Figures 18,19,20,21 show the pattern of the seismic stations shooted during the cruise (Table 19). The shooting mode was based on time (every 9-120 s) rather than on distance, since the compressor's air delivery could have been more easily controlled. Direct communications with INGV people monitoring the on-land network was able to provide information for fine tuning of the shooting lines.
OBS were recovered 2006-12-02. Appendix 7.4 reports the coordinates of release and overboard recovery of the OBS stations (Tables 16, 17) and the clock drifts (Table 18).
Figures 16, and 17 show the locations of OBS 1,2 and 3, derived from trilateration of range measurements by pinger.
figure
 |
 |
 |
 |
figure
The analysis of the recording of shots made by a seismometer on board (Fig. 22) and its comparison with shots table was able to evidence some problems in timing and signature quality for days 29 and 30 november. The data downloaded from each OBS after their recovery on board were checked for clock drifts and assembled in chunks of 40 s long records in the SEG-Y format, starting at the whole second of every Seismic Shot. Other SEG-Y files accomodated the entire OBS data set. Appendix 7.4 gives furter details and description of the conversion procedure to the SAC format ([SAC (2006)],[Goldstein et al. (2003)]). All OBS well recorded seismic data except for OBS06, due to hardware failure.
In Figure 24 the vertical component seismograms recorded by 9 OBS after shot 5/L39 (2006-12-02T01:46:38.090, ESE of Stromboli, fig.19 and 23) are presented. Planar distances from the source to the nearest (03) and the farthest (07) OBS range between few m (nearly vertical) and 11 km, while distance from source to the Stromboli craters area is about 8 km. See in table 14 the distances from the recording stations. Recording conditions (signal-to-noise ratio "S/N") were almost good enough to allow all OBS to detect the signals, while data quality from the different OBS is variable, depending on the distance from seismic source. OBS data close to source are generally of good quality (Fig. 24), left), and in several cases clear second and third (exceptionally up to fourth) arrivals can be identified (Fig. 25 , right). Most of the records display an impulsive first arrival recorded only at close distances from the shot site. On distant shots data quality is poorer, however picking procedure of incoming waves is still possible on the emergent phases by applying a filter. From the pattern of recorded seismograms, we may deduce that the data quality of each OBS is probably depending on the local geology and considerably affected by the presence of the volcanic structures. These latter are characterised by remarkable velocity variations in the lateral direction, and are likely to influence the spreading of seismic phases.
|
figure
 |
figure
figure
