Paper Summary
Acoustic FWI has been widely applied in seismic exploration. Nevertheless, acoustic FWI results could be negatively affected by the presence of elastic effects, although the data is dominated by compressional waves and elastic phenomena are less evident in certain cases than others, depending on the geological scenarios and the type of data acquired. These effects commonly manifest themselves as variations of amplitude and phase across the offset. In this study the authors address the technical challenges associated with building a high resolution velocity model by acoustic FWI focusing on a dataset exhibiting AVO variations at target reservoir depths. Through data decomposition in angle domain, the input data for FWI is selected such that certain phase effects (e.g., reversals) in AVO variations are avoided while the amplitude variations are handled well by dynamic matching FWI.