The Connection between Magma Chamber Processes and Effusive-Explosive Transitions in Silicic Volcanoes: Case Studies from the Aegean Arc

Abstract

Effusive-explosive transitions are widespread volcanic phenomena that consist in the changing of eruptive style from effusive to explosive, and vice-versa. Volcanoes extruding any type of magma can exhibit these transitions, given that sufficient volatiles are present in the melt. Sometimes, shifts in eruptive style occur during the same eruptive event and although predictable, they are difficult to forecast. In this case, the changes in eruptive behavior are closely coupled to conduit feedbacks that occur on short timescales. However, changes in eruptive styles often occur between distinct eruptions that can be separated by long periods: tens or even thousands of years. In this case, the transitions are more likely related to magma chamber processes that act over longer timescales to change the properties of the magmas. In both cases, changes in volcanic behavior are hazardous and can lead to loss of life. This work explores effusive-explosive transitions between independent eruptions. It aims at providing a deeper understanding on the relation between magma chamber processes and changes in eruptive behavior, which could lead to methods of forecasting such events. To this end, the thesis explores two very different calc-alkaline systems that are part of the South Aegean Volcanic Arc. The first case study is the Nisyros-Yali volcanic area, characterized by a large rhyolitic upper-crustal mush that has generated some of the biggest Quaternary eruptions of the Aegean Sea. The second case study is Methana volcano, characterized by a smaller upper-crustal intermediate mush that generated numerous, albeit smaller volume eruptions. The case of Nisyros-Yali reveals that effusive and explosive eruptions were generated by compositionally similar magmas. The main differences are related to the pre-eruptive storage conditions. For example, mineral thermometry and hygrometry indicate that magmas generating explosive events were stored at relatively warm temperatures (815- 850˚C) and were water-rich (4.2-4.6 wt% H2O). Paradoxically, magmas generating effusive eruptions were stored at colder temperatures (710-790˚C) and were water-richer (5.6-6.5 wt% H2O). A newly developed ATR-FPA-FTIR technique, which allows the measurement of dissolved water contents as well as the evaluation of the degassed or pristine quality of the melt inclusions, confirms the hygrometry results. Fluorine and chlorine measurements in melt and apatite inclusions indicate that the magmas generating effusive events were stored under water-supersaturated conditions. On the contrary, the magmas generating explosive events were water-undersaturated at the moment of the eruption-triggering. U-Th disequilibrium dating of zircon crystals reveals that effusive events were preceded by longer volcanic repose periods (12.5 – 45 ky) than explosive events (4 – 10 ky). Longer repose timescales are related to longer periods of magmatic differentiation, which favor water-supersaturation. The key to effusive explosive transitions at Nisyros-Yali is related to the presence (effusive) or absence (explosive) of exsolved volatiles during pre-eruptive storage. This effect is conceptually explored. The case of Methana reveals that effusive and explosive eruptions were generated by magmas ascending from the lower crust, which interacted differently with the upper-crustal intermediate mush of the volcano. Lower-crustal basaltic andesites were stored in the upper-crust, where they hybridized after ingesting <40 vol% intermediate material. The eruptions of these crystal-rich (~40-50 vol%) and relatively water-poor (1.5 – 4 wt%) andesitic-dacitic hybrids always behaved effusively. On the other hand, crystal-poorer (~ 30 vol%) and water-rich (~ 4 wt%) andesites ascended from the lower crust and had minimal interaction with the upper-crustal mush. After ingesting <5vol% material, these magmas erupted explosively. Effusive-explosive transitions at Methana are related to the ascent of magma batches from different lower-crustal sources. The differences in crystallinity, which affect the outgassing potential and ascent velocity are directly controlling the explosivity or effusivity of these magmas.