Spatial prediction filters for attenuation of seismic interference noise

The interest in the subject of time series analysis started in mid 1920s. The use of real valued prediction filters for predicting and attenuating temporal reverberations present in the recorded seismic signals, known as deconvolution, emerged in mid 1950s. The use of complex valued prediction filters along space axis on constant frequency seismic data for coherent signal prediction and random noise attenuation took place in mid 1980s. When coherent noise can be viewed from a dimension where it is incoherent then a similar filter can be used to attenuate such noise. Here we present the application of this idea to the attenuation of strong coherent noise, known as interference noise, generated by other seismic boats surveying the same area.