Primary Reflection

Key Takeaways

• The traveltime of a primary reflection in a horizontally layered medium is described by the hyperbolic NMO equation t^2(x) = t_0^2 + x^2/v_rms^2, where t_0 is the zero-offset two-way traveltime, x is the source-receiver offset, and v_rms is the root-mean-square velocity to the reflector; this hyperbolic moveout allows primaries to be distinguished from multiples (which have lower stacking velocities because they have traveled through the shallower, lower-velocity section multiple times) during velocity analysis, and the NMO correction flattens primaries across offset while leaving multiples with residual curvature, so that CMP stacking (summation over offset after NMO correction) enhances primaries and attenuates residual multiples; in practice, the velocity discrimination between primaries and multiples is incomplete (particularly for short-period multiples from closely spaced reflectors, or for deep multiples that have spent enough time in high-velocity sections to nearly match primary velocities), requiring supplementary multiple-attenuation processing after stacking; the NMO equation breaks down in anisotropic media (where velocity varies with direction of propagation) and in complex structural settings (salt flanks, thrust sheets, steep dips) where ray paths deviate significantly from the hyperbolic assumption, requiring more sophisticated moveout equations or full 3D prestack depth migration to correctly position primaries.
• Surface-related multiple elimination (SRME) is the most powerful data-driven primary-multiple separation algorithm for marine data, exploiting the fact that all surface-related multiples (those that have bounced off the sea surface at least once) can be predicted by convolving the data with itself: a water-bottom multiple arrives at the receiver after reflecting once off the water bottom and once off the sea surface, with a traveltime equal to the sum of the traveltimes of two shorter-path events; SRME computes the predicted multiple wavefield by summing convolutions of all pairs of traces whose combined traveltime equals the multiple's arrival time, without requiring any subsurface model (making it a data-driven, model-independent method); the predicted multiple wavefield is then adaptively subtracted from the measured data using a matching filter to correct for amplitude and phase differences between the prediction and the actual multiple, leaving the primary wavefield with reduced multiple contamination; SRME is highly effective for water-bottom and peg-leg multiples in marine data but requires dense trace sampling to avoid spatial aliasing artifacts, and is less effective for internal multiples (reverberations between subsurface interfaces) which require different prediction strategies such as inverse scattering series (ISS) internal multiple attenuation.
• The distinction between primary reflections and mode-converted reflections is important in multicomponent seismic acquisition: a primary PP reflection (compressional wave down, compressional wave up) arrives at a vertical geophone or hydrophone as a primary but a primary PS reflection (compressional down, shear up) arrives at a horizontal geophone after converting from P to S at the reflector; both are primaries (one reflection only) but the PS conversion has a different traveltime (because the upgoing shear wave is slower than the P wave), different moveout (asymmetric because the upgoing and downgoing legs travel at different velocities), and different NMO velocity (related to the S-wave velocity of the interval above the reflector); in OBC and OBN multicomponent acquisition, the PS data provides an independent measurement of the subsurface that is complementary to the PP data (sensitive to different rock and fluid properties, unaffected by the gas chimney masking that blanks PP reflections above gas reservoirs), making the identification and preservation of primary PS reflections (and the attenuation of PS multiples, which have their own multiple series) an important part of multicomponent data processing; the PS conversion point moves toward the reflector with increasing S/P velocity ratio, complicating CMP binning and requiring asymmetric CCP (common conversion point) gathers rather than symmetric CMP gathers for PP data.
• Internal multiple reflections (reverberations between subsurface interfaces, not involving the sea surface) are the most difficult class of multiples to attenuate because they cannot be predicted by SRME (which only handles surface-related multiples) and their moveout characteristics may closely resemble primaries from deeper reflectors: a first-order internal multiple generated between two closely spaced interfaces (such as the top and base of a high-impedance carbonate or evaporite layer) arrives at a traveltime greater than the base of the interval but less than the next primary below it, potentially masking or interfering with the primary from the underlying target formation; the inverse scattering series (ISS) method for internal multiple attenuation (developed by A.B. Weglein and colleagues at the University of Houston) predicts internal multiples by a series expansion of the seismic wavefield without requiring subsurface model information, but requires dense spatial sampling and careful amplitude scaling; Marchenko redatuming (a more recent approach based on focusing functions derived from the reflection response) provides an alternative framework for retrieving the primary response at depth without multiples, offering potential for superior primary-multiple separation in cases where the ISS approach is limited by aperture or spatial sampling constraints.
• Amplitude-versus-offset (AVO) analysis of primary reflections provides a direct hydrocarbon indicator (DHI) by measuring how the reflection amplitude changes with source-receiver offset: Zoeppritz's equations describe how the reflection and transmission coefficients at an interface vary with angle of incidence for P and S waves, with the gradient of the amplitude-versus-offset curve (the AVO gradient, related to the change in Poisson's ratio across the interface) providing sensitivity to pore fluid type (gas-saturated sands typically show a large negative gradient for a Class IIp or Class III AVO anomaly, while brine-saturated sands show a smaller or positive gradient); AVO analysis requires that the primary reflections are accurately preserved in amplitude (no residual multiples, no migration artifacts, no NMO stretch distortion) and that the processing sequence has maintained relative amplitudes from near to far offset, making amplitude-preserving processing (including careful multiple attenuation that does not damage primary amplitudes at the subtraction step) a prerequisite for reliable AVO interpretation; false AVO anomalies from residual multiples are a significant source of dry-hole risk in AVO-driven exploration, because a multiple from a shallow water-bottom reflection can overlap in traveltime with a primary from a deeper potential reservoir and mimic the AVO behavior of a gas sand.