One long term goal is an improvement of existing strombolian eruption models. The idea between the currently mostly accepted model is that a gas slug rises through the conduit to the surface where it burst (Vergniolle, 1996, Seyfried, 2000). The most important question, which also separates the different models from one another is the actual overpressure in the bubble before bursting. While Seyfried (2000) give evidence from lab experiments as well as theoretical considerations that there is a negligible amount of overpressure, Vergniolle (1996) through the analysis of acoustic data and also a theoretical model derive a significant overpressure before bubble rupture. Setting up a combined network of seismic arrays, pressure sensors (infra-sound) and doppler radar we want to prove the different models by evaluate the gas slug dynamics from the source region, where the gas slug is formed to the exit of the vent where it bursts. With our doppler radar data, by measuring the explosion velocities, we can actually try to determine the amount of kinetic energy involved in strombolian eruptions which will significantly improve the energy balance of strombolian eruptions and therefore help to constrain the amount of overpressure in the system. In particular, by comparing the amount of dynamic deformation and where it happened, measured by the broadband seismic array, and acoustic data with the doppler radar data, a quantitative measure will be achieved how much of the overpressure translates into seismic deformation and acoustic pressure, respectively.