Gather: Common Midpoint Sorting, NMO Velocity Analysis, and Seismic Imaging of WCSB Reservoirs

A gather is a display of seismic traces grouped together because they share a common acquisition parameter, and it is the working unit of nearly every step in seismic data processing. Raw seismic data arrives from the field organized by the shot that produced it, but a single shot record mixes energy reflected from many subsurface points at many different source-to-receiver distances, which makes it poorly suited for measuring velocity or building a clean image. Processors therefore re-sort the traces into gathers that isolate one variable at a time. The most important of these is the common midpoint gather, abbreviated CMP, which collects every trace whose source and receiver straddle the same surface midpoint. Because all the traces in a CMP gather sample very nearly the same subsurface reflection point, the only systematic difference among them is offset, the distance between source and receiver. That controlled relationship is what allows the two foundational operations of reflection seismology: velocity analysis and stacking. Other gather types serve specialized roles. A common shot gather holds all traces from one source, a common receiver gather holds all traces recorded by one geophone or hydrophone, a common offset gather holds traces with identical source-receiver spacing, and a common image point or common reflection point gather is the migrated equivalent used in modern depth imaging and amplitude versus offset work. In a CMP gather of a flat reflector, the reflection arrival time follows a hyperbola: the zero-offset traveltime is the apex, and traces at larger offset arrive later because the raypath is longer. The curvature of that hyperbola encodes the stacking velocity. Processors apply a normal moveout correction, or NMO, that flattens the hyperbola by removing the offset-dependent delay, and once the events are flat the traces are summed, or stacked, into a single high-quality trace at that midpoint. Stacking is powerful because coherent reflection energy adds constructively while random noise tends to cancel, so a sixty-fold gather, meaning sixty traces sharing one midpoint, can improve the signal-to-noise ratio dramatically. In the Western Canadian Sedimentary Basin, gathers are the foundation for imaging Cardium and Viking sand pinchouts, mapping Leduc and Nisku carbonate reefs, characterizing Montney and Duvernay resource plays, and steering horizontal wells, and the same gathers feed amplitude analysis that helps distinguish gas-charged sand from wet sand before a single dollar of drilling capital is committed.

Key Takeaways

  • Shared parameter defines a gather: A gather groups seismic traces that have one acquisition attribute in common. The common midpoint gather collects traces whose source-receiver pairs share the same surface midpoint, isolating offset as the only systematic variable so velocity and reflectivity can be measured cleanly without interference from changing reflection geometry.
  • Fold controls signal-to-noise: The number of traces in a CMP gather is the fold or multiplicity. Stacking a 60-fold gather sums 60 traces that share a reflection point, so coherent signal adds while random noise averages toward zero. Higher fold improves imaging of deep, faint targets such as Nisku reefs but raises acquisition cost per square kilometre.
  • NMO links curvature to velocity: In a CMP gather, a reflection traces a hyperbola whose curvature depends on the stacking velocity above the reflector. Velocity analysis scans candidate velocities to flatten the gather, producing the velocity field used to convert seismic time to depth, position events correctly, and tie reflections to well control in metres or feet.
  • Multiple gather domains serve different jobs: Shot, receiver, offset, and CMP gathers each expose a different aspect of the data. Common offset gathers help quality-control statics and ground roll, while migrated common-image-point gathers feed AVO and prestack inversion that estimate porosity and fluid in Montney and Duvernay intervals before drilling.
  • Residual moveout reveals problems: If events in a corrected gather are not flat, the residual curvature flags an incorrect velocity, anisotropy, or structural complexity. Interpreters use that signal to refine the velocity model, detect VTI anisotropy common in shale plays, and avoid mispositioning a horizontal landing point by tens of metres.

From Field Records to a Stacked Section

Field crews shoot thousands of source points into spreads of hundreds of receivers, so the raw volume is enormous and disorganized for imaging. Processing first applies geometry, assigning each trace its true source and receiver coordinates, then sorts the data from shot domain into CMP domain. A typical WCSB 3D survey is designed for nominal fold between 30 and 90, meaning each midpoint bin holds that many traces. After deconvolution and statics, velocity analysis is run on selected CMP gathers spaced across the survey, the NMO correction is applied, and the gathers are stacked to produce the brute and final stacked sections that interpreters load into mapping software to pick horizons and faults.

Gathers in AVO and Reservoir Characterization

Beyond imaging, the prestack gather carries information that the stack throws away. Amplitude versus offset analysis examines how reflection strength changes from near to far traces within a gather, because gas-charged sands often brighten or dim with offset in diagnostic ways. In the Montney and Duvernay, processors preserve true-amplitude prestack gathers and run simultaneous inversion to estimate P-impedance, S-impedance, and density, which map to porosity and fluid. A reliable AVO read depends on flat, noise-suppressed, true-amplitude gathers, so gather conditioning, including trim statics and residual NMO, is a critical step before any quantitative interpretation drives a drilling decision.

Fast Facts

The common midpoint method was patented by Harry Mayne in 1956 and transformed exploration economics overnight. Before CMP stacking, a single noisy trace had to carry the burden of imaging a reflector. By gathering dozens of traces that sample the same subsurface point and summing them after moveout correction, geophysicists achieved signal-to-noise gains roughly proportional to the square root of the fold, so a 64-fold gather improves signal by about eight times. That single idea made it economic to image deep carbonate and shale targets across the entire WCSB.

The gather sits at the centre of a web of seismic concepts. Common Midpoint names the most important gather type, the foundation of stacking, while Normal Moveout is the correction applied to flatten reflection hyperbolas inside a gather before summing. Stacking Velocity is the parameter recovered from gather curvature during velocity analysis, and Migration repositions the events to their true subsurface location, producing the image-point gathers used in modern amplitude analysis.

WCSB Field Scenario: Conditioning Gathers for a Montney AVO Drill in NE British Columbia

An operator targeting an over-pressured Montney interval near Dawson Creek, British Columbia, acquired a 3D survey designed for 72-fold and budgeted roughly CAD 1.1 million for prestack gather conditioning and simultaneous inversion. The processing team found that raw CMP gathers showed residual moveout from VTI anisotropy, which was flattening the near traces but leaving the far traces 6 to 8 milliseconds late. Left uncorrected, this skewed the AVO gradient and would have falsely brightened wet zones.

By applying anisotropic NMO and trim statics, the team produced flat, true-amplitude gathers that isolated a genuine Class III AVO anomaly over the best reservoir. The conditioned inversion guided the horizontal landing to a 4 m sweet spot, and the resulting well produced about 30 percent above the pad average, a return that paid back the gather-conditioning spend many times over.