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

Beware the interpretation-to-data trap

Some say there is an art to seismic interpretation. Art is often described as any work that is difficult for someone else to reproduce. It contains an inherent tie to the creator. In this view, it is more correct to say that seismic interpretation is art.

Subsurface geoscience in general, and seismic interpretation in particular, pre- sents the challenge of having long, complex workflows with many interim sections, maps, and so on. As the adage of treasure and trash goes; one person’s interpretation is another person’s data. The routine assignment of interpretations as data is what I call the interpretation–to–data trap, and we all fall into it.

Is a horizon an interpretation, or is it data? It depends on whom you ask. To the interpreter, their horizons are proud pieces of art, the result of repetitious decision making, intellectual labour, and creativity. But when this art is transferred to a geomodeller, for instance, it instantly loses its rich, subjective history. It becomes just data. Without fail, interpretations become data in the possession of anyone other than the creator. It is a subtle but significant concept. And consider the source from which the interpreters’ horizon manifests: seismic amplitudes. Stacked seismic data is but one solution from a choice of migration algorithms, which is itself an interpretive process. More data from art. To some extent, this is true for anything in the subsurface, whether it be the wireline log, production log, or well top.

There are a number of personal and social forces that deepen the trap, such as lack of ownership or lack of foresight. Disowning or disliking data is easy because it is impersonal; disowning your own interpretation however is self-sabotage. People are rarely blamed for spending time on something within their job description, but it is seldom anyone’s job to transfer the implicit (the assumptions, the subjectivity, the guesswork, the art) along with the explicit. It takes foresight and communication at a personal level, and it takes a change in the culture of responsibilities, and maybe even a loosening of the goals on a team level.

Because of the interpretation–to–data trap, we must humbly recognize that even the most rigorous seismic interpretation can be misleading as it is passed downstream and farther afield. If you have ever felt that a dataset is being pushed beyond its limits, that a horizon was not picked with sufficient precision for horizontally steering a drill bit, or a log-conditioning exercise was not suitable for an AVO analysis, it was probably a case of interpretation being mistaken for data. Sometimes these are inevitable assumptions, but it doesn’t absolve us of responsibility.

I think there are three things you can do. One is for giving, one is for receiving, and one is for finishing. When you are giving, arm yourself with the knowledge that your interpretations will transform to data in remote parts of the subsurface realm. When receiving, ask yourself, ‘is there art going on here?’ If so, recognize that you are at risk of falling into the interpretation–to–data trap. Finally, declare your work to be work in progress, not because it is a way to cop-out or delay finishing, but because it is an opportunity to embrace iteration and make it a necessary part of your team dynamic. 

Don’t neglect your math

Calibrate your intuition