Skip to content
Sanitas · Near-Surface Research

Discover the Very Near Surface

Getting the near-surface right is the difference between a stack that images the subsurface and one that doesn’t.

Explore Solutions Download Sanitas Brochure
  • Refraction statics
  • Cascade Eikonal tomography
  • Azimuthal anisotropy
Scroll to the near-surface problem

The problem is right at your feet.

The industry treats the near-surface as a solved problem. The statics say otherwise.

For four decades a quiet contradiction has run through land processing: delay-time statics beat tomostatics, consistently. Tomography produced beautiful models — yet the statics those models implied were almost never as good. If the statics are wrong, the model is wrong.

The reason is the Very Near Surface: a thin, at-the-surface disturbance that contaminates every first break and carries the vast majority of the static. Separate it from the model, and both come out right.

Equation: t-sr equals a-s plus b-r, plus T of x-m and h, plus A T cosine of 2 times theta-sr minus phi, plus epsilon-sr.
Equation: t-sr equals a-s plus b-r, plus T of x-m and h, plus A T cosine of 2 times theta-sr minus phi, plus epsilon-sr.
Pick time equals shot plus receiver statics, plus the turning-ray (LMO) trend, plus a time-scaled azimuthal TTI term, plus a residual. The azimuthal term is the horizontal-ray projection of a tilted TI axis (tilt χ, azimuth φ): A = ½ εTI sin²χ · vertical velocity: vvert ≈ vax (1 + εTI sin²χ) HTI is the χ = 90° special case; VTI (χ = 0) is the classic near-surface “vertical anisotropy” that upholes have always disagreed with refractors about.
Near-surface anisotropy

The anisotropic term is TTI — and it is explicit.

One TI medium in three regimes: VTI with a vertical symmetry axis at tilt 0°, TTI with a tilted axis between 0 and 90°, and HTI with a horizontal axis at 90°. Below, two curves against tilt: the azimuthal cos 2θ amplitude rising from zero at VTI to a half at HTI, and the uphole versus refractor mismatch falling from one to minus a half, the two crossing near 54.7°.

For seismic data in general

Near-surface TTI puts an azimuth- and offset-dependent shift on every arrival — it masquerades as statics plus velocity error, and it degrades stacking, AVO-with-azimuth, and any isotropic model built through it. Because it sits in the first breaks, Sanitas measures it deterministically: a surface-consistent station term (ms) and a per-cell medium term (A, φ) — an explicit field, separable from the data, not an interpretation.

For FWI in particular

FWI’s ability to resolve anisotropy is uncertain — worst in the near surface, where short wavelengths, cycle-skipping risk, and ε–velocity crosstalk collide. An explicit, measured expression for the anisotropy component of the input data changes the problem:

  • Correct the data remove the known near-surface anisotropic moveout so FWI does not map it into false shallow structure
  • Constrain the model the (vax, εTI, χ, φ) field is a starting/prior TTI model for the shallow section FWI resolves worst
  • QC the result FWI-recovered anisotropy has an independent, first-break-measured field to answer to

The near-surface anisotropy stops being FWI’s problem to resolve — it arrives measured.

Sanitas

Rethinking land seismic from the very near surface down.

  1. AMGRT · 2025

    VNS

    First breaks decompose into a surface-consistent shot term, a receiver term, and a turning-ray velocity term. The VNS terms are the statics — they never belonged in the model.

  2. 2026

    Cascade Tomography

    An interpretation-free hybrid that separates VNS statics from the tomographic model: delay-time-quality statics and the model you would expect from the surface datum, from a single workflow.

  3. 2026

    Concurrent Anisotropic Tomography

    VNS statics, azimuthal anisotropy, and the velocity model solved together — and the anisotropic correction applied through velocity analysis and stack. The first system to close that loop.

The near-surface is the foundation of the house. Every processing step that follows is only as sturdy as what it stands on.

Same data · same picks

Fix the foundation and the section sharpens.

Velocity-analysis semblance on one CMP supergather. With no statics the VNS smears the moveout and the picks are a guess. With Cascade 3-3 and the anisotropy correction the energy collapses into a focused bullseye — before a single processing parameter changed.

2.4×

semblance focus

No statics
0.016
Cascade 3-3, anisotropy corrected
0.039
See the stacked sections and full results
Velocity semblance panel with no statics applied: the energy is smeared and no clear peak is present. No statics
Velocity semblance panel after Cascade 3-3 statics with anisotropy correction: energy focuses into a single tight bullseye. Cascade 3-3, anisotropy corrected

Drag the handle, or focus it and use the arrow keys, to wipe between the two results.

Stack, Profile 1

Cascade 3-3 statics, anisotropy applied. Reflector continuity restored end to end — no smoothing, no interpretation, no hand-editing.

Raw stacked section on Profile 1: reflector continuity is broken across the line. Raw
The same stacked section after Cascade 3-3 statics with the anisotropy correction applied: reflectors are continuous end to end. Cascade 3-3 statics, anisotropy applied

Drag the handle, or focus it and use the arrow keys, to wipe between the two results.

Sanitas Trails™ · packaged expertise

The senior processor’s route — and the senior processor’s reasoning.

Any Sanitas workflow can be captured as a Trail: a guided sequence of the system’s own working windows — pick the trends, build the model, compute the statics, QC the stack — each step pared to its task, presented in order, branching where a decision changes the path.

Embedded in every Trail is SeisMentor™: ask why at any step and get answers grounded in the CT-VNS methodology and your project’s own numbers.

Experts author Trails; the library ships with the system; the expertise that usually lives in one person’s head becomes something the whole shop runs.

Sanitas Trails, step 1 of 6 in the Trend to Tomo trail, wrapping the system’s own spatial LMO trend picking window in the very near surface.
Step 1 of 6, Trend → Tomo. The trail banner and the step bar wrap the system’s own picking window — the real tool, focused on one task, with SeisMentor one click away.

One system, end to end

Sanitas is the newest life of a processing engine with a two-decade commercial pedigree in refraction statics — rebuilt around VNS, and carrying the correction all the way through velocity analysis and stack.

  1. Sanitas

    Eikonal tomography, cascade passes and concurrent azimuthal anisotropy in one processing system.

  2. Guided workflows

    Any workflow can be captured as a guided sequence of the system’s own working windows, in order, branching where a decision changes the path.

  3. SeisMentor™

    Ask why at any step and get answers grounded in the CT-VNS methodology and your project’s own numbers.

The Best bring us their Worst.

Start with the near-surface: read the full guide to both statics pathways, or bring us a dataset and talk it through with the people who built the system.

Download the Sanitas Brochure Talk to the SeisVantage team

SEG IMAGE 2026 · Houston Booth 846 Live demos daily