|
<< Click to Display Table of Contents >> Navigation: Migration > HTI delta estimation |
The HTI delta estimation module computes the anisotropic Thomsen delta parameter (δ) for a Horizontally Transverse Isotropic (HTI) medium on a layer-by-layer basis. In an HTI medium, the axis of symmetry is horizontal, meaning seismic velocities are slower in the horizontal propagation direction and faster in the vertical direction — a situation typically associated with vertically aligned fractures or fine vertical layering.
Delta is one of three Thomsen anisotropy parameters (alongside epsilon and gamma) that describe the degree of anisotropy in a medium. The delta parameter governs near-vertical P-wave moveout and is critical for anisotropic depth migration (PSDM). This module provides an alternative or complementary way to estimate delta outside of the PSDM imaging workflow, using only depth velocity fields and horizon picks.
The computation is straightforward: for each layer defined by the input horizons, the module measures the layer thickness as seen in the isotropic velocity model and the layer thickness as seen in the vertical (check-shot) velocity model. The ratio of these two thicknesses minus one gives the per-layer delta value. The result is written as a depth-domain volume where each sample within a layer carries the constant delta value for that layer. Use this module to build anisotropy models for depth imaging workflows when check-shot data and an isotropic velocity model are available.
The isotropic depth velocity field. This is a post-stack, depth-domain velocity volume in which each trace represents a 1D velocity profile at a given bin location. The module uses this field to determine the isotropic layer thicknesses between the input horizons. The data must be post-stack (one trace per bin), defined on a constant datum, and the datum must be positioned above the topographic surface. This field is typically the velocity model used in isotropic PSDM processing.
A set of depth horizons defining the layer boundaries as interpreted in the isotropic velocity model. Each horizon represents the base of one layer. The isotropic horizon set is used to compute the depth extent of each layer according to the isotropic earth model. The number of layers (horizons) in this set must exactly match the number of layers in the vertical horizons input — if the counts differ, the module will report an error and halt.
A set of depth horizons defining the same layer boundaries as seen in the vertical (check-shot) velocity model. These horizons capture the true vertical depth extent of each layer as measured by borehole check shots or vertical seismic profiles (VSP). The ratio of the isotropic layer thickness to the vertical layer thickness is the basis for computing delta in each layer. This set must have the same number of layers as the isotropic horizons.
This module has no user-adjustable processing parameters. All inputs are provided through the three data items above. The delta value for each layer is computed automatically from the ratio of isotropic to vertical layer thicknesses: delta = (isotropic thickness / vertical thickness) − 1.
The number of CPU threads to use when computing the delta volume. The module processes traces in parallel — increasing this value reduces run time on multi-core workstations. Set this to match the number of available physical CPU cores for best performance. For most workstations, values between 4 and 16 are appropriate. Setting a value higher than the number of available cores will not improve performance and may degrade it.
The output depth-domain volume containing the estimated Thomsen delta parameter. The volume has the same spatial geometry as the input depth velocity field. Within each layer (between successive input horizons), every depth sample is assigned the constant delta value computed for that layer. Samples outside the horizon-defined depth range are left empty. This delta volume can be used directly as input to anisotropic depth migration (PSDM) workflows that require a spatially varying delta model.