Is that bear fat or just chonky? Lasers may help scientists find out

Cusick was taking a break from work, standing on a platform along the Brooks River in Alaska’s Katmai National Park and Preserve — a world-famous spot for bear watching, as the giant mammals like to wade in the water to hang out and catch fish. Otis was a few hundred feet away, standing relatively still.

Cusick, who works for the National Park Service’s Alaska regional office, creates maps and trains people on GPS and the use of scanners in the field. He typically uses a laser scanner — specifically, a terrestrial lidar scanner — to measure the volume of stationary objects in the park like buildings and gravel piles. It’s a $70,000 industrial-grade tool that sits on a hefty tripod. That evening, Cusick aimed it at Otis, and took a scan.

Lidar is short for “light detecting and ranging” and is probably best known for its use in autonomous vehicles. A lidar scanner sends out millions of pulses of infrared light and measures how long it takes for them to return after hitting an object, such as Otis. These measurements form a point cloud that can then be used to build a three-dimensional map of the object.

In a matter of seconds, Cusick could see what looked like pinpoints comprising Otis’s rear on a tablet linked to the scanner. Computer software later processed the scan, creating a 3-D model that could be used to determine the width of the bear’s behind.

Cusick was excited; he didn’t think his experiment would work.

“I was like, ‘Wow, I got a return — I can measure the butt of Otis here!'” he recalled with a chuckle to CNN Business this week.

A point cloud of a bear that can be used to make a 3-D model.
That lightbulb moment for Cusick led to a multi-year effort to use a lidar scanner to unobtrusively estimate the volumes (and, through that, the weights) of some bears at Katmai as they fatten up for winter, which may help biologists understand more about the animals’ health (about 2,200 brown bears live in the park). In 2019 and 2020, bears were scanned around the time of Fat Bear Week at the park — an annual online competition, currently underway through Tuesday, where people can vote for the chunkiest (or chonkiest) brown bears as they get ready to hibernate for the winter. Though the lidar-assisted attempt at estimating the bears’ weights was halted this year due to the pandemic, which made fewer employees available to help with the process, Cusick hopes it will continue in the future.

It might sound simpler to use a scale to weigh bears, but it’s impractical in the wild, where it might get chewed on. (Plus, you’d have to lure the bears onto the scale and get them to stand still for about 11 seconds or so.) In spring, bears in the wild may be weighed by biologists who fly in via helicopter, tranquilize the bear, push it into a net, and lift it via a pulley system. But, in addition to its intrusiveness, the method may not be possible in fall, when bears have fattened up for the winter.

“A huge, huge upside of this method is it’s noninvasive; we don’t have to capture animals,” said Lindsey Mangipane, an Anchorage-based polar bear biologist for the US Fish and Wildlife Service. “It’s a lot less logistically challenging for us as well.”

Estimating a bear’s weight with lidar is much less disruptive, though it’s still very much a complicated guessing game. For instance, those taking the scans are only able to capture part of the bear, because there’s no way to put a scanner on the other side of the river, Cusick said. To make up for this, they may slice a bear scan in half and double the volume of the more complete side of the body to estimate the total volume of the bear. And while a 3-D scan can be used to determine the bear’s volume (how much space it takes up, in three dimensions), it doesn’t infer anything about its density, which is needed to help figure out how heavy it is.

Face recognition isn't just for humans — it's learning to identify bears and cows, too

To get a good guess at density, which may vary from bear to bear depending on gender, bone structure, time of year, and many other factors, Cusick spoke with biologists who study the animals. They figured it would be reasonable to estimate a bear is made up of 60% water and 40% fat, he said, so he and his colleagues used those percentages in 2019 and 2020 to tease out, from the scanned volumes, estimates of the weights for a number of bears at Katmai.

The overall process requires multiple people, Cusick said: one person performs the scan while another verifies the bear’s identity, for example. The weight estimates, which for a single bear may be made with numerous scans, are also made independently by separate people and compared.

A 3-D model of 747 is shown next to a picture of the bear; 747 was the 2020 winner of Fat Bear Week at Katmai National Park and Preserve, and had the highest estimated weight.

It’s currently impossible to know how accurate this method is, but Mangipane is hoping to validate the scanning technology by partnering with zoos to scan captive polar bears, whose weights are known since they can be trained to step on scales. Those known weights could be compared with weights derived from 3-D scans of the bears.

Accuracy aside, Cusick said the results have been consistent: Last year, for instance, he and his colleagues used about 10 scans taken over several days of a bear known as 747. Two people independently estimated 747’s weight from each scan, yielding 20 estimates that averaged 1,416 pounds. Each of the individual estimates were within about 100 pounds, Cusick said, and the average was an eight-pound difference from the bear’s heft in 2019.
747 was estimated to be the most voluminous bear last year, as well as the heaviest. He was also voted to the top of the bracket by Fat Bear Week voters.

While it will be impossible to know for sure which bear is truly the fattest this year, Cusick thinks 747 looks the biggest once more.

“We all have our favorites,” he said. “The anthropomorphism here is quite strong.”

Source link

Leave a Reply

Your email address will not be published. Required fields are marked *