Wavefield separation by energy norm Born scattering

Abstract

In Reflection Based Waveform Inversion, the gradient is computed by cross-correlating the direct and Born scattered wavefield with their adjoints applied to the data residuals. In this case, the transmitted part of the Born scattered wavefield produces high wavenumber artifacts, which would harm the convergence of the inversion process. We propose an efficient Energy Norm Born Scattering (ENBS) to attenuate the transmission components of the Born modeling, and allow it to produce only reflections. ENBS is derived from the adjoint of the Energy Norm (inverse scattering) imaging condition and in order to get deeper insights of how this method works, we show analytically that given an image, in which reflectivity is represented by a Dirac delta function, ENBS attenuates transmission energy perfectly. We use numerical examples to demonstrate that ENBS works in both the time and the frequency domain. We also show that in reflection waveform inversion (RWI) the wave path constructed by ENBS would be cleaner and free of high wavenumber artifacts associated with conventional Born scattering.