Model Building in Complex Geological Situations Using Low-Frequency Data from an Optimised Airgun Technology Based Source

Abstract

Summary The emergence of seismic inversion techniques such as full-waveform inversion (FWI) has provided a significant increase in the resolution that may be achieved in models of physical earth properties. It has become clear that achieving optimal results i.e. a broad band of model wavenumbers implies data with good azimuthal coverage, long offsets, full frequency coverage and, ideally, inversion of multiscattered energy. Modern ocean-bottom node surveys commonly provide good azimuthal and offset coverage, however such surveys often still do not provide good quality data at low frequencies. This is due to deficiencies in both low frequency energy generation by the source and low frequency energy detection at the receiver. Consequently, there has been considerable recent interest in improving the low-frequency performance of seismic sources. In this paper we discuss a survey recently conducted in the Western Gulf of Mexico in which a new seismic source based on airgun technology was utilized with an array of sparse ocean-bottom nodes. We show that the increase in signal-to-noise at low frequencies (1.5–4Hz) that is achieved by the new source (relative to a conventional multi-airgun array) leads to an improvement in the resulting velocity model derived using FWI.