GNGTS 2013 - Atti del 32° Convegno Nazionale

extreme cases: either for very smooth surface (as glass or tiles) for a narrow incidence angle close to the orthogonal direction, or for very rough surfaces (as coarse gravel) for dipping angles exceeding 15 degrees. For more realistic surfaces found in nature, e.g., sand, the signal strength was considerably above the minimum threshold required for reliable detection of Doppler shift. As part of the ongoing uncoupled sensor development and characterization program, field tests were followed by quantitative laboratory measurements on the most recent prototype. The particular measurements described herein consisted of comparisons between the performance of the new prototype and that of a conventional 10-Hz geophone, which is normally used for large scale seismic surveys for oil and gas exploration and production. Needless to say, this and other geophone types require planting by a spike. Fig. 2 displays the shaker table apparatus built and used at the Seismological Research Centre of OGS in Udine, able to generate controlled frequency signals in the range from 0.3 to 100 Hz (Ponton et al. , 2002). This setup was employed to calibrate the response of the uncoupled sensor for a range of spot frequencies. The calibration system hardware includes a shaker table, a laser interferometer, a custom- made interface board that allows for A/D conversion between the laser interferometer and a personal computer. An audio board and a M24 Lennartz data logger is also included (Fig. 2). The shaker table is mounted on a 500 kg reinforced cement base. The system software has been written using MatLab. An automatic procedure was devised to determine the response curves of the sensors. The input, broadband signal, used to drive the shaker table, is designed to fit the characteristics of the sensor type being calibrated (i.e., velocimeter or accelerometer), its nominal peak frequency response and the target frequency range. For short period instruments, reliable response curves can be estimated in the time span of a few minutes, whereas longer acquisition times are required for broadband sensors. Spectral response. Further calibration tests were carried out at the Oceanographic Calibration Laboratory of OGS in Trieste, because its accurate temperature control and good acoustic insulation can minimize the environmental noise. A 24-bit multi-channel acquisition system was used to collect data from a standard geophone and the uncoupled sensor. No filters were applied as our goal was to study the signal properties without distortions. Both sensors measured signals with a level well above the noise floor. The measurements carried out with this system are optimal above 5 Hz, because the shaker tabled used in that case was built and optimized for standard 10-Hz geophones, commonly used for seismic surveys. Instead, the calibration system available at the Centre of Seismological Research in Udine is targeting quite lower frequencies, up to 0.1 Hz, which are relevant for earthquake monitoring. The integration of these two frequency ranges was needed, therefore, for appreciating the capabilities of the new device. Fig. 3 shows the amplitude spectra of signals provided by a 10-Hz geophone and the uncoupled acoustic sensor. The spectra obtained from both receivers display a major peak at the expected frequency of 10 Hz, and a similar trend for the noise at other frequencies. Away from the peak the standard geophone shows essentially the ambient noise above the acquisition system noise floor. Assuming that the ambient noise is identical for both the geophone and the new sensor, the noise level of the latter is higher (about 20 dB) within the band of frequencies between 1 and 100 Hz. We expect that by Fig. 3 – Response at shaker table for 10 Hz from the uncoupled acoustic sensor (grey), a 10-Hz standard geophone (red) and the acquisition system noise (magenta). 79 GNGTS 2013 S essione 3.1