Roll-Along Data Format for 2D Cross Section

Data analysis starts with input seismic field files (also called
records) usually in
SEG-2 format, which is an engineering
standard.  The entire steps can be divided into pre- and main-
process steps as illustrated by a generalized flowchart on the

Pre-Process Step

This is an optional step that may or may not be required
depending on the specific analysis software being used, and also
on the actions taken during data acquisition (for example,
whether or not the acquisition geometry information was
previously encoded during data acquisition).  

  • Internal Format: Depending on the specific analysis
    software, the input data set in SEG-2 is often converted
    into a proprietary format of the software as the very first
    step of the processing.  This may take place internally with
    the software or requires a separate preparation of the
    converted data set.   

  • Source/Receiver (SR) Setup: All field files (records)
    have to possess correct information about acquisition
    geometry such as surface coordinates of source and
    receivers.  Although this information is often encoded by
    acquisition software during data acquisition in the field, it is
    usually encoded (or re-encoded) through a data
    processing step provided by the analysis software.
Main-Process Step

This is a required step that performs theoretical analysis of surface waves contained in input seismic data set with proper field geometry information.
  • Dispersion Analysis (Step 1 of 2: Generation of Dispersion Images): This generate one dispersion image (often called overtone) from one
    field record by using a proper 2-D wavefield transformation method (e.g., phase-shift method, tau-pi trasformation, f-k, etc.).  

  • Dispersion Analysis (Step 2 of 2: Extraction of Dispersion Curve): This extracts one fundamental-mode (M0) dispersion curve from each
    of the dispersion image (overtone) generated in the previous step.  This curve is called a "measured M0 curve."   

  • Inversion of Dispersion Curve (M0): This tries to find a 1-D shear-wave velocity (Vs) model (i.e., Vs variation with depth) whose theoretical
    dispersion curve (M0) best matches the measured M0 curve.

The final output is a
1-D velocity (Vs) profile for each field record.  If input data set contained field record(s) obtained by using a stationary receiver
array, then there will be only one 1-D Vs profile obtained as output.  On the other hand, if multiple field records obtained at successively different
locations along a survey line were used as an input data set, then multiple 1-D velocity (Vs) profiles will be generated at the end.  In this case, a 2-D
interpolation scheme is used to generate a
2-D velocity (Vs) cross section as output.