GNGTS 2019 - Atti del 38° Convegno Nazionale

GNGTS 2019 S essione 1.1 95 only during 2017 springtime. Other alignments of relevant superficial coseismic fractures, with centimeters displacement and connected to the SW dipping normal fault, are mapped along the Nottoria - Preci alignment. They run from the Norcia area through the Nera river valley up to the north. Geometrical analysis of these displacements versus distance of the fault planes and ground ruptures was analyzed along several cross-sections orthogonal to the fault strikes. The cross-sections highlight slip accommodation through linkage, which shows to be a common fault growth mechanism. Moreover, the analysis of coseismic ruptures shows that about 40% of the total surface displacement occurred as off-fault deformation, over a mean deformation zone width of a hundred meters. The rupture zone fabric and the o ff-fault deformation is mostly controlled by the structural complexity of the fault system, with a weaker correlation with the rheology of the ruptures materials. HOW DEFINING A QUALITY FACTOR TO CLASSIFY EARTHQUAKE LOCATIONS M. Michele 1 , A. Emolo 2 , D. Latorre 1 1 Istituto Nazionale di Geofisica e Vulcanologia, Rome, Italy 2 Dipartimento di Fisica ‘E. Pancini’, Università Federico II, Naples, Italy After computing an earthquake location, it is necessary to define the uncertainties related to the obtained estimate. Generally, despite of the method used to locate a hypocenter, the solution is validated a posteriori considering different reasonable range of the uncertainty estimators derived after the inversion, like the root mean square (RMS), the number of phases used (NPHS), the azimuthal gap (GAP) and so on. Although this is the standard way to classify the goodness of a seismic location, it is still an unsolved problem to establish an exhaustive and quantitative criterion that provides a simple and accessible classification of earthquake location. Aiming at giving an objective estimate of a seismic location’s quality, we analyze the behavior of different uncertainty estimators, founding that they are dependent on each other (Fig. 1), implying the need to evaluate all the estimators together. For this reason, we combined in an empirical formula the different uncertainty estimator of an earthquake location with the goal of determining a unique value depicting the reliability of the solution. The formula is applied after estimators normalization (so to legitimate the summation of different physical measurements) and a cleaning procedure based on the Chauvenet criterion application for outliers removal. The formula provides a value that we named q f (quality factor), ranging from 0 to 1. The smaller is the q f value, the more constrained the location is. By computing this value, a simple classification of the location’s quality can be done. We defined 4 evenly spaced goodness intervals for q f , corresponding to four quality classes, from A (best class) to D (worst class). To verify the efficiency of the described approach, we applied it to different case studies, sequence and no-sequence seismicity, located by means of different method (linearized and global). Our results confirm a reasonable quality classification for the statistical distribution of parameters, allowing to highlight, just using best quality classes locations, the active seismic structures. The uncertainty estimators used, common for both the cases study, are the root mean square (RMS), the number of phases (NPHS), the azimuthal gap (GAP), the formal errors of horizontal (ERH) and vertical (ERZ) component. Only for the locations derived with the global method, where the hypocenters are provided with a probability density function, additional parameters can be used for adding information about the uncertainty of the location. We added the distance between the pdf expected value and its maximum likelihood (LOCDIST) that