Efficient and accurate modeling of ocean bottom seismometer data using reciprocity

Abstract

Seismic experiments in which the number of sources is considerably larger than the number of receivers occur regularly. An important example is the collection of crustal scale seismic data using ocean bottom seismometers and marine sources. We describe a method to accurately and efficiently compute synthetic seismograms for such experiments by using finite differences and reciprocity. We show numerically how to decompose an explosive source into its equivalent body force components using the staggered-grid finite-difference technique with a fourth-order approximation for the spatial derivative and a second-order approximation for the temporal derivative. This decomposition results in a source configuration where the equivalent body forces are defined in 12 points around the point where the ex-plosive source is applied. We then use the derived equivalent body forces for the explosive source and seismic reciprocity theorems to convert the common receiver gather to a common shot gather. The method is tested on a structurally complex elastic model of the crust and the results show that it is accurate within floating point precision. The synthetic data are compared to data from a real ocean bottom seismometer experiment conducted across a continent-ocean transition zone. A good fit in terms of traveltime is observed for many of the prominent seismic phases. The amplitude fit of these arrivals is not always as good as the traveltime fit. This indicates that full-waveform modeling of such data can provide useful information about the subsurface that cannot be obtained from traveltime modeling. If enough data are available, the modeling method can be used in full-waveform inversion.