GNGTS 2017 - 36° Convegno Nazionale

30 GNGTS 2017 S essione 1.1 INTENSITY-ATTENUATION RELATIONSHIPS FOR STRONG APENNINES EARTHQUAKES: QUESTIONING THE STABILITY OF REGRESSION COEFFICIENTS OVER TIME AND SPACE BASED ON MCS AND ESI SCALES M.F. Ferrario, A.M. Michetti, F. Livio Università degli Studi dell’Insubria, Dipartimento di Scienza e Alta Tecnologia, Como, Italy Introduction. The reliable evaluation of historical earthquakes is pivotal for seismic hazard assessment, and macroseismic observations are the only data available for the estimation of location andmagnitude of historical events. Seismic parameters for pre-instrumental earthquakes included in the Italian catalogue (CPTI15, Rovida et al. , 2016) are estimated on the basis of well-constrained empirical relations between Io and Mw, and also through dedicated algorithms (i.e., Boxer code; Gasperini et al. , 1999, 2010). However, the application of traditional, damage-based macroseismic scales (i.e., MCS – Mercalli Cancani Sieberg; MMI – Modified Mercalli; EMS98 – European Macroseismic Scale) to contemporary earthquakes suffer some limitations such as the saturation at higher degrees and a strong dependence on the spatial distribution of human settlements. Most pertinent for this research, these scales are significantly influenced by the large variability of building vulnerability as a function of age and local economy. In this sense, the seismic sequence occurred in Emilia Romagna region in 2012 is particularly relevant. Seismic parameters estimated from the MCS and EMS survey do not match those instrumentally recorded or expected from the application of general empirical regressions (e.g., Galli et al. , 2012; Graziani et al. , 2015). This has been attributed to the concomitant role exerted by the local geologic setting and the characteristics of the building stock, or to overestimated intensity of strong seismic events in the historical catalogue. A complementary approach is provided by the ESI07 (Environmental Seismic Intensity) scale. This scale is based solely on Earthquake Environmental Effects (EEEs), which have the significant advantage of still increase in their dimensions also close to full scale and to respond consistently in time and space (Michetti et al. , 2007). The goal of this note is twofold: (i) evaluate the reliability of EEEs attenuation with distance using a preliminary dataset of Italian strong earthquakes with normal kinematic and (ii) pursue the integration of MCS and ESI scales, which result in a more comprehensive picture of the historical and modern seismic events. We focus on the MCS scale, for which a much larger database is available; arguably, the preliminary results described in the following apply also to the other damage-based scales. The investigated database. We gathered data of all the events in the Italian Apennines for which both MCS and ESI macroseismic data points (MDPs) are available; the database consists of 14 normal faulting earthquakes (Tab. 1) in the magnitude range 6.0 – 7.1, occurred between 1688 and 2016 AD, for a total of 6419 and 431 MDPs for MCS and ESI scales, respectively. MDPs range from intensity 2 to 11 for MCS and from 3 to 11 for ESI. Traditional intensities were compiled from the DBMI database (Locati et al. , 2016; for the Amatrice 2016 event: Galli et al. , 2016; 2017), whereas ESI values were collected from the EEE catalogue (Guerrieri et al. , 2011) and published literature (Serva et al. , 2007; Comerci et al. , 2015). Macroseismic epicenter . The macroseismic epicenter was calculated with Boxer algorithm (v. 3.3) for the 2 separate datasets, as the barycenter of MDPs with highest intensity (“Method 0” in Gasperini et al. , 2010). MCS-based epicenters were collected from CPTI database, while ESI-based ones were computed in this study. The ESI-derived macroseismic epicenter agrees well with the MCS-derived one, with mean differences of 12.2 ± 10.9 km. The epicenter calculated from ESI data should be more closely connected to the surface projection of the seismogenic source than the MCS epicenter. This proved true for at least the most recent events, i.e., those where the seismogenic source is well-known, and earthquake surface faulting has been observed (Colfiorito 1997; L’Aquila 2009; Amatrice 2016).