GNGTS 2014 - Atti del 33° Convegno Nazionale

94 GNGTS 2014 S essione 1.1 the number of the false events. In the second local configuration (Tab. 2) the minimum number of phases is higher (4 phases within � �� �� ��� � ������ ������ � ��� ��� �� ��� �������� ��� ≈ �� �� ��� � ������ ������ 50 km and 5 phases within � ��� ��� �� ��� ≈ 110 km) so the criteria for the identification are more restrictive; however in this case we utilize also the S phases of the stations close to the source and weighted by the appropriate function. In the last step the GENLOC inversion (Pavlis et al. , 2004) is performed on the local events adopting the parameters of the Tab. 2; in the inversion the Friuli velocity model (Bressan, 2005) is used because it is demonstrated to give the best results while the arrivals are weighted by applying the Huber method. It is worth to evidence that the GENLOC inversion assumes as starting model the results obtained from the grid association; so the inversion can be as an improvement of the starting grid result, but the GENLOC algorithm can not adjust a starting location that is totally wrong. Our system configuration (Tabs. 1 and 2) is tested by running the procedure on the signals recorded in the year 2011; the results are compared with the locations retrieved by NEI bulletin related to the year 2011 and the percentage of earthquakes detected is estimated. All events with M L ≥��� ��� ��������� ���������� ��� ������� �� ��� ��������� ��������� ����� ��� ���� 1.0 are correctly recognized and located by our automatic procedure (Fig. 1); this aspect is the most important in off-line automatic system that analyses the recordings to detect and locate the seismicity of a selected area. The false earthquakes are about 18% of all located events. Antelope ® recognizes and locates 80% more earthquakes than the published catalogue (Fig. 2): the new events are not distributed homogeneously on the studied area and also the magnitude ranges differ in various considered zones. A relevant microseismicity is located in FVG region and in western Slovenia, concentrated on the principal seismogenic structures as the piedmont zone in FVG, the Idrija and Ilirska Bistrica areas in Slovenia; the optimal detection of the microseismicity in this area is due to the dense and homogeneous spatial distributions of the recordings stations that make possible the recognition of the earthquakes with negative magnitude values. In Veneto the situation is more complex: the geometry of the recordings network is wider than the FVG region so the new events are fewer and, sometimes, their locations (and the related magnitudes) can have large uncertainties especially at the S-SW regional borders with Lombardia and Emilia-Romagna. Further many new events are related to the seismic sequence occurred at Langhirano (PR). Fig. 2 – The earthquakes that are located by the Antelope® procedure and not identified in the NEI bulletin.