Dynamics of magma supply, storage and migration at basaltic volcanoes: Geophysical studies of the Galapagos and Hawaiian volcanoes

Basaltic shields forming ocean island volcanoes, in particular those of Hawai'i and of the Galapagos Islands, constitute some of the largest volcanic features on Earth. Understanding subsurface processes such as those controlling magma supply, storage and migration at these volcanoes, is essential to any attempt to anticipate their future behavior. This dissertation presents a series of studies carried out at Hawaiian and Galapagos volcanoes.;InSAR measurements acquired between 2003 and 2010 at Fernandina Volcano, Galapagos, are used to study the structure and the dynamics of the shallow magmatic system of the volcano (Chapter 3). Spatial and temporal variations in the measured displacements reveal the presence of two hydraulically connected areas of magma storage, and the modeling of the deformation data provides an estimate of their location and geometry. The same dataset also provides the first geodetic evidence for two subvolcanic sill intrusions (in 2006 and 2007) deep beneath the volcano's flank. The lateral migration of magma from the reservoirs during these intrusions could provide an explanation for enigmatic volcanic events at Fernandina such as the 1968 caldera collapse without significant eruption.;Space-geodetic measurements of the surface deformation produced by the most recent eruptions at Fernandina, reveal that all have initiated with the intrusion of subhorizontal sills from the shallow magma reservoir (Chapter 4). A synthetic aperture radar (SAR) image acquired 1-2 h before the start of a radial fissure eruption in 2009 captures one of these sills in the midst of its propagation toward the surface. Galapagos eruptive fissures of all orientations have previously been presumed to be fed by vertical dikes, but these new findings allow a reinterpretation of the internal structure and evolution of Galapagos volcanoes and of similar basaltic shields elsewhere on Earth and on other planets.;A joint analysis of InSAR and groud-based microgravity data acquired at Kilauea volcano, Hawai'i, between 2009 and the end of 2012 (Chapeter 5), allows us to infer the location of a shallow area of magma storage beneath the summit caldera and detect a process of mass increase within the reservoir. This mass accumulation, however, occurred without a significant uplift of the surface and the volume change inferred from the modeling of the InSAR deformation data can account for only a small portion (<10%) of the mass addition responsible for a gravity increase. We propose that this discrepancy between gravity change and deformation could be explained by the replacement of gas-rich magma within the shallow reservoir with denser, outgassed magma. In fact, since 2008, the opening of a new vent within Kilauea's summit caldera allows magma to convect up to the surface, loose its volatiles content and sink back into the reservoir.;Finally, in Chapter 6 we use InSAR and GPS time-series of the surface displacement to characterize the storage system of the remaining five active volcanoes of the western Galapagos Islands and to estimate volumes and rates of magma supply to the archipelago during the past two decades (1992 - 2011). Together with Fernandina, four other volcanoes, Wolf, Darwin, Alcedo, and Sierra Negra, all have a shallow reservoir within 1-3 km depth, while at Cerro Azul magma is stored at greater depth (~6 km) and no evidence for shallower storage is found. Our results highlight that the rate of magma supply from the mantle hotspot to the Galapagos volcanoes may be an order of magnitude lower (~ 0.02 km3 yr-1) than that inferred at the Hawaiian volcanoes (0.1-0.2 km3 yr-1 ). The magma supply rate, however, largely varies through time and seems to be influenced by the occurrence of eruptive and intrusive activity at the volcanoes. On the other hand, eruptions during the past two decades have only occurred at those volcanoes showing the highest rates of magma supply (Sierra Negra, Fernandina and Cerro Azul). A positive feedback between the two processes is therefore possible. (Abstract shortened by UMI.).