AV����̳

Photo by Kelly Reynolds.
Santosh Panda on the University of AV����̳ Fairbanks Troth Yeddha' Campus

By Laura Weingartner

The Pretty Rocks landslide in Denali National Park and Preserve became national news in 2021 when it took out a large swath of the only road going deep into the park. 

The landslide, which has been active since at least the 1960s, has accelerated over the years. From a few inches per year before 2014 to almost an inch per hour in 2021, the road began to slide, causing the park to close the road near mile 45.4, roughly the midway point into the park. 

Pretty Rocks lies on ground long underlain by permafrost — soil and rock that remain frozen for two or more consecutive years. Warmer temperatures and heavy rainfall are believed to be causing the permafrost to thaw, which in turn is accelerating the landslide and taking the road with it. 

Pretty Rocks isn’t the only area in the park, or the state, where permafrost is degrading, an issue that affects the entire landscape through changes in water distribution, ground slumps, changes in vegetation and more. Knowing where that permafrost is and how likely it is to thaw is one of the first steps in minimizing or managing those changes.

In 2015, Santosh Panda, a professor of geographic information science at the University of AV����̳ Fairbanks' Institute of Agriculture, Natural Resources and Extension, produced detailed maps showing permafrost distribution and its vulnerability to thaw across seven AV����̳ national parks, an area of land similar in size to Florida.

Drawing together vegetation, soil, snow and climate data, he created a model capable of reliably predicting where permafrost is present, how cold it is and the thickness of the seasonally thawed active layer above the permafrost. By running the same models with projected climate scenarios, he produced maps of what near-surface permafrost — the layer just beneath the active layer — may look like in the coming decades.

Many assumptions are made to create models such as these maps. To ensure accuracy, the model results are compared with field observations collected by National Park Service scientists: summer thaw depths, the presence or absence of permafrost, and ground temperatures recorded at climate stations. 

Denny Capps, geologic hazards lead with the park service, uses these maps to inform infrastructure and management decisions in Denali. He said that despite these assumptions, one can still learn a lot from models.

“We never used it like it was the absolute truth — more like guidelines," Capps said. Then, through years of experience, we verified that much of it was correct.”

These maps draw a striking picture. In the 1950s, approximately 75% of Denali contained near-surface permafrost. By the 2000s, that number had dropped to 51%. By the 2050s, projections show just 6% of the park will be underlain by near-surface permafrost. 

Park geologists, like Capps, have put Panda’s maps to work, using them to anticipate where ground instability might occur and where the park may need to focus monitoring efforts. 

“I believe that permafrost thaw is the No. 1 threat to Denali’s resources (as we know them) and visitation, and I’m not alone," Capps said. "So, this map is a foundational part of our understanding of the park and its future."

For Panda, that practical impact is the heart of his work. His research always begins with a problem that affects people or the environment and ends with something anyone can use or benefit from, whether it’s an app, a website or a map.

“I like to create something, a tool, so that anybody who is interested in this kind of product or knowledge can use it. They don’t have to be a scientist,” he said.

The permafrost maps are one example. Another is a water depth map of the Colville River near Nuiqsut. The silty riverbanks along the Colville are held together by moisture and ice that start to slump when that ice melts. 

“When the ground is completely frozen, it's pretty much like concrete. As the ground thaws, that moisture will release out of the system and the ground starts collapsing,” Panda said, describing how slabs of land collapse into the river, sending sediment downstream and creating shallow water in unexpected places.

Channels that Colville River boaters once knew by memory were reshaped, and residents were increasingly getting stranded, damaging their boats or were cut off from tributaries previously used for hunting or fishing. Panda’s map shows shallow water sections and poor tributary connections, helping the community of Nuiqsut navigate more safely.

Another example is his work mapping forests. Working with doctoral student Sumana Sahoo and post-doctoral researcher Anushree Badola, Panda is also using remote sensing to detect aspen leaf miner outbreaks, identify hot spots of tree mortality and map wildfire fuels.

This work focuses on how AV����̳'s forests are transforming due to persistent warming, inconsistent precipitation, insect pests and intensifying wildfires. Panda and his team create products that enable land managers, decision-makers and the public to assess fire risk, prepare for vegetation shifts and understand how the landscape is changing. 

Remote sensing makes this possible. Using data collected from satellites and aircraft, remote sensing measures the electromagnetic energy that emanates from the sun and is reflected by objects on Earth back toward satellites and aircraft. Electromagnetic energy travels in the form of waves of various sizes, from radio waves with very long wavelengths to gamma rays, which are much shorter. Visible light, the small portion of energy the human eye can detect, is in the middle of that range. Panda's research uses the visible, infrared and microwave parts of the electromagnetic spectrum.  

Everything on Earth reflects, absorbs or transmits energy in a unique pattern, known as a spectral signature. Researchers can use spectral signature information to map tree species, insect damage and dying or dead vegetation. By combining these satellite and aircraft data with environmental variables, Panda and his team can build models that predict how the landscape might change. 

The models are then tested against data collected in the field, a process known as ground truthing, to validate their accuracy. For the aspen leaf miner project, for example, this means checking satellite-identified patches against on-the-ground observations of leaf damage made by these tiny moths, as well as the type of trees found in that patch, to ensure the models truly capture what is happening in the forest.

Panda’s journey to this work didn’t happen all at once. Growing up in rural India, he never felt overly enthusiastic about a specific subject, but clearly excelled in school. 

“I was just following the crowd,” he said about his pursuit of an undergraduate and a master’s degree in geology. 

His interest in research was sparked after graduate school when he was hired onto a project that involved working with local people to help reclaim coal mining sites in northeastern India. 

“The research was exciting, and it gave me exposure to applied research that is useful for people, for the environment,” he said."It also gave me the idea that maybe I should pursue a Ph.D.”

For the first time, Panda thought he should consider leaving India. After a bumpy start to a Ph.D program in India, researching thermal signals from earthquakes, he learned about a project at UAF mapping permafrost along the AV����̳ Highway. He applied, interviewed, was offered the position and moved to Fairbanks in January 2007, right after a big snowstorm. Despite the cultural shocks that he still experiences, he found that the move instantly felt right.

“Fairbanks has worked for me from day one. It made me a better person,” he said.

He’s been at UAF since, completing his Ph.D. and postdoctoral work before being hired as a professor in 2016.

In addition to research, Panda is an instructor committed to widening access to geographic information science. He teaches four credit-bearing GIS and remote sensing courses at UAF, leads the Geospatial Science graduate certificate program, and has created four massive open on the online learning platform EdX, which have reached more than 50,000 learners worldwide. 

“GIS is a very interdisciplinary tool, and I feel happy and satisfied with what we have been able to offer to UAF students and students outside the system,” he said. 

Panda speaks three languages: Odiya, the official language of the state of Odisha, where he grew up; Hindi, the most widely spoken language in India; and English, the language he used while attending school from 10th grade onward. He understands at least two more. 

Perhaps his research is just another language, and he works to translate the world into something others can use.