Application of Source Independent Full Waveform Inversion to Land Seismic Data: A Case Study of Donggi Field, Indonesia

Full-waveform inversion (FWI) is a data-fitting process based on seismic full-wavefield modeling that provides high-resolution images of subsurface geological structures from observed field data. However, its applicability to real-world scenarios, particularly in land data, is limited due to several factors. These factors include the lack of modeling of strong surface waves, converted waves, phase dispersion, near-surface weathering layer distortion, and wave phenomena such as attenuation energy and distortion because of complex geology. Additionally, the source wavelet information is often missing for land surveys, and recorded waveforms can vary significantly due to spatially variant shot characteristics such as source charge size, shot depth, and hole pattern, as well as the distortion caused by the near-surface weathering layer. In this study, we address the main challenges of seismic data with a low signal-to-noise ratio and an unknown correct source wavelet. To solve for the unknown source wavelet, we use a source-independent objective function and perform several preprocessing techniques on the seismic input to improve the signal-to-noise ratio without significantly changing the amplitude. We use velocity analysis data as a prior model to reduce the non-linear problem. We demonstrate the effectiveness of our technique by applying it to land seismic data from the Donggi Basin in Indonesia, and our results show that the velocity model improves and has a better correlation with well-log data.