Based in Mahone Bay, nova scotia, agile libre is an independent publisher of technical books about the subsurface.  

Practise smart autotracking

Horizon autotracking is one of the most critical tasks that we do when interpreting seismic data on a workstation. All interpretation systems contain functionality for this, but each is different in the sense that the user has varying degrees of control over tracking parameters. Three important considerationsfor using horizon autotracking are:

1. The data to be autotracked must have a sufficiently good signal-to-noise ratio so that you can reasonably expect autotracking to produce reliable results. It is your responsibility to assess overall data quality and signal:noise on a horizon-by-horizon basis before using autotracking.

2. Once you’ve decided to use autotracking, you should invest time in testing the sensitivity of autotracking output to changes in user-specified tracking parameters. These parameters include, but are not limited to: the time interval of the correlation window; the ‘search distance’ or amount of shift up or down along the seismic trace that the correlation window can be moved in making a pick; and the goodness-of-fit required to propagate the pick to the next traces. As you might guess, autotracking is iterative in the sense that you usually converge on an optimum set of tracking parameters as you evaluate tracking results and adjust input values. Having said this, you should also be aware that any given set of tracking of parameters will probably vary both laterally and vertically throughout a project because of changes in data quality and geology. Most autotrackers have functionality to allow the parameters to vary iteratively.

3. Within an interpretation workflow you should include time for quality control of tracking output. Autotracking results are acceptable only if the tracking algorithm makes its pick in the same place that you would. For the sake of both accuracy and efficiency you probably should not autotrack a horizon which you spend more time correcting than you would have spent, in the limiting case, picking every line in your dataset. Within the context and business constraints of your interpretation project you must decide on the degree of tracking accuracy you need to achieve acceptable results. This decision is not always obvious or easily made at the beginning of an interpretation, and can change as your work progresses.

In addition to these three considerations, you should remember that autotracking operates by correlating from trace to trace on the basis of similarity of reflection character. Accordingly, there are some geologic surfaces, for example the walls of incised channels, which by their nature do not lend themselves to autotracking and must be manually picked. Your judgment in when and how to autotrack horizons will develop as you gain experience in the art of seismic interpretation.

Further reading
Herron, D A (2000). Horizon autopicking: The Leading Edge 19, 491–492.

How dinosaur tracks were made