to H2O (e.g., Javoy and Pineau 1991; Dixon et al. 1995). As much as 90% of the initial CO2 dissolved in basaltic melts may be already degassed at crustal depths, whereas H2O remains dissolved because of its higher solubility such that H2O contents of basaltic magmas at crustal depths may reach a few percents. Most subduction-related basaltic magmas are rich in H2O (up to 6–8 wt%; Sisson and Grove 1993; Roggensack et al. 1997; Newman et al. 2000; Pichavant et al. 2002; Grove et al. 2005) compared to mid-ocean ridge basalts (<1 wt%; Sobolev and Chaussidon 1996; Fischer and Marty 2005; Wallace 2005). During magma movement towards the surface, exsolution of major volatile constituents (CO2, H2O) causes gas bubble nucleation, growth, and possible coalescence that exert a strong control on the dynamics of magma ascent and eruption (Anderson 1975; Sparks 1978; Tait et al. 1989). Gas bubbles have the ability to move faster than magma (Sparks 1978), particularly in low viscosity basaltic magmas. Bubble accumulation, coalescence and foam collapse give rise to differential transfer of gas slugs and periodic gas bursting (Strombolian activity; Jaupart and Vergniolle 1988, 1989) or periodic lava fountains (Vergniolle and Jaupart 1990; Philips and Wood 2001) depending on magma physical properties and ascent rate. It is also thought that strombolian and lava …