GNGTS 2014 - Atti del 33° Convegno Nazionale

GNGTS 2014 S essione 3.1 103 This site stands on the top of a 35 m high calcareous cliff (Fig. 1b), located in the eastern sector of Ispica village, showing instability conditions since the presence of deep fracturing and karst, as well as of several blocks about to fail. For the reconstruction of the stratigraphic succession, 3 geognostic mechanical drillings were performed in front of the monastery (Fig. 1c). The resulting stratigraphy is represented by a succession of calcarenites with different degree of fracturing; particularly, under a thin talus level, 5 meters of intensely fractured calcarenites are present; between 5 and 20 m b.g.l. the succession shows a lower degree of fracturing; from 20 to 30 m b.g.l. intensely fractured calcarenites were drilled. Deep fracturing, especially close to important tectonic structures, is one of the main causes of slope instability (Barbano et al. , 2014; Pappalardo, 2014). ������������ ��� ���� ������� Furthermore, the high seismic activity of the study area (category II according to the Italian Ordinance OPCM 3274) may be regarded as one of the potential triggers for rockfalls. A geophysical survey has been performed to obtain information on the degree of fracturing of the deep rock mass and to estimate the resonance frequency of the site. Geostructural surveys, kinematic and stability analysis allowed us to better understand the main instability features, as well as to evaluate the safety factors for the most unstable blocks. Methodologies. The study of the environmental microtremors (noise), through passive seismic recordings of single-station H/V (or HVSR), is a non-invasive technique for the investigation of the underground. It allows highlighting the frequencies at which the ground motion is amplified by stratigraphic resonance. The spectral ratios of the vertical and horizontal components of the ground motion, recorded at each measurement station, are taken into account by this technique (������� ��� ��������� ����� �������� ������ ��� ��������� ����� ���������� Nogoshi and Igarashi, 1970; Nakamura 1989). The amplitude ratio associated with the resonance frequencies of the site is an indicator of the “minimum” amplification expected at the site in case of earthquake. In a simple two-layer system, characterized by two different speeds (V 1 and V 2 ) and by two different densities (�ρ 1 and �ρ 2 ), the equation linking the resonance frequency “f” to the thickness “H” of the resonating layer, depends on the shear waves velocity, as shown by: f = nVs/4H (1) where n (= 1, 3, 5 ...) indicates the order of the mode of vibration (fundamental, first superior etc.), Vs and H represent the shear waves velocity and the thickness of the resonating layer respectively. Eq. (1) allows understanding how the H/V technique can also provide information on stratigraphic characters. Indeed, starting from a noise measurement providing f, once known the Vs of the coverage, the depth of the main seismic reflectors or vice versa can be easily estimated (Ibs-von Seht and Wohlenberg, 1999); each peak in the H/V graph corresponds to a possible reflector (seismostratigraphic level) that presents an impedance contrast compared to the neighbor level; the greater the impedance contrast, the greater the amplitude of the peak would be. However, this report does not follow a linear trend. The measurements of environmental microtremors were carried out through the use of a portable digital tromograph TROMINO (Micromed S.p.A.), equipped with three electrodynamic orthogonal sensors (velocimeters) responding in the band 0.1 ÷ 1024 Hz. Seismic noise has been acquired with sampling frequency of 128 Hz, and recorded for 20 minutes at 3 measurement stations (N1-N2-N3) (Fig. 2). These were placed close to the previous drilled boreholes (S1-S2-S3), to verify if such a survey could detect any difference in the physical-mechanical properties of the lithology (i.e. changes in the degree of fracturing), already pointed out by the stratigraphy successions (Fig. 2). The instruments were located on the ground, with the major axis oriented toward north. Recordings were processed through the software Grilla, to evaluate the relationship between the spectral components of the ground motion. Data were then processed by dividing the acquired traces in 20 seconds time-windows. Then, a triangular smoothing of 10% was applied, and the signal was “cleaned”, eliminating