A long time ago I used to study geology in a nice city called Cluj, in the middle of that interesting part of Romania known as Transylvania. This was a place and time where and when I learned about quartz, feldspar, species of coral and foraminifera in great detail, heard about sequence stratigraphy and turbidites for the first time, and went on some great geological field trips. Not to mention the half-liter bottles of beer that would be significant components of any decent geological trip or spontaneous philosophical discussion in the evening. The less pleasant part was that many of the classes I took involved brute-force memorization of fossils, minerals, chronostratigraphic names, and formations. Although the geological vocabulary that I picked up was pretty broad and proved useful as a good set of words, terms, and definitions to play with, I forgot many of the details by now. If you asked me what was the difference between granite and granodiorite, I would have to check. I don’t remember at all what fossils are characteristic of the Late Jurassic. And, despite doing some fieldwork myself over there, I cannot remember the stratigraphic nomenclature in Transylvanian Basin; I would have to look it up (probably in this paper).
After college, it took me about one year to realize with convincing clarity that there was a lot left to learn. I went on to grad school on the other side of the planet, at a well-known university. Many of the classes I took over there were – unsurprisingly – quite different; a lot more focus on laws, processes and the connections between geological things than the ‘things’ themselves. It was also there that I started to see the links between geology and physics and math. I picked up quite a bit of math and physics during high school, but then quickly relegated them to the status of “stuff that is rarely used in geology”. At grad school, it dawned on me that numbers and mathematical laws are not only useful in geology, but are in fact necessary for doing good earth science. Maybe I am stating the obvious, but here it goes anyway: geology deals with enormous variations in scale, both in space and time; and it is not enough to say that the river was deep (how deep?), the tectonic deformation was fast (how fast?), the sea-level highstand lasted long (how long?), or the sediment gravity flows were high-energy flows (I am not even sure what that means). One of the most important things I learned was an appreciation for physical and quantitative insight in geology, that is, having at least an idea, a feel for what are the scales and rates involved in the formation of the rocks you are looking at. I cannot say it better than Chris Paola, one of the important and influential advocates of moving sedimentary geology closer to physics and math:
“For the ‘modal’ sedimentary-geology student, it is not sophisticated computational skills or training in advanced calculus that is lacking, but rather the routine application of basic quantitative reasoning. This means things like estimating scales and rates for key processes, knowing the magnitudes of basic physical properties, and being able to estimate the relative importance of various processes in a particular setting. Understanding scales, rates and relative magnitudes is to quantitative science what recognizing quartz and feldspar is to field geology. Neither requires years of sophisticated training, but both require repetition until they become habitual.”
Developing these skills is a lot easier if one is not afraid of tinkering with simple computer programs. Want to really understand what Stokes’ law is about? There is no better way than typing the equation into an Excel spreadsheet or a Matlab m-file and see how the plot of settling velocity against grain size looks like. What about settling in a fluid with different viscosity? Change the variable, and compare the result with the previous curve. High-level programming languages like Matlab or Python* are a lot easier to learn than languages closer to ‘computerese’ and farther from English, and they are great tools for these kinds of exercises and experiments. As somebody interested in stratigraphic architecture, I have become especially fond of creating surfaces that vaguely resemble real-world landscapes and then see how the evolution of these surfaces through time – deposition over here, erosion over there – creates stratigraphy. Complex three dimensional geometry is a lot easier to grasp if you can visualize and dissect it on the computer screen.
Of course, numbers, diagrams and images that come from computer programs are only useful if they demonstrably say something about the real world. Data collection in the field and the laboratory are equally important. But nowadays we often have more data than we wished for, and quantitative skills come handy for visualizing and analyzing large datasets – and comparing them to models.
Not everyone is excited about the growing number of earth scientists who tend to see equations ‘in the rocks’. The logo of the Sedimentology Research Group at the University of Minnesota features the Exner equation carved into a pebble, allegedly as a response to the exclamation “I haven’t seen yet an equation written on the rocks!” There is some concern that many geology graduates nowadays do not get to see, to map and to sample enough real rocks and sediments in the field. Although I think this unease is not entirely unsubstantiated, I wouldn’t want to sound as pessimistic as Emiliano Mutti – one of the founding fathers of deepwater sedimentology – does in the last phrase of a review article:
“This approach raises a problem, and not a small one: in connection with data collection in the field, how many field geologists are being produced in these times of increasingly computerized geology; and how good are they?”
As far as I know, geological field work is still an important part of the curriculum in many departments of geology – as it clearly should be. The number one reason I have become a geologist was that I loved mountains, hiking, and being outdoors in general, way before I started formally studying geology. And I still take every opportunity to go to the field. But I cannot see the growth of “computerized geology” – and of quantitative geology in general – as a bad thing. Does dry quantification take away the beauty and poetry of geology? I don’t think so. Unweaving the rainbow, unfolding a mountain, and reconstructing a turbidity current only add to our appreciation of the scale and grandeur of geology.
* I will let you know later whether this is true about Python…
** I have started writing this post for Accretionary Wedge #38, mostly because I found the call for posts quite inspiring, but haven’t finished it in time. Read all the good stuff at Highly Allochthonous.
Totally agree with the sentiment.I've found that Python is excellent for science. Slightly cleaner syntax than Matlab, slightly fiddlier for typical science usage, but vastly more suitable once you are going beyond a simple plot or script. For example: I've found it easy to dispatch high-computation jobs to Amazon's cloud computing service (via PiCloud), which I believe is possible on Matlab too, but you have to fill out a form and probably spend $1000 to be told how it might work. It helps that Python is free, of course.
Thanks for the comment and the additional arguments for learning Python. The more I hear and read about it, the stronger the temptation.
Well I had the classical academic geology training, including and followed by quite a bit of fieldwork, for mineral exploration purposes.Then I became some sort of "computer geologist", in the days before PC's existed, and developed somehow into a change agent to develop and bring relevant software into mining/exploration companies. I know I had some colleagues (bosses, actually) who thought that any work in the office, without a hammer in hand, was not real geology work.Later I got into evaluation (geostatistics, etc.) and was once called a "numerate geologist" as if this was some sort of rare subspecies, or worse, an oxymoron.These days, in exploration, you would be dead without numerical and statistical understanding, and computer skills.The exploration coin needs both sides: field work, and numbers. Without both sides, there is no coin.