Plots and sections

The migrated and depth-converted sections are visualized as monochrome variable-area plots (with or without wiggle lines) and envelope sections. These two can also be overlaid with one another, producing a “two-attribute” picture. Two-attribute representation is typically used for near-surface data.

The variable-area plot is the classical way to visualize traces. The area between the baseline and the trace is filled in black whenever the polarity of the trace is positive. The trace itself (“the wiggle”) may or may not be drawn. The data is clipped at the 98th percentile before plotting to bring out the smallest features.

In envelope sections, each trace is replaced by the instantaneous amplitude (envelope) of the corresponding analytic trace, and the data is smoothed by averaging it to e.g. a 500-meter resolution. The data is represented as a seismic contour plot, with double linear interpolation performed between sample points. The colour map is a logarithmic greyscale or a colour gradient.

The OpenFIRE section viewer incorporates two types of envelope sections. The smaller-scale section represents migrated and depth-converted NMO stacks and is plotted in logarithmic greyscale. The depth range is from 0 km to 80 km. The medium-scale section is based on DMO stacks and is represented as a two-attribute picture, with a pale-coloured contour section (see Fig. 1 below) overlaid by a monochrome wiggle (+ variable-area) plot. The depth range is from 0 km to 40 km. A comparison of the two representations is in Fig. 2 (on the left).

Colour scale for FIRE sections

Fig. 1: The logarithmic colour scale used for OpenFIRE contour sections, extending from very pale blue (low values) to pale pink (high values).

The envelope sections were designed at the Institute of Seismology, University of Helsinki. They were directly written in PostScript code using a FORTRAN interface to the commercial FGL–AGL library (Release 7.5) included in the UNIRAS product family by Advanced Visual Systems Inc. The SCOLC subroutine draws a seismic colour contour section, and the SWIGG subroutine draws a seismic wiggle line and/or variable area.

The PostScript code was interpreted into a bitmap using GIMP by the GIMP Development Team, Ghostscript Suite by Artifex Software, and other similar GNU-licensed tools. The resolution was kept high to allow zooming and panning into the section.

A Matlab script for plotting greyscale envelope sections from stacked .su data is included as a reference and can be found in the OpenFIRE tools GitHub repository.

Fig. 2: (on the left) A comparison of the NMO (left) and DMO (right) envelope sections as they are represented in OpenFIRE. The NMO example is from FIRE 4B and the DMO example is from FIRE 3. The horizontal tick interval is 5 kilometers.

The most-detailed representations distributed in OpenFIRE are variable-area plots of the upper 8 kilometers of the crust. They were produced using a modified version of supswigb program included in the Seismic Unix collection by CWP. Variable-area plots are distributed as PDF books with 530 traces per page (~ 13 km) and 10 % of overlap between consecutive pages.

Coherence filtering has been used to enhance reflections. At each data point, we sweep over a range of slownesses (angles), and if the data point appears to be part of a coherent reflection, we enhance amplitudes locally. It is important to choose the range of sweep properly in order to avoid aliasing in the slant-stack domain. Aliasing should theoretically be avoided when | p | ≤ π / (ωmax Δx) but it might still appear in the PDF books, especially in areas with conflicting dips.

An implementation of the coherence filtering algorithm is included in the OpenFIRE tools GitHub repository.

In OpenFIRE, the plotting direction is always such that west and south are to the left, and east and north are to the right. This means that CMP numbers increase from left to right in FIRE 3 and 4 & 4A, but from right to left in FIRE 1, 2, 3A, and 4B.