On a cold night in November, I bundled up in my coat and hat and gloves, and lugged a telescope into the backyard. I’d been given a serious mission, but first I allowed myself a few minutes of fun. I used the telescope to observe the Cigar Galaxy, the Dumbbell Nebula, Orion’s Belt. There they were, gleaming brightly in gold and blue and pink, steady friends in a painfully unsteady winter.
Then I got to work. I moved the telescope across the sky until I found my target: a near-Earth asteroid. Scientists didn’t know much about it except that it would threaten life on our planet if it ever collided with us. So they’d asked amateurs like me, who happen to live somewhere with a view of the passing asteroid, to observe it and record data on it.
I was using a new smart telescope called the Unistellar eVscope, which digitally strips out light pollution to produce stunning images and allows backyard astronomers to upload their data to scientists at the Search for Extraterrestrial Intelligence (SETI) Institute. It was SETI that had sent me an alert asking me to please go out and look at the asteroid.
It didn’t look terribly impressive — just a blob tracing a line across my eyepiece as it flew through space. But I captured the data anyway and sent it to the scientists to analyze. It was after midnight when I finally went back inside and fell asleep wondering: What were the chances I’d discovered something that would actually help science?
I’m not alone in pursuing this kind of activity nowadays. The pandemic has driven a huge increase in participation in citizen science, where people without specialized training collect data out in the world or perform simple analyses of data online to help out scientists.
Stuck at home with time on their hands, millions of amateurs around the world are gathering information on everything from birds to plants to Covid-19 at the request of institutional researchers. And while quarantine is mostly a nightmare for us, it’s been a great accelerant for science.
:format(webp):no_upscale()/cdn.vox-cdn.com/uploads/chorus_asset/file/22215589/GettyImages_1221588449t.jpg)
Early in the pandemic, a firehose of data started gushing forth on citizen science platforms like Zooniverse and SciStarter, where scientists ask the public to analyze their data online. It’s a form of crowdsourcing that has the added bonus of giving volunteers a real sense of community; each project has a discussion forum where participants can pose questions to each other (and often to the scientists behind the projects) and forge friendly connections.
“There’s a wonderful project called Rainfall Rescue that’s transcribing historical weather records. It’s a climate change project to understand how weather has changed over the past few centuries,” Laura Trouille, vice president of citizen science at the Adler Planetarium in Chicago and co-lead of Zooniverse, told me. “They uploaded a dataset of 10,000 weather logs that needed transcribing — and that was completed in one day!”
Some Zooniverse projects, like Snapshot Safari, ask participants to classify animals in images from wildlife cameras. That project saw daily classifications go from 25,000 to 200,000 per day in the initial days of lockdown. And across all its projects, Zooniverse reported that 200,000 participants contributed more than 5 million classifications of images in one week alone — the equivalent of 48 years of research. Although participation has slowed a bit since the spring, it’s still four times what it was pre-pandemic.
Many people are particularly eager to help tackle Covid-19, and scientists have harnessed their energy. Carnegie Mellon University’s Roni Rosenfeld set up a platform where volunteers can help artificial intelligence predict the spread of the coronavirus, even if they know nothing about AI. Researchers at the University of Washington invited people to contribute to Covid-19 drug discovery using a computer game called Foldit; they experimented with designing proteins that could attach to the virus that causes Covid-19 and prevent it from entering cells.
Trouille said she’s not at all surprised to see so many people flocking to citizen science. Years back, Zooniverse surveyed participants to understand why they volunteered to spend so many hours sifting through data. Their top three reasons were wanting to contribute meaningfully to science, wanting to enjoy a pleasant distraction from everyday life, and wanting to be part of a supportive community.
Quarantine has intensified those drives. Many of us have longed to feel a sense of purpose, which experts on solitude say is vital for helping people get through isolation. We’ve often craved a distraction from the anxiety spiral of reading news about the coronavirus (not to mention politics and the economy), and looking at beautiful pictures of stars or animals can offer that. And we’ve all missed hanging out with people.
“People are just really needing to connect in meaningful ways,” Trouille said. “Welcoming online communities provide a wonderful outlet for that.”
If you’re new to citizen science, you might wonder how amateurs can start analyzing data without fear that they’re going to get it all wrong and ruin a scientific project. I definitely had this fear when I first started dipping my toes into it.
But it works because scientists break down complex tasks into micro-tasks that people without specialized training can do. Even if people get some classifications wrong, that’s fine because scientists aggregate many people’s classifications (Zooniverse gets 25 to 45 people to classify every image), relying on “wisdom-of-the-crowd” algorithms to arrive at an accurate result. A 2018 analysis showed that data collected by volunteers and scientists agreed 96 percent of the time.
So, for instance, when I opened the Zooniverse app on my phone and clicked on a project called Disk Detectives, it gave me a two-minute tutorial on how to identify stars with certain kinds of disks around them; these stars are more likely to have planets. Then I swiped through a bunch of images, jotting down whether I saw the disk or not. That’s all there was to it.
It’s so simple that at this point you’re probably wondering: Why don’t scientists just get computers to do this? Surely this is a task for AI?
Yes and no. Part of what people are doing when they classify data on these platforms is creating the training sets for AI — they’re teaching computers to recognize the patterns that they, the humans, can see. But, Trouille said, “Part of what the humans are doing that the AI will never be able to do is to identify the unusual or the weird.”
A delightful example of this arose on Galaxy Zoo, one of the astronomy projects on Zooniverse, soon after it launched in 2007. A participant noticed an unusual-looking bright green object and posted about it on the forum, asking if anyone had seen anything similar. Soon a whole discussion thread sprang up titled “Give peas a chance,” with participants calling themselves the “Peas Corps.” Together, they collected more than a hundred of these celestial objects.
Scientists eventually realized these were a special type of galaxy that had never been found before. They represent an important stage in galaxy evolution. In 2009, the scientists published a paper on them in the Monthly Notices of the Royal Astronomical Society, crediting 10 citizen scientists from Galaxy Zoo for their contribution to the research.
“Because the research team didn’t know that type of galaxy existed, there would have been no reason to train the computer to look for that,” Trouille explained. “So you would never have found that unless you had hundreds of thousands of people looking at all that data and just able to flag the unusual and bring it to the discussion forum.”
She believes AI will become increasingly integrated into citizen science platforms, but that human abilities will always remain a valuable complement to it.
Katie Cohen, who researches citizen science at the UK’s Centre for Science and Policy, is a bit more skeptical about the irreplaceability of human pattern recognition. “With these kinds of [online] crowdsourcing exercises, I don’t know if they necessarily have long-term potential because technology is improving so rapidly,” she told me. “Whereas I think the community-oriented programs might be where more of the potential lies — because lived experience can’t be programmed through AI.”
She was referring to the citizen science projects that ask volunteers to go out into the field, whether to count birds in their backyard, test air quality in their neighborhood, tag monarch butterflies migrating through their city, or observe asteroids zipping across their patch of sky.
Franck Marchis, a SETI scientist and co-founder of Unistellar, explained why it’s so valuable to have multiple eyes scanning the heavens, rather than relying only on the few telescopes that NASA uses to detect asteroids.
“If you only have a telescope in Hawaii and it’s cloudy there, you’re not going to have observations,” he said. “And that’s happened already. We had an asteroid that we basically discovered at the last minute because for a week it was cloudy in Hawaii and nobody saw it. So the idea of our program is to have a worldwide network that’s coordinated … a network that’s capable of observing the sky 24/7.”
At a time when public trust in experts has declined precipitously, some scientists are arguing that a more collaborative research model can help counter science skepticism. And governments are increasingly realizing that it may be worth investing in citizen science — not only to accelerate scientific research, but also to boost trust in it. Both the US and the European Union currently fund major programs in citizen science.
“I do think it has a lot of potential to really improve trust and relationships,” Cohen said. “That’s really why a lot of the policymakers that we work with are really interested in it across various UK government departments.”
The big question, she said, is what type of public involvement is most effective at improving trust. Most citizen science projects use volunteers for simple data collection or analysis aimed at helping professional researchers. But some seek to involve volunteers on a deeper level, putting them on almost equal footing with the researchers. One example of that, Cohen said, would be asking tobacco users to actually help design a study on how to minimize tobacco use.
An even more extreme example would be a “Citizens’ Assembly,” a group of randomly selected citizens that makes recommendations to their government on how to handle issues like the climate crisis. The idea of an assembly like this has gained ground in places like France and the UK. But this is definitely a more labor-intensive way to engage citizens than, say, a Zooniverse project, and it remains to be seen how well it’ll work.
In the meantime, many scientists, such as Marchis, don’t want the public to perceive them as elitists off in some ivory tower. “When science is not part of your life, it’s very easy to become against it because you don’t really understand exactly what those people called scientists are doing,” the astronomer told me. “My personal goal is that people consider science to be part of their life — not something you can do only if you have a PhD.”
I certainly don’t have a PhD in astronomy, but this winter I got to feel the thrill of participating in scientific discovery.
That night in November that I spent observing the near-Earth asteroid from the backyard? Turns out, 26 other amateurs — from America to Austria, from Finland to France — also peered through their telescopes in their backyards that night. Together, we obtained decisive data that allowed astronomers to model the asteroid’s shape. When they shared the shape with us in December — it looked like a potato — a huge grin spread across my face.
I was happy in part because I knew that scientists need this kind of data to divert asteroids from impacting Earth. But it was more than that. During the pandemic, I’ve barely been able to travel anywhere or see anyone. Yet somehow I’ve also traveled to an asteroid as it zooms past the planet, and to the Cigar Galaxy and the Dumbbell Nebula and Orion’s Belt, and I’ve connected with scientists and strangers who are trying to understand our universe a bit better and who are, despite the suffering that Covid-19 has wrought, finding this world to be full of beauty and awe.
Sign up for the Future Perfect newsletter. Twice a week, you’ll get a roundup of ideas and solutions for tackling our biggest challenges: improving public health, decreasing human and animal suffering, easing catastrophic risks, and — to put it simply — getting better at doing good.
source https://www.vox.com/future-perfect/22177247/citizen-science-amateur-backyard-birding-astronomy-covid-pandemic
Post a Comment