Orientation of Horizontal Components of Short Period Ocean Bottom Seismometers (SPOBSs) using P-wave Polarization.

Abstract:

We present and evaluate a technique to determine the orientation of the horizontal components of the Short Period Ocean Bottom Seismometers (SPOBSs) modified from the P-Polarity method used for teleseismic arrivals.The correct orientation of OBSs, which are generally deployed in free-fall, is crucial for various waveform-based seismic studies, such as focal mechanisms of local earthquakes, receiver functions analyses, SKS splitting and waveform tomography.

We applied principal component analysis to determine the particle motion directions of the first P wave arrival in 2D and 3D in multiple frequency bands, then calculated the orientation angle by constraining the difference between the real and instrument-determined estimated back-azimuths. As SPOBSs are generally insensitive to teleseismic arrivals, we used local events, whose back-azimuth uncertainty can be significant and must therefore be taken into account. Starting with datasets from Lucky Strike volcano (Mid-Atlantic Ridge) and Mayotte, we found that many stations have significant back-azimuth bias. We therefore expanded our study to public datasets from Axial Seamount (Juan de Fuca Ridge) and the TAG and Rainbow hydrothermal fields (Mid-Atlantic Ridge), covering 5 different regions and 5 different types of SPOBSs, in order to separate instrument-based and site-based bias. All of the regions/instruments exhibited some back-azimuth bias for local events, suggesting that at least some of the effect is site-based. We compile the success of the P-polarity method as a function of instrument and site and present recommendations for determining the instrument orientation and interpreting measured SPOBS arrival orientations.