Caldera collapse thresholds correlate with magma chamber dimensions.

Explosive caldera-forming eruptions eject voluminous magma during the gravitational collapse of the roof of the magma chamber. Caldera collapse is known to occur by rapid decompression of a magma chamber at shallow depth, however, the thresholds for magma chamber decompression that promotes caldera collapse have not been tested using examples from actual caldera-forming eruptions. Here, we investigated the processes of magma chamber decompression leading to caldera collapse using two natural examples from Aira and Kikai calderas in southwestern Japan. The analysis of water content in phenocryst glass embayments revealed that Aira experienced a large magmatic underpressure before the onset of caldera collapse, whereas caldera collapse occurred with a relatively small underpressure at Kikai. Our friction models for caldera faults show that the underpressure required for a magma chamber to collapse is proportional to the square of the depth to the magma chamber for calderas of the same horizontal size. This model explains why the relatively deep magma system of Aira required a larger underpressure for collapse when compared with the shallower magma chamber of Kikai. The distinct magma chamber underpressure thresholds can explain variations in the evolution of caldera-forming eruptions and the eruption sequences for catastrophic ignimbrites during caldera collapse.

Petrogenesis of isotopically enriched Quaternary magma with adakitic affinity associated with subduction of old lithosphere beneath central Myanmar.

We present a model for the petrogenesis of magma with adakitic affinity in an old subduction zone, which does not involve slab melting and is constrained by new geochronological and geochemical data for Mt. Popa, the largest of three Quaternary volcanoes in central Myanmar (Popa, Monywa and Singu). The edifice is composed of Popa Plateau (0.8-0.6 Ma) with high-K rocks and a stratovolcano (< 0.33 Ma) predominantly composed of medium-K rocks with adakitic affinity (Mg# 45-63, Sr/Y > 40). The distinct K contents indicate that the adakitic magmas cannot be derived from Popa high-K rocks, but they share trace-element signatures and Sr-Nd isotope ratios with medium-K basalts from Monywa volcano. Our estimation of water contents in Popa magma reveals that primary magma for medium-K basalts was generated by partial melting of wedge mantle with normal potential temperature (TP 1330-1340 °C) under wet conditions (H2O 0.25-0.54 wt%). Its melting was probably induced by asthenospheric upwelling that is recognized by tomographic images. Mafic adakitic magma (Mg# ~ 63, Sr/Y ~ 64) was derived from the medium-K basaltic magma in fractional crystallization of a garnet-bearing assemblage at high pressure, and felsic adakitic rocks (Mg# ~ 45, Sr/Y ~ 50) were produced by assimilation-fractional crystallization processes at mid-crustal depths.