An experimental comparison of three direct methods of marine source signature estimation

Three methods for estimation of the pressure wavefield generated by a marine airgun array are tested experimentally and compared. In the trial a variety of radiation angles and array configurations were used and some large synchronization errors were deliberately introduced. The source was equipped with near-field hydrophones and a subsource ministreamer. A tethered far-field hydrophone was used so that the three estimated far-field signatures could be compared with an independent measurement. The knowledge of the source signature is important for on-board source array QC, deconvolution, multiple attenuation, stratigraphic trap prediction, modelling and inversion, AVO analysis and reservoir monitoring. The methods perform very well and give estimates whose frequency-domain spectra match the measured spectra to within a few dB and within a few tens of degrees of phase over the tested bandwidth of 3.5-110 Ha. The time-domain error-energy is typically only a few per cent of the signal energy for radiation angles within about 30 degrees of the vertical. The third method proved to be sensitive to an experiment;al shortcoming leading to overloading of the ministreamer and meaningful comparison was not possible for some test configurations.