SEG machine learning contest: there's still time

Have you been looking for an excuse to find out what machine learning is all about? Or maybe learn a bit of Python programming language? If so, you need to check out Brendon Hall's tutorial in the October issue of The Leading Edge. Entitled, "Facies classification using machine learning", it's a walk-through of a basic statistical learning workflow, applied to a small dataset from the Hugoton gas field in Kansas, USA.

But it was also the launch of a strictly fun contest to see who can get the best prediction from the available data. The rules are spelled out in ther contest's README, but in a nutshell, you can use any reproducible workflow you like in Python, R, Julia or Lua, and you must disclose the complete workflow. The idea is that contestants can learn from each other.

View fullsize

Left: crossplots and histograms of wireline log data, coloured by facies — the idea is to highlight possible data issues, such as highly correlated features. Right: true facies (left) and predicted facies (right) in a validation plot. See the rest of the paper for details.

What's it all about?

The task at hand is to predict sedimentological facies from well logs. Such log-derived facies are sometimes called e-facies. This is a familiar task to many development geoscientists, and there are many, many ways to go about it. In the article, Brendon trains a support vector machine to discriminate between facies. It does a fair job, but the accuracy of the result is less than 50%. The challenge of the contest is to do better.

Indeed, people have already done better; here are the current standings:

	Team	F1	Algorithm	Language	Solution
1	gccrowther	0.580	Random forest	Python	Notebook
2	LA_Team	0.568	DNN	Python	Notebook
3	gganssle	0.561	DNN	Lua	Notebook
4	MandMs	0.552	SVM	Python	Notebook
5	thanish	0.551	Random forest	R	Notebook
6	geoLEARN	0.530	Random forest	Python	Notebook
7	CannedGeo	0.512	SVM	Python	Notebook
8	BrendonHall	0.412	SVM	Python	Initial score in article

As you can see, DNNs (deep neural networks) are, in keeping with the amazing recent advances in the problem-solving capability of this technology, doing very well on this task. Of the 'shallow' methods, random forests are quite prominent, and indeed are a great first-stop for classification problems as they tend to do quite well with little tuning.

How do I enter?

There is still over 6 weeks to enter: you have until 31 January. There is a little overhead — you need to learn a bit about git and GitHub, there's some programming, and of course machine learning is a massive field to get up to speed on — but don't be discouraged. The very first entry was from Bryan Page, a self-described non-programmer who dusted off some basic skills to improve on Brendon's notebook. But you can run the notebook right here in mybinder.org (if it's up today — it's been a bit flaky lately) and a play around with a few parameters yourself.

The contest aspect is definitely low-key. There's no money on the line — just a goody bag of fun prizes and a shedload of kudos that will surely get the winners into some awesome geophysics parties. My hope is that it will encourage you (yes, you) to have fun playing with data and code, trying to do that magical thing: predict geology from geophysical data.

---

Reference

Hall, B (2016). Facies classification using machine learning. The Leading Edge 35 (10), 906–909. doi: 10.1190/tle35100906.1. (This paper is open access: you don't have to be an SEG member to read it.)

Where is the ground?

This is the upper portion of a land seismic profile in Alaska. Can you pick a horizon where the ground surface is? Have a go at pickthis.io.

Pick the Ground surface at the top of the seismic section at pickthis.io.

Picking the ground surface on land-based seismic data is not straightforward. Picking the seafloor reflection on marine data, on the other hand, is usually a piece of cake, a warm-up pick. You can often auto-track the whole thing with a few seeds.

Seafloor reflection on Penobscot 3D survey, offshore Nova Scotia. from Matt's tutorial in the April 2016 The Leading Edge, The function of interpolation.

Why aren't interpreters more nervous that we don't know exactly where the surface of the earth is? I'm sure I'm not the only one that would like to have this information while interpreting. Wouldn't it be great if land seismic were more like marine?

Treacherously Jagged TopographY or Near-Surface processing ArtifactS?

If you're new to land-based seismic data, you might notice that there isn't a nice pickable event across the top of the section like we find in marine seismic data. Shot noise at the surface has been muted (deleted) in processing, and the low fold produces an unclean, jagged look at the top of the section. Additionally, the top of the section, time-zero — the seismic reference datum — usually floats somewhere above the land surface — and we can't know where that is unless it can be found in the file header, or looked up in the processing report.

The seismic reference datum, at a two-way time of zero seconds on seismic data, is typically set at mean sea level for offshore data. For land data, it is usually chosen to 'float' above the land surface.

Reframing the question

This challenge is a bit of a trick question. It begs the viewer to recognize that the seemingly simple task of mapping the ground level on a land seismic section is actually a rudimentary velocity modeling or depth conversion exercise in itself. Wouldn't it be nice to have the ground surface expressed as pickable seismic event? Shouldn't we have it always in our images? Baked into our data, so to speak, such that we've always got an unambiguous pick? In the next post, I'll illustrate what I mean and show what's involved in putting it in. 

In the meantime, I challenge you to pick where you think the (currently absent) ground surface is on this profile, so in the next post we can see how well you did.

x lines of Python: AVO plot

Amplitude vs offset (or, more properly, angle) analysis is a core component of quantitative interpretation. The AVO method is based on the fact that the reflectivity of a geological interface does not depend only on the acoustic rock properties (velocity and density) on both sides of the interface, but also on the angle of the incident ray. Happily, this angular reflectivity encodes elastic rock property information. Long story short: AVO is awesome.

As you may know, I'm a big fan of forward modeling — predicting the seismic response of an earth model. So let's model the response the interface between a very simple model of only two rock layers. And we'll do it in only a few lines of Python. The workflow is straightforward:

1. Define the properties of a model shale; this will be the upper layer.
2. Define a model sandstone with brine in its pores; this will be the lower layer.
3. Define a gas-saturated sand for comparison with the wet sand. 
4. Define a range of angles to calculate the response at.
5. Calculate the brine sand's response at the interface, given the rock properties and the angle range.
6. For comparison, calculate the gas sand's response with the same parameters.
7. Plot the brine case.
8. Plot the gas case.
9. Add a legend to the plot.

That's it — nine lines! Here's the result:

Once we have rock properties, the key bit is in the middle:

    θ = range(0, 31)
    shuey = bruges.reflection.shuey2(vp0, vs0, ρ0, vp1, vs1, ρ1, θ)

shuey2 is one of the many functions in bruges — here it provides the two-term Shuey approximation, but it contains lots of other useful equations. Virtually everything else in our AVO plotting routine is just accounting and plotting.

---

As in all these posts, you can follow along with the code in the Jupyter Notebook. You can view this on GitHub, or run it yourself in the increasingly flaky MyBinder (which is down at the time of writing... I'm working on an alternative).

What would you like to see in x lines of Python? Requests welcome!

What's that funny noise?

Seismic reflections are strange noises. Around 50 Hz, narrow band, very quiet, and difficult to interpret. It is possible to convert seismic traces (active or passive) into audible sound with a shift in pitch and a time stretch.

Made by the legendary Emory Cook, who recorded everything from steel bands to racing cars to ionospheric noises to this treatment of Hugo Benioff's earthquake recordings. Epic.

Curiously the audification thing has never really caught on in exploration geophysics — a bit surprising, given the fascination with spectral decomposition over the last 15 years or so. And especially so when you consider that our hearing has a dynamic range of about 100 dB, which is comparable to, indeed slightly greater than, our vision (about 90 dB).

Paolo Dell'Aversana of ENI wants to change that. Rather than listening to 'raw' seismic, he's sending it to a MIDI interface and listening to it as a piano roll. Just try to imagine playing seismic on a piano for a second, then listen to his weird and wonderful results — at 9:45 in this EAGE video:

In this EAGE E-Lecture Paolo Dell'Aversana discusses how digital music technology can support geophysical data analysis and interpretation. If you've read any of Dell'Aversana's articles, you'll know he has one of the most creative minds in exploration geophysics. Skip to 9:45 for the crazy seismic piano roll.

On the subject of weird sounds, one of my favourite Wikipedia pages is List of unexplained sounds. I especially love the eerie recordings of mysterious underwater noises, like this one called Upsweep:

No-one knows what makes that noise! My money's on a volcanic vent, but that doesn't explain the seasonality. Maybe we should do a hackathon on these unexaplained sounds some time. If you know of any others — I'd love tohear about them.

---

If you enjoy strange infrasound as much as I do, I recommend following these two scientists on Twitter:

• Milton Garces, aka Infrasound Hunter and director of the Infrasound LAb at the University of Hawaii.

• Stephen Gibbons, NORSAR seismologist who monitors nuclear testing with seismology.

---

If you really like strange noises, don't forget to check out the Undersampled Radio podcast!

Hooke's oolite

The new book

Hooke's engraving

Hooke's microscope

52 Things You Should Know About Rock Physics came out last week. For the first, and possibly the last, time a Fellow of the Royal Society — the most exclusive science club in the UK — drew the picture on the cover. The 353-year-old drawing was made by none other than Robert Hooke. 

The title page from Micrographia, and part of the dedication to Charles II. You can browse the entire book at archive.org.

The drawing, or rather the engraving that was made from it, appears on page 92 of Micrographia, Hooke's groundbreaking 1665 work on microscopy. In between discovering and publishing his eponymous law of elasticity (which Evan wrote about in connection with Lamé's \(\lambda\)), he drew and wrote about his observations of a huge range of natural specimens under the microscope. It was the first time anyone had recorded such things, and it was years before its accuracy and detail were surpassed. The book established the science of microscopy, and also coined the word cell, in its biological context.

Sadly, the original drawing, along with every other drawing but one from the volume, was lost in the Great Fire of London, 350 years ago almost to the day. 

Ketton stone

The drawing on the cover of the new book is of the fractured surface of Ketton stone, a Middle Jurassic oolite from central England. Hooke's own description of the rock, which he mistakenly called Kettering Stone, is rather wonderful:

I wonder if anyone else has ever described oolite as looking like the ovary of a herring?

These thoughtful descriptions, revealing a profundly learned scientist, hint at why Hooke has been called 'England's Leonardo'. It seems likely that he came by the stone via his interest in architecture, and especially through his friendsip with Christopher Wren. By 1663, when it's likely Hooke made his observations, Wren had used the stone in the façades of several Cambridge colleges, including the chapels of Pembroke and Emmanuel, and the Wren Library at Trinity (shown here). Masons call porous, isotropic rock like Ketton stone 'freestone', because they can carve it freely to make ornate designs. Rock physics in action!

You can read more about Hooke's oolite, and the geological significance of his observations, in an excellent short paper by material scientist Derek Hull (1997). It includes these images of Ketton stone, for comparison with Hooke's drawing:

Reflected light photomicrograph (left) and backscatter scanning electron microscope image (right) of Ketton Stone. Adapted from figures 2 and 3 of Hull (1997). Images are © Royal Society and used in accordance with their terms.

I love that this book, which is mostly about the elastic behaviour of rocks, bears an illustration by the man that first described elasticity. Better still, the illustration is of a fractured rock — making it the perfect preface. 

---

Now at Amazon.com — Amazon.co.uk — Amazon.fr — Amazon.de — Amazon.it — Amazon.es
Soon at Amazon.ca — Amazon.com.mx — Amazon.com.br — Amazon.co.jp — Amazon.com.au — Amazon.in

---

References

Hall, M & E Bianco (eds.) (2016). 52 Things You Should Know About Rock Physics. Nova Scotia: Agile Libre, 134 pp.

Hooke, R (1665). Micrographia: or some Physiological Descriptions of Minute Bodies made by Magnifying Glasses, pp. 93–100. The Royal Society, London, 1665.

Hull, D (1997). Robert Hooke: A fractographic study of Kettering-stone. Notes and Records of the Royal Society of London 51, p 45-55. DOI: 10.1098/rsnr.1997.0005.

52 Things... Rock Physics

There's a new book in the 52 Things family! 

52 Things You Should Know About Rock Physics is out today, and available for purchase at Amazon.com. It will appear in their European stores in the next day or two, and in Canada... well, soon. If you can't wait for that, you can buy the book immediately direct from the printer by following this link.

The book mines the same vein as the previous volumes. In some ways, it's a volume 2 of the original 52 Things... Geophysics book, just a little bit more quantitative. It features a few of the same authors — Sven Treitel, Brian Russell, Rachel Newrick, Per Avseth, and Rob Simm — but most of the 46 authors are new to the project. Here are some of the first-timers' essays:

• Ludmilla Adam, Why echoes fade.
• Arthur Cheng, How to catch a shear wave.
• Peter Duncan, Mapping fractures.
• Paul Johnson, The astonishing case of non-linear elasticity.
• Chris Liner, Negative Q.
• Chris Skelt, Five questions to ask the petrophysicist.

It's our best collection of essays yet. We're very proud of the authors and the collection they've created. It stretches from childhood stories to linear algebra, and from the microscope to seismic data. There's no technical book like it. 

Supporting Geoscientists Without Borders

Purchasing the book will not only bring you profund insights into rock physics — there's more! Every sale sends $2 to Geoscientists Without Borders, the SEG charity that supports the humanitarian application of geoscience in places that need it. Read more about their important work.

It's been an extra big effort to get this book out. The project was completely derailed in 2015, as we — like everyone else — struggled with some existential questions. But we jumped back into it earlier this year, and Kara (the managing editor, and my wife) worked her magic. She loves working with the authors on proofs and so on, but she doesn't want to see any more equations for a while.

If you choose to buy the book, I hope you enjoy it. If you enjoy it, I hope you share it. If you want to share it with a lot of people, get in touch — we can help. Like the other books, the content is open access — so you are free to share and re-use it as you wish. 

Q is for Q

Quality factor, or \(Q\), is one of the more mysterious quantities of seismology. It's right up there with Lamé's \(\lambda\) and Thomsen's \(\gamma\). For one thing, it's wrapped up with the idea of attenuation, and sometimes the terms \(Q\) and 'attenuation' are bandied about seemingly interchangeably. For another thing, people talk about it like it's really important, but it often seems to be completely ignored.

A quick aside. There's another quality factor: the rock quality factor, popular among geomechnicists (geomechanics?). That \(Q\) describes the degree and roughness of jointing in rocks, and is probably related — coincidentally if not theoretically — to seismic \(Q\) in various nonlinear and probably profound ways. I'm not going to say any more about it, but if this interests you, read Nick Barton's book, Rock Quality, Seismic Velocity, Attenuation and Anistropy (2006; CRC Press) if you can afford it. 

So what is Q exactly?

We know intuitively that seismic waves lose energy as they travel through the earth. There are three loss mechanisms: scattering (elastic losses resulting from reflections and diffractions), geometrical spreading, and intrinsic attenuation. This last one, anelastic energy loss due to absorption — essentially the deviation from perfect elasticity — is what I'm trying to describe here.

I'm not going to get very far, by the way. For the full story, start at the seminal review paper entitled \(Q\) by Leon Knopoff (1964), which surely has the shortest title of any paper in geophysics. (Knopoff also liked short abstracts, as you see here.)

The dimensionless seismic quality factor \(Q\) is defined in terms of the energy \(E\) stored in one cycle, and the change in energy — the energy dissipated in various ways, such as fluid movement (AKA 'sloshing', according to Carl Reine's essay in 52 Things... Geophysics) and intergranular frictional heat ('jostling') — over that cycle:

$$ Q \stackrel{\mathrm{def}}{=} 2 \pi \frac{E}{\Delta E} $$

Remarkably, this same definition holds for any resonator, including pendulums and electronics. Physics is awesome!

Because the right-hand side of that relationship is sort of upside down — the loss is in the denominator — it's often easier to talk about \(Q^{-1}\) which is, more or less, the percentage loss of energy in a single wavelength. This inverse of \(Q\) is proportional to the attenuation coefficient. For more details on that relationship, check out Carl Reine's essay.

This connection with wavelengths means that we have to think about frequency. Because high frequencies have shorter cycles (by definition), they attenuate faster than low frequencies. You know this intuitively from hearing the beat, but not the melody, of distant music for example. This effect does not imply that \(Q\) depends on frequency... that's a whole other can of worms. (Confused yet?)

The frequency dependence of \(Q\)

It's thought that \(Q\) is roughly constant with respect to frequency below about 1 Hz, then increases with \(f^\alpha\), where \(\alpha\) is about 0.7, up to at least 25 Hz (I'm reading this in Mirko van der Baan's 2002 paper), and probably beyond. Most people, however, seem to throw their hands up and assume a constant \(Q\) even in the seismic bandwidth... mainly to make life easier when it comes to seismic processing. Attempting to measure, let alone compensate for, \(Q\) in seismic data is, I think it's fair to say, an unsolved problem in exploration geophysics.

Why is it worth solving? I think the main point is that, if we could model and measure it better, it could be a semi-independent measure of some rock properties we care about, especially velocity. Actually, I think it's even a stretch to call velocity a rock property — most people know that velocity depends on frequency, at least across the gulf of frequencies between seismic and acoustic logging tools, but did you know that velocity also depends on amplitude? Paul Johnson tells about this effect in his essay in the forthcoming 52 Things... Rock Physics book — stay tuned for more on that.

For a really wacky story about negative values of \(Q\) — which imply transmission coefficients greater than 1 (think about that) — check out Chris Liner's essay in the same book (or his 2014 paper in The Leading Edge). It's not going to help \(Q\) get any less mysterious, but it's a good story. Here's the punchline from a Jupyter Notebook I made a while back; it follows along with Chris's lovely paper:

Top: Velocity and the Backus average velocity in the E-38 well offshore Nova Scotia. Bottom: Layering-induced attenuation, or 1/Q, in the same well. Note the negative numbers! Reproduction of Liner's 2014 results in a Jupyter Notebook.

Hm, I had hoped to shed some light on \(Q\) in this post, but I seem to have come full circle. Maybe explaining \(Q\) is another unsolved problem.

References

Barton, N (2006). Rock Quality, Seismic Velocity, Attenuation and Anisotropy. Florida, USA: CRC Press. 756 pages. ISBN 9780415394413.

Johnson, P (in press). The astonishing case of non-linear elasticity.  In: Hall, M & E Bianco (eds), 52 Things You Should Know About Rock Physics. Nova Scotia: Agile Libre, 2016, 132 pp.

Knopoff, L (1964). Q. Reviews of Geophysics 2 (4), 625–660. DOI: 10.1029/RG002i004p00625.

Reine, C (2012). Don't ignore seismic attenuation. In: Hall, M & E Bianco (eds), 52 Things You Should Know About Geophysics. Nova Scotia: Agile Libre, 2012, 132 pp.

Liner, C (2014). Long-wave elastic attenuation produced by horizontal layering. The Leading Edge 33 (6), 634–638. DOI: 10.1190/tle33060634.1. Chris also blogged about this article.

Liner, C (in press). Negative Q. In: Hall, M & E Bianco (eds), 52 Things You Should Know About Rock Physics. Nova Scotia: Agile Libre, 2016, 132 pp.

van der Bann, M (2002). Constant Q and a fractal, stratified Earth. Pure and Applied Geophysics 159 (7–8), 1707–1718. DOI: 10.1007/s00024-002-8704-0.

The sound of the Software Underground

If you are a geoscientist or subsurface engineer, and you like computery things — in other words, if you read this blog — I have a treat for you. In fact, I have two! Don't eat them all at once.

Software Underground

Sometimes (usually) we need more diversity in our lives. Other times we just want a soul mate. Or at least someone friendly to ask about that weird new seismic attribute, where to find a Python library for seismic imaging, or how to spell Kirchhoff. Chat rooms are great for those occasions, Slack is where all the cool kids go to chat, and the Software Underground is the Slack chat room for you. 

It's free to join, and everyone is welcome. There are over 130 of us in there right now — you probably know some of us already (apart from me, obvsly). Just go to http://swung.rocks/ to sign up, and we will welcome you at the door with your choice of beverage.

To give you a flavour of what goes on in there, here's a listing of the active channels:

• #python — for people developing in Python
• #sharp-rocks — for people developing in C# or .NET
• #open-geoscience — for chat about open access content, open data, and open source software
• #machinelearning — for those who are into artificial intelligence
• #busdev — collaboration, subcontracting, and other business opportunities 
• #general — chat about anything to do with geoscience and/or computers
• #random — everything else

Undersampled Radio

If you have a long commute, or occasionally enjoy being trapped in an aeroplane while it flies around, you might have discovered the joy of audiobooks and podcasts. You've probably wished many times for a geosciencey sort of podcast, the kind where two ill-qualified buffoons interview hyper-intelligent mega-geoscientists about their exploits. I know I have.

Well, wish no more because Undersampled Radio is here! Well, here:

• iTunes — subscribe in the app on your device. Best for iOS.
• Google Play Music — subscribe in the app. Best for Android.
• UndersampledRad.io — the website and RSS feed.

The show is hosted by New Orleans-based geophysicist Graham Ganssle and me. Don't worry, it's usually not just us — we talk to awesome guests like geophysicists Mika McKinnon and Maitri Erwin, geologist Chris Jackson, and geopressure guy Mark Tingay. The podcast is recorded live every week or three in Google Hangouts on Air — the link to that, and to show notes and everything else — is posted by Gram in the #undersampled Software Underground channel. You see? All these things are connected, albeit in a nonlinear, organic, highly improbable way. Pseudoconnection: the best kind of connection.

Indeed, there is another podcast pseudoconnected to Software Underground: the wonderful Don't Panic Geocast — hosted by John Leeman and Shannon Dulin — also has a channel: #dontpanic. Give their show a listen too! In fact, here's a show we recorded together!

Don't have an hour right now? OK, you asked for it, here's a clip from that show to get you started. It starts with John Leeman explaining what Fun Paper Friday is, and moves on to one of my regular rants about conferences...

In case you're wondering, neither of these projects is explicitly connected to Agile — I am just involved in both of them. I just wanted to clear up any confusion. Agile is not a podcast company, for the time being anyway.

In search of the Kennetcook Thrust

Behind every geologic map, is a much more complex geologic truth. Most of the time it's hidden under soil and vegetation, forcing geologists into a detective game in order to fill gaps between hopelessly sparse spatterings of evidence.

Two weeks ago, I joined up with an assortment of geologists on the side of the highway an hour north of Halifax for John Waldron to guide us along some spectacular stratigraphy exposed in the coastline cliffs on the southern side of the Minas Basin (below). John has visited these sites repeatedly over his career, and he's supervised more than a handful of graduate students probing a variety of geologic processes on display here. He's published numerous papers teasing out the complex evolution of the Windsor-Kennetcook Basin: one of three small basins onshore Nova Scotia with the potential to contain economic quantities of hydrocarbons.

John retold the history of mappers past and present riddled by the massively deformed, often duplicated Carboniferous evaporites in the Windsor Group which are underlain by sub-horizontal seismic reflectors at depth. Local geologists agree that this relationship reflects thrusting of the near-surface package, but there is disagreement on where this thrust is located, and whether and where it intersects the surface. On this field trip, John showed us symptoms of this Kennetcook thrust system, at three sites. We started in the footwall. The second and third sites were long stretches spectacularly deformed exposures in the hangingwall.  

Footwall: Cheverie Point

SEE GALLERY BELOW FOR ENLARGEMENT

The first stop was Cheverie Point and is interpreted to be well in the footwall of the Kennetcook thrust. Small thrust faults (right) cut through the type section of the Macumber Formation and match the general direction of the main thrust system. The Macumber Formation is a shallow marine microbial limestone that would have fooled anyone as a mudstone, except it fizzed violently under a drop of HCl. Just to the right of this photo, we stood on the unconformity between the petroliferous and prospective Horton Group and the overlying Windsor Group. It's a pick that turns out to be one of the most reliably mappable seismic events on seismic sections so it was neat to stand on that interface.

Further down section we studied the Mississippian Cheverie Formation: stacked cycles of point-bar deposits ranging from accretionary lag conglomerates to caliche paleosols with upright tree trunks. Trees more than a metre or more in diameter were around from the mid Devonian, but Cheverie forests are still early and good examples of trees within point-bars and levees.  

Hangingwall: Red Head / Johnson Beach / Split Rock

SEE GALLERY BELOW FOR ENLARGEMENT

The second site featured some spectacularly folded black shales from the Horton Bluff Formation, as well as protruding sills up to two metres thick that occasionally jumped across bedding (right). We were clumsily contemplating the curious occurrence of these intrusions for quite some time until hard-rock guru Trevor McHattie halted the chatter, struck off a clean piece rock with a few blows of his hammer, wetted it with a slobbering lick, and inspected it with his hand lens. We all watched him in silence and waited for his description. I felt a little schooled. He could have said anything. It was my favourite part of the day.

Hangingwall continued: Rainy Cove

The patterns in the rocks at Rainy Cove are a wonderland for any structural geologist. It's a popular site for geology labs from Atlantic Universities, but it would be an absolute nightmare to try to actually measure the section here. 

SEE GALLERY BELOW FOR ENLARGEMENT

John stands next to a small system of duplicated thrusts in the main hangingwall that have been subsequently folded (left). I tried tracing out the fault planes by following the offsets in the red sandstone bed amidst black shales whose fabric has been deformed into an accordion effect. Your picks might very well be different from mine.

A short distance away we were pointed to an upside-down view of load structures in folded beds. "This antiform is a syncline", John paused while we processed. "This synform over here is an anticline". Features telling of such intense deformation are hard to fathom. Especially in plain sight.

The rock lessons ended in the early evening at the far end of Rainy Cove where the Triassic Wolfville formation sits unconformably on top of ridiculously folded, sometimes doubly overturned Carboniferous Horton Rocks. John said it has to be one of the most spectacularly exposed unconformities in the world. 

I often take for granted the vast stretches of geology hiding beneath soil and vegetation, and the preciousness of finding quality outcrop. Check out the gallery below for pictures from our day.  

Cheverie Point

Lineations of tilted beds exposed only at low tide show the succession of strata at Cheverie Point.

Sketching on the horizon

John Waldron sketches Kennetcook Thrust system to describe the deformed packages on the horizon.

Footwall thrust

At Cheverie Point, a macro-sized thrust fault in sits in the footwall of the main Kennetcook Thrust system. 

Visualization on the rocks

Contemplating the geology map in-situ.

Upright trees

Carboniferous aged tree several inches in diameter standing up right in caliche paleosol in the Cheverie Formation.

Cheverie terrestrial red beds

Crosscutting sill

At site 2: A nearly 2 metre thick sill delaminates Horton strata and cross-cuts bedding in places.

Folds

Trevor McHattie for scale

Hangingwall folds

Site 2: Incredible folding at Red Head (Johnson's Cove)

Sketching lacustrine cycles

John sketches depositional sequence of lacustrine rocks at Red Head.

Triple trust

At site 3: a twice-duplicated series of thrust faults massively folded packages in the Hangingwall of the Kennetcook Thrust.

Mesozoic Unconformity

Dramatic unconformity between massively deformed, sometimes doubly overturned Carboniferous strata cut by gently dipping Mesozoic rocks of the Wolfville Formation.

Rainy Cove

I was quite enamoured with John's format. His field trip technologies. The maps and sections: canvases for communication and works in progress. His white boarding, his map-folding techniques: a practised impresario.

What are some of the key elements from the best field trips you've been on? Let us know in the comments.

Automated interpretation highlights

As you probably know by know, I've been at the EAGE conference in Vienna this week. I skipped out yesterday and flew over to the UK for a few days. I have already written plenty about the open source workshop, and I will write more soon about the hackathon. But I thought I'd pass on my highlights from the the Automated Interpretation session on Tuesday. In light of Monday's discussion, I made a little bit of a nuisance of myself by asking the same post-paper question every time I got the chance:

Can I use your code, either commercially or for free?

I'll tell you what the authors responded.

---

The universal character of salt

I especially enjoyed the presentation by Anders Waldeland and Anne Solberg (University of Oslo) on automatically detecting salt in 3D seismic. (We've reported on Anne Solberg's work before.) Anders described training eight different classifiers, from a simple nearest mean to a neural network, a supprt vector model, and a mixture of Gaussians classifier. Interestingly, but not surprisingly, the simplest model turned out to be the most effective at discrimination. He also tried a great many seismic attributes, letting the model choose the best ones. Three attributes consistently proved most useful: coherency, Haralick energy (a GLCM-based texture attribute), and the variance of the kurtosis of the amplitude distribution (how's that for meta?). What was especially interesting about his approach was that he was training the models on one dataset, and predicting on an entirely different 3D. The idea is that this might reveal the universal seismic characteristics of salt. When I asked the golden question, he said "Come and talk to me", which isn't a "yes", but it isn't a "no" either.

Waldeland and Solberg 2016. Salt probability in a North Sea dataset (left) and the fully tracked volume (right). The prediction model was trained on a Gulf of Mexico dataset. Copyright of the authors and EAGE, and used under a Fair Use claim.

Secret horizon tracker

Horizons tracked with Figueiredo et al's machine learning algorithm. The horizons correctly capture the discontinuities. Copyright of the authors and EAGE. Used under a Fair Use claim.

The most substantial piece of machine learning I saw was Eduardo Figueiredo from Pontifical Catholic University in Rio, in the same session as Waldeland. He's using a neural net called Growing Neural Gas to classify (aka or 'label') the input data in a number of different ways. This training step takes a little time. The label sets can now be compared to decide on the similarity between two samples, based on the number of labels the samples have in common but also including a comparison to the original seed, which essentially acts as a sort of brake to stop things running away. This progresses the pick. If a decision can't be reached, a new global seed is selected randomly. If that doesn't work, picking stops. Unfortunately he did not show a comparison to an ordinary autotracker, either in terms of time or quality, but the results did look quite good. The work was done 'in cooperation with Petrobras', so it's not surprising the code is not available. I was a bit surprised that Figueiredo was even unable to share any details of the implementation.

More on interpretation

The other two interesting talks in the session were two from Paul de Groot (dGB Earth Sciences) and Gaynor Paton (GeoTeric). Paul introduced the new Thalweg Tracker in OpendTect — the only piece of software from the session that you can actually run, albeit for a fee — which is a sort of conservative voxel tracker. Unsurprisingly, Paul was also very thorough with his examples, and his talk served as a tutorial in how to make use of, and give attribution to, open data. (I'm nearly done with the grumbling about openness for now, I promise, but I can't help mentioning that I find it a bit ironic that those scientists unwilling to share their work are also often a bit lax with giving credit to others whose work they depend on.)

Gaynor's talk was about colour, which you may know we enjoy thinking about. She had gathered 24 seismic interpreters, five of whom had some form of colour deficiency. She gave the group some interpretation tasks, and tried to gauge their performance in the tasks. It seemed interesting enough, and I immediately wondered if we could help out with Pick This, especially to help grow the sample size, and by blinding the study. But the conclusion seemed to be that, if there are ways in which colour blind interpreters are less capable at image interpretation, for example where hue is important, they compensate for it by interpreting other aspects, such as contrast and shape. 

Paton's research into how colour deficient people interpret attributes. There were 5 colour deficient subjects and 19 colour normal. The colour deficient subjects were more senstive to subtle changes in saturation and to feature shapes. Image copyright Paton and EAGE, and used here under a fair use claim.

That's it for now. I have a few other highlights to share; I'll try to get to them next week. There was a bit of buzz around the Seismic Apparition talks from ETHZ and Statoil, for example. If you were at the conference, I'd love to hear your highlights too, please share them in the comments.

References

A.U. Waldeland* (University of Oslo) & A.H.S. Solberg (University of Oslo). 3D Attributes and Classification of Salt Bodies on Unlabelled Datasets. 78th EAGE Conference & Exhibition 2016. DOI 10.3997/2214-4609.201600880. Available in EarthDoc.

M. Pelissier (Dagang Zhaodong Oil Company), C. Yu (Dagang Zhaodong Oil Company), R. Singh (dGB Earth Sciences), F. Qayyum (dGB Earth Sciences), P. de Groot* (dGB Earth Sciences) & V. Romanova (dGB Earth Sciences). Thalweg Tracker - A Voxel-based Auto-tracker to Map Channels and Associated Margins. 78th EAGE Conference & Exhibition 2016. DOI 10.3997/2214-4609.201600879. Available in EarthDoc. 

G. Paton* (GeoTeric). The Effect of Colour Blindness on Seismic Interpretation. 78th EAGE Conference & Exhibition 2016. DOI 10.3997/2214-4609.201600883. Available in EarthDoc.

A.M. Figueiredo* (Tecgraf / PUC-Rio), J.P. Peçanha (Tecgraf / PUC-Rio), G.M. Faustino (Tecgraf / PUC-Rio), P.M. Silva (Tecgraf / PUC-Rio) & M. Gattass (Tecgraf / PUC-Rio). High Quality Horizon Mapping Using Clustering Algorithms. 78th EAGE Conference & Exhibition 2016. DOI 10.3997/2214-4609.201600878. Available in EarthDoc.