Common Midpoint Method: CMP Seismic Acquisition and Processing

What Is the Common Midpoint Method?

Common midpoint method (also called the CMP method or CDP method) is the standard seismic data acquisition and processing technique introduced by W. Harry Mayne in 1962 in which multiple source-receiver pairs with different offsets are recorded such that each pair shares the same surface midpoint, then the traces from all pairs at a given midpoint are gathered and processed together using normal moveout (NMO) correction and stacking. The stacking of N traces reduces random noise by a factor of approximately the square root of N, producing a stacked seismic section with substantially improved signal-to-noise ratio that forms the foundation of modern reflection seismic interpretation.

CMP Geometry and Acquisition Design

In a 2D CMP survey, sources and receivers are deployed along a single line. The midpoint between any source-receiver pair is the horizontal location equidistant between them. By spacing sources and receivers symmetrically and at regular intervals, the geometry is designed so that multiple different source-receiver pairs share the same midpoint. For example, if the source interval and receiver interval are both 25 meters, a midpoint is shared by a near-offset pair (source 12.5 m west, receiver 12.5 m east), a medium-offset pair (source 37.5 m west, receiver 37.5 m east), and so on. All traces sharing a common midpoint are collected into a CMP gather for processing. The maximum offset in the gather is typically 1 to 1.5 times the depth to the target reflector to allow adequate velocity discrimination during NMO analysis.

In 3D CMP surveys, sources and receivers are deployed in a grid pattern and the midpoints fall within rectangular bins on a map grid. A typical 3D onshore bin size is 12.5 by 25 meters or 25 by 25 meters; offshore bins are commonly 6.25 by 12.5 meters. The fold varies across the survey area and is highest in the center of the acquisition template where the most source-receiver pairs share a given bin. The bin size determines the spatial sampling of the seismic data and controls the minimum wavelength of lateral geological variations that can be resolved. The azimuth distribution of traces within each bin is also important for azimuthal anisotropy analysis, which can detect fracture orientation in tight reservoirs.

The NMO correction is the key processing step applied before stacking. A reflection from a horizontal reflector at depth Z with velocity V arrives at offset X at time t(X) = sqrt(t0^2 + X^2/V^2), forming a hyperbola on the CMP gather where t0 is the two-way zero-offset travel time. NMO correction removes the offset-dependent time delay using a velocity function V(t) picked from semblance panels, flattening all traces in the gather to their zero-offset time. After NMO correction, the traces are summed (stacked) to produce a single stacked trace at each CMP location. The resulting stacked section simulates zero-offset recording but with far superior signal-to-noise ratio because constructive interference reinforces the signal while incoherent noise cancels across the stack.

CMP vs. CDP: A Historical Distinction

The earlier term common depth point (CDP) assumed that all rays from traces sharing a surface midpoint reflect from the same subsurface depth point, which is exactly true only for horizontal reflectors. For dipping reflectors, the actual reflection points migrate updip and are not common; they are common only at the surface midpoint. Mayne's 1962 formulation recognized this distinction and introduced the term common midpoint to correctly describe the acquisition geometry without implying that the subsurface reflection points coincide. In practice, the terms CMP and CDP are used interchangeably in the industry, and most processing software uses CMP. Post-stack migration, which moves reflectors to their true subsurface positions after stacking, corrects the dip-induced reflection point dispersal and restores the validity of the CDP approximation in the final migrated section.

Common Midpoint Method Synonyms and Related Terminology

The common midpoint method is also referred to as:

• Common depth point (CDP) method — the original terminology from early multichannel seismic practice, technically correct only for horizontal reflectors but still widely used as a synonym
• Multi-fold coverage — describes the acquisition strategy of recording multiple traces per midpoint without specifying the processing technique
• CMP stacking — refers specifically to the processing step of summing NMO-corrected traces within a common midpoint gather
• Mayne method — an informal reference to the inventor, used occasionally in historical or academic contexts

Related terms: normal moveout, seismic stacking, seismic migration, fold, reflection seismology

Frequently Asked Questions About the Common Midpoint Method

Why does stacking improve signal-to-noise ratio but not eliminate noise entirely?

Stacking improves S/N because coherent signal — the reflection arriving at a predictable time after NMO correction — adds constructively across all traces in the gather, while random noise (instrument noise, wind, traffic) adds incoherently and its amplitude grows only as the square root of the number of traces rather than linearly. For N traces, signal amplitude increases by N while noise amplitude increases by sqrt(N), so S/N improves by sqrt(N). However, coherent noise — noise that appears at consistent times across multiple traces, such as ground roll in land seismic or multiple reflections in marine seismic — also stacks constructively and is not suppressed by CMP stacking alone. Coherent noise must be addressed by other processing steps: f-k filtering for ground roll, surface-related multiple elimination (SRME) or parabolic Radon demultiple for water-bottom multiples.

What is the relationship between CMP fold, bin size, and survey cost?

Survey cost scales roughly linearly with the total number of source points required, which increases as fold increases and as bin size decreases. For a given 3D survey area, doubling the fold requires approximately double the source points (and therefore double the acquisition time and cost). Halving the bin size from 25 to 12.5 meters quadruples the number of bins in the survey and also requires approximately four times as many source points to maintain the same fold. Modern survey design optimizes the trade-off between fold (which controls S/N and multiple suppression), bin size (which controls spatial resolution), and offset-azimuth distribution (which controls velocity discrimination and anisotropy imaging) against the available acquisition budget. Compressive sensing and simultaneous source techniques introduced in the 2010s have partially decoupled fold from cost by allowing multiple sources to fire nearly simultaneously.

When is pre-stack depth migration preferred over post-stack time migration?

Post-stack time migration assumes that velocities vary smoothly and that lateral velocity variations are small enough that the stacking process is not significantly compromised by velocity heterogeneity. In structurally simple areas with gentle dips and laterally homogeneous velocity, post-stack time migration is adequate and computationally inexpensive. Pre-stack depth migration (PSDM) is required when strong lateral velocity gradients — caused by salt bodies, thrust belts, carbonates overlying shales, or gas clouds — create ray-bending that distorts the NMO hyperbola and corrupts the stacking velocity. PSDM processes each trace individually through a 3D velocity model before stack, correctly handling complex ray paths and producing accurate structural images beneath salt overhangs and in sub-thrust plays. PSDM is now the industry standard for deepwater exploration, sub-salt imaging, and complex fold-belt exploration, despite being 5 to 20 times more computationally expensive than post-stack methods.

Why the Common Midpoint Method Matters in Oil and Gas

The CMP method is the single most important technical innovation in exploration geophysics. Before Mayne's 1962 paper, single-fold seismic recording produced images so noise-contaminated that structural interpretation in challenging basins was unreliable and drilling success rates were low. The ability to stack dozens to hundreds of traces from a common midpoint, suppress random noise, and derive interval velocity functions from the moveout analysis transformed seismic exploration from an art into a quantitative science. Every modern 2D seismic line and every 3D survey cube is built on CMP acquisition and processing principles. The NMO velocity functions extracted during CMP processing are themselves used for depth conversion, pore pressure prediction, and reservoir characterization. The CMP method underpins the entire global exploration and production industry's ability to image subsurface structure and identify drilling targets with confidence.