GNGTS 2013 - Atti del 32° Convegno Nazionale

Results and Discussions. These results have to be considered as only preliminary (also given the limited time available), to be validated and tested, but they open the discussion on the usefulness of the technique on these types of applications. The WP 1 does not provide any interpretation of the results in terms of anomalies, due to the fact that the peaks falling outside of the measurement errors are few and not easily ascribable to seismic activity time changes (as by comparison of the temporal trends in Fig. 1). It is noteworthy, in fact, that within this project, we applied for the first time the seismic noise cross-correlation analysis on an area affected by earthquake swarms (Pollino); and on a floodplain affected by mild-strong earthquakes in 2012 (Emilia). Both situations are specific and involve several possible interpretations: Pollino: i) Seismic swarms (low-energy seismicity concentrated in time and space) may not be able to generate variations of the crustal parameters such as to be detectable with this technique which average the results over large volumes of sampled crust: the stations in this area are rather scattered, perhaps it would be more appropriate to analyze this kind of low- energy seismicity through more dense networks of stations. ii) It is intriguing that even a Ml 5 did not generated a significant crustal damaging: in Fig. 1A a velocity decrease begins before the occurrence of the event (indicated by the last red vertical line), and does not seem significantly different from other decreasing trends that can be observed throughout the whole time period. It may be necessary a stronger data manipulation as the use of adaptive filters or other techniques to increase the cross-correlation convergence. The velocity variations may be hidden by possible noise source changes: the geometry of the stations necessarily follows the trend of the geographic region, and the majority of the station pairs are oriented along the Fig. 3 – Averaged trends, in the period late 2012 to early 2013, of the parameters studied by the two WPs of UR2 in the Pollino area, compared with the trend of Vp/Vs in the same area. 113 GNGTS 2013 S essione 2.1