Hydrology, Cryosphere & Earth Surface News

Simplifying How (and When and Where) Snow Turns into Flow

A Montana researcher has developed a map for predicting how climate change may alter the water supply.

By Devi Lockwood

In Montana, the tiny streams that trace through mountains do more than feed the mighty Missouri and Mississippi Rivers. They could hold great power as a bellwether for climate change.

With scant summer rain, almost all of the precipitation that matters for watersheds comes from winter snowfall. And therein lies the key to its predictive power, explained Rebekah Levine, geomorphologist and assistant professor in the Environmental Sciences Department at the University of Montana Western in Dillon. “If we lose snow early, we also lose late-season base flow, so these streams will go dry. That’s what people are really worried about,” she said.

Dry streams in Montana mean low flows in some of the mightiest rivers in the country when summer and growing seasons for agriculture are in full swing. But what if there were a way to use snowfall patterns early in the year to anticipate dry conditions later?

Levine has been pondering these and other questions since 2012. And by simplifying her approach, she may have hit on some answers.

The Search for Quick and Simple

Levine’s research on this topic began when The Nature Conservancy approached her with a problem. Money for conservation is limited, so nonprofits like The Nature Conservancy want the most bang for their buck. For that, they often turn to hydrologic climate change models that pinpoint vulnerable areas.

But current hydrologic climate models, particularly in remote areas and at headwaters, involve a lot of complex inference and extrapolation. So relying on them to allocate resources involves a certain amount of risk, which donors want to avoid. Donors want to understand the future of water in the landscape, and to help interface with the public, they want answers to be quick and simple.

“So my idea was just exactly that. Can we make better predictions [of flow] based on things that we know really well? And how simple can we be?” Levine asked.

Levine likens her approach to Picasso’s The Bull, a series of 11 lithographs that start with a detailed, textured drawing of a bull and end with “the essence of the bull, which is really just a line drawing,” Levine said. “But all of us still know that it’s a bull.”

There’s a Map for That

To answer her questions, Levine worked with a team of researchers to create a map. She wanted the map to be based on simple, easy-to-measure characteristics of the mountains that can predict, to a significant level of accuracy, where snow will accumulate—and thus where streams will be fed by snow. Her goal was to pinpoint which parts of the watershed are most vulnerable to climate change, allowing conservationists to target their efforts toward species conservation and ways of increasing natural water storage.

Levine focused on the upper Missouri headwaters, encompassing the Continental Divide between the Missouri and Columbia drainage basins. The majority of the study area is within Beaverhead County, Mont., and contains portions of Gallatin and Madison Counties and Yellowstone National Park.

She chose three parameters—elevation, slope, and solar radiation—to map where snow is most likely to be. She isolated elevations above 2,400 meters, where precipitation typically falls only as snow, and slopes between 15° and 30°. On flatter slopes, there is more exposure to Sun and less shading; on steeper slopes, the snow is more likely to slough off in an avalanche.

Solar radiation is the amount of sunlight that hits any given patch of land, so Levine was searching for areas with low solar radiation. Mapping solar radiation might seem tricky, but for this, Levine again looked for simplicity. The Area Solar Radiation tool in ESRI ArcMap 10 already gives solar radiation values dependent on latitude and topography. The tool does take many hours of computing, but she saw no need to reinvent the wheel.

Then, using mapping software to categorize the accumulated solar radiation across the landscape, her team was able classify each watershed in the study area as having either high, intermediate, or low incoming sunlight during spring and summer. Then, looking at just the low category, her team summed areas that also had the appropriate elevation and slope for snow storage.

To put all this information on the map, Levine again sought simplicity. She chose a blue palette because users will readily associate it with cold. Then her team began to plot: The higher the sum of favorable areas within any given watershed was, the darker the blue they colored the watershed on the map. The resulting map is a patchwork quilt of blues, with darker colors indicating watersheds that are most favorable to later-season snow storage.

An area of Montana just northwest of Yellowstone National Park. Darker shades of blue indicate areas within watersheds that are likely snow refuges, where elevations, slopes, and solar radiation levels are favorable for the accumulation of snow. Credit: Rebekah Levine/TNC

Levine named these regions “snow refuges,” places where snow might accumulate and linger throughout the year. In other words, these areas would be less vulnerable to drying up by the end of summer.

Watery Proof

Were these mapped snow refuges indeed caches of snow? Streams within snow refuges, Levine’s thinking went, would keep flowing throughout the summer. So she assembled teams of interns and volunteers to collect streamflow measurements within these watersheds to ground truth her map.

Her teams selected 30 watersheds in Montana that represent variability across the landscape. For 5 years, they gathered late-summer discharge measurements from the watersheds’ streams.

The team found that the map was, to a significant degree, accurate at predicting which streams would retain late-season flow. As in any study, there were outliers that do not fit the trend. Levine is looking into what is different about these watersheds that makes the snow refuge area less important.

A Simple Challenge

Although there is still some variability that is not explained by the simple model, the map seems to be effective at giving a baseline to understand regional patterns of late-season flow that is based on immutable, measurable variables, Levine noted. Such simple baselines can help resource managers protect our water, she explained.

For example, the project’s Montana study area is the “ultimate headwaters of the Missouri and Mississippi system. Everybody lives downstream from us, right? Most of the nation,” Levine said. “If conservation decisions can be wisely made in the upper headwaters of the Missouri-Mississippi watershed, then those benefits may make a difference downstream,” she added. And downstream extends far away—all the way to the Gulf of Mexico.

Western Montana is also home to a unique ecosystem of elk, grizzly bear, and native fish. The Nature Conservancy is using Levine’s map to understand which watershed areas are most at risk and which of those have the most critical wildlife populations to protect.

“Rebekah’s work is really powerful because it starts with information that the stakeholders have already been observing,” said Sarah Marshall, a hydrogeologist for the federal government in Australia who saw a poster of the research at AGU’s Fall Meeting 2018 in Washington, D.C. Levine’s map uses this information “to constrain predictions about future climate and future water availability in a quantitative way that can then inform future water management strategies,” Marshall said.

And to best inform resource managers, scientists need to start with something simple, Levine noted. “I would challenge scientists to [ask], How much complexity do you need to explain these natural systems? Only use as much complexity as necessary to explain the world,” she said. “It’s always more complicated than what we can measure.”

—Devi Lockwood (@devi_lockwood), Graduate Program in Science Writing, Massachusetts Institute of Technology, Cambridge

Citation: Lockwood, D. (2019), Simplifying how (and when and where) snow turns into flow, Eos, 100, https://doi.org/10.1029/2019EO116689. Published on 20 February 2019.
Text © 2019. The authors. CC BY-NC-ND 3.0
Except where otherwise noted, images are subject to copyright. Any reuse without express permission from the copyright owner is prohibited.