Paper submitted to EAGE Annual 2026
Summary
3D Ultra-High-Resolution Seismic (3D UHRS) surveys are increasingly employed for offshore windfarm site characterisation, where accurate near-surface imaging is vital for foundation design and for reducing geotechnical uncertainty. These datasets require advanced processing workflows that move beyond conventional time-domain methods, critically incorporating velocity-model building (VMB) and depth imaging tailored to the distinct characteristics of ultra-high-resolution data (Limonta et al., 2024). Traditionally, site surveys have relied on sparse 2D lines or small 3D volumes, applying basic time-domain velocity analyses, typically semblance-based picking, to flatten common-midpoint (CMP) gathers and improve stacking response. While this approach yields RMS velocities, these values only approximate subsurface velocity variations and do not provide accurate interval compressional velocity (Vp) information, nor do they capture detailed soil property variations essential for engineering purposes. To overcome these limitations, TGS implemented a comprehensive 3D depth-domain velocity-model building workflow, combining Kirchhoff Pre-Stack Depth Migration (KPSDM), 3D tomographic inversion, and Dynamic Matching Full Waveform Inversion (DMFWI) (Huang et al., 2023) with frequencies up to 600 Hz. Interval compressional velocity (Vp) is particularly significant for quantitative interpretation (QI) and for estimating soil parameters such as uniaxial compressive strength (UCS) (Lindh and Lemenkova, 2022). We further emphasise the importance of validating seismic velocity models by direct comparison with geotechnical measurements, such as Seismic Cone Penetration Test (SCPT) results, to ensure both their accuracy and reliability.