Geoscience Reference
In-Depth Information
so picking the entire survey would take a year of solid mechanical effort, with no time
allowed for thinking about the meaning of the data. The practice therefore grew up of
interpreting only a proportion of the data, say every 10th inline and crossline. Often this
was sufficient for structural mapping; the benefit of having data correctly positioned in
space after 3-D migration was secured. However, the fine detail that was present in the
closely spaced data was lost.
The use of workstations for 3-D interpretation was therefore welcomed by inter-
preters. They offered several advantages:
(i) the ability to view sections through the data in any orientation,
(ii) automatic book-keeping of manually picked horizons: picks made on one line
would automatically be transferred to other lines or to map views,
(iii) semi-automated horizon picking,
(iv) calculation of pick attributes that can be used to extract additional information,
(v) ability to see the data volume in 3-D, not just as sections.
To achieve all this requires the use of fairly powerful workstations, and
Appendix 1
describes some of the hardware and data management requirements. Each of the topics
on the above list will now be addressed in turn.
3.2.1
Display capabilities
The 3-D seismic traces can be thought of as a volume of seismic amplitude values. In the
example discussed in the previous section, there would be 500 000 traces arranged on a
rectangular grid in map view, 500 inlines by 1000 crosslines. The two-way time on the
vertical axis might range from 0 to 4000 ms, sampled at 4 ms, giving us 1000 samples
this 'cube' of data. There are the obvious inlines and crosslines, but also horizontal
slices (time slices), and vertical sections at any orientation through the volume. These
'arbitrary lines' do not have to be straight; they might, for example, be constructed to
join up a number of well locations.
There are two possible modes of presenting seismic sections on the screen: as wiggle
traces or as 'variable intensity' displays. In either case, there are limitations imposed
by the screen resolution. This might, for instance, be 1024 by 1024 pixels. (A
pixel
is the smallest independently controllable element of a screen display. Software can
specify the brightness and colour of each pixel on the screen but cannot achieve any
higher (
x
,
y
) resolution than the pixel.) To get reasonable dynamic range on a wiggle
trace display, the trace would need to extend over, say, 10 columns of pixels. If the
traces do not overlap, this would imply that only 100 or so traces could be displayed at
any one time. Traces can be allowed to overlap in order to view more of them, but even
so a wiggle trace display will be limited to only a few hundred traces. This is suitable
for detailed work (e.g. well ties or study of lateral changes in loop character), but
makes it difficult to obtain an overview of the data. It is therefore often better to work in
