A comprehensive strategy for removing multiples and depth‐imaging primaries without subsurface information: Direct horizontal common image gathers without the velocity or “ironing”

Fang Liu, Arthur B. Weglein, K. A. Innanen, Bogdan G. Nita, Jingfeng Zhang2007 · DOI

TL;DR

This paper introduces a novel seismic imaging technique that accurately positions subsurface reflectors at their true depths, creating "flat" common image gathers. Crucially, this is achieved without needing to know the subsurface velocity model, a significant hurdle in traditional methods, and without compromising valuable amplitude information essential for AVO analysis.

What Problem Does This Paper Address?

Seismic imaging aims to create a picture of the subsurface. A key step in achieving a high-quality image is ensuring that reflections from the same geological layer appear at the same depth, regardless of the seismic wave's incident angle. This is visualized in "common image gathers" (CIGs), which ideally should be flat. However, in reality, CIGs are often "non-flat" because different incident angles lead to different apparent velocities, causing the same reflection point to be imaged at different depths. Traditional methods to "flatten" these gathers often involve "trimming" or applying velocity corrections, which can unfortunately distort or destroy crucial amplitude information. This amplitude information is vital for Amplitude Variation with Offset (AVO) analysis, a technique used to infer lithology and fluid content. The paper addresses the challenge of obtaining flat CIGs necessary for accurate depth imaging and subsequent AVO analysis, without relying on a known velocity model and without damaging the amplitude data.

Methods

The authors leverage a specific subseries of the Inverse Scattering Series (ISS) called the "higher-order imaging subseries" (HOIS). The ISS is a powerful framework for seismic data processing that can handle complex subsurface scenarios without requiring explicit knowledge of the velocity model. The HOIS, in particular, is designed to produce depth-migrated images. The core innovation here is that this particular subseries inherently produces flat common image gathers. This is achieved through a series of iterative calculations that effectively "undo" the wave propagation effects, placing reflectors at their correct depths. The method is velocity-independent, meaning it doesn't require an initial guess or estimation of the subsurface velocity field.

Datasets

The paper uses synthetic data generated from a salt model provided by Peter Traynin of ExxonMobil. This model is described as a "salt model" and is illustrated in Figure 3. The synthetic data includes shot gathers (Figure 4) that are then processed to generate the imaging results. No real-world seismic data is presented or used for evaluation.

Key Insights and Ideas

  • Velocity-Independent Imaging: The most significant insight is the demonstration that accurate depth imaging and flat common image gathers can be achieved without any prior knowledge or estimation of the subsurface velocity. This bypasses a major bottleneck in conventional seismic processing.
  • Amplitude Preservation: The HOIS method not only flattens CIGs but does so without compromising the amplitude information. This is crucial for subsequent AVO analysis, which relies heavily on subtle amplitude variations.
  • Intrinsic Flatness: The flatness of the CIGs is not an artifact of post-processing "ironing" but an inherent outcome of the HOIS algorithm itself, suggesting it's a more fundamental imaging approach.
  • Constructive Summation: By migrating all events to the same depth, the HOIS allows for the constructive summation of multiple images (generated at different angles), leading to an improved signal-to-noise ratio in the final image.

Key Results

The paper demonstrates through numerical examples that the HOIS method successfully produces flat common image gathers for various incident angles. Figures 5 and 6 show linear images for different angles, and Figure 7 illustrates a common image gather where only the water-bottom is flat, with underlying reflectors showing curvature. However, Figure 10, which shows a common image gather from the HOIS, clearly depicts all reflectors as flat. Figure 11 shows the sum of multiple HOIS images, highlighting improved signal quality. The authors state that this method produces common image gathers that are "automatically and accurately producing common image gathers, without the velocity, and with amplitude intact and ready for subsequent AVO analysis."

Limitations and Future Work

  • Truncation Artifacts: The authors note that for the highest incident angle tested (9°), truncation artifacts appear in the imaging results (below 2500m) due to data unavailability. This suggests that the method's performance might be limited by the extent and quality of the input seismic data.
  • Limited Angle Range: The study was conducted with relatively small incident angles (up to 9°). While the paper suggests the method is general, further investigation with a wider range of angles would be beneficial.
  • Acoustic Model: The theoretical framework presented is for a constant-density acoustic model. Extending this to more complex elastic or anisotropic media would be a valuable future direction.
  • Real Data Application: While the synthetic results are promising, the paper does not present results on real seismic data, which would be the ultimate test of its practical applicability.

Summary updated

Abstract

In AVO (Amplitude Variation with Offset) analysis, the amplitudes of reflected waves with different incident angles are studied to deduce lithology information beyond the structure map obtained by seismic imaging algorithms. The quantitative analysis of the amplitude, relies on common‐image gathers being flat (or equivalently, at the same depth). But the waves with different incident angles will have different apparent velocities, resulting in different depths for the same image point at different angles, or non‐flat common image gathers. In many scenarios, non‐flat common‐image gather was flattened by trim means at the cost of compromising zero‐crossing and polarity‐reversal information. This work presents a solution based on the seismic imaging subseries of the inverse scattering series (ISS) that flattens the common image gather without knowing or determining the subsurface velocity, and without any harmful amplitude consequencies.
BibTeX
@misc{liu2007comprehensive,
  author = {Fang Liu and Arthur B. Weglein and K. A. Innanen and Bogdan G. Nita and Jingfeng Zhang},
  title = {{A comprehensive strategy for removing multiples and depth‐imaging primaries without subsurface information: Direct horizontal common image gathers without the velocity or “ironing”}},
  year = {2007},
  doi = {10.1190/1.2792925},
}

Loading related content...

Figure