Delineating the crustal structure of central Java is crucial for understanding its complex tectonic setting. However, seismic imaging of the strong heterogeneity typical of such a tectonically active region can be challenging, particularly in the upper crust where velocity contrasts are strongest and steep body wave ray paths provide poor resolution. To overcome these difficulties, we apply the technique of ambient noise tomography (ANT) to data collected during the Merapi Amphibious Experiment (MERAMEX), which covered central Java with a temporary deployment of over 120 seismometers during 2004 May–October. More than 5000 Rayleigh wave Green's functions were extracted by cross-correlating the noise simultaneously recorded at available station pairs. We applied a fully non-linear 2-D Bayesian probabilistic inversion technique to the retrieved traveltimes. Features in the derived tomographic images correlate well with previous studies, and some shallow structures that were not evident in previous studies are clearly imaged with ANT. The Kendeng Basin and several active volcanoes appear with very low group velocities, and anomalies with relatively high velocities can be interpreted in terms of crustal sutures and/or surface geological features.

Our seismic tomographic images characterize, for the first time, spatial and volumetric details of the subvertical magma plumbing system of Merapi Volcano. We present P- and S-wave arrival time data, which were collected in a dense seismic network, known as DOMERAPI, installed around the volcano for 18 months. The P- and S-wave arrival time data with similar path coverage reveal a high Vp/Vs structure extending from a depth of ≥20 km below mean sea level (MSL) up to the summit of the volcano. Combined with results of petrological studies, our seismic tomography data allow us to propose: (1) the existence of a shallow zone of intense fluid percolation, directly below the summit of the volcano; (2) a main, pre-eruptive magma reservoir at ≥ 10 to 20 km below MSL that is orders of magnitude larger than erupted magma volumes; (3) a deep magma reservoir at MOHO depth which supplies the main reservoir; and (4) an extensive, subvertical fluid-magma-transfer zone from the mantle to the surface. Such high-resolution spatial constraints on the volcano plumbing system as shown are an important advance in our ability to forecast and to mitigate the hazard potential of Merapi’s future eruptions.

Volcanic heating of a tectonic fault changed the way it produces large earthquakes.