Vrancea seismic source calibration using a small-aperture array

A smalt-aperture seismic array (BURAR) was installed in 2002 in the northern part of the Romanian territory, in the Eastern Carpathians (Bucovina area). Since then, the array has been in operation under a joint cooperation programme between Romania and USA. The array consists of 10 stations installed in boreholes (nine short period instruments and one broadband instrument) with enough high sensitivity to properly detect earthquakes generated in Vrancea subcrustal domain (at about 250 km epicentral distance) with magnitude M-w above 3. Our main purpose is to investigate and calibrate the source parameters of the Vrancea intermediate-depth earthquakes using specific array techniques and data provided by BURAR as a stand alone station. Thirty six earthquakes with magnitudes between 3.4 and 6.0 were selected, including the recent events of September 27, 2004 (45.70 degrees N, 26.45 degrees E, h = 166 km, M-w = 4.7), October 27, 2004 (45.84 degrees N, 26.63 degrees E, h = 105 km, M-w = 6.0) and May 14, 2005 (45,66 degrees N, 26.52 degrees E, h = 146 km, M-w = 5.1), which are the best ever recorded. earthquakes on the Romanian territory. Empirical Green's function deconvolution and spectral ratio methods are applied for pairs of collocated events with similar focal mechanism. Stability tests are performed for the retrieved source time function using the array elements, Empirical scaling and calibration relationships are also determined. The scaling relationships assessed on the basis of the resulted source parameters reveal a behaviour close to the typical behaviour for other source areas with a significant dependence of the stress drop on the seismic moment (non-constant stress drop source model). Our study shows the capability of the BURAR array to determine the source parameters of the Vrancea intermediate-depth earthquakes as a stand alone station and demonstrates that the recordings of this array provides reliable and useful tools to efficiently constrain the source parameters and consequently source scaling properties.