GNGTS 2019 - Atti del 38° Convegno Nazionale

GNGTS 2019 S essione 2.2 349 of horizontal ground motion in the slope direction or perpendicular to the ridge crest (Spudich, 1996; Paolucci, 2002). During periods of higher amplitude background noise, generally related to weather conditions, we observe the most relevant variations in HVSR curves. To establish a relation between the noise amplitude and the variations of the HVSR curves we computed the rms of continuous signals over 4 minute sliding window, after an appropriate bandpass filtering. Hereafter we use “weak noise” expression to indicate low amplitude seismic noise during good weather periods, and “strong noise” expression to indicate higher amplitude seismic noise during low-pressure periods. From these observations we obtain a first important result: the most striking variations of HVSR occur during strong noise periods, in fact at all the sites considered the HVSR shows the appearance of one or more peaks of relevant amplitude during strong noise periods, that are not present during normal weak noise periods. The most interesting results, that means the most variable HVSRs, have been observed at the seven sites shown in Fig. 1. Here we give a detailed description for only two sites for shortness reasons. Fig. 2 shows the results of data recorded at site TP02 by a temporary seismic station (equipped with Lennartz 1 Hz seismometer and 24 bit data logger) installed for some months in the basement of a church in the historical center of Tortora village. This site is located on a narrow N-S ridge where few meters of Pliocene sands and conglomerates overlie the Fig. 2 - Results of the analysis of seismic noise at site TP02. Plot a shows the rms computed for 6 days of signals bandpass filtered in the frequency band: 1.3 - 4 Hz. Plots b and c show the plain and directional H/V spectral ratio of weak signal computed in the 8 hours time window shown by background color in plot a. Plots d and e show the plain and directional HVSR of strong signal.

RkJQdWJsaXNoZXIy MjQ4NzI=