500km from Sudbury

Impact craters are ubiquitous features on solid bodies in the solar system, but on Earth they’re rather annoyingly quickly eroded. The impact crater at Sudbury, Ontario is 1.85 Byr old with an initial diameter of between 150 and 260 kilometers. Sudbury remains an oblong scar on the Canadian shield, but little remains (or has been discovered/identified) of the ejecta. A recent paper by Cannon, et al. (2010) identified several localities of ejecta deposition in Michigan.

I had the opportunity in the Spring and Fall of 2011 to participate in a field trip to the Marquette area of Michigan, where there was supposed to be 40m of impact breccia in what the authors designate the McClure locality.

Google Earth mapWell I would walk 500 kilometers…

Estimates of crater ejecta distribution rely on a limited dataset of “small” nuclear explosions, a couple well-preserved terrestrial craters (Ries, Barringer Meteorite, Lonar, etc), and observations of lunar craters. Ejecta thickness is seen to decay following a power law where thickness t = 0.14 * (R^0.74) * [(r/R)^-3], where R = distance from crater center to site, and r = primary or transient crater radius. Using appropriate values here (R = 500km; r = 100km) gives a calculated ejecta thickness of ~30m at the Marquette area, a surprisingly spot on result.

But don’t get too bogged down by wishes and estimates and dry writing. The Sudbury impact didn’t deposit 40 meters of ejecta at this sole locality. Everywhere, in all directions at over 500 kilometers away would have been inundated with enough material to bury a ten story building!

Our first trip out in Spring 2011 had temporary success. We searched the area for a good hour or so, and our colleagues from Northwestern University eventually drove a bit further down the road and found a small outcrop. Hooray! Unfortunately, it was in an open pit of someone’s ongoing basement construction. Despite some fast talking by professors, the group was swiftly kicked out. A bit disheartened and rather hungry, we gave up for the trip and decided to look at a map before heading out on our next trip. At GSA we talked to THE ejecta guy, the one who found the outcrop and has published several papers on samples collected therefrom. Turns out we turned left where we should have gone right…

In October we returned and had almost immediate success!

Lots of big cherty bits in there! Click to view larger in Picasa album from the trip.

This was the first convincing piece of ejecta I found, and afterward I was too excited hunting around to take more photos. We returned the next day and found a nice – but highly weathered – continuous cropping out of ~4 meters of ejecta (still no photos), but we couldn’t find the upper or lower contacts or clear-cut glassy lapilli (supposedly concentrated near the lower contact). Close to the crater, high volumes of ejected material essentially buries the surrounding terrain. Distal (far away) ejecta has less volume but higher energy, and so it interacts with the surface more during deposition and incorporates more local terrain. I believe starting at ~2.5 crater radii from the impact, local material starts to make up the majority of the “impact” layer. Although the exact transition distance is fuzzy, at ~5 crater radii at Marquette we would certainly expect extensive incorporation of locally derived material. The impact layer at the McClure locality is supposed to lie directly over a banded iron formation, though there’s also a chert layer down there. I almost forgot to sample the site and had to grab a quick couple of samples on the way out:

Ejecta boulder, ~25x15x8 cm.  Centimeter scale.

Cut face. Larger clasts appear to be predominantly locally-derived chert

To access to the locality we sampled, turn right off of County Road 510 and park on the side of the road leading to the old bridge (no longer in service). The field photo was taken in the woods to the east of the road, where there are several large boulders and some really nice, smaller outcrops. The main outcrop (and where my sample was taken) is located near the arrow in the map below. I am not sure if the land is privately owned, so be courteous if you visit. In looking closer at the Cannon paper, it appears we weren’t at the exact McClure area, which they place at 46°33′N, 87°33′W (GMap link). Maybe we’ll try to check that out next fall. This was our sampled area:


W. F. Cannon, K. J. Schulz, J. Wright Horton Jr., and D. A. Kring (2010) The Sudbury impact layer in the Paleoproterozoic iron ranges of northern Michigan, USA. GSA Bulletin, v. 122, pp. 50-75; doi:10.1130/B26517.1

F. Hörz, R. Ostertag, and D. A. Rainey (1983) Bunte breccia of the Ries: Continuous deposits of large impact craters. Reviews of Geophysics and Space Physics, v. 21, pp. 1667-1725.

T. R. McGetchin, M. Settle, and J. W. Head (1973) Radial thickness variation in impact crater ejecta: Implications for lunar basin deposits. Earth and Planetary Science Letters, v. 20, pp. 226-236.

Warren Dunes, briefly

In July my friend visited Notre Dame to run in the Sunburst (Half-)Marathon, and while he was around we took a trip to Warren Dunes State Park. Dunes occur over a broad area in the eastern shore of Lake Michigan (Sleeping Bear, Silver Lake, Cowles Bog, etc.), with Warren Dunes being one of the largest and most popular southernmost parks.

Peak ridge of a sand dune

This was my second trip to a sand dune field (my first was nearby Cowles Bog), and I’m consistently surprised by the variety of environments in such a small area. Some near-shore areas are almost all sand:

People scrambling up, also note group of three on horizon at right

But a few hundred meters down the shore, heavily vegetated sand dunes appear to be securely anchored in place:

Of course, looks can be deceiving and a closer inspection revealed exposed roots in some areas – a result of dune migration, or just short-term erosion?

Other than a few birds, trace fossils, and a heck of a lot of mosquitoes in the boggy inland area, there wasn’t much visible life on the dunes. That is, until we saw this guy (he was hard to miss)

The mosquitoes were so bad that we took the first opportunity we came across to get out of the bog and back into dune territory. Unfortunately, it turned out to be a steep face of a sand dune. We had to take a few breaks on the way to the top…

Here I used photo-documentation as an excuse to pause.

It was so steep that in some areas, even though he was so far away his footsteps would cause sand near me to collapse. My best estimate from maps of the area are that it was ~70 ft to the top, with this photo being taken ~halfway up. We did a quick 4 mile hike around the park to see some highlights, but I’m looking forward to a return trip next summer to learn a bit more about the area and spend some time on the beach.


What did you do today? Yesterday? The past three years?

Tracing crystals. It’s one of the boulders I’ve been pushing uphill since the beginning of my graduate research. Why? What is it all for? Three little words: crystal size distributions. CSDs are a method to quantify rock textures from thin sections. Let’s look at one of my samples, Apollo 17 basalt 75015,52:

You can readily describe the overall texture of a thin section, and estimate phase proportions and morphology, etc. The coarse-grained nature and large crystals are obviously the result of slow cooling. Ilmenite (FeTiO3) makes up the majority of the opaque phase and varies in morphology from generally euhedral, lath-like crystals to skeletal and almost amoeboidal in some areas. CSDs are a way to (attempt to) address questions such as:
– Whether these phases all one population of crystals (i.e. a single crystallization sequence with larger crystals being the earliest formed) or a mixed population (perhaps large skeletal ilmenite crystals formed prior to eruption or elsewhere during flow and were later mixed with smaller euhedral ilmenite crystals) or were they affected by some other process;
– How fast the sample cooled, and whether it was constant or variable
– Whether this basalt is related to other basalts from the same area.

Crystal size distributions are an objective method of assessing crystal size, shape, abundance, and distribution. It’s basically a three step process. Step 1 is to make a mosaic of the thin section (as above). Step 2 involves tracing individual crystals (we use a tablet), which takes anywhere from a couple of days to a couple of weeks. I trace different phases on separate layers in Photoshop, with the final result looking like this:

Gray is the sample area, blue is pyroxene, yellow is plagioclase, and black is ilmenite.  For this sample I traced ~1500 individual crystals.

The rest is easy. We have a slew of programs to measure length, width, area, and position of the crystals (NIH ImageJ), estimate crystal habit short/intermediate/long axis dimensions, and population density (CSDCorrections).

See the website of M.D. Higgins for more information on quantitative analysis by CSDs.