normal moveout, Oil & Gas Glossary

Normal moveout, abbreviated NMO, is the increase in reflection traveltime that occurs as the distance between the seismic source and receiver grows, for energy reflected from a flat, horizontal interface. NMO correction is the seismic-processing step that removes this offset-dependent delay so that the same reflection, recorded at many source-receiver separations within a common-midpoint (CMP) gather, is aligned to its zero-offset arrival time before the traces are summed, or stacked. The physics is straightforward: a wave travelling from a source to a deeper reflector and back to a distant receiver follows a longer slant path than a wave going straight down and back, so it arrives later. For a single horizontal layer the relationship is hyperbolic, with the traveltime at offset x given by t squared equals t-zero squared plus x squared divided by velocity squared, where t-zero is the two-way zero-offset time and the velocity is the NMO or stacking velocity. By estimating that velocity through a velocity analysis, the processor computes how much each trace must be shifted upward in time and flattens the hyperbola so the reflection becomes a horizontal event across the gather. Stacking the corrected traces then sums the genuine reflection energy in phase while random noise, multiples, and out-of-plane scatter add destructively, raising the signal-to-noise ratio of the final image dramatically. The most important by-product to manage is the NMO stretch effect: the correction stretches the waveform most at large offsets and shallow times, lowering the apparent frequency content in a time- and offset-dependent way. Left unchecked, stretched, low-frequency wavelets smear the stack, so processors apply a stretch mute that removes the badly distorted far-offset, shallow data before stacking. Choosing the correct velocity is critical because too low a velocity over-corrects and too high under-corrects, both leaving residual moveout that degrades the stack and, downstream, the migration. In the Western Canadian Sedimentary Basin, accurate NMO and velocity analysis are essential for imaging thin, stacked targets such as the Cardium, Viking, and Montney, where bed thicknesses of a few metres demand that high frequencies be preserved and reflections aligned precisely. NMO is a foundational concept that feeds directly into stacking, follows from the common midpoint acquisition geometry, and precedes migration in the standard processing sequence.

Key Takeaways

Offset-dependent traveltime: NMO is the extra time a reflection takes to reach a far receiver versus a zero-offset receiver, because the slant raypath is longer. For a flat single layer the moveout is hyperbolic, t squared equals t-zero squared plus x squared over velocity squared, the basis of every NMO correction.
Alignment before stacking: NMO correction shifts each trace in a CMP gather to its zero-offset time so the same reflector flattens across all offsets. Summing the aligned traces reinforces real signal and cancels random noise, the core mechanism that gives a stacked seismic section its strong signal-to-noise ratio.
Stacking velocity drives the fix: The correction depends on an estimated NMO or stacking velocity from velocity analysis. Too low over-corrects (events smile upward at far offset), too high under-corrects (events frown), and both leave residual moveout that blurs the stack and the subsequent migration.
The stretch effect and muting: NMO correction stretches wavelets most at large offsets and shallow times, lowering their frequency. Processors apply a stretch mute to discard the worst-distorted far-offset shallow samples before stacking, trading some fold for preserved resolution.
WCSB thin-bed imaging: Resolving thin, stacked plays such as the Cardium, Viking, and Montney depends on preserving high frequencies and aligning reflections precisely, so careful velocity picking and stretch-mute design directly control whether a few-metre reservoir is visible on the final section.

Velocity Analysis and Residual Moveout

The processor estimates stacking velocity by testing a range of velocities against the data, typically with a semblance or velocity-spectrum panel that lights up where a chosen velocity best flattens each reflection. Picking these maxima down the section builds a velocity function used for the NMO correction. After correction the gather is inspected for residual moveout, any leftover curvature signalling a mispick. Higher-order, non-hyperbolic moveout terms are added for long-offset or anisotropic data, common in shale plays, where the simple hyperbola no longer fits and a four-term equation better describes the Montney or Duvernay response.

Managing the Stretch Mute

Because NMO stretch worsens with offset and shallowness, a fixed time shift would ruin shallow far-offset data. The stretch mute defines a maximum allowable stretch percentage and zeroes samples beyond it. Set it too aggressively and the shallow stack loses fold and signal-to-noise; set it too loosely and low-frequency smear degrades resolution. Processors tune the mute against the target depth, accepting reduced shallow fold to protect the high-frequency content needed for thin WCSB reservoirs, then verify the choice on stacked test panels before finalizing.

Fast Facts

The common-midpoint method that NMO correction enables was patented by Harry Mayne in 1956 and transformed exploration seismology by making stacking practical. Before CMP stacking, a single noisy trace had to stand alone; afterward, summing dozens of NMO-corrected traces from the same subsurface point could improve signal-to-noise by roughly the square root of the fold, so a 36-fold WCSB survey gains about six times the effective signal strength purely from correctly aligning and adding the data.

Normal moveout is the correction that makes stacking possible, the summation step that defines modern reflection seismology. It operates on traces gathered by common midpoint geometry, where many source-receiver pairs share one subsurface reflection point. Once the gather is NMO-corrected and stacked, migration repositions the energy to its true subsurface location, completing the path from field record to interpretable image.

WCSB Scenario: Cardium Velocity Pick at Pembina

A processing team handling a 3D survey over the Cardium at Pembina targets a 6 m (20 ft) sand at about 1,650 m (5,415 ft). Initial velocity picks leave residual moveout that smears the thin reflector on the stack. Reanalysis with a denser semblance grid and a tightened stretch mute at 25 percent recovers the high frequencies, and the reprocessing costs roughly CAD 95,000 against the original survey.

The refined NMO correction flattens the Cardium event cleanly across offsets, and the restacked volume resolves the sand and its lateral pinch-out. The operator uses the improved image to place a horizontal well that lands inside the thin reservoir, a result the original processing could not have supported.