|
<< Click to Display Table of Contents >> Navigation: Refraction statics > GLI Refraction static solver |
The GLI Refraction static solver is a legacy (deprecated) module that computes refraction-based static corrections using a Generalized Linear Inversion (GLI) approach. Starting from a layered subsurface model, the solver iteratively traces refracted rays between source-receiver pairs, compares computed traveltimes against observed first-break picks, and updates the refractor depth at each grid cell to minimize the traveltime residuals. The result is a set of static correction values that compensate for the effect of near-surface velocity and depth variations on seismic reflection data.
Note: This module is deprecated and retained only for backward compatibility with legacy projects. For new refraction static workflows, use the current refraction tomography and static solver modules available in the Refraction Statics processing group.
The inversion iterates until the sum of squared traveltime residuals across all ray paths falls below a convergence threshold, or until the user stops execution. At each iteration the module updates the display of the current depth maps, the ray fold map, and the convergence error curve so the user can monitor progress in real time.
This deprecated module does not accept interactive input data connections through the processing flow. The first-break pick data and the initial subsurface model are loaded internally from fixed file paths and hardcoded model parameters. No user-configurable input data items are exposed in the module interface.
This deprecated module has no user-configurable parameters. All inversion settings — including the number of model layers, grid spacing, damping factor, convergence tolerance, and minimum fold threshold — are fixed internally. To control these settings, use the current refraction statics workflow modules.
The primary output is a set of static correction values derived from the inverted refractor depth model. These corrections account for the travel-time delay caused by low-velocity near-surface layers and can be applied to seismic traces to improve the alignment of subsurface reflection events.
The module provides several real-time display panels that update after each inversion iteration:
A map view showing the starting depth model for each refractor layer before the inversion begins. Use this panel to verify that the initial model is reasonable before running the iterative solver.
A map view of the current inverted refractor depth for each layer, updated at the end of every iteration. The depth values (in metres) represent the distance from the surface to each refractor interface at each grid node. Monitor this panel to observe how the depth model evolves as the inversion converges.
A map showing the number of refracted ray paths that pass through each depth-model grid cell (ray fold). Areas with high fold are well-constrained by the data; areas with low fold may produce unreliable depth estimates. Cells below the minimum fold threshold are excluded from the model update.
A map view showing the geographic positions of all source-receiver pairs used in the inversion. Use this panel to confirm that the input pick geometry covers the area of interest.
A curve showing the sum of squared traveltime residuals plotted against iteration number. A steadily decreasing error curve indicates a stable, converging inversion. If the error curve flattens or oscillates, the inversion has reached its practical limit and can be stopped manually.