It used to be that a warming Arctic was primarily of interest to environmentalists, and then to maritime navigators due to changes in sea ice. As knowledge has expanded, so has the understanding that thawing permafrost can affect things NGA cares about, including the stability of infrastructure in Russia and China.

By combining satellite and other sensor data into algorithms created under an NGA R&D grant, GEOINT analysts will be able to remotely monitor changes in permafrost and hence anticipate impacts to airstrips, roads and infrastructure sitting on top of it.

“We know that permafrost covers more than 15% of the Northern Hemisphere, and projections indicate half of it will thaw over the next 25 years,” said David J. Blauvelt, Ph.D., lead researcher and program manager for Frostbyte, a remote monitoring capability that aims to provide information on the amount of ice below the ground surface in permafrost regions. The Synthetic Aperture Radar-based capability will enable assessments of mobility and infrastructure risk in polar regions.

Approximately two-thirds, or 65%, of Russia sits atop permafrost, as do China’s three northern-most provinces, where mining is conducted for nickel and other key manufacturing resources. In the United States, 85% of Alaska sits on permafrost, including approximately 40% of the state’s interior.

Why does this matter? Because when permafrost thaws, everything on top of it starts to sink and collapse, affecting building, transportation, availability of electricity, water quality, crops, and human and animal health. Erosion and flooding can occur. If changes are drastic enough, humans must look for new land to inhabit, and societal migration brings its own set of concerns and potential threats. Of particular interest to a combat support agency such as NGA, crumbling infrastructure, docks and runways, and flooding and erosion, can translate into significant impacts on navigation, energy infrastructure and troop movements.

Permafrost is ground made up of soil, rocks and sand that stays frozen, or more specifically, has stayed at 0 degrees Celsius for at least two consecutive years. According to geological and geophysical surveys conducted by Alaska’s Department of Natural Resources, permafrost extends as deep as 2,000 feet below the state’s coastal plain, providing a stable building environment even when the topsoil above it thaws. If the permafrost warms and thaws, however, the top layer of soil will degrade and become unstable.

Even worse, scientists have discovered that thawing permafrost releases carbon, methane and other gasses, heavy metals and pathogens that have been stored in the permafrost for thousands of years. Once released into the topsoil, water and air, many of these releases are toxic to plants, animals and humans.

Current remote methods monitor changes only in the ground surface.  Assessment of terrain instability caused by permafrost thaw – under the ground surface – is limited to costly onsite field surveys that require human intervention and large-scale modeling efforts.

The idea for how to remotely estimate how much ground ice is thawing, and at what rate, originated in a grant application submitted to the NGA R&D grants program by Simon Zwieback, Ph.D., of the University of Alaska, Fairbanks. NGA awarded the grant to enable Zwieback to explore the potential to use interferometric SAR to estimate how much ground ice is present near the surface.

SAR satellites can remotely measure displacement of a ground surface and provide information down to a very high resolution, specifically the 2- to 10-centimeter level, according to NGA research scientist Rachel Bernstein, Ph.D., who facilitated the grant supporting Zwieback’s research. By continually passing overhead, SAR can monitor the ground surface during the thaw season. Bernstein explained that Zwieback’s algorithm combines the rate of displacement observed throughout the year and incorporates additional factors such as moisture, geology and vegetation to infer the change in ice content near the surface.

“Examining permafrost regions throughout a season with SAR may identify which areas may be likely to be less – or more – stable over time, and predict how that might affect structures, utilities and transportation,” said Blauvelt.

Zwieback’s field experiments were due to start in 2020 but were delayed by the COVID-19 pandemic. In subsequent years, he validated his satellite data with permafrost cores in Alaska and produced the first ground ice maps using the algorithm in 2023.

The NGA program, named Frostbyte, is independently validating Zwieback’s work. Through collaboration with the U.S. Army’s Cold Regions Research Engineering Laboratory and Los Alamos National Laboratory, NGA is monitoring changes below/on/above the ground in three regions in the Arctic. These validation efforts require gathering data in the spring, summer and late fall to understand what is happening beneath the ground that might be causing displacement on the ground surface, and to see if these changes are accurately captured by SAR. 

The Frostbyte team is composed of specialists representing six agencies, including the U.S. Geological Survey, Sandia National Laboratory and the Geospatial Research Laboratory, as well as subject matter experts from NGA. Team members interrogate different landcover types using permafrost coring geophysical techniques, geodetic ground surveys, and LiDAR on unmanned aerial vehicles.

“NGA’s geodetic surveys team plays a key role in the collection and processing of accurate survey data, ensuring the different agencies collect data according to NGA standards,” said Blauvelt. “The Frostbyte team must balance extreme weather conditions, cold weather challenges to equipment and personnel, and the occasional polar bear, to ensure that data is properly collected within the short mission windows at remote locations.”

This year, the Frostbyte team completed spring campaigns in tundra environments at Prudhoe Bay and Utqiagvik – formerly known as Barrow – as well as in a forested environment outside of Fairbanks. They completed a second campaign in June and have plans to return for the third later this year.

The data collected in these field locations in 2024 is expected to validate the Frostbyte model in these landcover types. If validated, the model could be applied to larger regions, providing insight into the amount of ground ice in these areas, and assisting with planning and mitigation for infrastructure and mobility studies.