The mobile observatory is manned by 100 scientists who hope to show how the West can get a better handle on where and when water will be available
Oct 24, 2021
CRESTED BUTTE – Eight white shipping containers, instruments spouting from the tops of some and a generator humming away in another, sit in the East River valley, on the outskirts of this mountain town, pulling data out of the air.
The containers, a “mobile atmospheric observatory,” will gather bits of information over the next two years about the winds and clouds and rain and snow and heat and cold above the silvery and serpentine waterway as it slides past the gray granite dome of Gothic Mountain on its way to the Colorado River.
“It is like a satellite, but on the ground looking up,” said Heath Powers, who oversees the atmospheric observatory program operated by the U.S. Department of Energy. “It’s a traveling scientific carnival.”
Traveling, indeed. The last assignment for the observatory, now in the old mining town of Gothic, 9 miles north of Crested Butte, was on the deck of a German research vessel icebound in the Arctic.
“We’ve been to all seven continents with these observatories,” Powers said. “It is surprising to find such a remote and challenging place here in the old USA.”
The observatory, while in demand all over the world, is the centerpiece in an unprecedented effort to understand how — and how much — water moves from the sky to the rivers of the West. Three separate teams, nearly 100 scientists in all, are in the East River valley studying every facet of the question.
The researchers are employing an equally large array of instruments, from balloons to drones to aircraft to multiple kinds of radar to cloud chambers and flux sensors to stream gauges and rain buckets.
The goal is to better understand the “water story” so that water managers across the West can, from year to year, have a better handle on how much water will be available.
Those systems, however, are not well understood, hobbling forecasting. “We know the list of physical, chemical and biological processes that affect water,” Feldman said. “The question is how do they fit together?”
It is more than just a theoretical question. As the climate changes, and the world gets warmer, the Rocky Mountain snowpack, which provides 75% of the water for the Colorado River Basin, has already declined by a fifth in the past 30 years and by 2050 the flow of the river, supplying water to 40 million people, could drop by as much as 20%.
“We are moving into a no-analog future, where the past doesn’t tell the future,” Feldman said. “We are moving far and fast away from the past.”
And so, Feldman is leading a group of scientists in the Surface Integrated Atmosphere Laboratory project (SAIL), while Gijs de Boer, is heading the National Oceanic and Atmospheric Administration’s Study of Precipitation, the Lower Atmosphere and Surface for Hydrometeorology (SPLASH).
Both are seeking to better understand the atmospheric dynamics — clouds and rain, wind and snow.
The longest-running of the projects, dating to 2015, is the Watershed Function Science Focus Area (no neat acronym, just SFA), which is tracing what happens to the snow and rain once it falls to Earth.
“The goal is to improve water forecasting and water accounting,” said Ken Williams, lead researcher for the watershed project.
Can studying a single, small watershed — with measurements from the size of raindrops to the amount of water finding its way deep into bedrock — tell the tale for the 1,450-mile-long Colorado River and its 246,000-square-mile basin?
“The East River shares characteristics with the vast majority of headwaters in the Rocky Mountains,” Williams said. “What we are learning in the East River will be translatable to other mountain systems.”
“We are used to working in places where you can’t run down to the hardware store.”
The switch was flipped on at DOE’s mobile observatory Sept. 1 and it will gather data through the next seven seasons.
During the winters the three technicians operating the site will be snowbound, save for a once-a-week snowmobile run to town.
“Lots and lots of ramen noodles,” Powers said. The observatory also comes with its own workshop and supply of spare parts. “We are used to working in places where you can’t run down to the hardware store.”
Overseeing the operation is John Bilberry, 43, the lead project manager for SAIL. “I run the circus,” he said. Bilberry was with the mobile observatory in the Arctic (he had to hitch a ride on a Russian icebreaker to get there) and got stranded onboard by the outbreak of the COVID-19 pandemic.
In his previous life Bilberry was a recording engineer for a record label and went on tour with the industrial metal band Ministry. “This is a lot like being on tour,” he said. “You’re given all this expensive equipment and you have to make sure it works.”
SAIL, which is being run under the auspices of the Lawrence Berkeley National Laboratory, has deployed about 50 different instruments, some on the roofs or inside the shipping containers, some on valley hillsides.
The project also releases weather balloons twice a day and has a larger tethered balloon with an array of instruments that will be trucked around the watershed.
Those devices will gather detailed data on eight elements that affect the water cycle: the fine particles floating in the air called aerosols, clouds, rain and snow and the winds that drive them, sunlight, thermal energy and temperatures.
The total sky imager is tracking the horizontal distribution of clouds, microwave radiometers are measuring the water content of those clouds, doppler lidar radar is gauging the direction and speed of the wind, and a nephelometer is measuring the behavior of aerosols.
Other instruments will log ozone levels, the water content of falling snow, how much snowpack is lost to evaporation (known as sublimation) and the surface energy balance — heat coming in from the sun and that radiating back into the air.
Every hour a bank of computers, linked to the sensors, collects all the data and uploads it to the internet for use by SAIL and researchers around the world. “It is a virtual machine,” Bilberry said.
Each of these bits of information are like tiles in a mosaic. “Here we have an opportunity to piece these things together,” Feldman said.
Fitting the data into a big picture will be a challenge as the behavior of any one element can be complex.
Aerosols, for example, can, in the form of soot, warm the air, while sulfate aerosols can cool it. Dust covering the snowpack leads to a quicker melt. Aerosols create the nucleus around which moisture in the air forms rain and snow. Too little aerosol, no rain, too much and the moisture is disbursed and again there is no rain or snow, until it builds up and leads to really heavy downpours or snows.
“Aerosols have all these different effects that they are exerting on these mountainous watersheds,” Feldman said. “Aerosols are impacting the way water is delivered downstream.”
While SAIL efforts are centered in Gothic, NOAA’s SPLASH gear will be arrayed over more than 10 miles and will be focused on gathering data to help improve the administration’s forecasting tools.
These include the Unified Forecast System, which makes up to 14-day forecasts, the Rapid Refresh Forecast System, which provides hourly updates, and the National Water Model, which predicts stream flows.
“SPLASH was born out of a desire to build upon SAIL and tune things to be more specific to NOAA needs,” de Boer said. “That has turned into a very significant investment from NOAA.”
The project is being led by NOAA’s Physical Science Laboratory in Boulder and the University of Colorado, in collaboration with about a dozen other institutions, including Colorado State University and the National Center for Atmospheric Research.
Among SPLASH’s installations will be a 33-foot tower to measure winds, turbulence, radiation and temperatures. It will also deploy three drones to measure things such as soil moisture and snow reflectivity.
“When combined, SPLASH and SAIL provide what may be the most comprehensive study of the physics of the lower atmosphere and exchange with the surface, including water, ever conducted in areas of complex terrain,” de Boer said.
Some water near Gothic has been underground for 2,000 years
On a late summer morning, the SFA’s Williams was up on Snodgrass Mountain drilling a deep well into the mountaintop — SAIL’s white shipping containers could be glimpsed down below.
Granite dust billowed from the hole as the drill pounded away searching for groundwater.
Williams, a Berkeley Laboratory geologist, has drilled wells across the East River valley — into the shale beneath Aspen forests, the loose landslide deposits of Alpine meadows and hard granite of conifer forests — in search of groundwater.
That mixture of granite, shale and soils from mountainside erosion, and the spruce, aspen and evergreen forests, along with Alpine meadow sitting atop them, is a terrain widely shared by Rocky Mountain watersheds.
“The work we are doing is broadly representative of the Rocky Mountains in general,” Williams said, “and will enable us to get a handle on the structure of that system and how physical processes play out in that system.”
Williams’ wells have hit groundwater 15 to 20 feet below the surface, but in the well atop Snodgrass Mountain they found no water even at 300 feet. A dry hole. Williams lowered a borehole camera and found only fractures with seepage. Still, they are being monitored. “All data is useful data,” he said.
Once the water is found in a well, sensors are lowered to measure the soil moisture content at different depths. Samples are also taken for geochemical analysis, such as water dating. Some of the groundwater SFA has found has been down there for as long as 2,000 years.
Williams’ team of 55 scientists, buttressed by collaborators at universities around the country, is trying to write the last chapter in the mountain water story, how a mountainous watershed retains and releases water and how much actually gets to the river.
SFA researchers are trying to measure every drop from tree top to bedrock, down to the role microbes play.
“SAIL and SPLASH are providing a much higher resolution understanding of how and where precipitation is falling,” Williams said. SFA is “taking that handoff” and tracking the water flows.
“This is the first study going from the atmosphere to bedrock,” he said. “It has never been done before in a mountainous system.”
Among the questions Watershed Function is trying to answer is how much of the precipitation is lost to trees and plants sucking it up. In one experiment flux meters have been attached to trees to chart the water flowing from roots to leaves and out as water vapor.
Another question is how much water ends up in aquifers and how long does it stay there? While snowpack runoff feeds the river in the spring, by late summer more than 50% of the East River’s flow is coming from ground water, Williams said.
All the SFA data is also being put up on the internet — so far 69 data sets containing millions of data points — although not by the hour.
Data for modeling for everything from next week’s weather to climate change
The tools for understanding the massive amounts of data being collected by the three projects are computer models that aim to reflect everything from how much water flows in a stream, to next week’s weather, to the future impact of climate change on the world.
The models, however, are vulnerable in two ways. First, they are based on assumptions about how the world works — how much water vegetation absorbs or how snow gathers on mountainsides — and then they are only as good as the data they crunch. “Garbage in, garbage out” is an idiom in computing that goes back, in idea if not the exact words, to Charles Babbage, the 19th century father of the computer.
“There is a critical linkage between measurement and modeling,” Williams said. “The models need to be informed by the data being collected, to show they are anchored in reality.”
“It is data gathering not for the sake of data gathering, but to assure that our predictive models are as accurate as possible,” he said. Scientists call it “ground truthing.”
The data can aid in refining the assumptions and algorithms that run the model. “They can help improve our knowledge of the chemistry and physics of how the world works,” said Alejandro Flores, associate professor of geoscience at University of Idaho and a SAIL researcher focused on models.
Mountains have been particularly difficult to model.
“We have a big blind spot in terms of precipitation and how the models retain and release water,” Flores said. “We need to get a handle on precipitation in mountain landscapes which controls that precipitation.”
SPLASH, de Boer said, is seeking a better understanding of the “physics of key processes,” such as sublimation of snow, snow crystals and rain-on-snow events, that govern how much water ends up in the river.
Those data and insights will be used to evaluate the performance of the Weather Service forecasting and other NOAA models.
Ultimately, the data and knowledge of chemical, biological and physical processes gleaned from the East River could inform the Earth Systems Models that project the world’s climate.
“We currently do not have a good ‘truth’ (for these models), since we don’t have the ability to verify the projections as we do with weather models,” de Boer said.
Getting the model right is a bit like getting the recipe for a cake right, Powers said. “You need to know and understand the ingredients, the proportions,” he said. “If you get it wrong the cake is too sweet or it collapses.”
And it is not just a question of what happens in the West. Between 60% and 90% of the world’s water comes from mountainous watersheds. “Mountain environments are important and they are changing rapidly,” Flores said. “This is an important part of the world and it is important to focus on it.”
“Understanding the physical properties of the East River will help us understand what is happening across the Rockies and all the way to the Urals in Russia,” he said. “It will help anywhere there are mountains and people depend upon mountain snow for water.”