Drilling: what, why, and how

As mentioned, I spent a significant part of last weekend was spent on a paleomagnetic sampling project with collaborators from the University of Michigan. On Friday, this was our field area:


That’s the south slopes of Old Rag Mountain, a popular Blue Ridge hiking destination because unlike many Virginia peaks, when you get to the top, you see some rocks instead of 100% trees:


But we didn’t come here for the view. We came here for the dikes. Here’s the edge of one, with a pen for scale:


These are dikes of basalt and meta-basalt of the Catoctin Formation which are presumed to be feeder dikes pumping mafic lava to the surface of Virginia around 570 million years ago, during the breakup of the supercontinent Rodinia and the opening of the Iapetan Ocean basin. The dikes cut across the Grenville-aged basement rocks, in this case the Old Rag Granite of about 1000 million years age. The Old Rag area is especially great because the dikes are less metamorphosed than they are in other parts of the Blue Ridge province, where the Catoctin has been cooked into greenstone. Here’s an annotated view of the previous photograph:


As far as this project goes, we are interested in these dikes for the information that they (potentially) contain about the orientation of the Earth’s magnetic field in Virginia at the time of the supercontinent Rodinia’s breakup. By sampling these dikes and then analyzing the samples at their paleomagnetism lab back in Ann Arbor, Fatim and Matt hope to put some constraints on the question of paleo-Virginia’s latitude when these dikes cooled into solid rock.

As a reminder, you are not allowed to sample any rocks in any national park unless you have first applied for and been granted a research sampling permit by the National Park Service.

Close to the planet’s surface, the Earth’s magnetic field is shaped like a torus (or, in less technical terms, a doughnut, but one of those donuts with a pinched up midsection, and more of a dimple than a hole). It exits at the south magnetic pole, wraps north around the Earth, and plunges back into the inner core at the north magnetic pole:


A magnetically-sensitive mineral forming in a modern rock would have an upward-oriented high-angle magnetism if it formed at high southerly latitudes, a moderate-angled upward orientation at moderate southerly latitudes, a horizontal, northward-pointing orientation at the magnetic equator, and then the reverse as you head towards the north pole: a moderate-angled downward orientation at moderate northerly latitudes, and a downward-oriented high-angle orientation if it formed at high northerly latitudes, just like the red arrows show in the above image.

Of course, the flow of the magnetic field occasionally reverses direction (emerging at the north magnetic pole instead, and flowing south), but the shape of the field doesn’t change:


So the angle of inclination of a fossil magnet should be the same regardless of whether it’s poking up or plunging down, relative to the surface of the Earth. In this way, paleomagnetism can reveal the approximate latitude (but not longitude) at which a rock formed.

But wait, is it really so simple? No, of course not. Check out the map below, showing the positions of the north geomagnetic pole over the past 2000 years, with numbers showing the position of the pole in a specific year CE. It moves! The circles around geomagnetic poles at 900, 1300, and 1700 CE are 95% confidence limits on those geomagnetic poles; the mean geomagnetic pole position over the past 2000 yr is shown by the square with stippled region of 95% confidence. These data were compiled by Merrill and McElhinny (1983) and plotted by Butler (1982).


So this map shows us that even though the magnetic pole does wander about a bit, 2000 years of data is enough to generate an average which is more or less coincident with the geographic pole. And therefore a statistically significant batch of data (spread over a 2000-year-or-greater spread of time) will also reflect that average pole position.

Meert, Van der Voo, and Payne (1994) made a first attempt at constraining the paleomagnetics of the Catoctin Formation. Four of their 32 sites were feeder dikes, sills, and host rock (Grenvillian basement complex). One of the things these authors did was that they performed a “contact test” on two of their dikes. A contact test is a way of using an igneous contact (as with a dike) to determine whether the whole region has been magnetically reset, perhaps by thermal activity accompanying contact metamorphism. Consider this situation:


You sample a dike and its surrounding host rock, at several distances away from the dike. You find that they all give you the same magnetic orientation. This suggests you have the magnetic signature of a later overprinting, not the original orientations of dike and host rock.

Now what if you found this, instead?


Here, your dike shows a distinct signature that is different from the host rock, and the host rock shows a uniform orientation except right next to the dike, where the heat of the intrusion has partially reset the (older) host rock’s magnetism. If I were to annotate this up (with color coding!), it would look something like this:


Passing the contact test is critical to tying the two rocks’ magnetic data to their age data. It’s only with a positive contact test that you can use this data to say anything about where Virginia (and thus ancestral North America, often dubbed “Laurentia”) was when the Catoctin dikes were intruded.

The contact test is something that our team wanted to repeat, with more dikes than just the two that were featured in the Meert, et al. (1994) paper. We also wanted to double-check their results, and verify, reject, or modify them as our data warranted.

The key to constraining the magnetic orientation of these rocks as precisely as possible is to constrain the current orientation of the samples as precisely as possible. We measured the strike and dip of the surface of each sample very carefully, before we extracted it from the bedrock. At Old Rag Mountain, we were not allowed to drill (Old Rag is a wilderness area with no motorized equipment allowed), so we were collecting oriented hand samples.

Here’s Fatim Hankard writing orientation data in her field notebook while Matt Domeier takes a strike and dip reading in the background, using his Brunton compass:


Because these rocks are inherently magnetic (that’s why we’re sampling them, after all!), we have to control for the possibility that the rocks themselves might be throwing off our Brunton compass needles. A second compass is employed to control for any magnetic field coming off the rocks themselves. This is a solar compass. If you know exactly where you are (note Fatim’s GPS unit in the above photo), and when you are taking the measurement, you can use this solar compass to double-check the orientation you get from the Brunton compass.

Here’s Matt’s solar compass butted up against one of our Old Rag samples. Note the shadow being cast by the compass’s nomen, and also note the “arrow with a prong” strike and dip symbol that we wrote directly onto the face of the sample with a Sharpie:


Next, take a look at a photo of a sample once extracted. We label it redundantly, not only in terms of the orientation lines, but also in terms of the sample’s identity. That way, we’re less like to find a bunch of scratched-up but un-identified and un-orientable rock samples once the van gets back to Michigan:


While poking around, I found this interesting feature at the edge of one of the dikes. I’m hoping one of my more petrologically-inclined readers may be able to offer me some kind of interpretation of this pattern:


What I noticed is that in the first few mm of the dike, right up against the contact with the host rock, there are no white lathes of plagioclase feldspar. These relatively large feldspar crystals are phenocrysts, big chunky crystals that grow in the magma when it’s cooling relatively slowly underground, but then entrained in the flow as it moves upwards into the dikes, whereupon the surrounding liquid chills rapidly to make fine-grained basalt. So there are no phenocrysts right at the edge of the dike, then there are a bunch, all aligned with one another (but with no preferred sense of imbrication, so far as I can tell), and then there are more phenocrysts in the bulk of the dike, but they are (a) less concentrated, and (b) lack any preferred orientation. Let me annotate it for you, then go back and take another look at the unannotated version, so you can see what I’m referring to:


Okay, petrologists, I want to hear from you: How should I interpret this?

Back to the paleomag… On Saturday, we went to another location to sample. This one was much more convenient because (a) it was right on the side of the road, and (b) it wasn’t a wilderness area, so drilling was allowed. This was at the lovely selection of Catoctin dikes downhill (north) from the Little Devils Staircase overlook, on Skyline Drive in Shenandoah National Park. Here’s a charismatic dike with Matt acting as a sense of scale:




We unpacked the gasoline-powered diamond-grit-tipped drills and hooked them up to the water pump. We put on ear- and eye-protection, and got to work:


One the sample has been drilled out, you’re left with an empty hole. The white liquid is the cooling water with suspended dust from the abraded rock. This hole is about 3 cm in diameter:


The core (2.5 cm diameter) that came out of that hole:


In our field area, a core this size of the dike rock takes about ten minutes to extract. Basement rock (host rock) takes longer, as it’s made of harder minerals.

One worry is that the core will snap loose while you are drilling it out. If this happens, it may start rotating in the hole, and you will lose all sense of how it was originally oriented, which means you’ve just wasted a lot of time for no gain in data. To protect against this possibility, we used a technique of scoring a second circle with the drill bit, partially overlapping our actual core like a Venn diagram:


This way, if the core snaps off, you can line up its arc with the rest of the circle inscribed on the outcrop next to the hole. Whew! Core saved!

Fatim extracting another core:


After the core is drilled out (but still in its hole), Fatim oriented it. Notice the new array here – it’s a stand with slots into the drill-hole, then has a Brunton compass atop it with a solar compass atop that:


As you can see with this example, the solar compass is just about to become useless as the afternoon shadows advance! Next up, record all the orientation information (trend and plunge of the cylinder’s axis), and then score the core with a line:


Fatim and Matt sampled for two more days after I had to leave them due to other obligations, like teaching. They are headed back to Michigan today. Soon, hopefully, we’ll see whether our sampling campaign yields any meaningful results… Stay tuned!

As a final note, I would like to point out that this collaboration was born when Fatim read my blog post on feeder dikes and then proposed that we combine her paleomag skillz with my dike-location knowledge. It’s not the first time that my blogging has yielded a great opportunity, but it seems to be a shining example of how virtual connections online can lead to tangible work in the real world. The blog-curious should take note.


References cited:

R.F. Butler. PALEOMAGNETISM: Magnetic Domains to Geologic Terranes. Originally published by Blackwell in 1984, 248 pp. Updated online 2004. Retrieved September 15, 2010, from http://www.pmc.ucsc.edu/~njarboe/pmagresource/ButlerPaleomagnetismBook.pdf.

J. G. Meert, R. Van der Voo, and T.W. Payne. “Paleomagnetism of the Catoctin volcanic province: A new Vendian-Cambrian apparent polar wander path for North America,” March 10, 1994. Journal of Geophysical Research 99, No. B3, pp. 4625-4641.

R. T. Merrill and M. W. McElhinny, The Earth’s Magnetic Field, Academic Press, London, 401 pp., 1983.

Falls of the James II: fractures

In my previous post, I introduced you to the Petersburg Granite, as it is exposed south of Belle Isle, at the falls of the James River in Richmond, Virginia. I mentioned that it was fractured, and I’d like to take a closer look at those fractures today.

The geologically-imparted fractures were exploited by human granite quarriers, and in some parts of the river bed, you can see the holes they drilled to break out big slabs of the rock. Some of these block-defining perforations exploited pre-existing fractures.


This is also evident on the north side of Belle Isle itself, where there are several large abandoned quarries now mainly utilized as a rockclimbing locale. There are two dominant fracture sets in the area: one which parallels the schlieren (magmatic fabric), striking NNE; and a second which strikes ENE.

The meaning of these fractures are one of the problems Chuck Bailey (my host at Belle Isle) and his students have been considering. Under Chuck’s tutelage, James McCulla examined these fractures and reported his findings at the NE/SE GSA section meeting in Baltimore last March.

One of the first things Chuck showed me when we got to Belle Isle is some offset schlieren, like these:




Let’s annotate those up, so you can orient yourself:




So clearly, that looks like a right-lateral offset, right? Of course, it could just be an apparent right-lateral offset, as perhaps the inclined schlieren have been offset in a vertical sense, then exposed by erosion on the same horizontal section. We need to determine the true offset direction. If we look at a vertical exposure of the fracture surface itself, will slickensides back that up? Here’s one…


Yep, the slicks are very gently plunging (close to horizontal) and agree with the right-lateral offset we thought we saw in the horizontal exposures in the earlier photos. These are in fact true right-lateral offsets. Chuck is currently dating some muscovite that appears on these surfaces as a method of constraining the timing of deformation.

The other fracture set (NNE-trending, parallel to the schlieren) shows very little in the way of telling fracture-surface anatomy. There may be some weakly-developed steps facing to the upper left, but these surfaces are neither gouged nor mineralized:


Chuck and James therefore interpret the NNE-trending fractures as extensional fractures and the ENE-trending fractures as faults with small offsets. It is worth noting that the NNE-trending extensional joint set is parallel to extensional faults in the Richmond Basin, a Triassic rift valley 15 km upstream.

So which came first? Here’s a confounding exposure:


Allow me to lighten that up and annotate it for you:


We have two different relationships exposed here, less than a foot apart. At left, we see the NNE-trending joints truncating against the ENE-trending “fault.” At right, we see that the NNE-trending fracture steps to the right as it crosses the ENE-trending fracture. The left example suggests that the ENE “fault” is older, and the NNE joint came later, propagating to the pre-existing discontinuity but no further. The right example suggests that the NNE-trending joint was there first, but was then broken and offset (ever so slightly) in a right-lateral fashion, like the offset schlieren in the photos earlier in this post. In other words, the ENE “fault” is younger.

“Geology isn’t rocket science.” We know what’s going on with rockets — we built those suckers! This, on the other hand, is a bit more complicated!

Anyhow, Chuck and James have been over these rocks like gravy on rice, and they have documented many other instances of cross-cutting relationships. As James’ GSA abstract notes, they found enough exposures to feel confident interpreting the ENE-oriented set to be the older set to have formed as a result of WNW-directed contraction during the Alleghanian Orogeny. The NNE-oriented extensional fractures are the younger set, and are interpreted to have formed during Mesozoic extension accompanying the breakup of Pangea.

Next up, we should take a quick look at the James River itself, and the imprint it has left on this stupendous field site… Stay tuned…

Falls of the James I: pluton emplacement

Last Friday, NOVA colleague Victor Zabielski and I traveled down to Richmond, Virginia, to meet up with Chuck Bailey of the College of William & Mary, and do a little field work on the rocks exposed by the James River.

Our destination was Belle Isle, a whaleback-shaped island where granite has been quarried for dimension stone for many years. The island has also served as a Confederate prison for captured Union soldiers during the U.S. Civil War, and later for various industries. Today, it is preserved as park land, utilized by a wide swath of Richmond’s populace for recreational activities, both licit and non.

Fortunately, a large area of the James’ river bed south of Belle Isle is kept relatively dry by a long low diversion dam upstream. As a result, there are some mighty fine horizontal outcrops of rock:


The dam fed water into a hydroelectric power generation station, but that station has been abandoned for some time now:


The power plant dam has yielded enough exposure that some bedrock mapping is possible for those with the curiosity and fortitude to attempt it. Here’s a simplified geologic map of the area, authored by Chuck and his student James McCulla:


So you can see that most of the area is covered by sedimentary deposits of both modern and early Cenozoic vintage. Our goal, however, was the more interesting stuff beneath that. (All due respect to my sedimentological colleagues; the Coastal Plain just doesn’t get my juices flowing like ‘crystalline’ rocks do!)

So here’s what we came to see, the Petersburg Granite:


This is an Alleghanian pluton, ~320 Ma, and quite large: it extends for tens of kilometers north and south (Petersburg, the namesake locality, is to the south). It disappears beneath the Coastal Plain to the east, and beneath the Richmond Basin (a Triassic rift valley) to the west.

You can see from the photo above that in some places the Petersburg Granite is massive and equigranular, and in other places it’s “foliated,” with long dark lines running through it. These lines are schlieren, curtainlike zones of differing mineral ratios: more mafics than felsics, for instance. The schlieren (German for “lines”) are usually interpreted as magmatic flow structures as higher-temperature-crystallizing mafic crystals raft together in a more felsic flow. At Belle Isle, the schlieren are steeply dipping and trend NNE.

In places, there were also pegmatite bodies that were concordant (~parallel) with this overall magmatic fabric. Here’s an example of that texture:


And here’s a really big crystal of K-feldspar set amid finer-grained granitic groundmass. I guess you could call this a “megacryst”:


Another thing we saw a lot of were dark-colored inclusions in the granite. These were dark due to lots and lots of biotite mica in them. Here’s an example; notice how the schlieren wrap around it:


And another, with its long axis oriented parallel to the strike of the schlieren, suggesting alignment in the magma chamber before the granite set up:


How should we interpret these mafic inclusions? Are they xenoliths; fragments of country rock that were broken off and included in the intruding granitic magma? Or do they represent a plutonic emplacement process — perhaps an earlier stage of crystallization, or an immiscible bolus of mafic magma floating like a lava lamp blob in the surrounding felsic melt? When they’re fine grained and lacking internal structures, as with the above examples, it’s really hard to make that call.

On the other hand, this one clearly shows fragmentation along the right edge, suggesting to me that it was a coherent xenolith at the time the enveloping granite set up into solid rock:

That rules out the fluid-blob-within-another-fluid hypothesis, but is it country rock?

This one suggests that it is indeed country rock, as it is both foliated and kinked internally:

Here’s a heart-shaped inclusion which also suggests that it is a genuine xenolith. As with the previous example, it displays internal foliation that has been folded:


Victor ponders these xenoliths, as well as a dense clot of biotite (dark steak next to the yellow field notebook – not Chuck’s shadow, but parallel to it and closer to the photographer’s vantage point):


The photo above also shows how the schlieren wrap around these xenoliths. Here’s an example where the schlieren “tails” leave the xenolith “higher up” on the left side than the right side, suggesting a sinistral (counterclockwise) sense of magma-flow kinematics:


This one is a beauty. It’s almost perfectly circular in cross-section, though with little flanges coming off the upper left and lower right. However, the “tails” are both on the same side of the xenolith, so I don’t really feel like I’ve got a good bead on its kinematics:


A few more shots of these xenoliths:



This one is a cool one…


… because when you zoom in on the edge, you can see it has some ptygmatic folding inside it. Like the foliation and the broader folding we observed earlier, this internal structure suggests that these are genuine xenoliths; fragments of pre-deformed country rock.


Another xenolith, also showing this internal deformation of ptygmatically-folded granite dikes:


…And this one shows internal boudinage:


Chuck examines a small vertical surface to get a sense of what these xenoliths are doing in the third dimension:


This next bit was a real treat for me. It’s no secret that I’m a huge fan of boudinage, that brittle-ductile phenomenon that separates a more competent rock type into sausage-like chunks while a less competent rock type flows into the void between those chunks. Here’s some schlieren that evidently became thick enough slabs of biotite that they were able to behave as semi-coherent sheets, subject to boudinage:


…Not only that, but if you back out and follow these boudinaged schlieren along strike, you can see that they are folded, too! Check out these sweet isoclinally folded, boudinaged schlieren:


Biotite-rich inclusions which I interpret as similar “scraps of schlieren” which became entrained in later magmatic flows:



While everything I’ve talked about so far has been concordant with the dominant schlieren orientation (and thus reflective of main-stage magmatic flow in the Petersburg Granite), there are also some discordant features, like dikes, which cut across the regional fabric.

Here, for example, is an aplite dike:


Aplite is very felsic and displays a “sugary” fine-grained texture. This aplite dike is quite a nice feature, traceable over a long distance across the outcrop. We followed it a ways to a spot that Chuck was particularly eager to show us: a spot where the aplite dike crosses an earlier pegmatite dike, and then both dikes are cut by a right-lateral fault and a fracture set which parallels the schlieren. Check it out in outcrop (note the positive relief on the aplite dike):


And here’s a sketch of this outcrop (above photograph from the perspective of the lower right corner):


What a fine spot to bring students and have them suss out the order of events! First came the massive granite, then the pegmatite dike, then the aplite dike, then sometime later under very different P/T conditions, the rock was fractured and we get fractures: some of which show an apparent right-lateral offset (faults; oriented ENE), and others where no offset is apparent (joints). This second set appears to be utilizing the schlieren as zones of weakness, as it is parallel to the schlieren (NNE) and often occurs along their biotite-rich traces.

Whether the faulting or the jointing came first is a question we’ll examine in the next episode

Easter egg

Searching through my photo archives this morning for something suitably “Eastery”… something in pastel colors, perhaps? … a petrified lagomorph? … how about an egg, or something egg-shaped?

This is as close as I got:


This is in the Owens Valley of eastern California, showing a boulder of the Mesozoic Sierra Nevada Batholith bearing a faulted xenolith. I love outcrops like this, with a combination of primary structures (like the xenolith) and secondary structures (like the fault). And the fault surface appeared to have hosted some fluid flow, encouraging epidotization (hydrous metamorphism) along its surface. How appropriate, considering both the “cracked egg” implication of the round xenolith and the pastel tones of the green epidote.

I’ll annotate it up for you, because I know you love it when I do that:


Happy Easter, folks. Focus on the bunnies and candy, and not the zombies.

Folds of New York

Thursday is ‘fold day’ here at Mountain Beltway.

Let’s take a look at some folds I saw last weekend in New York City. We’ll start with a bunch seen in the Manhattan Schist in Central Park. Here’s an example of the foliation in the schist. It’s got finer-grained regions and coarser, schistier regions with big honking muscovite flakes. Metamorphic petrologists: Does this correspond to paleo-bedding? (i.e. quartz-rich regions that metamorphose less spectacularly, and mud-rich regions that converted more totally to muscovite during metamorphism?)

Anyhow, here’s what it looks like when it’s folded (accented with a small granite dike):


And another, with some boudinage thrown in for flavor:


This was one of the best outcrops I saw that weekend (on the edge of the ‘lake’), but it was inaccessible to closer photography. Sorry about all the branches in the image. What you’re looking at here is a series of folds with axes plunging at ~45° towards the lake:


Crudely annotated version:


Granite dike:


Boudinaged granite dike:


Folded and boudinaged granite dike #1:


Folded and boudinaged granite dike #2:


Lastly, here’s a couple of folds from inside the American Museum of Natural History. A metaconglomerate:

A little model mountain belt made out of compressed sand layers:


The thing that really struck me about this sand model is the folds visible in the green and yellow central part of the mountain belt: There are refolded folds there. The lower-central antiform with dark green atop yellow is the best example. I had the idea in my head that two generations of folds meant two generations of deformation, but here you’ve got two generations of folds resulting (presumably) from a single episode of ‘mountain building.’

Such beautiful complexity! I want a sand model like this for my lab.

Rusty weathering rind

On a granite block

Piedmont rocks exposed in a creek

One of the cool things about being the local geoblogger is that people get in touch with you about local geology. Sometimes this even leads to meeting up for field trips. Here’s two quick photos from a recent (January 2010) field trip to a creek near Springfield, Virginia.

My host was Barbara X, a local aficionada of Piedmont geology. She has lived in this particular neighborhood for many years, and is very familiar with the local woods and drainages through decades of dog-walking there.

Her main question for me was “Could the geologic map of this area be wrong?” She showed me the map, and then took me out to an outcrop which clearly was of a different rock type than the map indicated it “should” be.

The offending intruder, a meta-basalt with two prominent joint-sets:


A short distance downstream, a cut bank revealed some saprolitic rock that is more typical of the Piedmont province:


I think we’re seeing bodies of schist/ gneiss (highly foliated in cross-section), as well as coarse-grained, lighter-colored bodies of granite. All of them have been weathered to hell: you can scoop handfuls of this “rock” out of the outcrop if you want. If you’re a plant, you can plunge your apical meristem right into it, and let the roots follow.

This is typical “outcrop” around here: though the mid-Atlantic region has a fascinating story (including the Appalachian mountain belt, like these rocks), the wet climate has rotted most rock away. The only other thing that’s worth mentioning about this particular outcrop are the upper-left-to-lower-right brown lines: those are fracture traces decorated with rust. The fractures serve as plumbing to move fluids around in the subsurface, and their dissolved cargo of elements can then react with the rock on either side of the fracture.