GNGTS 2023 - Atti del 41° Convegno Nazionale

Session 1.1 GNGTS 2023 are not recomputed but obtained from previous available catalogs when this information is present. Magnitudes are generally local magnitudes (ML) whenever it is possible and within the range of the ML reliability. This is for the sake of homogeneity and completeness in a variety of studies. For events of the period 1981-2002 the CLASS magnitudes are inherited from the CSI catalog (Castello et al., 2007). For events without magnitude estimation in CSI, this information is extracted from the ISIDe database (ISIDe Working Group, 2007), giving priority to the Md (duration) INGV reference magnitude at that time. For events of the period 2003-2018, the CLASS magnitudes are extracted from the ISIDe database, which also includes all the solutions from the BSI bulletin, giving priority to the ML and within the range of the ML reliability. Results: Earthquake Hypocenter Locations and Quality Assessment The validity of the hypocenter locations then is assessed “a posteriori” by analyzing the inversion solutions. NLL provides a complete description of the hypocenter location uncertainties, whose main estimators are included in the quality analysis proposed by Michele et al. (2019). For each earthquake we compute a quality factor Qf by applying the empirical formula, in which the subscript n means “normalized”, “wj” are the weights and Nest the number of estimators. The empirical formula from Michele et al., 2019 applied for CLASS is ( ) = 1 ( ) 2 + 2 ℎ ( ) 2 + 3 ( ) 2 + 4 ℎ ( ) 2 + 5 ( ) 2 + 6 ( ) 2 + 7 ( ) 2 where the quality estimators of the hypocenter location uncertainty are: rms (root mean square error from P and S arrival time residuals), nphs (number of arrival time picks used in the hypocenter location inversion), sdist (minimum station epicentral distance), errh (hypocentral error on horizontal coordinates computed from the 68% ellipsoid error), errz (hypocentral error on vertical coordinates computed from the 68% ellipsoid error), gap (maximum azimuthal angle between two successive stations around the source epicenter), Rpdf (radius of a sphere having volume equivalent to the PDF volume) The Quality factor Qf is a normalized value from 0 to 1. In order to set up a simple way for selecting events in our catalog, we associated each location quality factor to a Quality Class having code from A (best location with quality factor from 0 to 0.25) to D (worst location with quality factor from 0.75 to 1). Map views (Fig. 2) show the distribution of the earthquakes selected by Quality Class (A-class, B-class, C-class, and D-class). We observe that that the hypocenter location quality is excellent for some 24% of the catalog (~101,900 events, A-class), good for 42% (~178,500 events, B-class), poor for 21% (~87,600 events, C-class), and bad for 13% (~54,500 events, D-class). We will present the main characteristics of the CLASS catalog, the procedure we followed to manage both heterogeneous datasets for hypocenter location and finite difference parametrization for travel time computation in 3D velocity models at large scale. We will finally discuss some limits and potentialities of our results.