A revolution in seismic acquisition?

We're in warm, sunny Calgary for the GeoConvention 2013. The conference feels like it's really embracing geophysics this year — in the past it's always felt more geological somehow. Even the exhibition floor felt dominated by geophysics. Someone we spoke to speculated that companies were holding their geological cards close to their chests, but the service companies are still happy to talk about (ahem, promote) their geophysical advances.

Are you at the conference? What do you think? Let us know in the comments.

We caught about 15 talks of the 100 or so on offer today. A few of them ignited the old whines about half-cocked proofs of efficacy. Why is it still acceptable to say that a particular seismic volume or inversion result is 'higher resolution' or 'more geological' with nothing more than a couple of sections or timeslices as evidence?

People are excited about designing seismic acquisition expressly for wavefield reconstruction. In a whole session devoted to the subject, for example, Mauricio Sacchi showed how randomization helps with regularization in processing, allowing us to either get better image quality, or to lower cost. It feels like the start of a new wave of innovation in acquisition, which has more than its fair share of recent innovation: multi-component, wide azimuth, dual-sensor, simultaneous source...

Is it a revolution? Or just the fallacy of new things looking revolutionary... until the next new thing? It's intriguing to the non-specialist. People are talking about 'beyond Nyquist' again, but this time without inducing howls of derision. We just spent an hour talking about it, and we think there's something deep going on... we're just not sure how to articulate it yet.

Unsolved problems

We were at the conference today, but really we are focused on the session we're hosting tomorrow morning. Along with a roomful of adventurous conference-goers (you're invited too!), looking for the most pressing questions in subsurface science. We start at 8 a.m. in Telus 101/102 on the main floor of the north building.

O is for Offset

Offset is one of those jargon words that geophysicists kick around without a second thought, but which might bewilder more geological interpreters. Like most jargon words, offset can mean a couple of different things: 

• Offset distance, which is usually what is meant by simply 'offset'.
• Offset angle, which is often what we really care about.
• We are not talking about offset wells, or fault offset.

What is offset?

Sherriff's Encyclopedic Dictionary is characteristically terse:

Offset: The distance from the source point to a geophone or to the center of a geophone group.

The concept of offset only really makes sense in the pre-stack world — to field data and gathers. The traces in stacked data (everyday seismic volumes) combine data from many offsets. So let's look at the geometry of seismic acquisition. A map shows the layout of shots (red) and receivers (blue). We can define offset and azimuth A at the midpoint of every shot–receiver pair, on a map (centre) and in section (right):

Offset distance applies to traces. The offset distance is the straight-line distance from the vibrator, shot-hole or air-gun (or any other source) to the particular receiver that recorded the trace in question. If we know the geometry of the acquisition, and the size of the recording patch or length of the streamers, then we can calculate offset distance exactly. 

Offset angle applies to specific samples on a trace. The offset angle is the incident angle of the reflected ray that that a given sample represents. Samples at the top of a trace have larger offset angles than those at the bottom, even though they have the same offset distance. To compute these angles, we need to know the vertical distances, and this requires knowledge of the velocity field, which is mostly unknown. So offset angle is not objective, but a partly interpreted quantity.

Why do we care?

Acquiring longer offsets can help undershoot gaps in a survey, or image beneath salt canopies and other recumbent features. Longer offsets also helps with velocity estimation, because we see more moveout.

Looking at how the amplitude of a reflection changes with offset is the basis of AVO analysis. AVO analysis, in turn, is the basis of many fluid and lithology prediction techniques.

Offset is one of the five canonical dimensions of pre-stack seismic data, along with inline, crossline, azimuth, and frequency. As such, it is a key part of the search for sparsity in the 5D interpolation method perfected by Daniel Trad at CGGVeritas. 

Recently, geophysicists have become interested not just in the angle of a reflection, but in the orientation of a reflection too. This is because, in some geological circumstances, the amplitude of a reflection depends on the orientation with respect to the compass, as well as the incidence angle. For example, looking at data in both of these dimensions can help us understand the earth's stress field.

Offset is the characteristic attribute of pre-stack seismic data. Seismic data would be nothing without it.

Brittleness and robovibes

Day 3 of the SEG Annual Meeting was just as rammed with geophysics as the previous two days. I missed this morning's technical program, however, as I've taken on the chairpersonship (if that's a word) of the SEG Online Committee. So I had fun today getting to grips with that business. Aside: if you have opinion's about SEG's online presence, please feel free to send them my way.

Here are my highlights from the rest of the day — both were footnotes in their respective talks:

Brittleness — Lev Vernick, Marathon

Evan and I have had a What is brittleness? post in our Drafts folder for almost two years. We're skeptical of the prevailing view that a shale's brittleness is (a) a tangible rock property and (b) a function of Young's modulus and Poisson's ratio, as proposed by Rickman et al. 2008, SPE 115258. To hear such an intellect as Lev declare the same today convinced me that we need to finish that post — stay tuned for that. Bottom line: computing shale brittleness from elastic properties is not physically meaningful. We need to find more appropriate measures of frackability, which Lev pointed out is, generally speaking, inversely proportional to organic content. This poses a basic conflict for those exploiting shale plays. 

Robovibes — Guus Berkhout, TU Delft

At least 75% of Berkhout's talk went by me today, mostly over my head. I stopped writing notes, which I only do when I'm defeated. But once he'd got his blended source stuff out of the way, he went rogue and asked the following questions:

1. Why do we combine all seismic frequencies into the device? Audio got over this years ago (right).
2. Why do we put all the frequencies at the same location? Viz 7.1 surround sound.
3. Why don't we try more crazy things in acquisition?

I've wondered the same thing myself — thinking more about the receiver side than the sources — after hearing about the brilliant sampling strategy the Square Kilometer Array is using at a PIMS Lunchbox Lecture once. But Berkhout didn't stop at just spreading a few low-frequency vibrators around the place. No, he wants robots. He wants an autonomous army of flying and/or floating narrow-band sources, each on its own grid, each with its own ghost matching, each with its own deblending code. This might be the cheapest million-channel acquisition system possible. Berkhout's aeronautical vibrator project starts in January. Seriously.

More posts from SEG 2012. 

Speaker image is licensed CC-BY-SA by Tobias Rütten, Wikipedia user Metoc.

Fold for sale

A few weeks ago I wrote a bit about seismic fold, and why it's important for seeing through noise. But how do you figure out the fold of a seismic survey?

The first thing you need to read is Norm Cooper's terrific two-part land seismic tutorial. One of his main points is that it's not really fold we should worry about, it's trace density. Essentially, this normalizes the fold by the area of the natural bins (the areal patches into which we will gather traces for the stack). Computing trace density, given effective maximum offset X_max (or depth, in a pinch), source and receiver line spacings S and R, and source and receiver station intervals s and r:

Cooper helpfully gave ballpark ranges for increasingly hard imaging problems. I've augmented it, based on my own experience. Your mileage may vary! (Edit this table)

Traces cost money

So we want more traces. The trouble is, traces cost money. The chart below reflects my experiences in the bitumen sands of northern Alberta (as related in Hall 2007). The model I'm using is a square land 3D with an orthogonal geometry and no overlaps (that is, a single swath), and 2007 prices. A trace density of 50 traces/km² is equivalent to a fold of 5 at 500 m depth. As you see, the cost of seismic increases as we buy more traces for the stack. Fun fact: at a density of about 160 000 traces/km², the cost is exactly $1 per trace. The good news is that it increases with the square root (more or less), so the incremental cost of adding more traces gets progressively cheaper:

Given that you have limited resources, your best strategy for hitting the 'sweet spot'—if there is one—is lots and lots of testing. Keep careful track of what things cost, so you can compute the probable cost benefit of, say, halving the trace density. With good processing, you'll be amazed what you can get away with, but of course you risk coping badly with unexpected problems in the near surface.

What do you think? How do you make decisions about seismic geometry and trace density?

References

Cooper, N (2004). A world of reality—Designing land 3D programs for signal, noise, and prestack migration, Parts 1 and 2. The Leading Edge. October and December, 2004. 

Hall, M (2007). Cost-effective, fit-for-purpose, lease-wide 3D seismic at Surmont. SEG Development and Production Forum, Edmonton, Canada, July 2007.

The power of stack

Multiplicity is a basic principle of seismic acquisition. Our goal is to acquite lots of traces—lots of spatial samples—with plenty of redundancy. We can then exploit the redundancy, by mixing traces, sacrificing some spatial resolution for increased signal:noise. When we add two traces, the repeatable signal adds constructively, reinforcing and clarifying. The noise, on the other hand, is spread evenly about zero and close to random, and tends to cancel itself. This is why you sometimes hear geophysicists refer to 'the power of stack'. 

Here's an example. There are 20 'traces' of 100-digit-long sequences of random numbers (white noise). The numbers range between –1 and +1. I added some signal to samples 20, 40, 60 and 80. The signals have amplitude 0.25, 0.5, 0.75, and 1. You can't see them in the traces, because these tiny amplitudes are completely hidden by noise. The stacked trace on the right is the sum of the 20 noisy traces. We see mostly noise, but the signal emerges. A signal of just 0.5—half the peak amplitude of the noise—is resolved by this stack of 20 traces; the 0.75 signal stands out beautifully.

Here's another example, but with real data. This is part of Figure 3 from Liu, G, S Fomel, L Jin, and X Chen (2009). Stacking seismic data using local correlation. Geophysics 74 (2) V43–V48. On the left is an NMO-corrected (flattened) common mid-point gather from a 2D synthetic model with Gaussian noise added. These 12 traces each came from a single receiver, though in this synthetic case the receiver was a virtual one. Now we can add the 12 traces to get a single trace, which has much stronger signal, relative to the background noise, than any of the input traces. This is the power of stack. In the paper, Liu et al. improve on the simple sum by weighting the traces adaptively. Click to enlarge.

The number of traces available for the stack is called fold. The examples above have folds of 20 and 12. Geophysicists like fold. Fold works. Let's look at another example.

Above, I've made a single digit 1 with 1% opacity — it's almost invisible. If I stack two 2s, with a little random jitter, the situation is still desperate. When I have five digits, I can at least see the hidden image with some fidelity. However, if I add random noise to the image, a fold of 5 is no longer enough. I need at least 10, and ideally more like 20 images stacked up to see any signal. So it is for seismic data: to see through the noise, we need fold.

Now you know a bit about why we want more traces from the field, next time I'll look at how much those traces cost, and how to figure out how many you need. 

Thank you to Stuart Mitchell of Calgary for the awesome analogy for seismic fold.