GNGTS 2017 - 36° Convegno Nazionale

GNGTS 2017 S essione 2.1 223 different methods are used (least-square and highest-similarity) and, correspondingly, two “best-fit” indexes are introduced. As a result, two distinct 2D GDs for each FFM are obtained. The 2D GDs are centred on the FFM cell with maximum slip. The 2D GD is characterised by an elliptical shape distribution field, depending on three parameters, namely the standard deviations along the two axes and the angle between the along-strike axis and the ellipse major axis. The optimal parameters have been determined by a least-squares procedure applied over the entire 3-parameter space scanned at regular steps. To quantify how well these distributions are able to mimic the original slip heterogeneity, we compare the co-seismic vertical displacement component computed starting from the SRCMOD FFM, taken as reference case (Fig. 1.1), with the same displacement component computed from the following alternative slip distributions: - a homogeneous fault model (���� ����� Fig. 1.4); - a heterogeneous fault model, namely the so-called “SmoothClosureCondition” distribution (���� ����� �� ����� ��� ������������� ������� ���� �� ����� ������� ��� �������� ����� Fig. ����� �� ����� ��� ������������� ������� ���� �� ����� ������� ��� �������� ����� 1.3), in which the heterogeneity depends only on depth (Freund and Barnett, 1976; Geist and Dmowska, 1999); - the two GDs (���� ����� Fig. ����� 1.2). Fig. 1 - Different source models for the 12/09/2007 Indonesia earthquake, with Mw = 8.5, which is one of the FFMs included in the SRCMOD database, are shown in the above graphs in a normalised version. The original FFM is from Gusman et al. (2010). The choice of the first two is justified by the observation that they have been, and are still widely used in many studies dealing, for example, with the deformation fields induced by earthquakes, with earthquake and tsunami hazard assessment, and with the simulation of earthquake-induced tsunamis (see, e.g., Geist and Dmowska, 1999; Tinti et al. , 2005; Gutscher et al. 2006; Babeyko et al. , 2010; Tonini et al. , 2011, and many others). For a given slip distribution, the surface displacement field, and in particular its vertical component, is computed by means of the widely used analytical formulas by Okada (1992), in which a rectangular fault is buried in a perfectly elastic, homogeneous and isotropic elastic half-