The new buzz: Scientists are using weather radar to devise detailed bug maps

As meteorologists are all too aware, not everything that shows up on their radar is related to weather. Sometimes, it's a flock of geese or a traffic jam. In other instances, the suspicious storm turns out to be a mass migration of grasshoppers, millions of mayflies hatching at once or an angry horde of flying ants.

While forecasters normally try to remove the bugs from their data, a group of meteorologists are now joining forces with insect researchers to study them. The collaboration, called BioDAR, has an ambitious goal: Monitor levels of flying insect activity in near real-time across the entire United Kingdom. If things go well, they hope to scale this effort up to other countries.

By bringing the observational powers of modern meteorology to bear on bugs, the researchers hope to create a new pipeline of data that can inform basic science, pest management, and conservation efforts.

The idea for the project came about in December 2016, when University of Leeds insect ecologist Christopher Hassall and atmospheric scientist Ryan Neely were chatting at a gathering for academics from different disciplines.

"I work on bees, and Neely said, 'I see bees in my radar data all the time and we throw that data out,'" Hassall recalled.

"That moment of horror for me is I think what formed the beginning of the collaboration," Hassall said in an interview.

Neely agreed that it would be better to put the buggy radar data to use than to ignore it.

Three years later, the two have amassed a dozen collaborators and partners at institutions around the world. They've received nearly a million dollars in funding from the Natural Environment Research Council in the U.K. to spend the next three years developing algorithms that use radar data to characterize and track insect populations.

With some additional seed funding from the Bill & Melinda Gates Foundation, the researchers are also partnering with the national weather agencies of Rwanda, Mali and South Africa to try to use local radar data to track crop pests like fall armyworm.

In your typical radar setup, an antenna is used to shoot invisible microwave radiation through the air. As those microwaves strike particles, like raindrops, they bounce, creating echoes that are picked up by the radar. Those echoes reveal information on the number and size of the particles and their speed, all of which is fed through algorithms that tell forecasters what type of weather event they're looking at.

As Neely puts it, it's not such a giant leap to go from algorithms that classify rain, sleet and snow to ones that can characterize different types of insectoid blobs.

The work to achieve that is proceeding along several tracks at once. Insects come in many shapes and sizes, and to get a better handle on what all that diversity looks like on a radar display, Hassall's PhD student Thomas Dally is conducting microCT-scans of dozens of different specimens housed at the Natural History Museum in London.

"That allows us to create a 3-D map of the exoskeleton of the insect which we can then use to inform the algorithm they're using to train the radar systems," Dally said.

- - -

The researchers are also planning to catch bugs in the wild with the aid of a Helikite, a hybrid of a kite and a weather balloon. The kite, which will be launched from various locations across the UK over the next few summers, flies about 3,000 feet off the ground on a long tether.

A series of nets affixed along the length of the tether will catch insects from dawn to dusk. By matching what the Helikite ensnares with local radar scans, the researchers will be able to double check their classifications and learn more about the types of insects present at different heights.

It's also crucial to understand what radar will see if a big insect swarm emerges. This is why, in what has been aptly dubbed the "biblical apocalypse" stage of the project next summer, researchers will release about 100,000 flies into the air, training their radar instruments on the horde as it's unleashed.

"This is the key test of the method," Hassall said. "It'll allow us to see what the radar sees when a large number of animals enter the air column." While 100,000 flies may sound like a lot, Hassall assured that it's really just "a few shoe boxes" full, and that the environmental impacts of the bug drop are expected to be negligible. The worst-case scenario he could imagine was all of the flies dying during transport.

Neely's team will be feeding all of this field data into algorithms that tell insect researchers what their radar scans mean. While they're not going to be able to identify bugs down to the species level, Freya Addison, a weather radar PhD student working with Neely, will be using the size and shape of insect clusters to estimate the total biomass buzzing through the air - a vital and understudied aspect of insect ecology.

With the UK radar network scanning the nation's skies every five minutes, the goal is to produce a continuously-updated map of flying insects across the country.

"The ultimate aim is a UK bug map, a weather map of insects," Hassall said. The stretch goal? Handing off their algorithms to other countries to make that map global.

The researchers hope that such maps can assist in monitoring insect declines, which scientists have observed around the world in recent years and which have been tied to everything from the overuse of pesticides to climate change.

Christie Bahlai, an insect ecologist at Kent State University, said recent studies reporting declines often "talk about very specific taxa and extrapolate out."

"This would get more at a community level," Bahlai, who isn't involved in the project, said of BioDAR. "And so I'm really intrigued by this."

Manu Saunders, an insect ecologist at the University of New England in Armidale, Australia, cautioned that there's no silver bullet when it comes to understanding insect declines, and that there's only so much insect diversity we can monitor from the air. Saunders is not involved in the project.

"So, this method is mostly useful for detecting highflying insects that fly in large groups, for example, mass emergence of aquatic insects, or migratory insects like locusts, moths or hoverflies," Saunders said in an email. "It won't tell us anything about very small insects, solitary, ground-dwelling, subterranean, or flightless insects."

But Saunders did think the project had the potential to "contribute more knowledge of spatial and ecological aspects of insect migration," which could help researchers understand the value of these ecosystems.

Ultimately, the collaboration might even help forecasters sleuth out the weird shapes that pop up on their radar screens. Neely is crowdsourcing the weather community for examples of various insect clusters so that he can start building out a catalog.

"We have a meteorologists' handbook for hail, snow, and rain," he said. "Hopefully we can now say if radar ranges are this, this is definitely butterflies, ants, or bees."

Perhaps, this will even allow the National Weather Service to issue bug warnings so that people can take cover before a grasshopper storm descends.