In the decade-plus that University of Alaska, Fairbanks, professor Matt Nolan has studied the impacts of a changing climate on Alaska’s McCall Glacier, he’s grown accustomed to the logistics of researching glaciers: cold and short field seasons, inclement weather, and more.
Operating efficiently in the field at as low a cost as possible is paramount to Nolan and his colleagues. So last summer he devoted several weeks to engineering technology to generate more accurate, detailed maps of the region faster for a fraction of the price. Specifically, he spent several weeks in the skies above McCall Glacier testing a new methodology he developed to create topographic maps of the area using DSLR cameras.
“I’ve figured out how to use standard DSLR cameras—like a Nikon D800, or whatever you like—to help create very high-precision, low-cost topographic maps from photos,” Nolan explained. “One big advantage to my method is being able to map with cheap cameras.”
The National Science Foundation supports his research.
Saving Time and Money
Time-series measurements and maps are important tools for scientists tracking glacier change. But often collecting such data involves expensive technologies like airborne LiDAR, a laser remote-sensing technology that scientists use to map landscapes and observe changes.
However, integrating LiDAR data collection in to a research project can cost anywhere from $400,000 to $1 million to purchase a unit suitable for mapping mountainous landscapes. Once acquired, the technology requires additional ground control data collection. This involves surveying fixed points—a boulder or a building, for instance—to tie the data to real world coordinates. This part of the LiDAR mapping process is labor intensive and adds yet another layer of expense.
Nolan’s innovative camera system can create a more accurate, higher-resolution map that does not require on-the-ground accuracy testing. Better yet, it costs roughly $15,000, an order of magnitude cheaper than LiDAR. It’s also a faster process; in the past, commercial vendors took months, or even years, to deliver LiDAR data of McCall Glacier to Nolan, who can now do it in-house in 24 hours, including data collection, processing and analysis.
“In some sense it’s just doing the same thing but better and more efficiently. But because it is so much better and more efficient it allows us to do brand new things,” he said.
A Test Run Over McCall
This summer, Nolan cut a small hole in the floor of his plane and mounted the camera to the bottom with the lens facing downward. He then connected the camera to an intervalometer, a device that activates the camera to take pictures at a set interval.
The next step was to plan the flight route to make sure he covered the area thoroughly and upload the points he wanted to cover in a hand-held GPS stationed on the dash of the plane.
“The other key part of the system is the camera is hooked up to a high-resolution GPS—a survey-grade GPS that has an antennae on the roof of the plane—and every time the camera fires it puts a marker on the GPS so that I know exactly where the camera was when the image was taken,” Nolan explained. “This is an essential part of the whole work so that I know each photo location in the air to within about 5 cm. And that is what allows me to do all this without any ground control.”
It took Nolan about an hour to fly over McCall Glacier and collect the photos he needed to make a map (by contrast, the flight to simply reach the glacier took four hours one way from Fairbanks). He started the data processing that night, and by morning it was completed.
Old Versus New: Comparing the Maps
A key test for the system was to see how the new maps produced by the photos taken with a DSLR camera compared to older maps of McCall Glacier that Nolan created using LiDAR data.
“The most exciting thing for me this summer was when I processed the data from the flight over McCall and compared it to LiDAR we had from five years ago. The results came back perfectly,” he said. “The glacier is changing, so to compare the accuracy you have to pull out the rock areas that don’t change. I did that and it was actually better than some of the LiDAR data I had from other years.”
One challenge Nolan faces with his method is managing the massive data files. But that challenge is not holding him back. Nolan is still testing and fine-tuning the camera system both in the air and on land whenever he has the opportunity.
Tests completed after the September trip to McCall Glacier at a nearby open-pit gold mine show just how well these maps can show change on a variety of time scales, even day-to-day changes at an active mine.
Nolan hopes his methods of making cost-effective, highly accurate maps can be applied to a variety of scientific pursuits ranging from coastal erosion and glacier retreat to tracking animal populations.
This fall and winter Nolan will test the system to see if the imagery it generates can measure snowfall. He also sees this tool being useful to National Park Service managers in Alaska who are always looking for ways to locate and track caribou. In fact, he recently made topographic maps of caribou herds with the camera system and sent it to park rangers and mangers, who were astounded.
“This is really game-changing in terms of the science. I can go out there now and make a map of McCall Glacier once a week!” Nolan said. —Alicia Clarke
For more information on Matt Nolan’s work at McCall Glacier and to see some of the maps created with this new methodology, visit: http://www.drmattnolan.org/photography/2013/