Spatial concentration of coda energy in the summit region of volcanoes
In general the energy of the seismic coda becomes uniformlydistributed in space at some late lapse time, a fact thatis used for example in the coda normalization method. Somerecent observations at strato volcanoes, on the contrary,indicate an abnormal spatial concentration of coda energyin the summit regions of these volcanoes. This observedcoda localization can be explained by the leakage of energyfrom the strongly scattering volcanic edifice into the muchmore homogeneous underlying earth crust. The leakage resultsin an inhomogeneous distribution of energy in space, wherethe energy is low near the volcano-crust boundary and largeinside the strongly scattering volcano far from that boundary.We present two models for this observation: The first one isbased on an analytical solution of the diffusion equationfor a scattering cylinder (representing the volcano) embeddedin a homogeneous half-space (representing the surroundingcrust). The second model is based on a Monte-Carlo simulationof the acoustic equation of radiative transfer. In thissimulation we take into account multiple scattering insidethe volcanic edifice as well as leakage at the bottom of thevolcano into the less heterogeneous crust. Additionally, inthis model we also consider the true topography of the volcanoby simulating reflections at the free surface, where we use adigital elevation model of the volcano and the Kirchhofftangent plane method. Both models can explain the observedcoda localization. We compare theoretical seismogram envelopesof both models to data of shallow volcano-tectonic earthquakespreceding the 1998 eruption of Merapi volcano (Indonesia).