Velocity Image | Oil Authority

A velocity image in petroleum geophysics is a two-dimensional or three-dimensional spatial representation of seismic wave velocities throughout a subsurface volume — derived from the analysis of seismic reflection traveltimes and converted to velocity values through inversion, tomography, or velocity analysis algorithms — that displays how seismic compressional (P-wave) or shear (S-wave) velocity varies with location and depth, providing a quantitative map of subsurface lithology, fluid content, and formation pressure that serves as both a tool for depth migration of seismic data and an independent geophysical interpretation product for reservoir characterization and pore pressure prediction.

Key Takeaways

Velocity images are constructed from seismic data through a multi-step processing workflow: stacking velocity analysis (fitting hyperbolic moveout curves to common midpoint gathers to derive NMO velocities), conversion to interval velocities using the Dix equation (which converts NMO velocity versus depth to the velocity within each stratigraphic interval between reflection horizons), and then further refinement through tomographic inversion that iteratively adjusts the velocity model until the predicted traveltimes match the observed traveltimes in the raw shot gathers — the final tomographic velocity model is the velocity image used for pre-stack depth migration and geological interpretation.
The velocity image is the foundation for pre-stack depth migration (PSDM) — converting seismic data from the time domain (where reflections are recorded as two-way travel time from source to reflector to receiver) to the depth domain (where reflectors are mapped at their true geological depths and positions); an incorrect velocity image displaces reflectors from their true positions (imaging errors) and degrades the spatial resolution of the migrated seismic, making accurate velocity model building the most critical step in the seismic processing workflow for complex geological environments with lateral velocity variations such as salt bodies, steep fault zones, and gas clouds.
Full waveform inversion (FWI) is the most advanced method for constructing high-resolution velocity images — FWI iteratively minimizes the difference between observed seismic waveforms and synthetic waveforms predicted by a velocity model, updating the model at each iteration to reduce the residual; FWI produces velocity images with spatial resolution approaching one-third of the seismic wavelength (2 to 10 meters at typical seismic frequencies), far finer than conventional tomography, making it the preferred method for imaging salt base, subsalt reservoirs, and near-wellbore anomalies where the velocity structure controls imaging accuracy.
Pore pressure prediction from velocity images uses the empirical relationship between seismic velocity and formation pore pressure — overpressured formations (where pore fluid pressure exceeds the hydrostatic gradient) have anomalously low seismic velocities compared to normally pressured rocks at the same depth, because excess pore pressure reduces the effective stress on the rock frame and keeps the porosity higher than normal compaction would allow; converting the low-velocity zones in the velocity image to equivalent pore pressure values using compaction trend velocity models (Bowers, Eaton) provides a seismic-based pore pressure volume that guides drilling engineers in selecting casing points and mud weights before the well is drilled.
Anisotropic velocity images account for the directional dependence of seismic velocity in layered or fractured rock — most sedimentary formations are vertically transversely isotropic (VTI), meaning velocity is lower in the vertical direction than horizontal due to the layered structure of bedding; ignoring anisotropy in the velocity image causes systematic depth errors in imaging and incorrect pore pressure predictions from velocity, while including anisotropy parameters (Thomsen parameters epsilon and delta) in the velocity model corrects these errors and produces more accurate depth-migrated seismic images.

Fast Facts

The concept of velocity analysis for seismic data processing was developed in the 1950s by the CONOCO research group and formalized by Ken Dix in 1955 with the publication of the Dix equation relating stacking velocity to interval velocity — a relationship still used in every seismic processing workflow today. Tomographic velocity model building was introduced to exploration seismology in the 1980s, adapted from medical CT-scan tomography mathematics. Full waveform inversion (FWI) was mathematically formulated by Lailly and Tarantola in 1984 but only became computationally feasible for industrial 3D datasets in the 2010s with advances in GPU computing and parallel processing. Today, velocity image construction from modern ocean-bottom node (OBN) surveys using FWI can produce velocity models with centimeter-scale resolution in shallow water environments, approaching the resolution of well log data.

What Is a Velocity Image?

Seismic surveys measure the time it takes for acoustic energy to travel from a surface source down to subsurface reflectors and back to surface receivers — but converting those traveltimes to accurate depths requires knowing how fast the seismic waves traveled through each rock unit they traversed. The velocity image provides this spatial velocity information: a three-dimensional volume (or two-dimensional cross-section) in which every point is assigned a seismic wave velocity value derived from analysis of the seismic data itself.

The velocity image serves two distinct roles in the exploration and production workflow. As a processing tool, it is the velocity model used in depth migration algorithms that reposition reflection energy from its apparent location in time-domain seismic data to its correct position in depth and space — without an accurate velocity image, depth migration cannot produce a reliable structural image, and the depths and positions of geological features (reservoir tops, fault throws, salt body edges) will be systematically incorrect. As an interpretive product, the velocity image provides a quantitative map of rock properties that can be used directly for geological interpretation — identifying lithology changes, mapping gas accumulations (gas sands have anomalously low seismic velocity), predicting pore pressure, and characterizing reservoir quality before the first well is drilled.

The challenge of velocity image construction is that the velocity is both an input needed to accurately process the seismic data and an output that the processed data reveals — a circular dependency that is resolved iteratively. Initial velocity estimates from well logs and stacking velocity analysis provide a starting model; the seismic data is migrated with this model and the residual moveout (mismatch between model predictions and data observations) is used to update the velocity model; the process repeats until the model is consistent with the data at acceptable residual levels.

Velocity Image Construction Methods

Stacking velocity analysis remains the foundation of velocity image construction for most 2D and 3D surveys. The analyst examines common midpoint (CMP) gathers — collections of seismic traces at different offsets (source-to-receiver distances) recording the same subsurface reflection point — and fits hyperbolic moveout functions to each reflection event. The velocity that best flattens the reflection hyperbola is the NMO velocity, which represents a weighted average of velocities from surface to the reflector. By repeating this analysis at multiple CMP locations and multiple reflection times throughout the dataset, a velocity field is built that covers the entire survey area and the full depth range of the data.

Tomographic inversion upgrades the stacking velocity model to a spatially detailed interval velocity image by back-projecting the residual moveout signals in the migrated gathers along ray paths through the velocity model, computing the velocity update needed to minimize the residual at each point, and applying the update to produce an improved model. Multiple tomographic iterations converge toward a velocity model that correctly predicts the observed traveltimes throughout the dataset. Reflection tomography (using reflections from mapped horizons) provides well-constrained velocity updates in the depth range of available reflectors; refraction tomography (using the direct and head wave arrivals at near offsets) constrains the near-surface velocity structure that is critical for proper statics corrections.

Well log velocity ties integrate borehole data into the velocity image construction — vertical seismic profiles (VSPs) and check-shot surveys in exploration wells provide direct measurements of seismic velocity at specific locations, which are used to calibrate the surface-seismic-derived velocity model and to correct systematic biases from anisotropy, dispersion, or incorrect assumptions in the inversion. The integration of well and surface seismic velocity data produces a hybrid velocity image that is more accurate than either source alone, particularly near the wellbore where the high spatial resolution of borehole measurements compensates for the limited resolution of surface seismic velocity analysis.

Velocity Images Across International Jurisdictions

Canada (AER / WCSB): WCSB seismic surveys in the Alberta Deep Basin and Foothills thrust belt present complex velocity imaging challenges — the Foothills thrust sheets have strong lateral velocity variations as carbonate-dominated thrust sheets overlay lower-velocity shales and sandstones, creating severe velocity contrasts that deflect seismic ray paths and distort structural images if not properly handled by the velocity model. Velocity images for Foothills exploration programs use anisotropic tomography and pre-stack depth migration to correctly image the steeply dipping thrust anticlines and the sub-thrust exploration targets that contain the Foothills' gas reserves. Companies including Tourmaline, Canadian Natural Resources, and ConocoPhillips Canada invest heavily in velocity model building for Foothills 3D surveys where correct depth imaging of sub-thrust targets is directly tied to exploration well location accuracy.

United States (API / BSEE): Gulf of Mexico deep water velocity imaging is among the most technically demanding in the world — the Sigsbee Escarpment and allochthonous salt bodies in the GOM create extreme velocity contrasts (salt has P-wave velocity of approximately 4,480 m/s versus surrounding sediments at 1,500 to 3,000 m/s) that require accurate salt geometry in the velocity model to correctly image subsalt reservoirs. BSEE requires operators to submit well prognosis and location justification data that implicitly relies on accurate depth imaging from well-calibrated velocity models. ExxonMobil, Shell, Chevron, and BP have spent tens of millions of dollars on GOM velocity model building programs using FWI with ocean-bottom node data to improve imaging of subsalt Wilcox and Norphlet reservoir targets.

Norway (Sodir / NORSOK): Norwegian North Sea velocity imaging focuses on the chalk, salt, and overburden velocity heterogeneity that affects depth imaging of Ekofisk, Valhall, and Eldfisk chalk reservoirs. Equinor's Heidrun and Åsgard fields in the Norwegian Sea require anisotropic velocity models for correct depth migration of the deep Jurassic targets below the gas-charged Cretaceous overburden that causes strong velocity pull-up effects on structural maps. Sodir requires operators to document the seismic processing and imaging methodology used for structural mapping in exploration license applications, ensuring that depth uncertainty from velocity model limitations is appropriately communicated in resource estimates.

Middle East (Saudi Aramco): Saudi Aramco uses velocity images derived from 3D seismic surveys over the Arabian Platform to map the Arab Formation structure and characterize carbonate reservoir quality through velocity-porosity relationships. The relatively simple geology of the Arabian Platform (gentle dips, minimal faulting) allows high-confidence velocity model building from conventional tomography and well-calibrated velocity analysis, producing depth images of the Arab Formation with meter-level accuracy at depths of 2,000 to 3,500 meters. Aramco's exploration programs in the Rub' al Khali basin use velocity images from long-offset 3D seismic surveys and FWI to map deeper Paleozoic targets where velocity complexity from deep salt and gas accumulations requires advanced imaging technology beyond conventional tomography.