Unsolved problems

One of the recurring dreams I've had this year is about unsolved problems. I've always loved these lists, the best known of which is perhaps the German David Hilbert's 1900 list of 23 unsolved problems in mathematics. There are several published versions of the list; take look at a later manuscript describing some of the problems.

A year or two ago, I read this meta-list in Wikipedia. Natch, I immediately wanted to create a list of unsolved problems in geoscience. It could help researches find big, interesting problems. It could help software developers focus their talents. It might just be a bit of fun. However, articles in Wikipedia need something to reference^{[citation needed]}, so even if I were capable of such a thing, one can't just sit down and hack one out. 

But you can try. Earlier this year, I drafted a proto-list for geophysics, drawn mostly from chats with friends. Please feel free to vote on the list, or add problems of your own. It is, I admit, a bit biased towards problems in seismology in pursuit of hydrocarbons. The list should be much broader, but I'm not yet the polymath I strive to be and quickly get out of my depth!

Here are the top five (per today) from my Google Moderator list of unsolved problems in geophysics:

• How can we represent and quantify error and uncertainty from acquisition, through processing and interpretation, to analysis?
• What useful signal or information can we extract from what we usually call 'noise' (multiples, refractions, reverberations, etc)?
• How can we exploit the full spectrum in acquisition, processing, interpretation, and analysis?
• Is there a 'best practice' for tying wells; if so, what is it?
• What exactly is AVO-friendly processing?

What might a list of unsolved problems in geology look like? My likely-ignorant outlook suggests some:

• Is it possible to predict the location, severity, and/or timing of earthquakes?
• Do mantle plumes exist?
• How do magnetic reversals happen?
• Are mass extinctions cyclic?
• Do the earth's physico-chemical systems mostly drive, or mostly respond to, changes in climate?
• Does eustatic (global, synchronous, uniform) sea-level change happen, or does the ubiquity of local tectonism obviate the concept?
• What exactly was the sequence of events that resulted in the end-Permian extinction? The end-Cretaceous?

I am proposing a workshop on the topic of unsolved problems in exploration and development geophysics at SEG next year in San Antonio. Ideas welcome.

Setting up shop

My family and I finally moved into our new old house on Sunday. We are slowly gaining the upper hand in the battle of the boxes. The kids have got some of their toys back and are finding their way around the place. I can find most things by looking in only about three cupboards. 

And my new office is almost finished. The last occupant of this place was a boat-builder, and he had a great workshop in the back yard (see pic). I would love to have kept it as-is, but a couple of days of trying to work in the house made it clear that I need a workspace outside the house. 

So I've partitioned off half the space, added a long landscape window, and a floor, and a built-in desk, and it should be comfortable enough to work in for a few months at least, maybe even years.

Even starting with some constraints, it was surprisingly hard to put my finger on exactly what would make my ideal workspace. Here are some of the things I decided I wanted, in no particular order:

• Some sort of view
• Natural light, ideally from a north-facing skylight
• A long desk with no cable clutter
• Plenty of space to tidy hard drives, printers, routers, etc.
• Plenty of space for books and papers
• Somewhere comfortable to sit and read
• Somewhere to work on non-computery stuff (like another desk)
• A wood-burning stove
• A kitchenette and a toilet

The skylight, stove, and sanitation will have to wait, but I think I got most of the rest checked. I'll post a follow-up when it's nearing completion.

Update on 2010-12-30 14:11 by Matt Hall

The office is finished, more or less, and I have moved in, more or less... still haven't figured out lighting and heating, but am working effectively. More or less.

Resumé 2.0

Click for PDF

Is your resumé boring?

When I was a confused postgraduate researcher (if there's any other kind), I applied for a job in marketing and product development. The company asked for a pitch, rather than a straight curriculum vitae. I made something up in Adobe's Illustrator drawing tool, with some product development ideas (shower gel you could shave with, IIRC), and got an interview. After that, I made slightly weird CVs for a few other jobs, including ones at the British Antarctic Survey and EnCana (see image, right). Nowadays I tend to think that if I can't get a job with a weird CV, then it's not the sort of job I want.

Recently I have started just giving people my LinkedIn profile, which contains most of the information you'd usually put on a resumé. 

I have also looked at VisualCV, an online resumé tool, but never actually taken the time to try it. I'm not convinced it lets you be as creative as you might want to be. 

If you need more inspiration, check out this list of efforts. Most of them are for what are sometimes called 'creative' jobs like graphic design, but I would argue that geology and geophysics are creative jobs too!

I don't know how much I really need a resumé any more, but I've enjoyed maintaining this geological timetable since 2005; click to download the PDF. [Click here for 2016 version, now with fewer typos!]

Update

on 2010-12-21 11:35 by Matt Hall

LinkedIn just added a resumé builder to their Labs... Click here to try it out. It seems to work quite nicely, though the output is quite conservative, and I've run into a couple of bugs. Worth a look.

Giftology and giftophysics

Geologists are not difficult to buy gifts for. In fact, you could do worse than just filling a shoe box with rocks from your garden. But if you want to, you can excite and inspire a geologist with some new kit, a nice map, or a piece of meteorite.

Geophysicists might be slightly trickier to please. A book on Fortran? A couple of ki's of dynamite? Best thing is to accidentally on purpose treat them like a geologist. After all, it's the thought that counts!

Compasses

Features to look for include clinometer, declination adjustment, and a sighting mirror. A bubble level and a scale bar are nice to have. Seasoned field geologists will already have a favourite, so steer clear unless you know what they need.

1. Good — Silva Expedition 15TDCL, about $60
2. Better — Suunto Tandem with declination adjustment, about $220
3. Best — Brunton GEO Pocket Transit, about $820

Hand lenses

Features to look for include German manufacturer, metal housing, glass lenses, triple lens configuration, no chromatic aberration (this property is sometimes called achromatic), no spherical aberration (aplanatic). The 'gold standard', as it were, is the Bausch & Lomb Hastings Triplet, which usually sells for about $40 to $50 (for example, here). But there are others out there, like these:

1. Good — BelOMO Triplet Loupe, about $35
2. Better — Celestron LED illuminated loupe, about $40
3. Best — Harald Schneider triplet loupe, about $280

Random stuff

You can't go wrong with any of these excellent gifts. 

• Hammer — Estwing pick with leather grip, about $40
• Holster — Estwing pick holster, about $10
• Landforms — Extremely cool models of landforms, about $340
• Moon rocks — Well, almost... terrestrial analogs to some lunar rock types, about $160
• Map — A geological map of the world, €10
• T-shirt — Brilliant T-shirt, $cheap but sold out so please vote for its return!

For the geologist who has everything

These gifts speak for themselves. Joy guaranteed.

1. Awesome — UGOBE PLEOrb robot dinosaur, about $470
2. Awesomer — Andy Paiko glass seismograph, about $5000
3. Awesomest — Triceratops horridus skull, about $70 000

Scale

One of the most persistent themes in geoscience is scale. Some properties of the earth are scale independent, or fractal; the shapes of rivers and coastlines, sediment grain shapes, and fracture size distributions might fall into this category. Other properties are scale dependent, such as statistical variance, seismic velocities (which are wavelength dependent), or stratigraphic stacking patterns.

Scale independent phenomena are common in nature, and in some human inventions. For example, Randall Munroe's brilliant comic today illustrating the solutions to tic-tac-toe (or noughts-and-crosses, as I'd call it). It's the optimal subset of the complete solution space, which shows its fractal nature completely.

The network of co-authorship relationships in SEG's journal Geophysics is also scale-free (most connected authors shown in red). From my 2010 paper in The Leading Edge.

3 ways to be Agile*

Building on last week's post, I think that not only the principles of agile development could apply to subsurface science, I think some of the tactics employed might also benefit us geoscientists. For example:

Ship and iterate: get maps, sections, velocity models, even geomodels, made early. Don’t wait until everything is perfect (it never will be). Making these things will help reveal the weaknesses in the data and the uncertainties interpretation, and you can be more strategic about what you spend time on. Do everything you can to make the iteration faster (use macros, write scripts, outsource).

Daily scrums: subsurface teams get together on a daily basis, for no more than 10 or 15 minutes. Everyone gives their two or three headlines, quick things are dealt with, other things are flagged for follow-up. And everyone can get on with their day... no more 1 hour meetings!

Pair interpretation: seismic interpreters sit together to interpret, with one picking the lines and looking at waveform character, the other taking a wide-angle view, looking for consistency, nearby well ties, or thinking about the geological setting. Slower, probably, but maybe better (in programming, this technique produces fewer bugs).

Signs and symbols

I made this sheet of signs and symbols when I was an explorationist at Statoil. I had always used Macs until then, and wrote my thesis in LaTeX; both make it very easy to write scientific signs and symbols. It seemed to be a struggle on the PCs we had at Statoil, so I made a cheatsheet of keyboard shortcuts for myself.

I sent it around to some people when I left ConocoPhillips and some said they liked it. So I thought I'd put it here for anyone to take. The idea is to print it out and put it on your wall. Next time you catch yourself writing 12 km2 or 25oC or not bothering with the pesky acute accent in resumé, just glance at the list. On Windows you would hold down the Alt key, then with NumLock on, type out the number indicated on the number pad (the numeric keys above the alphabet keys won't work).

What is Agile*?

The name Agile* is connected to a concept I want to promote in my work as a geoscientist: agile interpretation. What the heck is agile interpretation?

It's borrowed, or stolen if you prefer, from the agile software development philosophy. The idea is that you take an idea or strategy and, above all, stick to it. It drives everything that happens next. Then you do something, anything. This is anti-perfectionist and quite counter to many people's intuition (including mine). Then you iterate, as quickly as possible.

This might sound obvious, but it's not how most projects work. Most projects, of almost any kind, build a plan, preferably with a Gantt chart, then slowly but surely execute that plan, then... fail. Or if they don't fail, they at least are late and over-budget. This is not agile.

Unless you're building a nuclear reactor, I think it might be wiser to start small but bold, do good things quickly, fail often, and become adaptable and resilient. This is agile.