GFZ

N/A

Il Monte Amiata, ubicato nella porzione sud-occidentale della regione Toscana, è un edificio vulcanico che si è strutturato durante la parte finale del Pleistocene medio (350 - 200 ka; Laurenzi et al., 2015; Principe et al., 2018) al di sopra delle unità tettoniche strutturatesi durante le fasi mio-plioceniche dell’orogenesi appenninica. La distribuzione dei centri eruttivi sembra essere controllata da una zona di debolezza strutturale plio-pleistocenica, orientata circa NE-SW, che interessa sia i depositi vulcanici che le unità strutturali sottostanti (Brogi & Fabbrini, 2009, Brogi et al., 2015; Piccardi et al., 2017, Principe et al., 2018). Il gradiente geotermico è caratterizzato da valori molto alti (fino a 15°/100m), rendendo l’area particolarmente idonea per la produzione di energia geotermica. La produzione geotermica iniziò a partire dal 1960. Attualmente, gli impianti produttivi di ENEL- Greenpower di Bagnore e Piancastagnaio (Fig. 1), sfruttano un serbatoio geotermico collocato tra i 2000 e i 3500 metri di profondità rispetto al piano campagna. Il Catalogo Parametrico dei Terremoti Italiani (CPTI; Rovida et al., 2016 riporta, tra il 1287 e il 1940, 13 terremoti con una magnitudo equivalente compresa tra 4.5 £ Me £ 5.3 che hanno causato danneggiamenti fino al grado VIII MCS (Fig. 1), evidenziando un’attività sismica naturale e capace di causare seri danneggiamenti, ben prima dell’inizio dello sfruttamento geotermico dell’area. La sismicità recente, registrata dalla rete sismica nazionale dell’INGV (Castello et al., 2006; http://cnt.rm.ingv.it), riporta meno di 150 terremoti nell’area amiatina negli ultimi 25 anni, di cui 35 eventi con ML ≥ 1.5. Tra questi, il terremoto del 1.4.2000 (Md=4.0; http://cnt.rm.ingv.it/event/1132509) causò danni ad oltre 50 edifici, e la prossimità dell’epicentro con l’impianto di produzione di Piancastagnaio sollevò l’ipotesi di una sua natura antropogenica (Mucciarelli et al. 2001). Braun et al. (2018) hanno ricalcolato ipocentro e meccanismo focale di questo evento, collocandolo ad una profondità prossima al serbatoio di produzione, giungendo però alla conclusione che non sia possibile, per questa via, discriminare la sua natura antropogenica o meno. In generale, rispetto alle profondità tipiche della sismicità crostale osservata in Toscana (tra circa 5 e 13 km) gli ipocentri degli eventi sismici registrati nell’area amiatina hanno delle profondità simili a quelle di produzione (< 5 km). La bassa densità della rete di monitoraggio INGV in quest’area del territorio nazionale (Fig. 1) è causa, comunque, di una bassa capacità di rilevazione (detection) sismica e di una altrettanto bassa capacità di risoluzione ipocentrale. Per migliorare le capacità di detection e di monitoraggio sismico nell’area del Monte Amiata, nel periodo 2015 - 2018 abbiamo installato una rete locale composta da 8 stazioni in vicinanza delle centrali di produzione geotermica di Bagnore e Piancastagnaio. L’obiettivo dell’esperimento era quello di abbassare la magnitudo di completezza e di comprendere meglio l’origine della sismicità in vicinanza degli impianti di estrazione, cercando di discriminare tra sismicità naturale e eventi sismici antropogenici. A questo scopo, abbiamo applicato una metodologia di analisi automatica, scansionando l’enorme dataset con un nuovo e robusto approccio di detection e localizzazione, chiamato waveform beam-forming grid search approach (LASSIE; Heimann et al., 2017). In uno step successivo, gli eventi sismici associati vengono rilocalizzati con un waveform-based locator (LOKI: Grigoli et al. 2014). Il catalogo sismico così ottenuto, aggiornato e molto più completo rispetto a quanto mai ottenuto prima in termini di Magnitudo di completezza (Mc), rappresenta la base per definire criteri di discriminazione, ad esempio attraverso la correlazione spazio-temporale della sismicità osservata con i parametri di produzione geotermica. I risultati ottenuti e le potenzialità di tale approccio saranno oggetto della presentazione qui proposta.

We apply the Bakun and Wentworth (Bull Seism Soc Am 87:1502–1521, 1997) method to determine the location and magnitude of earthquakes occurred in Central Asia using MSK-64 intensity assignments. The attenuation model previously derived and validated by Bindi et al. (Geophys J Int, 2013) is used to analyse 21 earthquakes that occurred over the period 1885–1964, and the estimated locations and magnitudes are compared to values available in literature. Bootstrap analyses are performed to estimate the confidence intervals of the intensity magnitudes, as well as to quantify the location uncertainty. The analyses of seven significant earthquakes for the hazard assessment are presented in detail, including three large historical earthquakes that struck the northern Tien-Shan between the end of the nineteenth and the beginning of the twentieth centuries: the 1887, M 7.3 Verny, the 1889, M 8.3 Chilik and the 1911, M 8.2 Kemin earthquakes. Regarding the 1911, Kemin earthquake the magnitude values estimated from intensity data are lower (i.e.MILH=7.8 andMIW=7.6 considering surface wave and moment magnitude, respectively) than the value M=8.2 listed in the considered catalog. These values are more in agreement with the value MS=7.8 revised by Abe and Noguchi (Phys Earth Planet In, 33:1–11, 1983b) for the surface wave magnitude. For the Kemin earthquake, the distribution of the bootstrap solutions for the intensity centre reveal two minima, indicating that the distribution of intensity assignments do not constrain a unique solution. This is in agreement with the complex source rupture history of the Kemin earthquake, which involved several fault segments with different strike orientations, dipping angles and focal mechanisms (e.g. Delvaux et al. in Russ Geol Geophys 42:1167–1177, 2001; Arrowsmith et al. in Eos Trans Am Geophys Union 86(52), 2005). Two possible locations for the intensity centre are obtained. The first is located on the easternmost sub-faults (i.e. the Aksu and Chon-Aksu segments), wheremost of the seismicmoment was released (Arrowsmith et al. in Eos Trans Am Geophys Union 86(52), 2005). The second location is located on the westernmost sub-faults (i.e. the Dzhil'-Aryk segment), close to the intensity centre location obtained for the 1938, M 6.9 Chu-Kemin earthquake (MILH=6.9 and MIW=6.8).

No abstract

In this work, we investigate the site amplification effects observed in the Norcia plain, Central Italy. Data from 30 selected local earthquakes (2 ≤ Ml ≤ 4.1) recorded by a temporary seismic network composed by 15 stations, are analyzed to determine the spa- tial variability of site effects. Both the Horizontal-to-Vertical spectral ratio and the Standard Spectral Ratio techniques are applied to estimate the site amplification effects. The results show that most of the sites in the valley are affected by strong amplifications (up to a factor of 20) in the frequency range 0.5–5 Hz. The value of the fundamental frequency of resonance is strictly dependent on the location within the basin and on the sediment thickness. Strong amplifications also affect the vertical components. The time-frequency analysis performed on a station located inside the basin shows the presence of a large spectral amplitudes after the S-wave phase, not observed on a station located on the bedrock, suggesting the pres- ence of locally generated wave trains. Then, in agreement with earlier observations for other alluvial basins in Central Italy, 2D–3D effects play an important role in determining the site amplification effects in Norcia.

Kyrgyzstan, which is located in the collision zone between the Eurasian and Indo-Australian lithosphere plates, is prone to large earthquakes as shown by its historical seismicity. Hence, an increase in the knowledge and awareness by local authorities and decision makers of the possible consequence of a large earthquake, based on improved seismic hazard assessments and realistic earthquake risk scenarios, is mandatory to mitigate the effects of an earthquake. To this regard, the Central Asia Cross-Border Natural Disaster Prevention (CASCADE) project aims to install a cross- border seismological and strong motion network in Central Asia and to support microzonation activities for the capitals of Kyrgyzstan, Uzbekistan, Kazakhstan, Tajikistan, and Turkmenistan. During the first phase of the project, a temporary seismological network of 19 stations was installed in the city of Bishkek, the capital of Kyrgyzstan. Moreover, single-station noise recordings were collected at nearly 200 sites. In this study, the site amplifications occurring in Bishkek are assessed by analyzing 56 earthquakes extracted from the data streams continuously acquired by the network, as well as from the single-station noise measurements. A broadband amplification (starting at ∼0:1 and 0.2 Hz), is shown by the standard spectral ratio (SSR) results of the stations located within the basin. The reliability of the observed low-frequency amplification was validated through a time–frequency analysis of denoised seismograms. Discrepancies between horizontal-to-vertical spectral ratio and SSR results are due to the large amplification of the vertical component of ground motion, probably due to the effect of converted waves. The single-station noise results, once their reliability was assessed by their comparison with the earthquake data, have been used to produce the first fundamental resonance frequency map for Bishkek, whose spatial variation shows a good agreement with the presence of an impedance contrast within the Tertiary sedimentary cover.

Providing quantitative microzonation results that can be taken into account in urban land-use plans is a challenging task that requires collaborative efforts between the seismological and engineering communities. In this study, starting from the results obtained by extensive geophysical and seismological investigations, we propose and apply an approach to the Gubbio basin (Italy) that can be easily implemented for cases of moderate-to-low ground motion and that takes into account not only simple 1D, but also more complicated 3D effects. With this method, the sites inside the basin are classified by their fundamental resonance frequencies, estimated from the horizontal-to-vertical spectral ratio applied to noise recordings (HVNSR). The correspondence between estimates of the fundamental frequency from this method and those derived from earthquake recordings was verified at several calibration sites. The amplification factors used to correct the response spectra are computed by the ratio between the response spectra at sites within the basin and the response spectra at a hard-rock site using data from two seismic transects. Empirical amplification functions are then assigned to the fundamental frequencies after applying an interpolation technique. The suitability of the estimated site-specific correction factors for response spectra was verified by computing synthetic response spectra for stations within the basin, starting from the synthetic recording at a nearby rock station, and comparing them with observed ones.