Drilling Magma for Science, Volcano Monitoring, and Energy

Magma chambers are central to understanding magma evolution, formation of continental crust, volcanism, and renewal of hydrothermal systems. Information from geology, petrology, laboratory experiments, and geophysical imagery has led to little consensus except a trend to see magma systems as being crystal-dominant (mush) rather than melt dominant. At high melt viscosities, crystal-liquid fractionation may be achieved by separation of melt from mush rather than crystals from liquid suspension. That the dominant volume has properties more akin to solid than liquid might explain the difficulty in detecting magma geophysically. Recently, geothermal drilling has intersected silicic magma at the following depths and SiO2 contents are: Puna, Hawaii, 2.5 km, 67 wt%; Menengai, Kenya 2.1 km, 67 wt%; Krafla, Iceland, 2.1 km, 75 wt%. Some similarities are: 1) Drillers encountered a "soft", sticky formation; 2) Cuttings or chips of clear quenched glass were recovered; 3) The source of the glass flowed up the well; 4) Transition from solid rock to recovering crystal-poor glass occurred in tens of meters, apparently without an intervening mush zone. Near-liquidus magma at the roof despite rapid heat loss there presents a paradox that may be explained by very recent intrusion of magma, rise of liquidus magma to the roof replacing partially crystallized magma, or extremely skewed representation of melt over mush in cuttings (Carrigan et al, this session). The latter is known to occur by filter pressing of ooze into lava lake coreholes (Helz, this session), but cannot be verified in actual magma without coring. Coring to reveal gradients in phase composition and proportions is required for testing any magma chamber model. Success in drilling into and controlling magma at all three locations, in coring lava lakes to over 1100 C, and in numerical modeling of coring at Krafla conditions (Su, this session) show this to be feasible. Other unprecedented experiments are using the known location and properties of magma to calibrate geophysics (Brown et al, this session) and understand signals of "unrest". How can we not make such observations when there is so much to learn, so much at stake in correctly monitoring volcanoes, and such a need for clean, renewable energy?