Receiver Deghosting of High-Resolution Shallow-Tow Seismic Data

The increasing demand for offshore wind farms and the need for detailed sub-bottom geotechnical information has led to an increase in shallow-tow high-resolution or ultra-high-resolution seismic acquisition. However, the resulting data presents several processing challenges, particularly in receiver deghosting.

Shallow-tow streamers and sources are more susceptible to wave interaction, leading to source variability and receiver datum issues. Additionally, the commonly used flat surface model with a constant reflection coefficient may be inadequate for approximating acoustic waves reflecting from an undulating water surface. While trace-by-trace deghosting is effective for deeper-towed streamers, it may present challenges for shallow-tow data with a single visible ghost notch.

To address these challenges, we propose a multi-layered approach to deghosting shallow-tow high-resolution seismic data. The first layer involves streamer-by-streamer sparse estimation to calculate ghost-free waveforms and ghost delays in short windows around the first arrival times for each trace. The results are used as initial values for the non-linear second layer, where multiple shots are analysed together to yield time- and offset-dependent delay times, reflection coefficients and signal spectra. The final layer utilises this information to deghost the original seismic data, providing improved results compared to traditional receiver deghosting techniques.