Testing How Earth Works

Chelsea Harvey
Thursday, July 2, 2015

The Washington Post reports on experiments testing Taylor Perron's proposed explanation of evenly spaced ridges and valleys: "This is how the Earth’s hills and valleys were made — and what will shape them in the future."

Read this story in the Washington Post

Stand atop a ridge among Oregon’s Painted Hills, and you’ll find yourself looking out over a sea of rolling hills and valleys, all of them more or less the same size as one another. Viewed from above, you would see an organized pattern of crisscrossing, evenly spaced valleys carving up the landscape. Next, visit California’s Gabilan Mesa, and you’ll see a similar sight. The hills might be bigger than the ones in Oregon, the valleys longer: But within the region, they’re all roughly equally sized and spaced. It’s the same story all over the globe.

This is a feature of the planet that has both puzzled and fascinated geologists for years. Hills and valleys nearly always occur in organized, evenly spaced networks — but their scale, or how big the ridges and valleys are within each network, changes from one location to the next.

This pattern “suggests that, on the one hand, there is a fundamental similarity in the processes that are shaping different landscapes that create ridges and valleys of a characteristic size,” said Taylor Perron, an associate professor of geology at Massachusetts Institute of Technology. “But on the other hand, there must be something that’s different between all of these different landscapes that have such different scales.”

In recent years, scientists have started researching the natural processes that carve out hills and valleys in an attempt to figure out what factors influence their size and spacing. Understanding the origin of such landforms can give us insight into the kinds of natural events that might have affected the planet in the past and how they might shape Earth in the future.

A recent series of laboratory experiments, described Thursday in the journal Science, demonstrated that hill and valley formation is dictated by two main processes. The first is sediment transport, or “soil creep,” which is when sediment on an uneven landscape is disturbed by an event like a tree falling, an animal burrowing or water splashing on the ground and starts to tumble downhill. This tumbling sediment tends to fill in uneven places on the landscape, smoothing out the ground and covering up indentations where valleys might form. The second process is runoff, which is when a river or stream cuts away at the landscape, carving out valleys.

In the new paper, the researchers’ goal was to test out the idea that competition between these two processes determines the spacing of hills and valleys — a theory that Perron, the MIT geologist, helped propose in a 2009 paper in the journal Nature. The basic idea is that if soil creep is a stronger force in a region than flowing rivers, it will smooth out more of the landscape and result in bigger hills and fewer, more widely spaced valleys. On the other hand, if runoff, or stream flow, is more dominant, more valleys can be cut into the landscape. Although the theory was already out there, nobody had ever tested it in a lab before.

Testing the idea in an experimental setting has been difficult until now for several reasons. Landscapes form over long time periods, sometimes thousands of years, meaning that observing them change in real time is basically impossible. Additionally, landforms are so big that it’s not usually feasible to conduct experiments on them. But the scientists involved in the new study devised a way to get around these difficulties.

Led by Kristin Sweeney, a doctoral candidate in geological sciences at the University of Oregon, the researchers set up a series of experiments using miniature landscapes — small sandboxes, essentially — in order to simulate the formation of hills and valleys in real time. “Our idea was to speed up these processes and generate the processes themselves we thought were responsible for landscapes,” Sweeney said. That meant simulating both soil creep and runoff in the mini-landscapes.

The researchers generated soil creep using a “drip box,” which forcefully shot water onto the landscapes to disturb the sediment, and they generated runoff using a mister, which created little rivulets in the sandboxes. They conducted a series of five experiments, varying the amount of mister and drip box used in each one, to see what would happen to a landscape when it experienced more runoff vs. more soil creep.

Their findings showed that the competition theory holds up in a controlled lab setting: The more the drip box was used — in other words, the more soil creep a landscape experienced — the wider and more broadly spaced the valleys were. Less drip box and more mister caused runoff to become the dominant force, and the landscapes to became more highly carved with many close-set valleys.

“My overall reaction is that this was an ingenious set of experiments,” said Perron, who was not involved in the study. “Their ability to turn that dial and adjust the relative strength of those two competing mechanisms is what really makes this study unique.” Having experimental evidence of the way runoff and soil creep can affect a landscape, rather than simply looking at a landscape that already exists and making inferences about how it formed, is a kind of confirmation that current theories about the formation of hills and valleys are accurate.

The knowledge could be useful for predicting how Earth’s landscapes will be affected as its climate changes — over geologic timescales, that is. Sweeney stressed that landscapes are often shaped over hundreds or thousands of years, so the study can’t necessarily be used to make predictions about short-term climate changes, such as those that might occur in the next century as a result of human-caused global warming. “I wouldn’t feel comfortable making any sort of predictions about short-term climate change,” Sweeney said.

But Earth’s climate will also shift in the long term, maybe partly due to anthropogenic influences, but also as a result of natural planetary processes like glacial cycles. In these cases, she said, “as climate changes, precipitation should change, and that should change the way that these processes work.” For example, if precipitation becomes heavier over the next few hundred (or few thousand) years, landscapes will be pelted with more water. But an increase in precipitation may also lead to an increase in vegetation, which could then house more burrowing animals — both of which could lead to greater disturbances in a landscape’s sediment. Each of these climate-related factors has the potential to influence the formation of landscapes in the future.

And insights from the study can also be used in reverse, Perron added, as a way of looking back in time and making assumptions about what processes were happening on Earth thousands of years ago to create the landscapes we see today. “Records of past environmental change on the continents can be challenging to construct,” said Perron. “So any clues that we can extract from landscapes that can help us understand what’s been happening in terms of longer-term changes in climate and biological communities are really valuable.”

All practical applications aside, the study’s experimental evidence helps shed some light on one of geology’s most intriguing questions, a satisfying outcome for many a puzzled Earth scientist. “One of the most salient characteristics of landscapes is the size of the ridges and valleys,” Perron said. “Those are the most prominent topographic features that one recognizes when one looks at a landscape. And so understanding what controls that size is really a fundamental problem in the study of Earth’s surface.”

Returning to those Painted Hills we started with, or to the Gabilan Mesa, or to any of the world’s picturesque hilly landscapes, we now have a deeper understanding of the events that caused them to arise in the first place — and perhaps a better glimpse of the processes that may continue to shape them long after we’re gone.