By Seamus McAfee
In the wilds of northern Alaska and Canada’s Northwest Territories, as Arctic spring approaches, the winter’s deep snowpack finally begins to soften under the bright sun. A herd of caribou rests beneath wind-swept black spruce. One morning, without warning, they move out, undertaking a sweeping journey hundreds of miles long towards their seasonal calving grounds.
This migration, an ancient rhythm, has played out for millennia among the Bathurst caribou herd, which takes its name from Bathurst Inlet near the herd’s traditional calving grounds. But scientists have long puzzled over one crucial question: What signals the caribou that it is time to leave?
A study published in Remote Sensing, led by scientists from Lehigh University, the National Snow and Ice Data Center (NSIDC), and the SUNY College of Environmental Science and Forestry (ESF), may have finally cracked the code: snowmelt. The breakthrough came from combining satellite data on snowmelt collected hundreds of miles above Earth with caribou location data from the Government of the Northwest Territories. By analyzing these data sets together, the team—working closely with co-author and caribou expert Eliezer Gurarie of ESF, who developed the migration timing code—found that the Bathurst herd begins its migration just days after snowmelt begins.
Collared coordinates
Researchers focused on the Bathurst herd in part because it is one of the largest in North America and undertakes a vast 2,000-mile migration through rapidly changing Arctic terrain. Many female caribou in the herd have been fitted with GPS collars, providing detailed movement data over time.
At the heart of the study is a trove of snow data from NSIDC. Using satellite-derived brightness temperatures—measurements of upwelling longwave radiation from Earth—NSIDC scientists have developed improved long-term records of surface conditions across the Arctic and globe: the Calibrated Enhanced-Resolution Brightness Temperature (CETB) data set. From these records, the team developed a continental view of melt patterns, providing pixel-by-pixel, day-by-day coverage thanks to its enhanced spatial resolution and consistent calibration.
With records stretching back nearly 40 years to 1987, researchers selected ten years of CETB data that overlapped with caribou GPS collar data. They then applied an algorithm that detected day and night differences in microwave brightness temperatures to derive precise snowmelt timing across the landscape.
The CETB data set is hosted and distributed through NASA’s NSIDC Snow and Ice Distributed Active Archive Center (NSIDC DAAC). Part of NASA’s Earth Science Data Systems program, the NSIDC DAAC archives and shares cryospheric data with researchers around the world, enabling discoveries like this one.
Microwaves and meltwater
Mary Jo Brodzik, retired senior scientist at the National Snow and Ice Data Center (NSIDC), spent her career developing satellite data sets to track changes in Earth’s frozen regions. In this study, she played a key role in applying the CETB data set to detect snowmelt by analyzing passive microwave brightness temperature. “What scientists are able to do is look at a given pixel over time,” Brodzik explained. “They create a long-time series and step through it twice daily—morning and evening. Generally, brightness temperatures stay low until there’s some liquid water in the snowpack.”
Liquid water makes snow appear brighter in passive microwave data because it emits more microwave radiation than frozen snow or ice. As snow begins to melt, even small amounts of liquid water in the snowpack cause a noticeable increase in brightness temperature. In contrast, frozen snow remains relatively dark because it scatters and absorbs microwave energy rather than emitting it. This contrast allows scientists to detect the onset of snowmelt by tracking changes in brightness over time.
Distinguishing between daytime and nighttime data was also important because snowpack properties typically change throughout the day. For example, they can become frozen overnight and begin melting in the afternoon, which affects microwave brightness temperatures. The research uses these differences to capture snowpack dynamics.
Joan Ramage, a geoscientist at Lehigh University, led the snowmelt and migration analysis. By working closely with Brodzik and others, Ramage helped develop the snowmelt indicators that ultimately revealed a strong link between the timing of snowmelt and the migration patterns of caribou.
The telltale thaw
What made the finding especially striking was its consistency. Across early and late snowmelt seasons, with 10 years of data, the same pattern emerged: caribou appeared to respond to the timing of snowmelt.
“We compared the caribou departure date and the snowmelt onset date and found that there’s a relationship,” said Joan Ramage, a geoscientist at Lehigh University and co-author of the study. “We do not know why... but it was really interesting to see that there is actually a temporal relationship.”
Unlike optical instruments, which rely on reflected visible light, satellite microwave sensors are sensitive to longwave emissions directly from Earth. This capability allows scientists to “see” through clouds and during periods of darkness, which is essential for Arctic research where daylight is limited and weather conditions often obscure visible imagery.
“Microwaves are sensitive to the liquid water… the microwave signal is very different when those things are changing,” Ramage explained.
This highlights the importance of using data collected frequently over time, rather than averaged over longer periods, to detect meaningful environmental changes—especially ones that influence wildlife behavior.
Ramage, who has worked with microwave data since the 1990s, said spatial resolution has improved dramatically on data sets like the CETB. The previous passive microwave imagery had a coarse resolution of around 25 kilometers (16 miles). Using multiple daily satellite overpasses and a refined algorithm, the data were synthesized to finer resolutions over time, enabling much more detailed spatial analysis today.
“At 3 kilometers, it’s 64 times finer spatial resolution than the 25-kilometer grid,” said Ramage.
Finer resolution makes it possible to observe local patterns of snowmelt with much greater spatial accuracy, enabling researchers to match satellite observations more closely with animal movement across complex landscapes.
Implications in a warming world
The study comes at a time when Arctic snowmelt patterns are changing faster than ever. Climate models show that spring melt in the north is occurring earlier, more erratically, and with consequences that ripple across food webs. For caribou, earlier melt may seem like a good thing—until it is not. If migration is triggered too soon, before plants emerge, cows may take too long to trudge through snow, reach calving grounds without sufficient food, or even give birth and perish with their calves along the way.
“If they have their calf too early on the migration route, there’s a very high probability of calf, if not calf and mother, mortality,” said Ramage. “So how the migration proceeds is really important for the success rate of the animals.”
So, precise timing is critical. The caribou’s survival depends on a delicate alignment between their movement, the seasonal emergence of food, and the physical conditions of the terrain they traverse. If snowmelt acts as a cue to migrate, disruptions to that signal—caused by climate change, mining, road development, and other human activities—could throw off this balance and endanger both calves and adults.
The exact reasons behind the timing of the snowmelt-induced migration are still being investigated. “We don’t know if migrating is easier or there’s something about access to food changing, or they’re unrelated but correlated in time,” Ramage said. “But for starters, it was really interesting.”
Ramage’s point reflects the complexity of ecological systems: the observed correlation between snowmelt and migration does not yet reveal causation. Still, it opens the door to deeper insights.
Across the Arctic, caribou herds like the Porcupine, Central Arctic, and Teshekpuk each navigate unique landscapes shaped by climate, geography, and human development. These differences can lead to varied responses to environmental change. Because caribou are highly social animals, often moving in coordinated groups, their decisions—such as when and where to migrate—are likely influenced by both environmental cues and herd dynamics.
Studying the Bathurst herd offers a window into broader impacts on caribou and sets the stage for future research on how other herds may be adapting across the Arctic. The research team plans to expand their study to an additional caribou herd to explore whether the pattern of migration holds across these different populations. “If this turns out to be a robust relationship,” Ramage said, “it might allow us to predict their movement or barriers to their movement.”
That kind of foresight is becoming increasingly important as Arctic ecosystems face mounting pressure. Some herd populations are down by more than 98 percent, according to Brodzik. While scientists do not fully understand why caribou migrate when they do, they are getting closer—thanks to decades of satellite data and the tools now available to interpret it.
A new lens on old patterns
The caribou study joins a growing body of research that blends wildlife biology with satellite data—an approach that is becoming increasingly powerful thanks to open data policies and interdisciplinary collaboration. NSIDC’s commitment to free, publicly accessible data sets has enabled researchers around the world to tap into decades of cryospheric observation. By making these large-scale data sets available to anyone, NSIDC is equipping ecologists and wildlife scientists with the tools to ask complex questions about ecosystems—and find meaningful answers.
“I think there’s a tremendous benefit to Earth observation for wildlife monitoring,” Ramage said.
This benefit is magnified by recent technical advances. The open accessibility of trusted data, combined with innovations like enhanced-resolution processing, is transforming remote sensing into both a discovery engine and a storytelling tool. Scientists can now detect finer patterns, draw stronger conclusions, and communicate findings that resonate far beyond the research community.
Although the research in this case centered on caribou, the data it draws from also offers significant insights for humans. As Ramage says, “All kinds of spatial data are incredibly valuable for understanding the environment that the animals are in. It’s really important for understanding human impacts, like mines, urbanization, and road widening.” She emphasized that earth observation offers a better understanding of human systems, enabling improved awareness of environmental changes, and even aiding disaster management efforts.
“We can use it for snowmelt, for hydrology, for mobility, for climate,” Ramage said. “These data have tons of potential for other projects.”
But high-quality satellite data does not just appear—it requires specialized and meticulous work behind the scenes to both acquire and maintain. As Brodzik recalled, her team submitted a proposal four times over six years before finally securing funding in 2012 to develop the improved passive microwave data set. This painstaking process involved adapting complex algorithms to handle vast amounts of data—tens of terabytes spanning decades—and coordinating across multiple research centers to ensure ongoing data collection and quality.
Since finally taking on the data set at NSIDC’s Distributed Active Archive Center (DAAC), staff continue to curate, calibrate, and maintain the integrity of the data, ensuring it can be trusted for use in peer-reviewed ecological studies and a wide range of other applications. Without this work, the insights drawn from remote sensing would not be possible.
In the case of the caribou, that well-curated, high-resolution data helped uncover a subtle but powerful truth: long before roads, fences, or climate models, the animals already knew to wait for the snow to melt. Thanks to sensors flying on satellites, scientists like Brodzik and Ramage—and the trusted, open-access data stewards from NASA’s NSIDC DAAC—now we know too.
Access data through the NSIDC DAAC
NASA's NSIDC DAAC manages, distributes, and supports a variety of cryospheric and climate-related data sets as one of the discipline-specific Earth Science Data and Information System (ESDIS) data centers within NASA's Earth Science Data Systems (ESDS) Program. User Resources include data documentation, help articles, data tools, training, and on-demand user support. Learn more about NSIDC DAAC services.
NASA NSIDC DAAC data highlighted in this article include:
Brodzik, M. J., Long, D. G., Hardman, M. A., Paget, A. & Armstrong, R. (2016). MEaSUREs Calibrated Enhanced-Resolution Passive Microwave Daily EASE-Grid 2.0 Brightness Temperature ESDR. (NSIDC-0630, Version 1). [Data Set]. Boulder, Colorado USA. NASA National Snow and Ice Data Center Distributed Active Archive Center. https://doi.org/10.5067/MEASURES/CRYOSPHERE/NSIDC-0630.001.