|

Deconvolution Explained

Byadmin
 

Deconvolution Explained

Introduction

Seismic data is a blurred version of the subsurface. Every reflection is smeared by the seismic wavelet — a combination of the source signature, earth filtering, and recording system effects. Deconvolution is the process that sharpens this wavelet, improving vertical resolution and making geological boundaries easier to interpret.

This article explains what deconvolution is, why it matters, and how it fits into the seismic processing workflow.

1. What Is Deconvolution?

Deconvolution is a signal‑processing technique that removes the effects of the seismic wavelet from recorded data. The goal is to recover a reflectivity series that more closely represents true subsurface layering.

In simple terms:

Raw seismic = reflectivity × wavelet + noise Deconvolution ≈ removing the wavelet

The result is a sharper, cleaner seismic trace.

2. Why Deconvolution Matters

The seismic wavelet limits vertical resolution. Without deconvolution:

  • Thin beds blur together

  • Reflectors appear smeared

  • Multiple energy contaminates primaries

  • AVO analysis becomes less reliable

Deconvolution improves:

✔ Vertical resolution

Sharper reflectors reveal subtle stratigraphy.

✔ Wavelet stability

A more consistent wavelet improves interpretation.

✔ Multiple suppression

Predictive deconvolution reduces short‑period multiples.

✔ Attribute quality

Attributes like instantaneous frequency and amplitude become more meaningful.

3. Types of Deconvolution

A. Spiking Deconvolution

Attempts to compress the wavelet into a spike. Pros: Improves resolution. Cons: Sensitive to noise

B. Predictive Deconvolution

Predicts and removes periodic events such as multiples. Pros: Effective for short‑period multiples Cons: Requires careful parameter selection

C. Surface‑Consistent Deconvolution

Applies corrections per:

  • Source

  • Receiver

  • Offset

  • Midpoint

Pros: Corrects for acquisition variations Cons: Requires large datasets

D. Gabor Deconvolution

Time‑frequency method that adapts to non‑stationary wavelets. Pros: Excellent for broadband data Cons: Computationally intensive

4. Deconvolution Workflow

  1. Data conditioning

  2. Wavelet estimation

  3. Operator design

  4. Application of deconvolution

  5. QC using spectra and gathers

  6. Parameter refinement

QC is essential — over‑deconvolution can create artifacts.

5. Challenges

  • Noise contamination

  • Non‑stationary wavelets

  • Poor wavelet estimates

  • Over‑whitening

  • Multiples overlapping primaries

Conclusion

Deconvolution is a foundational step in seismic processing. Sharpening the wavelet and improving vertical resolution enhance the clarity and interpretability of seismic data, supporting better geological decisions.

Similar Posts