Monitoring structural volcanic changes in an objective way

It is well known that the occurrence of a tectonic event sufficiently close to a volcano and/or sufficiently energetic can trigger a change in its eruptive activity. Examples include Ambrym (Vanuatu), where tectonic earthquakes have triggered a significant increase of the permanent lava lakes activity; Teide (Spain), where local tectonic events change the characteristics of the seismic noise during the current phase of possible unrest; Tungurahua (Ecuador) and Villarrica (Chile), where an increase of volcanic activity was triggered by the occurrence of tectonic earthquakes. Timely identification of a change following an earthquake has of course serious implications in terms of hazard. However, little effort has been carried out so far to formalize the identification of these structural changes with a statistical approach. Moreover, this is only a particular case of the more general problem of identifying transitions between significantly different volcanic regimes at active volcanoes, at geothermal fields and/or dormant volcanoes. In this work a statistical approach is applied to test for structural changes in volcanic tremor. We considered different fluctuation-type tests using the free source package strucchange in R language [Zeileis et al., J. Statist. Software, 2002]. Stromboli volcano is a particularly interesting case of study, because previous works [Falsaperla et al., J. Volcanol. Geotherm. Res., 2003] had suggested that tectonic events have no influence on volcanic activity, while most recent results [Carniel and Tarraga, Geophys. Res. Lett., 2006] has suggested that they can have it. Other test cases currently under study include Ambrym volcano in Vanuatu and Nisyros geothermal field in Greece.