A preview of things to come…

My favorite part of doing geology is the travel. Fieldwork is a different sort of travel than the more vacation-y trips I’ve taken (which I also enjoy very much). But odds are, you’d never find the destinations of most field geologists in a travel guide. You sleep on the ground or in cheap motels. You go to the local grocery store, the local pub, and get to know the people who live and work nearby. I’ve been extremely fortunate to have seen so much of this beautiful and diverse country through fieldwork in college and grad school, and with each new place I go, I gain a greater appreciation and deeper understanding of the processes that have shaped it. So far though, my field experience has been limited to the US. But that’s about to change.

This summer, I was fortunate enough to be awarded a Graduate Research Opportunities Worldwide (GROW) grant from NSF to spend 8 months in Bergen, Norway between January and August 2020. There, I will work with researchers in the geography department of the University of Bergen to study canyons carved by glacial lake outburst floods. I am incredibly excited about this opportunity to explore new scientific questions and to broaden my personal and professional horizons. But wait, it gets even better!

My friend, hydrogeology badass, and fellow UMass PhD student Sarah McKnight, applied for and was awarded a prestigious Boren fellowship, which will send her to Israel for the same 8 months. There, she will be working with the Israeli Geological Survey to map and model flows of the highly saline groundwater in the arid country. Sarah and I started as MS/PhD students at the same time, and she’s become one of my closest friends as we’ve celebrated the high points and kept each other going during the low points of our graduate experiences. Once we found out we would both be doing international research at the same time we started scheming about ways to stay in touch while we were away and, inspired the Letters to a Prescientist outreach program, came up with a plan to correspond with each other and practice our scientific communication skills at the same time.

Screen Shot 2020-01-01 at 3.01.14 PM

Me n’ Sarah enjoying one of our classic “Fancylunches”

Sarah and I will be writing letters to each other during our time abroad, and posting them on our blogs for all to read. We’re shooting for a total of 8 letters, 4 apiece, spaced more-or-less evenly over the 8 months we’ll be away. We want to share the messy, fun, sometimes-frustrating process of geologic research from our own perspectives as two women challenging ourselves to do science out of our comfort zones. And yeah, there will be lots of desert and fjord photos. So keep checking back–I leave this weekend, so the first letter will be posted soon! You can also follow us on twitter–I’m at @KarinInACanyon and Sarah’s at @sarahvmcknight. Now I’ve got to get back to packing.

In the Sandboxing Ring

Geology is slow, and geologists are impatient. We deal with processes that occur over huge areas, at barely-detectable rates, or far below the earth’s surface. Nevertheless, these processes have a huge impact on our lives–think about the gradual creep of the earth’s tectonic plates resulting every so often in an earthquake, sometimes a deadly one. Since we can’t speed up geology, pause it, or rewind, we model the earth’s behavior in a number of different ways so that we can control for different things and see things you normally couldn’t. For instance, you could build a sandbox that mimics two plates colliding by pushing the material inside it together, with transparent sides to show the layers which would normally be obscured underground, and the whole thing takes 20 minutes instead of 20 million years. Or you could write a computer program to incorporate all the physics we know about how rocks break and use it to predict ground motion during an earthquake–much safer than the real thing. Both of these examples, by the way, are ongoing projects at UMass by Dr. Michele Cooke and her awesome research group (https://www.youtube.com/watch?v=h6v-TvzwtWM). Models are great for visualizing geologic change, predicting future events, and testing how well we actually understand geologic processes, but making a good model that actually answers the questions you’re asking is very complicated. So I was very excited to spend ten days in Minneapolis at the Summer Institute on Earth-Surface Dynamics (SIESD) workshop hosted by the University of Minnesota last summer, learning all about what makes a good model, and what they’re good for.

The SIESD class of 2018 enjoying the sunshine outside the St. Anthony Falls hydraulic lab (SAFL). I had so much fun getting to know everyone in this group of smart, diverse students!

Setting up for a sandbox model of stream erosion at one of the many facilities for physical modeling at SAFL!

At the end of the workshop, we had to put together projects applying some topics we learned to a new problem. I decided to use this opportunity to tackle a question that I’ve been trying to wrap my head around since geomorphology class: how do you measure soil production rates using exposure dating?

Soil production? What’s that? And exposure dating sounds a little raunchy for a geology blog. I’ll get to all that shortly, but really, I was trying to understand why this diagram works. My labmate and I have spent many hours at the bar trying to make sense of it, napkin flowcharts and all, but there are so many components to this system that after a while we just give up and order another round. But after an intense ten days of learning the ins and outs of modeling earth systems, I decided what I needed was a way to visualize and play with this process, piece by piece, to understand how it all fits together.

The more I look at this diagram from Granger & Riebe, 2014, the more I realize how complex soil really is!

Here’s what’s going on behind that diagram. Cosmic rays from supernovae produce slight changes in chemistry at the earth’s surface by occasionally knocking a helium atom out of an oxygen atom, to create a new atom of beryllium called 10Be with one more neutron than the typical nine. This has been going on since nearly the beginning of the universe at a very steady rate (https://www.space.com/32644-cosmic-rays.html). The oxygen in the top of the atmosphere is what actually gets impacted with cosmic rays, but that starts a chain reaction down to the lower atmosphere and into the top layer of rock exposed at the surface, with exponentially fewer of these reactions happening the deeper you go (oxygen is one of the main components of many minerals). The longer a rock sits around exposed at the surface, the more 10Be accumulates compared to the 9Be the rock originally formed with. So measuring the ratio of two different isotopes of beryllium is a sneaky way to figure out how long a certain piece of the Earth has been sitting on the surface. But as I hope you’ve gathered if you’re a fan of the blog, rock rarely stays put. Bedrock gets weathered into smaller pieces by different physical, chemical, and biological processes, and then mixed around with organic matter to produce soil. At the same time, geologic processes can erode rock off the surface, or deposit new sediment on top of it.

10Be from collisions with cosmic rays and atmospheric oxygen causes chain reactions in the lower atmosphere and top few meters of the earth at a predictable rate, so we can use the concentration of 10Be to figure out how old rocks and dirt are! Diagram from http://1.bp.blogspot.com/-sIQM5lwHq08/U7O9ehrqG9I/ AAAAAAABKaQ/6ZZv3hf9fnQ/ s1600/meteor_Be.jpg

I incorporated all of these processes into a Matlab model, where I could make each process go as quickly or slowly as I wanted, to see how it changed the concentration of cosmically-altered atoms with depth. I started with a solid column of bedrock, made of a grid of cells, that was old enough to have an ideal profile of decreasing 10Be with depth. Then I started “weathering” the bedrock; a cell at the weathering front (initially the surface of the bedrock) could either stay put, swap places with the one below it, or swap places with any cell above it and then shift the whole column down, all with equal probability. This algorithm, called a fast-path random walk, was inspired by a lecture from Dr. Vaughan Voller at the University of Minnesota, who uses this method to model the growth of deltas.


A figure from my model showing hypothetical 10Be concentrations after 5,000 years, assuming weathering and soil mixing but no erosion.

My model is far from accurate. It massively simplifies many complex processes, doesn’t take into account horizontal mixing, completely ignores isotope decay, and takes many other shortcuts. But that’s ok, because it allowed me to watch quantifiable natural processes interact in a way that I’d never be able to do in nature. These things happen over very long timescales, their actual mechanisms are not necessarily continuous, and if I’m being honest all dirt kind of looks the same to me. Not to mention that to actually date a single sample costs hundreds of dollars and many hours of field and lab work! Modeling the interaction of geologic processes helped me better understand the complexity of cosmogenic exposure dating, and gave me a way to fiddle with nature’s controls.

My ten days in the world of numerical modeling left me completely exhausted, but wonderfully excited about new questions and new ways to answer them. As the statistician George Box and many scientists since him have insightfully noted, “all models are wrong, but some are useful”.

The Great Wall(s) of New England: Part II

Note: if you haven’t read Part I of this blog post, you can find it here

What started as a simple observation about the ubiquity of stone walls in New England had turned into a neat research question. How fast have hillslopes eroded since European colonists started modifying the landscape for agricultural purposes, and can we use the stone walls that they built around the same time to estimate this post-settlement erosion rate? The elevation data my lab partner and I measured by a cemetery wall in Western Massachusetts revealed a change in the hillslope profile just uphill of the wall in Leverett, which I guessed might be from material eroding off the hillslope getting blocked at that point

After solid rock weathers into smaller sediment and soil, it moves downhill (it’s pretty hard to move sediment uphill!) by diffusive processes. These are processes that are slope-dependent, so they tend to smooth out points and fill in pits. Lots of transport mechanisms fall under this category, such as raindrops throwing grains around when they hit the ground, trees falling over and pulling up dirt, and critters digging burrows and holes. While each of these processes involves the movement of just a few particles, over a wide area and over a long time they strongly influence the shape of the landscape. It’s impossibly difficult to try and track down the motion of each little sediment particle, but thankfully these processes have similar physics and are fairly easy to model in aggregate. All these processes are primarily slope-dependent, the net movement of material is downwards, and at points further downhill you have to transport all the sediment at that point as well as everything coming from uphill. In fact, these processes are what gives hills their characteristic “rolling” shape; since diffusive transport is slope-dependent, slopes steepen the further you are from the top to handle the increased load further downhill. 

But back to walls! I had found an interesting feature, so I wanted to try and unravel the processes responsible for making it. I wrote a short Matlab script to see if I could get a similar-shaped pile starting with a smooth, wall-free hillslope. To get the two hill shapes I needed, I grabbed elevation data from the high-resolution LiDAR dataset along two parallel lines from the same cemetery site–one line crossed the wall, but the other, just a few meters away, didn’t.


I used line B for the topography across the wall, and line C for the topography next to the wall, and subtracted one from the other to see how the wall changed the topography uphill of the cemetery wall.


The model I wrote was to find the diffusion coefficient, a measure of how efficiently the hillslope erodes. My approach was to reconstruct the profile predicted by the diffusion equation for different values of the diffusion coefficient D, and see which one resulted in a shape closest to the pile uphill from the wall after 238 years.


dz/dt is the rate of change in elevation, or erosion/aggradation at a point, per unit time. d2z/dx2 is the second derivative of elevation along the profile, or the curvature. Essentially, this equation says that the more extreme curvature you have somewhere, the more it will erode (or fill in, if it’s a hole instead of a peak). So this equation describes landscapes going from bumpy to smooth.

I had everything else I needed from my measurements; the headstone survey gave me the change in time, the elevation came from the LiDAR data, and the curvature was easy to calculate from the profile. So I hit go and watched the hillslope undergo 238 years of erosion in a few seconds!


I used a line fit to the profile uphill of the wall to find the hillslope shape along line B without a wall.

I had programmed the model to stop once the area of the modeled wedge barely exceeded the area of the real one, and it stopped at a diffusivity of 0.0038 m3/m/yr (1). Which is a pretty reasonable value for till-covered fields, falling somewhere between coarse soil (0.004 – 0.006 m3/m/yr) (2) and sandy gravel (0.002 m3/m/yr) (3).


I evolved the topography along line C according to the rules of diffusion for 238 years using increasing values of D. The different shades of brown on the hill show the final profile at the end of the model. The model stops as soon as the area of the material eroded from the top of the hill on the left side (which equals the area of material accumulated at the bottom of the hill) exceeds the wedge area of 1.002 m2.

Now the real fun could begin. I had a diffusivity value calibrated from the Leverett hill, which I could use to predict erosion in any other place given just a topographic map of the area. First, I tried this on a map of the same cemetery as before, for the same length of time–238 years. I modeled it using a python toolbox called LandLab developed to model all sorts of processes in geomoprhology.

After 238 years, it looked like my predictions matched my observations pretty well! A depositional wedge had indeed formed uphill of the wall with about the same width as the Matlab model, and erosion was happening downhill of the wall. But since the starting shape of this area already included a wall and the resulting sediment backlog from 238 years of hillslope erosion, I needed to find a new, wall-free spot to test my hypothesis.


Elevation change in the Leverett cemetery after 238 years of erosion at a diffusivity value of 0.0038 m3/m/yr. Red shows areas that were lowered by soil erosion; blue shows areas that were raised by deposition. The result is a depositional wedge uphill of the wall about the same size as the one we measured in the field!

I picked a hillslope with a similar shape and size across the street from the cemetery hill for this last experiment, and “built a wall” at about the same height on the hillslope as the first wall, perpendicular to the curvature of the hill. It wasn’t a physical wall, but a virtual one; all I had to do was raise the elevation along the line I wanted by a little bit. Then I threw that back into LandLab to undergo 238 years of erosion.

…and found the same topographic change patterns! A nice depositional wedge had formed uphill of the wall, again with about the same size as the calibration wall. There also happened to be a real wall oriented in the direction of hillslope curvature, which didn’t develop a depositional wedge at all, indicating that the deposition is coming from material uphill that gets “caught” by the wall. I gave the entire landscape the same diffusivity value, including the wall, so instead of an erosional wedge downhill from the wall there’s another depositional wedge. Since I was mostly interested in the material coming from uphill of the wall that was fine, but if I had more time it would be neat to vary the erosion rate to include non-eroding manmade structures.


The wall I “built” across the street from the cemetery shows similar sediment accumulation on the uphill side. The fake wall is modeled to have the same properties as the surrounding hillslope, so there’s a hotspot of erosion along the top of the wall itself. Just goes to show that rocks are better than dirt for making permanent structures!

So what do these patterns tell us about how we’ve been using land and modifying our landscape since Europeans arrived? The hilltop in Leverett eroded at a rate of 21 cm/yr, which is awfully close to, but slightly higher than, the offshore sediment deposition rate of 0.18 cm/yr over the last 1000 years (4). I would need to measure a lot more sediment buildup behind a lot more walls, but from this one example it looks like farming on the hillslopes has been making it easier for sediment to move downhill, until it eventually ends up offshore or gets stuck on floodplains, in ponds, or behind dams, sometimes for centuries (5).

It’s a challenge to study something that’s not there anymore, like the sediment that has been removed from hillslopes by erosion. But sometimes there are enough clues left in the landscape to piece together processes that happened a long time ago. Part of the fun of being a geologist is finding those clues and coming up with new ways to figure out what they mean!

  1. This funny unit is the volume of soil (m3) moving past a unit width of the hillslope (m) in a given length of time (yr)
  2. Fernandes and Dietrich, 1997
  3. Nash, 1984
  4. Hubeny et al, 2009
  5. Ly 1980; Syvitski et al., 2005; Slagel and Griggs 2008; Meade and Moody, 2010; Gupta et al. 2012; Woodruff et al., 2013, Yellen et al., 2017; Kasprak et al., 2013

The Great Wall(s) of New England: Part I

If you’ve ever been to New England, this is a familiar sight:

Related image

A stone wall in the woods of Holliston, MA, a suburb of Boston. Image from http://mapio.net/pic/p-67347133/.

I find these small, unassuming walls lining roads and backyards all over Western Mass. Occasionally on a hike I’ll even pass one in the middle of the woods, miles away from any house, and think “what’s that doing there??” It turns out that these cute, scenic walls are part of a massive land-use change that was brought about by European colonists as they began converting forests to farmland in short order.

Image result for new england stone wall

Surface geology of New England–everything was either eroded by glaciers, or deposited by them! Image from https://www.earthmagazine.org/article/history-science-and-poetry-new-englands-stone-walls.

Much of the New England surface is covered in glacial till, sediment composed of a wide range of sizes from clay and sand grains to big gnarly boulders, deposited during the last ice age as glaciers retreated northward. The boulders in this mix get pushed to the surface by water expanding and contracting in the soil, a process called frost heave. When the colonists set their plows to the dirt, they found not the peaty, loamy soils of England, but giant stones left and right. Nonetheless, farming was what they were familiar with, so rather than adapting their practices, they adapted the landscape, and diligently set about digging up the pesky rocks. But what was to be done with all these rocks coming out of the soil? Some were used for making cellars, and others were just left in giant piles, but most ended up taking the form of stone walls. Walls were great for marking property boundaries, enclosing livestock, and showing off your status with that extra touch of landscaping.

All this rock removal, along with the clear-cutting involved in creating farm- and pastureland and the tilling of the farms themselves, had an immediate effect on the land. Sediment core chemistry off the coasts suggest that erosion rates increased after the arrival of European colonists, even though dams and mills being constructed at the time kept a lot of the newly-mobilized sediment from making it all the way out to sea. Check out the cool research my friend Justin Shawler (https://justinshawler.wordpress.com/research/) is doing to figure out anthropogenic influences to sediment delivery offshore!

Watching the walls go by on the bus ride to work one day, and thinking about the colonists who put them there, I started to wonder if those walls could be useful for more than just pretty fencing. If they’re all from about the same time period, and were placed on the sides of newly-deforested hills, could we use them to figure out how quickly hillslopes have eroded since the agricultural era of New England, and see if terrestrial erosion rates increased similarly to the offshore sediment deposition due to the land-use practices of settlers?

First, I had to find some walls. I pass the ones along the road every day, but I wanted walls that had been relatively unmodified since their construction, and were further away from places with a lot of infrastructure. Lucky for me, Massachusetts has LiDAR imagery covering the entire state. I could write a whole separate blog post about LiDAR (and maybe I will!) and how useful it is for geologists, but it’s a method of remote sensing that can see past trees and buildings, revealing only the solid surfaces beneath. And stone walls show up clear as day in it.


Strip away the trees in this forested part of Leverett, and stone walls are clearly visible as straight, intersecting ridges! Read more about how to Katherine Johnson and Will Ouimet at UConn are finding stone walls in LiDAR imagery here: https://geomorphology.uconn.edu/research-projects/lidar-and-southern-new-englands-historical-landscape/

After mapping out the stone walls in Leverett, MA, a nearby town in Western Mass where I had previously seen many walls, I selected one to visit and test my hypothesis. 


Looking downhill into the cemetery in Leverett from the uphill side of the wall

I figured that if a wall was constructed on the surface of a hillslope at about the same elevation everywhere and left alone for a while, material would accumulate on the uphill side of the wall, and erode on the downhill side. The wall I picked was uphill from the road a ways and oriented perpendicular to the slope of the hill, and even better, was enclosing a cemetery. Cemeteries have dates, which are hard to come by in geology–a survey of the headstones in this one showed the oldest date to be from 1780, making it 238 years old


Surveying the hillslope by reading the measurement on the meterstick (held by my lab partner Evan) through the autolevel on the tripod

We took an autolevel survey tripod out to the wall, and measured the surface height in a few places across the wall. The autolevel is a fun device; one person holds a measuring pole at different locations along a line, and the other person reads the height of a stationary scope. It’s a neat way to map out low-resolution surface elevations. Sure enough, the sediment gets backed up on the uphill side to form a wedge that’s flush with the top of the wall, and there’s a missing wedge on the downhill side (although this area is part of the cemetery, so perhaps that has more to do with the missing sediment than the normal hillslope processes do).


Three surveys across the wall (grey box) seem to show some sediment deposition on the uphill side and erosion on the downhill side! Can we use that volume to figure out how fast hillslopes are eroding since European colonization?

Our field observations were promising, so I headed back to the lab to see if I could replicate this with some numerical models. Check back for those results in the next post!

(If you want more cool facts about stone walls, check out this great book by Robert Thorson: https://robertthorson.clas.uconn.edu/writing/books/stone-by-stone/)

The Geomobile

Field work involves a lot of driving. There’s no avoiding it; many field sites are in remote areas, and geologists tend to have a lot of gear and scientific equipment in tow. As anyone who’s taken a road trip knows, the car you drive makes a huge difference in terms of your comfort, stress, and general enjoyment of the experience. So as we drive into the new year, here are a few of the field vehicles that have been part of my geology adventures…for better or for worse!

  1. The Swaggin’ Wagon

Taken together, I think I’ve spent several weeks in the white 12-passenger vans we used for department field trips at William & Mary. So many enthusiastic geowallies in one place makes even the longest drives interesting and fun. You can expect intense car games like contact, cows and graveyards, story building, and the ever-popular alphabet game, punctuated with enlightening discussions of the local geology. The long windows are great for checking out landscapes and towns as you roll to your destination. I would always try to claim the window seat behind the driver, so that I could catch all the conversations going on, or take a break and watch the rocks go by. 


A view from the cramped but cheery Swaggin’ Wagon following a dip in Mono Lake, CA. Photo by Ryan Flesch.

While a pain to maneuver, these are great for pulling over beside interesting rock outcrops, and can even be used as map holders in a pinch! The “Swaggin’ Wagon” nickname for our oldest department van was conceived on a 3-day sedimentary geo camping trip through western Virginia. As our department grew to more students than we had van seats we rented extra wheels for our larger trips, which got stuck with the unfortunate name of “Space Turd” due to its sleekly pretentious and futuristic exterior.

blog cars7

Chuck Bailey, Structural Geology professor at William & Mary, ingeniously uses the magnetic side wall of the “Space Turd” to orient us in the Virginia Piedmont on a Southeast GSA field trip 


  1. blog cars2

    The Truck, parked in our campsite and ready for a full day of roaming the Washington Scablands!

    The truck

Getting a pickup truck for our first trip to the scablands of eastern Washington was definitely a mistake. We couldn’t leave our expensive computers, drills, and other equipment unlocked, so everything–camping gear, food, equipment, and bags of rock samples–had to go in the backseat. To make things worse, the music system was entirely controlled by an app neither of us could download, since we were out of range of wifi or any cell networks. We were stuck with the radio–which would have been great if we could have found a radio station that played anything other than The Eagles…



  1. The Spaceship
blog cars3

The Spaceship waits patiently among the wildflowers that grow in the Scablands basalt as we collect samples. Photo by Madison Douglas.

Determined not to repeat mistakes of the past, we rented a sizable SUV when we traveled back to the scablands the following year. It had plenty of space, and handled the potholed dirt roads beautifully. But this car was real advanced. The rental office handed me a key with a dozen different buttons on it, and I must have tried all of them before finding the one that opened the doors! Everything in the car was controlled by a button too, which made driving the car feel a bit like flying a spaceship, and once I figured out what all of them were for, it was actually kind of fun. Unfortunately I didn’t remember this when I went to sleep in my tent with my car key in my pocket and woke up to find that I had opened the trunk during the night. Sure enough, the squirrels had gotten in during the night and shredded our granola bars all over the backseat.


And then there are the cars you see while you’re in the field. There was a car with a giant toothbrush on its roof in California, the rust bucket we found in the middle of a volcanic crater in Texas, and the object in eastern Washington that I had marked as a boulder on the satellite imagery, but which turned out to be the back half of a truck. You never know what you’ll see on the road–or off it!


blog cars4

See the giant red toothbrush strapped to the roof of the car on the left? Wild things happen in the Alabama Hills, CA!

blog cars5

When you’re hoping to find olivine xenoliths in a volcanic crater (Kilbourne Hole, TX), but find a rusty car instead…







So what’s the ideal field car? It depends on where you’re driving, how long you’ll be in the car, how many people are with you, and how much gear you’re hauling. But here are my basic requirements: lots of space, sturdy tires, large windows, and not too complicated to use or drive. And it has to be easy to clean, inside and out! Because even the best field car will inevitably end up looking like this after long hours on a dusty road:

blog cars6

Geo-grafitti on the Texas field camp van! The learning never stops…except at red lights.

Geology by day…astronomy by night

Did you catch the Great American Eclipse? Millions of americans hit the road to catch this spectacular event last month, braving heavy traffic and long drives to witness a total solar eclipse. Astronomy, much like geology, draws people to places that aren’t your typical vacation destination–Carbondale, IL attracted a whopping 50,000 people on August 17, which is about double the city’s population of 25,500! [1]

fed eclipse

The path of totality in the August 2017 solar eclipse could be seen from many of our spectacular national parks and natural forests, not to mention plenty of state and local parklands…which also often happen to be ideal places to study geology! Image from http://mediad.publicbroadcasting.net/

An eclipse whose path spans from coast to coast (like the August 2017 one did) covers mostly rural areas, since cities comprise just 3.5% of the United States’ land area [2]. But the incentive to venture into remote areas doesn’t end there–for the most impressive stargazing you need clear skies free of light pollution, and that means getting as far away from cities as possible. There’s a lot of overlap between the regions ideal for astronomy viewing and geologic field work, and one of my favorite parts of any field trip is looking up at the night sky after the sun goes down.

Some places are better suited to amateur astronomy than others. The International Dark Sky Association has published a handy tool at http://darksitefinder.com/maps/world.html to find the amount of light pollution in any part of the world. The data comes partially from permanent monitoring stations, but much of the recent data has been collected through citizen science projects; you can find out how you can contribute data about the night sky near you at https://www.globeatnight.org/ The light pollution map reports an average level of absence of manmade lighting, but other factors like cloud cover impact the quality of the night sky for viewing too.

dark sky

On the Dark Site Finder map, major population centers in China, India, Europe, and the United States light the map up like a Christmas tree, and coasts all around the world highlight the continents with their shipping, industrial, and residential illumination.

According to the Dark Site Finder tool, the field site I’ve been to with the darkest skies is Big Bend, Texas, narrowly beating out Devils Tower, WY and Craters of the Moon, ID, and boasting one of the few places left in the United States with completely dark skies. After a long day in the field we would sit outside our tents and pool our collective astronomy knowledge as we pointed out the constellations we knew, and sometimes caught glimpses of meteorites streaking across the sky. As someone who grew up near a large city I was blown away by how many stars I could see, and I recall the sense of awe I felt realizing that this was the view of the cosmos the ancient civilizations around the world would have seen as they grouped these stars into constellations and incorporated them into their stories and cultures.


A stunning view of the Milky Way through the star-spangled sky over Big Bend National Park, TX. Photo from http://topworldresort.com/nature/big-bend-national-park/.

Humans have been watching the night skies for thousands of years. Telescopes, with all their accessories, can be a great way to enhance the stargazing experience–especially when you point them at the moon and planets–but when trying to pack light for doing geologic field work I usually have to leave mine at home. I don’t really mind though, because there’s so much you can see with just the naked eye. There are lots of apps that use a phone’s GPS to help you identify night sky objects, or you can go old school and print out a star wheel for the date and location you’re heading for (just be sure to put a red filter over your flashlight so you don’t mess up your night vision). If you’re savvy with a camera, take the opportunity to experiment with long-exposure photography. Or just find a spot on the ground, grab a pillow, and look up!

[1] City-data.gov

[2] Census.gov

Field foliage–you can’t escape it!

It’s summer, and here in Massachusetts that means plants. Look up on a hike and you’ll be hard-pressed to find any patches of blue between the thick tree canopy. Every street seems to have a farm stand overflowing with produce. And I’ve heard plenty of jokes about needing to lock houses and cars, not to keep out unwanted people, but to avoid vegetation taking up root in them. It’s always interesting to learn about the native vegetation wherever I go, and often necessary to take the flora into account when preparing for a day in the field.

Vegetation and geology are more closely related than you might think. On Mars, the lack of a concealing canopy of trees makes it easy to measure geologic surface features, but the absence of plants leads to significant differences in the shapes of features like rivers, which on Earth rely on plants to hold their banks together. Our photosynthesizing plants have even changed the chemistry of Earth’s atmosphere, which before the development of plants consisted mostly of carbon dioxide, sulfur, and methane. Vegetation is an important factor in the stability of beaches and marshes, hillslope erosion, and water chemistry, just to name a few things. But nowhere is this connection more obvious than when doing fieldwork, from chopping our way through dense thickets in Virginia with the Buckmarlson Banshees to pulling cactus spines out of my shoes, legs, and backpack in Texas. And on my last field trip to Washington, we were able to spot patches of basalt bedrock from far away by looking for the weedy red grass that prefers to grow in that rock. All of this begs the question…how do you measure something like vegetation?

The answer is…depends what you’re trying to find out. On smaller scales, one might be interested in the plant density to calculate water infiltration rates, root depth as a proxy for soil competence, or biodiversity to indicate ecosystem health. If your study area is a fairly uniform environment, like a single meadow or watershed, you can do what this Oregon State University research group is doing and use a quadrat.


OSU students using a quadrat to measure biomass

This simple square frames out a representative area, allowing researchers to obtain the biomass, species distribution, percent coverage, or just about anything else you’d want to know about this small patch, and then extrapolate it to the size of the area it represents. Perfect! But what if I wanted to compare these factors between some of my field sites, which are scattered all over the country?

To compare vegetation in distinct regions, or to find an average over a large area, we have to turn to remote sensing. NOAA’s Advanced Very High Resolution Radiometer (AVHRR) is one of several space-based instruments currently measuring vegetation quality and quantity around the globe. This instrument measures the wavelengths of sunlight that bounces off the earth, and is especially sensitive to the wavelengths reflected by healthy plants (0.4 to 0.7 microns). Its measurements are taken 1 km apart, and each point on earth gets re-measured every 1-2 weeks, providing a wealth of information about ever-changing conditions such as drought extent, fire damage, crop yields, and even illegal logging.

Each one of these measurements returns a spectrum, or a distribution of the intensity of reflected light over a range of wavelengths. To use these numbers in a meaningful way, remote sensing experts have created indices that relate the reflected light waves to qualities of vegetation. Healthy, lush plants absorb all visible light waves except green (which gets reflective, and is why plant leaves look nice and green to our eyes), while unhealthy plants and bare ground reflect more of the visible light colors and look brown. Happy plants also reflect a lot of the near-infrared light that comes in from the sun, while unhappy plants absorb a larger fraction of these wavelengths.


Plants reflect light differently based on whether or not they are actively photosynthesizing. Image from NASA’s Earth Observatory

The NDVI, or Normalized Difference Vegetation Index, uses both of these tendencies to calculate a quantitative vegetation score for any plant area. Here’s the formula:


A plant with no green leaves will score about 0 on the NDVI. Other factors, such as evapotranspiration rates, incoming sunlight, and human population density certainly influence (and/or are influenced by) the degree of vegetation in a region, and these global data sets allow researchers to correlate these factors over wide areas.


new ndvi

NDVI of North America from AVHRR composite imagery. Image from NASA’s Earth Observatory

Looking at this map of the nationwide NDVI, it’s obvious that the east coast is the place to be a plant. We have just enough sunlight, and plenty of water to go around. The variability in precipitation, temperature, soil type, and lots of other factors between my field sites is reflected in a wide range of NDVI values. Here are just a couple; how do you think your hometown compares with its NDVI value?

crater2Craters of the Moon, ID: NDVI = 0.2

Most of the green stuff growing on these basalt lava flows are lichens, which break down rock into soil and allow larger plants to eventually colonize

wilson2Carrizo Plain, CA: NDVI = 0.3

Not much grows in this dry region of south-central California for most of the year, but some bushes can tough it out in gullies, where their roots are closer to groundwater

yos2Yosemite, CA: NDVI = 0.3

A decent annual precipitation can support a beautiful evergreen community, but the vegetation cover is limited by the soil availability on these steep, glacially-eroded surfaces

gc2Grand Coulee, WA: NDVI = 0.2

Plenty of sun here, but low precipitation makes for patchy shrubs and grass only. The water held in the Banks Lake reservoir off to the right is used to irrigate the rich glacial soil, allowing farming in otherwise unusable land

mult2Multnomah Falls, OR: NDVI = 0.6

About 3 hours west of Grand Coulee the landscape changes dramatically; water precipitates on the coastal side of the Cascades, resulting in thriving plant communities (and gorgeous waterfalls!)

(photo by Mike Larsen)

howrd2Howardsville, VA: NDVI = 0.8

The thick vegetation cover in the Virginia Piedmont is quite effective at weathering the old basement rock…it can be quite a challenge sometimes to see the rock through the plants!


Devils Tower, WY: NDVI = 0.4

The moderate vegetation cover includes species from both coasts; you can find both western ponderosa pines and eastern oak trees at our first National Monument

News, updates, and of course…a map!

If you’ve just stumbled upon my collection of field geology stories, welcome! If you’ve been checking the blog for a while now wondering when I was going to write a new post (hey mom)…welcome back! It’s been ages since my last post (well, the blink of an eye in geologic time), but I’ve been sitting on a lot of good stuff that I can’t wait to write about. In an effort to post more frequently I’m going to try out a new posting format. The old format read more like a diary because I keep an informal field journal in addition to my official research notes, and I would simply write those up day by day when I returned from a gig. Made sense with my original intention for the blog; whenever I’d return from fieldwork I would be asked by many friends and family members how it went, and I’d have to repeat all my tales from the trip, inevitably forgetting things each time. Now however, people with widely ranging geologic backgrounds are interested in hearing about geologic fieldwork, research and general shenanigans, which is terrific! With that in mind, I’m going to be changing the format of the posts from a journal dump to a more topical style, with each post focused around a particular subject. You won’t get to hear what I had for breakfast every morning (although perhaps I’ll do a separate post on field food…there’s an idea…), but you will get a punchy and hopefully cohesive scoop out of many aspects of the fieldwork, labwork, and other components of geologic research that have become part of my experience as a geology student. It will also be nice to draw from multiple field sites in one post and do some geographic comparisons around a single subject…but I’m getting ahead of myself.

So that’s the update on the blog, but a quick update on myself is also called for. Last time I posted I was a rising senior at the College of William & Mary. Following all those adventures in Texas (which you can read about in gruesome detail in the previous eleven posts), I completed an internship at NASA Goddard Space Flight Center, which included some sweet fieldwork and got nicely wrapped up into an honors thesis. After graduating from William & Mary I spent a summer working as a Geoscientist-In-The-Park at Devils Tower National Monument, WY, a position sponsored by the National Parks Service, the Geological Society of America, and Americorps. Now, I am at the University of Massachusetts Amherst busily churning up a Masters thesis. All of these projects have taken me to many different places over the past two-and-a-half years, and since this is a field blog I’ll put all those places on a handy dandy map.


Geographic distribution of Karin’s field trip and fieldwork sites as of June 2017. Map from https://antarctic.files.wordpress.com/2014/10/7.gif

There you have it! Stay tuned for lots more posts in the future, and if you have requests feel free to leave me a comment. I’m already writing my next post, and I promise there will be rocks in it. And also, things that are not rocks! Hype!!