GNGTS 2019 - Atti del 38° Convegno Nazionale

GNGTS 2019 S essione 1.1 121 The attenuation curves describing the decay of macroseismic intensity of these three events, elaborated by using a distance-binning approach similar to that used by Gasperini 2001, showed a marked difference in the near-field, which seemed to be due to their different hypocentral depth. Analyzing other earthquakes included in the HSIT database, occurred in the same region, we found a correlation between the depth of the earthquakes, and the slope of the attenuation curve within 50 km of the epicenter, regardless of their magnitude. We used a straight line to fit the intensity data within the first 50 km from the epicenter because, as already noted by Gasperini 2001, the attenuation curves of Italian earthquakes can be approximated by a bilinear function showing a change in slope between 40 and 50 km. An abrupt change in the slope of the attenuation curve was also evidenced at about the same epicentral distance by using a numerical simulation of the attenuation of the PGA (Fah and Panza, 1994). Our experimental observations arising from the analysis of the attenuation curves of the HSIT dataset shows that generally the change in slope is closer to 50 km. We calculated the absolute value of the slope of the best-fitting straight lines for a total of 20 earthquakes and plotted each one against its hypocentral depth. The resulting distribution of data points is well fitted by a logarithmic function. The trend of slope of the lines, fitting intensity attenuation with distance versus depth, agrees with both an Intensity Prediction Equation (IPE) and a Ground Motion Prediction Equation (GMPE), thus confirming the reliability of experimental behaviour. In particular we show that the values of slope computed through the IPE elaborated with HSIT intensity data and those obtained by the transformation of PGA in MCS (Faenza and Michelini, 2010), the former predicted by using the GMPE of Bindi et al. (2014), are consistent with our experimental function. We then used such function as a modern “Rosetta Stone” to uncover the hypocentral depth range of 20 selected pre-instrumental earthquakes ranging in age from the 1570 “Ferrara” to the 1972 “Appennino settentrionale” events, and with a magnitude range of 4.9 to 6.5. Our statistics show that the entire approach is independent of the magnitude of the earthquake being considered but depends only on its focal depth (Fig. 2), and as such it is ideal for application to pre-instrumental earthquakes. Besides, knowing the focal depth of these earthquakes also allows for recalculating their equivalent magnitude by inverting for the magnitude the IPE proposed byTosi et al. , 2015. By applying this formula, we obtain M L magnitude values, which we converted into M W using the equation proposed by Grünthal and Wahlström (2003). The resulting equivalent magnitudes are consistently larger for deeper events (i.e. depth larger than 30 km) than those calculated using traditional methods based on intensity alone. We present case histories referred to selected events and discuss the seismotectonic implication. As an example, our study showed the occurrence of several earthquakes in the Po Plain having hypocentral depths larger Fig. 2 - attenuation curves of three events occurred at similar hypocentral depth but having different magnitude. The slope of the best-fitting line in the first 50 km is similar among the three, attesting its independence from the magnitude.