ASPEN — On a sunny day in mid-April, scientist Jeff Deems strapped on his backcountry skis, tromped out to a snow-covered field and began digging.
“I was expecting there to be a little more than 70 centimeters here,” he said, standing in the middle of a snow pit that looked vaguely like a shallow grave. Somewhere overhead, out of sight, his team was in an airplane over the Roaring Fork Valley mapping its snowpack from above.
Each year, Deems and his partners at Airborne Snow Observatories Inc. use lasers, planes and their own two feet to calculate how much water is in the mountain snowpack, where most of the state’s water supply is stored before it melts each spring. Their monitoring technique, developed at NASA, offers better accuracy than other tools and can gather snow data where other monitoring programs are blind.
That’s particularly helpful now when two decades of drought have tightened water supplies in the Colorado River Basin.
“It used to be, we could accept a 15-, 20-, 30% error on runoff forecasts,” said Deems, co-founder of Airborne Snow Observatories. “Because supplies are getting more limited, or at the very least more variable, and because demands are greater than when the forecasting systems were built, then that error becomes more critical.”
In Colorado, 83% of the state’s water supply comes from surface water fed by winter snowpack and spring runoff, according to the state’s 2023 water plan. Colorado’s snowmelt also flows downstream to millions of other users in the Colorado River Basin, which spans seven Western states, 30 Native American tribal lands and two states in Mexico.
Like Deems, Westerners have been hauling themselves and their gear up to the mountains to try to measure the snow and predict runoff for more than a century. Having the most accurate snowpack measurement possible is vital for water agencies, which use the data to figure out how much ends up in home faucets and on farms for irrigation.
Researchers have learned how to put gamma rays, cosmogenic rays, passive microwaves, radar and more to use in measuring and modeling techniques. Some techniques are more experimental than others, and some aren’t as helpful in mountainous topography, said Noah Molotch, a geography associate professor at the University of Colorado Boulder.
“All those techniques I mentioned, for the most part, are not providing data at the accuracy level that’s needed for the water management community,” said Molotch, who researches another new approach involving satellites to gather snowpack data. He sometimes uses ASO data in his work but is not affiliated with Airborne Snow Observatories Inc.
The search for new, more accurate ways to measure snowpack is on. In March, the Bureau of Reclamation announced it had almost $12 million available to fund emerging snow monitoring techniques, and NASA is experimenting with remote sensing technologies in its SnowEx program.
New technologies could add to the data gathered by the longstanding snow-telemetry, or SNOTEL, network, which has been operated by the Natural Resources Conservation Service for decades and is a primary source of information for water managers.
The ASO program is a much younger method. NASA’s Jet Propulsion Laboratory started the program in 2013, and after eight years of development, Deems and his partners wrapped up their research at NASA in 2019 and launched Airborne Snow Observatories Inc.
The aerial technique’s laser sensors and imaging spectrometer create highly accurate, watershed-level data. It’s a big transformation, Molotch said. The West’s water cycle has become less predictable in recent decades with a changing climate, and the techniques that have been used to forecast streamflow and water supplies are becoming less accurate, he said.
“The new models need better data, and they need data that covers the landscape in its entirety, so that’s why we need measurements like the Airborne Snow Observatory,” Molotch said.
“It’s going to help us fill in gaps”
When ASO flies over a watershed, lidar technology on the bottom of the airplane sends pulses of laser light to the ground to take snow depth measurements every 10 feet across all elevations in the watershed.
From there, the team determines the snow-water equivalent — the amount of liquid water in snow — by measuring snow density on the ground and using modeling techniques. The combination gives Deems a high-resolution image of the snow’s depth and the volume of water in the snow across the watershed.
“Really I just can’t wait to see that everywhere, because every range, every basin has its own idiosyncrasies, and it’s just fascinating to see all of them,” Deems said.
The ASO data is helping to fill in gaps in information from the SNOTEL network, which has been around since the 1970s. Widespread manual measurements have been taken since the 1930s or before. The network is made up of over 900 stations in the West, 114 of which are in Colorado. Sensors at each station take continuous measurements at one location in a watershed, providing data on the snow-water equivalent, precipitation, soil moisture and more.
Forecasters use 30-year averages of this data to model the spring runoff and help predict water supply. SNOTEL site locations are chosen carefully to make sure they will be indicative of the greater basin, said Brian Domonkos, the Colorado snow survey supervisor for the Natural Resources Conservation Service. Too low in elevation, and the snow will melt too early leaving the site with no snowpack to record. Above treeline, and winds can carry some snow away. Most sites are below treeline, and in Colorado, are between 9,000 and 11,600 feet in elevation, he said.
But that means SNOTEL sites miss snow below 9,000 feet and above 12,000 feet. They are a point measurement: They don’t, and can’t, actually measure an entire basin. Deems said it would be like trying to map Denver traffic using just a few cameras, while ASO is more like mapping traffic using Google Maps.
“That’s where ASO data is going to come in and do a great job. It’s going to really help us fill in those gaps,” Domonkos said. “Melding those two will hopefully provide an improvement in understanding of distributed snowpack.”
ASO flights capture all elevations in a watershed, instead of just one point. Snow melts more quickly at lower elevations, so knowing where the snow is across different elevations helps forecasters assess spring runoff. That’s key information for water managers trying to determine when to release water from reservoirs to maximize their annual storage.
“You’re hearing us capture variability throughout this snowpack right here. Now think about the entire watershed,” said Deems in mid-April over the scrape and clang of metal as he took measurements with his density cutter, a metal scoop-like tool that measures volume. “You’ve got different depths, different densities, different layering. Trying to figure out the volume of water contained in the basinwide snowpack from a single measurement is impossible. Literally.”
In 2019, ASO flight data in the Blue River watershed above Dillon Reservoir proved to be “invaluable” to water management decisions, which serves about 1.5 million people in and around Denver, said Taylor Winchell, a senior planner and climate adaptation specialist at Denver Water.
The SNOTEL sites above Denver Water’s Dillon Reservoir in Summit County were mostly melted out, but the ASO data showed more than 100,000 acre-feet of water available in soon-to-melt snow at high elevations in the Blue River watershed.
“So if you were going by SNOTEL stations, you would think there’s almost no snow in the basin, essentially,” Winchell said. “With (the ASO) information, we increased the outflows at Dillon Reservoir to prepare to capture that snow as it melted off.”
ASO flights also measure the reflectivity of the snow, or its albedo. Just like black, leather car seats get hotter in the sun, darker snow absorbs more energy and warms more quickly into liquid water.
On April 3, a layer of dust blew through Colorado and landed on the snow. Days later, when Deems made his trip out to the snowpack above Aspen, the snow’s surface was still covered with it, and its internal structure had been gutted from rapid melt. It was basically like pouring gasoline on the snowpack and lighting it, Deems said, after post-holing on skis repeatedly during the short hike through the snow.
“I would imagine based on our prior research experience into dust on snow that it’ll shorten our melt season by about a month,” Deems said. “That’s never really a great thing. The snowpack is our biggest reservoir, so the longer we can keep it on the mountains, the better off we are in terms of that water being available later in the summer.”
With the continuous data from SNOTEL sites and a growing basinwide data set from aerial measurements, water agencies can have a near-complete understanding of how snow is behaving in their watersheds. The two measurement techniques go hand-in-hand, Deems said.
“Having ASO data is a fascinating new perspective into the dynamics of our mountain snowpack and our biggest reservoir. We’re seeing differences year to year that aren’t detectable any other way,” he said. “That gives us an important capacity to adapt to changing conditions and really enhance our ability to manage and respond to those changing conditions.”
What does ASO show about this year’s snowpack?
Airborne Snow Observatories Inc. flew over more Colorado watersheds this year than any other year since Denver Water’s inaugural Blue River flight in 2019.
The company first flies over a basin without snow to provide a base map and then flies one or more times during the snow season to measure the snow depth. The basin data sets are in their earliest stages: In most cases, there isn’t enough year-to-year data to spot snowpack trends, and water managers are still working out how to use the new information in their reservoir operations.
Denver Water worked with ASO and multiple funding partners to map the Upper South Platte watershed for the first time this year. The ASO measurement gave them more confidence in how much runoff to expect, Winchell said.
Streamflow forecasts can offer a wide range of possible runoff, which means water managers plan for a variety of scenarios. For example, the NRCS April 1 forecast, which uses SNOTEL data, showed that the South Platte River just downstream of the ASO flight area could receive anywhere from 47,000 acre-feet to 171,000 acre-feet, with a mid-estimate of 109,000 acre-feet, he said.
“With the ASO measurement of 108,000 acre-feet, that allows us to focus our planning closer to that number,” he said. “It takes a wide range of possibilities and narrows it down so that we can make more informed decisions.”
Emily Carbone, water resources specialist at Northern Water, said ASO flight data has shown how the snow depth has shifted between elevations year to year. For example, she noted a 269% increase in the snowpack between 8,000 and 9,000 feet — where SNOTEL stations are sparse — from 2022 to 2023 in the Granby Basin of the Colorado River headwaters.
“That’ll be useful over time to see, what does it do to runoff timing when you’ve got more in the lower elevations compared to your middle elevations and higher elevations?” she said. “Once we get a longer record, the ASO data will get even more valuable.”
The city of Aspen doesn’t even have a SNOTEL station in its primary watersheds, Castle Creek and Maroon Creek, said Steve Hunter, Aspen’s utilities resource manager. That means city staff rely on SNOTEL data from other places in the Roaring Fork Valley.
“We tend to use Independence Pass or Schofield Pass to try to make a rough correlation between their snow-water equivalent and streamflow for us, which is really, really rough,” he said. “Those aren’t even in the same basin where we’re pulling water from.”
Plans to add a SNOTEL station to the city’s watersheds are in the works. In the meantime, Aspen is using ASO data for a better glimpse into its water resources, particularly in order to maximize the water stored at Leonard Thomas Reservoir. The reservoir only holds 10 acre-feet of water — one day of water supply during peak demand in the summer irrigation season. Most municipalities have three years of water stored, Hunter said.
The city has some backup storage and plans to add more, but if a wildfire or some other disaster hits its main watersheds, hydropower production and irrigation could be shut off entirely, he said.
This year, Hunter noted that the Maroon Creek watershed had about 73,400 acre-feet of snow-water equivalent in mid-April, about 11,000 acre-feet more water in its snowpack than in April 2019. Castle Creek had less water this year: 66,700 acre-feet compared with 70,301 acre-feet in April 2019.
At around 12,000 feet in elevation, which is higher than most SNOTEL stations, there was almost 125,000 acre-feet of water in the entire Roaring Fork watershed in April, he said.
“We’re seeing a lot of snow that you would never know was up there looking at the SNOTEL alone,” Hunter said.
A lot of the basins with ASO flights this year saw good low elevation snowpack, lower than the typical SNOTEL station elevations. The Dolores Basin, for example, had a lot of big storms and a lot of snow, Deems said.
“That has contributed since the time we’ve flown to some of the flooding that we’ve seen around the state,” Deems said.
Will it last?
The aerial snowpack data will be even more useful to water operators when they have five, 10 or more years worth of information, Northern Water’s Carbone said.
So far, the ASO data has generated enough interest among water operators, agencies and nearly 100 Colorado stakeholder groups that they banded together to form the Colorado Airborne Snow Measurements working group to try to grow the program to sustainable levels. The cost of flights and the scramble to find funding for the whole program are big challenges.
“That’s something we’re all trying to figure out: How is this sustainable, or is it? We’re still in the early phases of working with this data,” Carbone said. “Definitely, I’m finding value in it now, but it’s something that we continue to evaluate every year.”
Water districts, like Northern Water, are paying for the ASO flights out of their own budgets or with assistance from agencies like the Colorado Water Conservation Board.
A single ASO flight survey can measure a basin up to 3,500 square kilometers, about the size of the entire Roaring Fork River watershed. Each flight, along with data processing, costs between $120,000 and $150,000. That doesn’t include flight costs for base-layer mapping when there is no snow, according to a 2022 CASM report.
“It’s a burden to small entities that are currently supporting it,” Deems said. “Pretty much a universal consensus is that a sustained program needs buy-in from local entities, but really needs sustained state and federal support in order for that program to be something that exists long enough for folks to get used to using it and doesn’t represent a burden on local communities.”
The stakeholder group is trying to work out whether it would become a fully fledged program within the Colorado Water Conservation Board or whether the board would act in an advisory role for a third-party entity, said Erik Skeie, the board’s special projects coordinator, in an email to The Colorado Sun.
The amount and frequency of the board’s continued investment would depend on its future role, but it would take a combination of federal, state and local funding to sustain a statewide effort, he said.
ASO also operates in California, where the flight program is much more extensive. The state appropriated $12 million per year for its program starting in 2023, which supports four flights per basin each year for most of the Sierra Nevada and north into the Yuba, Feather and Lake Shasta basins. The state added flights this year in response to record snowpack and emergency conditions.
“That’s been a really important capacity for operators to understand exactly how much snow is in their watershed and allow them to manage their reservoirs both for flood control — which is the key management target right now — as well as making sure those reservoirs are full enough for a water supply later in the season,” Deems said.
The working group projected a fully functional ASO flight and forecasting program would cost around $26 million per year, depending on how quickly the program grows, according to the 2022 report from the working group. That would fund 214 flights, enough for six across all major headwaters in Colorado.
“We really see it as a climate change adaptation tool. Climate change directly impacts the snowpack and being able to better measure that snowpack will help us to adapt to changing climate conditions,” said Winchell, who is on the working group’s planning team. “That’s what ASO allows us to do.”