Multi-scale analysis of multiparameter geophysical and geochemical data from active volcanic systems

Persistently active volcanoes present short and long term variation in their magmatic, hydrothermal, and/or hydrogeological systems. As magma is rarely accessible on the surface, investigation of the dynamic behaviour of the hydrothermal system is an indirect approach to study the underlying magmatic activity. Variations in volume, mass and flow direction of the water are expressed through change and generation of local disturbances of potential-fields, such as gravity or Self-potential. The source generating the potential-field signal is a non-unique solution, making it very difficult to model. This study uses a modern signal analysis technique, Multi-scale wavelet tomography, to accurately determine the depths of these sources. The accuracy of Multi-scale wavelet tomography on Self-potential data was tested on three volcanoes (Masaya, Stromboli, Waita) in comparison with water depths calculated by more traditional geophysical methods. Traditional inverse gravity modeling is also used to better constrain the non-unique solution of potential-fields. This study also investigates two persistently active volcanoes, Masaya and Kawah Ijen, through time-series and spatial surveys to monitor change occurring within them. This study shows that a well established and mature hydrothermal system can show limited surface expression as at Kawah Ijen volcano, while an apparently low intensity hydrothermal system can have an extensive and complex system beyond its active crater, such as at Masaya volcano. The hydrothermal system of Masaya is spatially controlled by a ring fault structure and has been stable between 2006 and 2009. In contrast, on Kawah Ijen, the intense and well-established hydrothermal system is completely self-sealed within the upper volcanic edifice and can only release the pressurized fluids and gas through the active crater. Nevertheless, between 2006 and 2008, the hydrological system showed significant vertical change due to seasonal effects. By integrating a wide variety of distinct, complementary techniques to a number of persistently active volcanoes, over an extended period of time, it is possible through accumulation of baseline information, to characterize the components of signals detected on volcanoes. A more accurate understanding of the volcanic system as a whole, through accurate constraint of the different volcanic signals, is fundamental in improving volcano monitoring and hazard mitigation.