3-D Multi-Component Reverse Time Migration Method for Tunnel Seismic Data

Peng Guan, Cuifa Shao, Yuyong Jiao, Guohua Zhang, Bin Li, Jie Zhou, Pei Huang2021 · DOI

TL;DR

This paper introduces a 3D multi-component Reverse Time Migration (RTM) method specifically designed for tunnel seismic data. By leveraging all three components of seismic waves and employing a full wavefield migration, this approach significantly improves imaging accuracy compared to traditional methods, especially in challenging tunnel environments with limited data and small offsets.

What Problem Does This Paper Address?

Tunnel seismic data acquisition is inherently challenging due to confined spaces, resulting in small datasets with limited offsets. Traditional seismic imaging methods, like Kirchhoff migration, struggle with these limitations. Kirchhoff migration, based on ray theory, has a restricted migration aperture and low accuracy, particularly for small offsets. This makes it difficult to accurately image geological anomalies like fault zones, karst caves, and water-rich areas that pose risks to tunnel construction and safety. The paper aims to overcome these limitations by developing a more robust 3D imaging technique that can effectively utilize the available multi-component data.

Methods

The core of this research is the development and application of a 3D Multi-Component Reverse Time Migration (RTM) method. RTM is a full wavefield migration technique that uses the wave equation to model seismic wave propagation both forward and backward in time. This allows it to overcome the aperture limitations of ray-based methods.

Key aspects of their method include:

  • 3D Finite-Difference Time-Domain Equation: The RTM algorithm is implemented using a 3D acoustic wave equation solved via a second-order finite-difference method in time and a twelfth-order method in space. This ensures accurate wave propagation modeling.
  • Multi-Component Data Integration: Unlike single-component methods, this approach utilizes all three components (X, Y, and Z) of the seismic wavefield. This allows for a more comprehensive understanding of the subsurface and enhances the imaging of anomalous interfaces.
  • Forward and Reverse Continuation: The process involves two main steps:
  • Forward Continuation: Simulating wave propagation from the source.
  • Reverse Continuation: Propagating the recorded wavefield backward in time from the receivers.
  • Cross-Correlation Imaging Condition: The imaging is achieved by cross-correlating the forward and reverse propagated wavefields at each time step.
  • Laplace Filter: A Laplace filter is applied to remove low-frequency noise from the final migration results.
  • Velocity Model: The velocity model is derived from the first arrival times of the direct wave, which is a standard practice in tunnel seismic data processing.

Datasets

The paper utilizes two types of datasets:

  • Modeled Data: A synthetic 3D tunnel model was created to simulate karst cave and interlayer scenarios. This model had dimensions of 50m x 200m x 50m with a grid spacing of 0.5m. It included two geophones (R-1 and R-2) and 24 shot points. This synthetic dataset allowed for controlled analysis of the RTM method's performance.
  • Measured Data: Real seismic data collected from a water diversion tunnel in Chongqing, China. This data was acquired using two three-component geophones and 24 shot points. The tunnel environment presented challenges like limestone and argillaceous limestone, karst caves, and potential mud inrush disasters.

Key Insights and Ideas

  • RTM's Superiority for Small Offsets: The study strongly demonstrates that RTM, by considering the full wavefield, effectively overcomes the limitations imposed by small offsets inherent in tunnel seismic data, leading to improved imaging.
  • Multi-Component Advantage: Integrating all three seismic components significantly enhances the energy of anomalous interfaces and improves the resolution of geological features compared to single-component methods.
  • Suppression of False Anomalies: The RTM method, particularly the multi-component version, is more effective at suppressing false anomalies (artifacts) that can plague Kirchhoff migration.
  • Enhanced Resolution: The ability of RTM to utilize multiple waves leads to a higher resolution of anomalous interfaces, making it easier to delineate geological structures.
  • Practical Validation: The real-world tunnel data and subsequent excavation results provide strong validation for the effectiveness of the proposed 3D multi-component RTM method in accurately predicting geological hazards.

Key Results

  • Modeled Data: The 3D multi-component RTM method showed superior performance in terms of Signal-to-Noise Ratio (SNR) compared to Kirchhoff migration and single-component RTM. The SNR values for the modeled data were:
  • Single-component Kirchhoff migration: 169.48
    • Single-component RTM: 210.10
    • Multi-component Kirchhoff migration: 209.17
    • Multi-component RTM: 217.15
  • Measured Data: Similar improvements were observed with the real tunnel data:
  • Single-component Kirchhoff migration: 279.10
    • Single-component RTM: 314.01
    • Multi-component Kirchhoff migration: 300.54
    • Multi-component RTM: 414.80
  • Qualitative Results: The multi-component RTM profiles provided clearer delineation of anomalous interfaces (karst caves, interlayers) and better distinguished between different geological features compared to other methods. The geological interpretation based on tunnel excavation results showed a strong correspondence with the multi-component RTM profiles.

Limitations and Future Work

  • Velocity Model Accuracy: While the paper uses a standard method for deriving the velocity model, the accuracy of the RTM results is still dependent on the quality of the velocity model. Errors in the velocity model can lead to migration artifacts.
  • Computational Cost: RTM is computationally intensive. While not explicitly detailed as a limitation in this paper, it's a known challenge for RTM methods, especially in 3D.
  • Data Volume: Despite the multi-component approach maximizing the use of limited data, the inherent small data volume in tunnel seismic surveys can still pose challenges for resolving complex geological structures with high certainty.

Future work could focus on:

  • Developing more robust methods for velocity model building in complex tunnel environments.
  • Exploring computational acceleration techniques for RTM to reduce processing time.
  • Investigating the application of RTM to other types of tunnel seismic data and geological scenarios.
  • Further research into the interpretation of multi-component RTM results for more precise geological hazard prediction.

Summary updated

BibTeX
@misc{guan2021d,
  author = {Peng Guan and Cuifa Shao and Yuyong Jiao and Guohua Zhang and Bin Li and Jie Zhou and Pei Huang},
  title = {{3-D Multi-Component Reverse Time Migration Method for Tunnel Seismic Data}},
  year = {2021},
  doi = {10.3390/s21093244},
}

Loading related content...

Figure