Modelling, Migration, and Inversion for Angle Domain Common Image Gathers

Ziguang Su

This thesis explores the feasibility, performance, and interpretability of Distributed Acoustic Sensing (DAS) seismic data, acquired in an experimental, shaped, deployment. This research addresses one of the main limitations of DAS-its directionality and lack of sensitivity to broadside seismic energy-by testing a novel, low-cost, permanent, multi-component DAS sensor array referred to as the Croissant. The Croissant consists of three sensors, each consisting of fiber wrapped around two 1x1m vertical plastic frames set at right angles. In this thesis I investigate whether the integration of compact, point-based multicomponent sensors of this kind into otherwise single-component DAS systems can overcome directional limitations and enhance seismic monitoring, while preserving the cost-effectiveness of the technology. The study is based on multi-year field data recorded between 2023 and 2025 at the Carbon Management Canada (CMC) Newell County facility using a P-wave vibrator along a circular source line. This experimental dataset enables the analysis of the Croissant's response to a wide range of azimuths and offsets, allowing for an evaluation of its directional sensitivity. Initial observations showed consistently strong vertical responses, leading to the hypothesis that refracted waves may be arriving with dominant vertical energy. To test this hypothesis, a geometric DAS model was developed by parameterizing the Croissant's fiber trajectory according to its physical design and wrapping pattern. Synthetic shot gathers were generated using both an analytical model and a 3D elastic finite-difference simulation (ElasWave3D) and compared with co-located geophone measurements for validation. Results showed strong agreement between synthetic and field data: the Croissant reliably captured vertical and horizontal strain, and amplitude decay followed expected trends. Synthetic geophone data supported the interpretation of refracted energy in the experimental DAS recordings. The findings support the Croissant's effectiveness as a point-based, multicomponent DAS sensor. Improvements in signal clarity observed from 2023 to 2025 likely reflect improved sensor coupling due to soil compaction. While each sensor currently uses ~28 meters of fiber using a gauge length of 7 meters, early results suggest that could be reduced to ~14 meters under certain conditions. Co-located geophones provided critical reference data, reinforcing the directional trends and validating the Croissant DAS wave field interpretations.