Balloon Science
UM staff and students traveled 6,500 miles to southern Chile in the middle of a pandemic, during finals, to launch 100 weather balloons in bad weather to study a total solar eclipse. Could anything go wrong?
By Jacob Baynham
It wasn’t the easiest time to travel. Before she could board her flight for Santiago, Chile, last December, UM student Hannah Woody first had to visit the Chilean consulate in San Francisco to get documents allowing her to travel between regions. When she got back to Missoula, she had to self-isolate for 10 days and finish her finals remotely. Then came the mandatory COVID-19 test, 24 hours of N-95 masked flights to Chile, another COVID test in a Santiago hotel and one more day of isolation until she got her results.
Only at that point could Woody and the rest of her research team — a dozen students and five staff members from UM and three other universities — pack their rental vans and drive nine hours to their final destinations: two remote towns in southern Chile that lay in the direct path of a total solar eclipse, the only total eclipse in 2020. The team’s mission was to launch hourly weather balloons before, during and after the eclipse to detect and observe stratospheric gravity waves — a meteorological phenomenon that happens when air cools and contracts, creating ripples in the atmosphere that can change the reliability of weather forecasts.
Nine months earlier, all of that was unknown to Woody, a senior biology major with a minor in astronomy. In the spring of 2020, a professor encouraged her to apply for an internship with the Montana Space Grant Consortium to get better at coding.
“I applied not really knowing what it was,” she recalls. “Most people hadn’t heard about atmospheric gravity waves. I certainly hadn’t.”
In the first weeks of the internship, Woody and the other interns pored over papers on fluid dynamics, atmospheric science and gravity waves. Woody learned how to write scripts of code. She studied up on common terrestrial and meteorological sources for gravity waves — things like mountains, storm fronts, convection and wind shear.
Then came 10 weeks of hands-on experience launching weather balloons on the UM Oval. The balloons are 4 feet in diameter on the ground and rise to about 100,000 feet –halfway through the stratosphere — before they burst. They carry a palm-sized scientific instrument called a radiosonde that relays meteorological data to a computer on the ground.
The culmination of the internship was a two-week field campaign to Chile for which the team had received almost $700,000 from the National Science Foundation. But when the pandemic hit, all bets were off. Summer turned to fall, and the pandemic only worsened. Woody and the other students assumed there would be no way they could go down to Chile to apply their new skills during the eclipse.
“Things just fell apart,” Woody says. “Throughout the fall, I told my partner, I’m definitely not going to Chile. Then the University approved our travel, and Chile opened its borders. It all came together about a month before we left.”
If the pandemic thought it was going to derail a rare opportunity to detect eclipse-driven atmospheric gravity waves, then it had underestimated the resourcefulness of two key people at UM: Jen Fowler and Carl Spangrude.
“We’re both optimists, probably unreasonably so,” says Fowler, who directed UM’s Autonomous Aerial Systems Office before leaving the University in October to take a job with NASA. “We started writing up safety protocols. We started asking the Chilean government, what would it take to allow us to do this?”
Spangrude, a 2019 UM graduate who now serves as UM’s remote sensing director for the Montana Space Grant Consortium, wasn’t going to be deterred. He worked with the U.S. Embassy in Santiago and the Chilean Consulate in San Francisco. He even explored the possibility of launching balloons from a U.S. Naval ship. Finally, the Chilean ministry of science and the Ministry of Health approved their entry.
“We were not going to stop trying,” Spangrude says. “Jen and I thought, how are we going to navigate this? We can problem solve. We can make it work. People who achieve great things are the ones who do that. They figure out a way to make it work when everyone else is telling them they’re crazy.”
Scientifically, the stakes were high. Fowler and Spangrude had been in Chile for an eclipse in 2019, launching weather balloons from the southern edge of the Atacama Desert. During that field campaign, they documented the first eclipse-induced atmospheric gravity waves. The concept of eclipses producing gravity waves had been around since 1970, but they had never been definitively recorded in the stratosphere. Now they wanted to do it again.
“Without replicating that, there’s no way to say it wasn’t a fluke,” Spangrude says. “Our methods were rigorous, but the sample size was one campaign.”
That’s why it was imperative to get back to Chile for the 2020 eclipse, which would also pass from west to east, crossing an ocean and then the Andes, in almost the same geographical location as in 2019. Back-to-back eclipses with such similar attributes wouldn’t happen again for another 27 years.
“The 2019 and 2020 eclipses were the one opportunity for us to do this work in our careers,” Spangrude says.
So with binders full of contingency plans, Spangrude, Fowler and the rest of the group flew down to Chile in December to conduct their research. They booked lodging through Airbnb and arranged to have 5-foot helium tanks purchased and delivered through the University of Santiago. They were in communication with the Chilean version of the Federal Aviation Administration to make sure the balloons didn’t upset any flight plans.
“We had this planned down to the minute,” Spangrude says. “There was no detail too small. There were so many things that could’ve gone wrong. If we got shut down it was going to be because of something beyond our control.”
By the time the actual eclipse drew near, the two teams were in position, one on the coast and the other at the feet of the Andes. Even though the students came from four different universities, they had all trained to launch the balloons in the same way. The teams divided into night and day shifts.
“We call it game day when the launches start,” says Fowler. “Filling balloons with helium is pretty loud. We’re yelling at each other. The launch site is separate from the sleeping quarters and food. Every 30 minutes you’re doing something.”
The process itself was pretty repetitive. The launch crew consisted of three different positions. One person prepped the radiosonde by entering initial weather data and ensuring it was connected by radio signal to a laptop on the ground. The second position filled the balloon to a precise pressure and then released it. The third position recorded current meteorological measurements on the ground station to help initialize the radiosondes. Once the balloon burst at the end of each flight, the radiosonde would fall back to earth on a parachute. The team did not retrieve them — all the data were transmitted via radio signal — but information on the back of the devices enables people to return them to UM if they are found.
“We did musical chairs on what position we were doing,” Woody says. “Switching roles helped keep you awake. Going outside on a 40-degree night would kind of wake you up a little more.”
Fowler, Spangrude and professors from other universities were acting in support roles, in case anything went wrong. In the mornings, Fowler fried up dozens of eggs to feed the night shift that was coming off work and the day shift that was starting. At both sites, the students themselves filled and launched the balloons and recorded the data, one launch each hour for a nine-hour shift.
Periodically Fowler and the others would take a computer down to the local internet cafe to upload data to the cloud to be quality checked by the team’s mission control, a group of students holed up with coffee, fruit and granola bars at the University of Idaho. That’s where Graham Moss, then a UM senior in physics, spent the eclipse. He couldn’t get his passport in time to go down to Chile, so he took advantage of high-speed internet to provide the teams with weather forecasting and early data analysis.
“We spent basically all hours of the night and the eclipse campaign waiting for data to come in,” Moss says. “We’d look at data to see if there were any problems with it, and we were able to run that data through some of the gravity wave detection algorithms.”
Woody, who was working the night shift at the coastal site, woke up to witness the eclipse itself. It was the first eclipse she’d seen, and the experience was profound. Although the day was cloudy, there was a tiny break in the clouds right at totality, so the team could look directly at the eclipse crescent without dark glasses.
“All the frogs and crickets started chirping as it got really dark,” Woody recalls. “It got really still. The wind stopped. You just feel this immense force of this eclipse moving over you. Then it passed. The animals quieted down, it got breezy again, and it was back to business.”
On the day of the eclipse, windy rainy weather rolled in. The extra moisture changed the amount of helium the balloons needed to rise at the ideal rate of 5 meters per second. And the wind sometimes whipped the balloons over the towering Andes so quickly that the radio signal was lost before the balloon flight was complete, creating holes in the data.
Once Fowler and the team returned to the U.S., they realized that the weather was caused by a rare phenomenon called an atmospheric river, a narrow band of heavy moisture that happened to coincide with the totality of the eclipse. Very little data exists on atmospheric rivers in South America. They are more difficult to predict ahead of time than eclipses.
“We got an entire data set of hourly radiosondes from the evolution to the dissipation of an atmospheric river,” Fowler says. “It added a whole other dimension to our data.”
Currently, the team is busily modeling their 2020 data to detect eclipse-induced atmospheric gravity waves. So far they haven’t definitively identified one, but they have several likely candidates. The research is leading to at least six different student-written academic papers, and some of the student participants — Hannah Woody and Graham Moss included — were invited to present at the next American Geophysical Union conference, one of the world’s largest scientific gatherings.
What’s more, two upcoming eclipses in 2023 and 2024 offer additional opportunities for atmospheric gravity wave research. Spangrude is a key leader of the NASA-funded $7 million Nationwide Eclipse Ballooning Project, which will include funding for UM student researchers.
In the end, Fowler and Spangrude were able to do what seemed impossible — conduct rigorous scientific research in another country during a pandemic. The team’s data provides a valuable contribution to atmospheric science. Weather forecasts will improve when these waves are better understood. Spangrude is inspired by the fact that the effort was all student-driven.
“It was a proof case to have undergraduates lead the research,” he says. “We were demonstrating that undergraduate students have much more capability to do rigorous meaningful science.”
Meanwhile, students like Woody walked away with a foundational scientific experience of a lifetime. Woody, who plans to do graduate work in astrobiology, also did what she set out to do: improve her coding skills.
“I really enjoy being able to code,” Woody says. “I don’t have to sweat knowing coding as I move into a STEM field. I’m going to be able to keep up.” •
