Robert “Rocky” Rockwell just completed his 47th summer studying snow geese and their impact on the fragile Hudson Bay coastline in Wapusk National Park near Churchill, Man., but this year’s field season took the research to new heights.
As part of the Hudson Bay Project, a collaborative research program that includes partners from the U.S. and Canada, staff and students from the UND Biology Department spent several weeks at Rockwell’s research camp on La Perouse Bay east of Churchill demonstrating how unmanned aircraft can be used to monitor the overabundant geese and the damage they’re causing to the remote tundra landscape.
Where researchers historically have surveyed the area on foot, walking transect lines to count geese and measure vegetation while risking encounters with polar bears and grizzly bears, the UAV provided detailed imagery in a manner that not was only safer, but more efficient with less impact on the wildlife, Rockwell said.
UAS is the common name for unmanned aircraft systems, while UAV refers to unmanned aerial vehicles.
“I was amazed,” said Rockwell, an ornithologist at the American Museum of Natural History and professor at City University of New York. “This is the wave, it really is. We can do (research) more extensively, multiple times during the season.”
Leading the UND team on its trip to the tundra were Susan Ellis-Felege, a wildlife ecologist and assistant professor; Robert Newman, an associate professor and director of graduate studies; Christopher Felege, biology instructor; graduate student Andrew Barnas; undergraduate student Sam Hervey; and Michael Corcoran, a veteran military helicopter pilot and UAS expert who served as operations manager for the UND unmanned aircraft team.
Also part of the project were Travis Desell of the UND Computer Science Department and Marshall Mattingly, a graduate research assistant in the Computer Science Department.
Ellis-Felege said the Hudson Bay collaboration began more than a year ago when Mike Johnson, a longtime waterfowl biologist for the North Dakota Game and Fish Department who retired earlier this year, helped arrange a trip for her to Rockwell’s research camp.
Ellis-Felege and Newman had been exploring options for using UAS in their research, and the remote Hudson Bay coastline seemed like a perfect setting.
“She came up for 10 days, and by the time she left, she was full of more ideas than she will ever be able to do,” Rockwell said. Key among them, he said, were UAS and placing video cameras on goose and eider duck nests to get video and audio recordings of the birds’ behavior and how they responded to the unmanned aircraft.
“I was intrigued by both, and we managed to put funding together,” he said.
Ellis-Felege said the project has been a big collaboration.
“Between nest cams and unmanned aircraft, we ended up planning this project out,” she said. “Our team rose to the challenge.”
The UND Biology Department in the past year purchased a Trimble UX5 UAV, which Ellis-Felege describes as a 5½-pound Styrofoam aircraft with an expensive brain and a Sony camera inside the nose bay. The survey-grade UAV launches using a catapult and can be pre-programmed to fly transect lines and take photographs over specific areas at one-second intervals, stitching the imagery together into a mosaic that provides a detailed picture of the ground below.
The UAV had to be within view at all times, she said, and was in the air 35 to 40 minutes for each flight. Each mosaic covered about half a square kilometer, and the aircraft flew at heights ranging from about 245 feet to 395 feet.
“Essentially, you take a computer and pre-program it and say, ‘I want it to fly this box, and I want it to start here,’ ” Ellis-Felege said of the UAV. “It launches with a catapult, and it swings the aircraft out, and it goes and flies these transects, predetermined white lines that allow it to take a picture about every second and with a high amount of overlap so it can fix together a map with all the images.
“It then comes back and lands in a predefined location.”
Operating the UAV requires certification, and Ellis-Felege, Newman, Barnas, Christopher Felege and Corcoran in March took a weeklong course from Trimble in North Carolina to learn how to operate the aircraft and become certified in using it.
Rockwell, as a lead researcher on the Hudson Bay Project, helped secure the required paperwork from Canadian officials to operate the UAV in a national park and alleviate their concerns about how the aircraft would be used.
“Susan convinced me to start with a science plan with UAV,” Rockwell said. ” Not just fly it because it’s cool—let’s have a plan of action.
“My bigger contribution was helping her work through the various hurdles to convince a national park that’s a little dubious about unmanned vehicles. Most national parks have had issues with people flying quadcopters in national parks because they can chase wildlife.”
The Trimble isn’t a quadcopter and can’t hover over animals.
With a background as a military helicopter pilot and UAS expert, Corcoran came to the project from a different perspective than the UND Biology Department crew. Working at La Perouse Bay, he programmed the flight paths and provided technical support, cross-training the biologists to be flight crew members while they trained him to be a biologist.
During two trips to Hudson Bay in June and July, the UND crew conducted 87 flights, logging more than 54 hours of flight time and producing more than 80,000 aerial images.
“You could see a gosling. You’re seeing small rocks on the ground,” Corcoran said. “We did that for the duration of our time up there.”
Because of the aircraft’s limited flying capacity, the researchers collaborated with a helicopter company to transport the UAV and launcher to more distant locations, including an area about 50 miles southeast of camp that Rockwell described as “dicey” because of its high bear population.
While the UND crew worked with the UAV, Rockwell stood guard with a firearm in case bears came too close.
“That experiment was totally successful,” Rockwell said. “It’s an area becoming actively degraded, and we got some really good mosaics from down there.”
The UAV appeared to have no impact on the geese, which largely ignored the aircraft when it passed. It’s only speculation for now, Rockwell says, but it’s possible the geese don’t see the UAV as a predator and therefore not a threat.
“The geese just sit there and watch it,” he said. “They don’t care. It was utterly amazing.”
He said the response bodes well for using UAS to learn more about nesting and nest failure, especially the increasing predation on geese and nests by bears.
“Normally, we go out and stomp around the colony, which disturbs things, and make a guess based on ancient data that most nest failure occurs during early incubation,” Rockwell said. “But we’re getting higher rates of nest failure later in the incubation, and we know that is the polar bears and grizzly bears. With this UAV approach, we can monitor that and fly the aircraft over certain study areas, say, every five days. We can clearly see nests, birds sitting on their nests and see eggs without any trouble.”
Ellis-Felege said it’s too early to draw any conclusions about the impact of the UAV on nesting waterfowl, but the success of this year’s inaugural experiment ensures unmanned aircraft and UND’s role in that research will remain a key tool on the tundra.
Besides sampling waterfowl and vegetation, the UAV shed light on polar bears, fox dens and areas of historic and cultural significance.
“For the first time, I think we have the proof of concept this is a tool that can be valuable to survey waterfowl populations and understand some things about their habitat,” Ellis-Felege said. “I think it has potential for a lot of mammals, even though it wasn’t our primary target.”
Corcoran, who works with defense and remote sensing programs at the UND Energy and Environmental Research Center, said this summer’s initial UAV effort on Hudson Bay ranks among the most satisfying projects on which he’s worked. The collaboration with the Biology Department, the UND College of Arts and Sciences and the Hudson Bay Project, he said, allowed him to enjoy his job at UND for the first time in several years.
“Not only did we achieve what we initially set out to as project goals, we discovered that systems like this are extremely capable of producing relevant data for wildlife and environmental assessments, and it’s less intrusive and safer than sending that biologist or manager into the park,” he said. “My hope is also simple—all of the departments at UND should look at this as an example of what ‘right’ looks like for UAS research programs.
“If we really want to lead the nation with the use of UAS, this is how it will happen. I know that Biology Department is just getting warmed up.”
Others on the UND team shared that sentiment.
“It was an amazing experience,” Ellis-Felege said. “We were able to answer a lot of questions, and certainly we are only at the beginning of what you can do.”
Barnas, the UND graduate student, described working on the tundra and the ever-present risk of bear encounters as a “surreal” experience.
“But at the end of the day, we are working in the bear’s territory, and it all comes down to a mutual respect between bears and humans,” Barnas said. “The entire focus of this operation is to be as noninvasive as possible, and that is a theme we take very seriously when working in bear country.”
Rockwell said UAS will allow the researchers to extend their field season into August and September when bear problems increase, helping them “answer some important questions that have never been answered.”
He credits Johnson, the retired North Dakota Game and Fish Department biologist, for arranging the UND Biology Department collaboration.
“This really is a fulfillment of one of his dreams—he’s the one who put this marriage together,” Rockwell said. “I’m just totally impressed with the system and certainly impressed by Susan.
“I think Grand Forks and the university there should be totally proud of the Biology Department and Susan and her team.”