Volunteer users will watch a short video on the left. Then, based on the shape and timing of the colored bubbles, they will be asked to pick the corresponding category from among the five options on the right. Credit: IceCube Collaboration/Zooniverse
Every second, about 100 trillion neutrinos pass through your body unnoticed. At the South Pole, the National Science Foundation-funded IceCube Neutrino Observatory uses a kilometer-wide array of sensors to detect these elusive particles. The team of more than 300 international researchers who are working to identify their astronomical origins to help unlock mysteries of the universe. Such an undertaking requires a massive amount of data collection, with one terabyte of data recorded daily by the observatory. But organizing the data can be labor intensive. This is where the public can help.
Starting today, volunteers from anywhere can participate in the “Name that Neutrino” project led by IceCube researchers at Drexel University's College of Arts and Sciences, which asks users to help categorize IceCube data. Through the Zooniverse platform, volunteers can participate using their own computer or smartphone. “Name that Neutrino” is open to everyone and will run for about 10 weeks.
“We created this project so that people can get involved in our research and learn about our exciting work,” said Christina Love, PhD, an associate teaching professor in the College. “We are really hopeful that this project can help us uncover a new way of looking at the sky: instead of the light from the stars, we are seeing the particles!”
When a neutrino interacts with a molecule in the ice, secondary charged particles emit light that creates a signal or light pattern that can be used to determine the neutrino’s energy and direction. The IceCube sensors collect this light, which is subsequently digitized and time stamped.
Volunteers will first be given a brief tutorial of the different categories of the signals and what they might expect to see in the video. Next, users will watch a short video and pick one category (out of five) matching that signal. By categorizing the data, volunteers are helping IceCube researchers sift through the background signals and determine what type of particle caused each signal.
This project will offer valuable information about how to better understand what the universe looks like in neutrinos. IceCube has recently developed machine learning algorithms that can identify the different types of signals happening in IceCube. The machine learning algorithms, and data analysis, will improve with the assistance of citizen scientists simultaneously processing the data and eventually comparing it to the performance of the machine learning algorithm.
“We use AI and machine learning to try our best, but I'm excited to see if citizen scientists can help us even more,” said Naoko Kurahashi Neilson, PhD an associate professor of Physics in the College.
The volunteer classifications will then be compared to machine learning predictions, which will improve the fidelity of the data and provide even better results.
“I believe that citizen science and AI can jointly contribute incredible things to research,” said Elizabeth Warrick, a graduate student in the College. “This project from IceCube is another opportunity to showcase how much we can learn about the universe from all perspectives.”
Want to get involved? Here’s how:
- Click on the link: https://www.zooniverse.org/projects/icecubeobservatory/name-that-neutrino
- Click “Get Started” to begin.
- Click “Tutorial” to learn about how to classify signals.
- Watch the brief video and pick one of the five categories for signals.
- Check out the "Field Guide" for more examples and information.