Wave Equation: Acoustic Propagation, Seismic Migration, and Reverse Time Imaging
The wave equation is a second-order partial differential equation that describes how a disturbance, such as an acoustic or elastic wave, propagates through a medium as a function of space and time. In its simplest scalar acoustic form it relates the second spatial derivatives of wave displacement to the second time derivative, scaled by the square of the wave velocity, so that the sum of the spatial curvatures of the wavefield equals one over velocity squared times the acceleration of displacement. In seismic exploration this compact expression is the physical foundation of nearly everything an acquisition and processing crew does, because a seismic survey is fundamentally an experiment in generating controlled waves, recording their reflections from subsurface layers, and then solving the wave equation to reconstruct where those layers actually sit. When a vibroseis truck sweeps the ground surface in the Montney fairway or an air gun array fires over the Flemish Pass, the resulting energy travels downward, reflects off changes in acoustic impedance at formation boundaries such as the top of the Nisku or the base of the Duvernay, and returns to geophones or hydrophones where amplitude and traveltime are recorded. Turning that raw record into an accurate depth image requires migration, and the most physically complete migration algorithms solve the full wave equation rather than approximating ray paths. Reverse time migration, or RTM, propagates the recorded wavefield backward in time through a velocity model and cross-correlates it with the forward-modelled source wavefield to place reflectors in their true positions, handling steep dips, overturned beds, and complex velocity structures that ray-based methods cannot. The wave equation also underlies seismic velocity analysis, full waveform inversion, and forward modelling used to design surveys before a single source is fired. Because velocity varies with rock type, porosity, fluid content, and pressure, the velocity term inside the wave equation is where geology enters the mathematics, and building an accurate velocity model is often the single hardest and most valuable step in imaging a Western Canadian or offshore East Coast prospect. Mastery of the wave equation, and of the numerical schemes used to solve it on large computer clusters, is what allows modern processing to resolve stratigraphic traps and thin pay zones that earlier approximations blurred beyond recognition.
Key Takeaways
- Second-Order Partial Differential Equation: The scalar acoustic wave equation sets the Laplacian of the wavefield equal to one over velocity squared times the second time derivative of displacement. This single relation governs how seismic energy spreads, reflects, and diffracts, and every migration and modelling algorithm used in exploration is ultimately a numerical scheme for solving it across a discretized earth model.
- Velocity Is Where Geology Enters: The velocity coefficient inside the wave equation encodes rock properties, since acoustic velocity depends on lithology, porosity, cementation, pore fluid, and pressure. A gas-charged Montney sand slows compressional waves relative to a tight limestone, and building an accurate velocity model is the most decisive and error-prone step in producing a correctly positioned depth image.
- Reverse Time Migration: RTM solves the two-way wave equation to propagate recorded energy backward in time and correlate it with the forward source wavefield, imaging steep dips, salt flanks, and overturned beds with no dip limitation. It is computationally expensive but delivers the most complete image beneath structurally complex WCSB thrust belts and offshore Grand Banks salt.
- One-Way Versus Two-Way Solutions: One-way wave equation migration propagates energy in a single dominant direction and is cheaper but cannot image turning waves or steep overhangs. Two-way schemes such as RTM honour the full equation including back-scattered and multiply reflected energy, trading much higher compute cost for fidelity in the hardest imaging settings.
- Forward Modelling and Survey Design: Running the wave equation forward through a known velocity model generates synthetic seismograms, letting geophysicists test whether a proposed survey geometry will actually illuminate a target such as a deep Duvernay pinch-out before committing millions of CAD to acquisition. This predictive use prevents costly surveys that would leave the objective in a shadow zone.
From Ray Tracing to Full Wave-Equation Migration
Early seismic imaging approximated wave propagation with ray theory, tracing energy along high-frequency ray paths much like light through a lens. Ray methods are fast and intuitive but break down where velocity changes sharply or dips exceed roughly 60 degrees, exactly the settings found in the Rocky Mountain thrust belt west of the WCSB and beneath offshore salt. Full wave-equation migration replaced rays with numerical solutions of the equation itself, honouring diffraction and multipathing. The payoff is measurable: a Foothills gas prospect that ray migration smeared into an unusable blur can resolve into a coherent, drillable structure under RTM, changing a 15 million CAD well from a gamble into a defensible bet.
Velocity Model Building and Imaging Accuracy
Because the wave equation's velocity term controls where reflectors land, migration is only as good as the velocity model driving it. Geophysicists iterate using migration velocity analysis, checking whether imaged events are flat across common-image gathers; residual moveout signals velocity error and prompts correction. In the Duvernay and Montney, subtle velocity variation tied to overpressure and organic richness can shift a mapped horizon by tens of metres, which is the difference between landing a horizontal well inside the target window or drilling out of zone. Full waveform inversion now refines these models by matching modelled and recorded waveforms directly through the wave equation.
Fast Facts
The wave equation was first derived by French mathematician Jean-le-Rond d'Alembert in 1747 to describe a vibrating string, nearly two centuries before reflection seismology existed. Reverse time migration was proposed in the early 1980s but sat largely unused for two decades because no computer could afford it; only when cluster computing matured in the 2000s did RTM become routine. A single modern RTM job for a large 3D survey can consume millions of core-hours, making it one of the most compute-intensive tasks in all of applied earth science.
Related Terms
The wave equation is the engine behind migration, the processing step that repositions dipping reflectors to their true subsurface locations. It depends on seismic velocity, the coefficient that carries all the rock physics into the mathematics. Its solutions are analyzed through Fourier synthesis, since wavefields are routinely decomposed into and reconstructed from frequency components, and it produces the reflection events studied in amplitude analysis for direct hydrocarbon indicators.
Real-World WCSB Scenario: Imaging a Foothills Thrust Sheet
An operator targeting a sub-thrust gas play in the Alberta Foothills near the Brazeau area shoots a 3D survey over structurally complex terrain where the Cardium and deeper carbonates are folded and faulted at dips exceeding 50 degrees. Standard Kirchhoff ray migration returns a smeared image that cannot distinguish the crest of the thrust sheet from its overturned limb, leaving the drilling target ambiguous. The team reprocesses with reverse time migration on a velocity model refined through migration velocity analysis, a job that runs for several days on a rented compute cluster at a cost near 250,000 CAD.
The RTM image resolves the thrust geometry cleanly, relocating the structural crest more than 300 metres from where ray migration had placed it. The operator repositions the surface location and well trajectory accordingly, drills the 14 million CAD well into the true crest, and encounters commercial gas that a mis-tied ray image would have missed entirely.