Any geometry student (at least those who pay attention) can explain the concept of dimensions. One dimension is a line, two is a plane, etc. In our three-dimensional world, it’s hard to get a concept of a fourth spatial dimension, but many people consider time as the fourth dimension.

In seismic this makes sense — a 4-D survey is simply two or more 3-D surveys shot in the same place at different times to measure differences in the reservoir. In recent years many have come to favor the term “time-lapse” seismic to avoid giving the impression that since 3-D seismic offers better resolution than 2-D seismic, 4-D resolution is comparably better than 3-D.

So when I saw the term “5-D” crop up in a press release recently, I shook my head in wonder. How many more Ds are there going to be? If the fourth dimension is time, what is the fifth dimension?

Well, the only ready answer I had was a 1960s quintet that had a big hit with the song “Age of Aquarius” from the musical “Hair.” Something told me that this was not the right answer.

So I called the experts (and the people who put this term into their press release) at Landmark. Turns out what you learned in geometry has a different twist when applied to Landmark’s data processing and interpretation.

“We’re not trying to confuse anyone,” said John Basche, seismic processing product manager with Landmark. “We’re looking at dimensions slightly differently than you would typically look at them.”

No kidding. Basche said that standard 3-D seismic data is broken down by inlines, crosslines, and the velocity of the sound waves. Traditional 4-D adds the component of changes over time. But 4-D is not just time-lapse at Landmark.

“I’d ask you to put that limited concept aside,” he said. “We’re looking at ways to build frameworks around our seismic data so we can do more powerful things to it, whether it’s noise suppression or depth imaging.”

Basche said Landmark’s processing product supports a prestack data format called JavaSeis. Much of the work in the JavaSeis project is based on designs originally worked out in a joint project between Sun Microsystems and Arco. “The idea is that you can take typical seismic data and build a ‘smart wrapper’ around it that allows you to orient the data, transpose it, sort it, and look at it from different perspectives,” Basche said. “Ultimately you can apply true 3-D operators or noise-suppression tools on it.”

He added that while 3-D processing includes multiple domains, JavaSeis puts multiple domains at the user’s fingertips in a 5-D context, not just one domain at a time or multiple domains sequentially. For example, in a 3-D marine survey, conventional axes might be the inline, crossline, velocity, and offset. “Then for a 5-D dataset, maybe you would consider azimuth,” he said. “Because we have this JavaSeis format and it’s designed to run on cluster computer environments, we’re able to organize the signal to run more powerful algorithms on it.”

Typically, he said, 3-D seismic processing utilizes 2-D algorithms interpolated to consider 3-D geometry. “For us, by using this JavaSeis framework, we can write true 3-D algorithms and ultimately have a better spatial understanding of the seismic data to better utilize it in things like wide-azimuth surveys.”

Is the world ready for another dimension? Basche said that his company has been working for the past year or so with its user groups to incorporate some of the JavaSeis tools. “There are companies that are in tune with this,” he said. “But there’s a bigger learning curve to understand how you process this data. It requires you to understand more about your seismic data than you’ve had to in the past because studying these frameworks isn’t a push of a button all the time. You have to understand the context of your data.

“Ultimately, you’re looking at faster processing execution and more accurate subsurface pictures, so the payoff is there,” he added. “But we’re facing a learning curve.”
How many more “dimensions” are out there? “It depends on how you slice and dice your data,” Basche said. “But right now we only support 5-D frameworks.”
That’s probably enough.