Better data and better models to predict the effects of climate change
A significant part of the project is focused on improving how satellite data is interpreted, since more accurate interpretations can lead to more accurate climate models, which are used to predict everything from temperature changes to greenhouse gas emissions. That information can help communities and governments plan for impacts from increasingly severe and frequent weather events that may impose infrastructure damage in Arctic regions, shifting ecosystems and increased carbon release. The techniques developed through the project could also be applied more broadly, Lanagan said, including monitoring soil moisture for agriculture, managing water resources and improving environmental tracking in remote or hard-to-access areas.
“Permafrost holds about 1,500 billion tons of carbon, almost double what’s already in our atmosphere,” Rahman said. “As the Arctic rapidly warms, this trapped carbon is released as greenhouse gasses such as methane and carbon dioxide, which drives global warming even more.”
Tracking those changes across remote Arctic landscapes is not easy. Scientists rely heavily on satellites and drones, which scan the surface from above using electromagnetic signals.
“When the satellites, or drones, fly over, they can capture information about very remote places where people just can’t be on the ground,” Lanagan said.
He explained that the instruments send signals down to the surface and measure how they reflect back. The property of the soil — such as wet, dry, frozen or not frozen — impact how a signal is reflected back to a satellite or drone, enabling scientists to infer what is happening below the surface.
By combining lab experiments, computer models and fieldwork in the Arctic, researchers are learning how the permafrost changes over time. They use this data to build models that can interpret satellite and airborne imagery, allowing them to monitor remote regions without needing constant on-site measurements. This work helps track environmental change, protect infrastructure, and support operations in polar regions, while also advancing powerful new remote sensing technologies.
Bringing the Arctic into the lab
Instead of traveling to the Arctic, Penn State researchers recreate permafrost conditions in the lab. They build samples using combinations of sand, silt and clay provided by St. Louis University, then vary the amount of water inside, mimicking the various kinds of permafrost in different regions and depths. The samples are cooled to below freezing, simulating the freeze-thaw cycles that occur in nature during seasonal changes in temperature.
“We are able to cool the soil samples from the room temperature to around negative 10 to negative 15 degrees Celsius,” said Mingjin Lu, a computer engineering major who led a Penn State College of Engineering senior capstone team on the project. That’s about 14 degrees Fahrenheit to 5 degrees Fahrenheit.
As the temperature changes, the team measures how the soil samples respond to electromagnetic — or microwave — signals.
“Really, what the capstone project is doing is providing a way to simulate the permafrost and get the microwave response in a lab setting,” Lu said. “And then use the correlation between temperature and the microwave data to help support the remote sensing.”
The setup was built as part of Penn State’s engineering capstone program, where students design and construct real-world systems. The permafrost capstone project focused on designing a system to freeze and thaw the samples while measuring how they response to electromagnetic signals. Data from the measurement system are then used to develop a model for predicting permafrost changes from satellite imagery.”
“We really felt that we are able to contribute to a work that is going to be a larger project and a long-term project, and it will really be a really impactful one,” Lu said.
Connecting permafrost structure, satellite signals and climate
Inside each soil sample, the arrangement of particles and water changes as it freezes or thaws. Those structural changes affect how the material interacts with electromagnetic waves and ultimately how signals from satellites and drones reflect from the permafrost surface.
“So, in material science, we always look at the structure of the material and how it relates to the properties,” Lanagan said. “We’re always correlating structure and properties.”
To study that structure, the team uses a range of advanced imaging tools across Penn State, including CT scanning, electron microscopy and magnetic resonance imaging.
Agustin Harte, a senior in engineering science and mechanics working on the project, focuses on analyzing how soil structure changes with different water levels and compositions.
“It’s really important to quantify and really understand how the structure really pertains to the changes in the dielectric properties,” Harte said.
As the soil freezes or thaws, it changes its dielectric properties also known as permittivity, meaning that it will respond to an electromagnetic wave — a satellite signal — differently, which the satellite or drone can detect.
By linking structure and properties, the researchers can better understand how reflected signals from satellites are shaped by what lies beneath the surface, Lanagan said.
“So, when satellites fly over, they can do that correlation,” Lanagan said. “That’s what this is, this area is frozen, or this area has a lot of water. Then, we can interpret that to mean the permafrost is still frozen or thawing.”