DECO and IceCube summer 2015 student internships

Monday, August 24, 2015 - 11:00am

The Distributed Electronic Cosmic-Ray Observatory (DECO) is a citizen science project that allows users around the world to detect cosmic rays and other energetic particles with their cell phones and tablets. Four high school students joined the DECO development team at WIPAC for a seven-week internship. This internship was supported by the UW–Madison QuarkNet Center hosted at WIPAC.

A fifth student worked as an intern for IceCube with support by a UW–Madison Hilldale professorship.

DECO interns
Summer 2015 interns and mentors (left to right): David Schultz, Peter Karn, Ilhan Bok, Noah Cagnazzo (back), Silvia Bravo (front), Felipe Campos (back), Heather Levy (front), Alex Diebold, Matt Meehan, Justin Vandenbroucke, Paolo Desiati. Image: Jamie Yang, WIPAC

While the five of us worked in the same environment, individually we had separate tasks and objectives. Four of us worked collectively on the DECO project: Ilhan Bok, senior at West High School, Felipe Campos, junior at Collegiate School in Richmond, VA, Alex Diebold, senior at Evansville High School, and Heather Levy, senior at Middleton High School. One of us, Noah Cagnazzo, junior at James Madison Memorial High School, partnered with senior scientist Paolo Desiati to work with IceCube cosmic-ray data.

We all began the internship in June with varying levels of knowledge about DECO and IceCube, computer programming, and particle astrophysics. Some of us had begun programming as young children, while others knew nothing coming into the first day.

students
Felipe, Ilhan, and Heather examine the properties of particles detected by the DECO app. 
Image: Jamie Yang, WIPAC

Our work on the DECO project covered two main areas: classification of particle-interaction events, such as cosmic-ray muons or electrons from radioactive decays, and development of the iOS app for DECO.

So far, the Android version of the DECO app has produced tens of thousands of events, and classifying all of them by eye would prove both inefficient and tedious. With an effective algorithm developed by Ilhan, an efficient machine learning program built by Alex, and a filtering system for hotspots and noise made by Heather, we can begin to analyze these candidates with high accuracy. However, there is still much work to do in finalizing such a program for use on large portions of data.

As for the iOS development, there were two primary goals beyond having DECO run on iPhones: to increase livetime in the current version in order to increase the detection of cosmic-ray events and to provide high-quality, thoroughly analyzed images in order to make server-side analysis a more efficient process. Felipe made fantastic progress, achieving livetime over 99 percent and implementing efficient and effective device-based image processing, analysis, and filtering. The application is now ready for beta testing.

Students working
Noah and Paolo discussing cosmic ray anisotropy maps in preparation for a group presentation. Image: Jamie Yang, WIPAC

For IceCube, the project consisted of building sky maps of the arrival directions of cosmic-ray muons detected in IceCube to show how the cosmic-ray anisotropy changes with time. Noah produced maps that allowed looking at two different co-existing components of the cosmic-ray anisotropy: the one produced by the movement of the Earth around the Sun, which is well understood and varies throughout the year, and the one produced by magnetic fields in the Solar System and beyond, called sidereal anisotropy and whose origin is unknown. The variations of this second component are one of the hot topics in high-energy cosmic-ray research. These two sources of anisotropy interfere with each other over the course of a year due to the different reference systems they originate from. The study of yearly modulations of the observed cosmic-ray anisotropy can shed light on time variations that are not related to such interference. The maps produced by Noah showed that the sidereal anisotropy remains very stable throughout the year.

As we neared the end of the internship, we each took a moment to sum up our experience.

Noah: “I decided to work at IceCube because I love science. I had heard about the experiment before and I thought it was very interesting. During my five weeks at IceCube, I learned how to program and make sky maps of the cosmic-ray anisotropy in different reference systems. I also met a lot of nice people during my time here. It was an amazing experience and I hope to do it again.”

Heather: “I was familiar with the DECO project from my previous internships with WIPAC. When my advisors presented the opportunity for me to work at WIPAC over the summer, I was thrilled! I had little experience with programming, so I was excited to learn Python during the first week. Throughout the seven weeks of my internship, I got to discuss ideas with intriguing post-docs and scientists, present my research in a formal setting, and work simultaneously with some of the smartest high schoolers that I know. This project sparked my interest in exploring a degree in physics and research. To all of the sponsors of our grants and the staff at WIPAC: thank you for allowing us to have this incredible learning experience!”

Ilhan: “DECO was my first experience that gave me a taste of the difficulties and rewards of scientific experimentation in the real world. Working on event classification at DECO was also my first experience doing paid work, teaching me the merits of collaborative work, even though I worked on my own for most of the time. I started the internship with a good amount of knowledge in programming overall, but minimal knowledge in both Python and bash, which rapidly grew through the course of the program. I want to thank the organizers of this internship for providing me with skills I will continue to use regardless of my career path.”

student presenting
Alex during the presentation of the DECO and IceCube internship projects at WIPAC. Image: Jamie Yang, WIPAC

Alex: “I was originally a part of the team to make the iOS version of the DECO app, but eventually transferred over to the classification of events. I focused primarily on utilizing a form of artificial intelligence called machine learning to efficiently and effectively classify the events. I had some prior programming knowledge before this internship, taking Pascal and C++ programming classes offered at my school and attending a two-week summer camp at Purdue University to learn Python. However, the vast amount of experience I gained on this project using the Mac terminal and Python is exponentially greater than my previous knowledge. I am deeply grateful for this internship as it has taught me many skills needed to be a programming asset for companies.”

Felipe: “Having held a deep interest in physics and programming from a young age, I decided to apply for the DECO internship. With my current programming experience spanning five languages (including Python, C#, and C++), interest in gaining experience in Apple’s new language, Swift, along with understanding how the DECO application worked on a hardware and software level, I worked on developing DECO on the iOS platform. The freethinking and open environment at WIPAC allowed me to spend my free time expanding my knowledge and discussing topics of interest with others around the workspace. The seven weeks I spent at WIPAC provided me with an amazing experience, which deepened my love and appreciation for scientific thought. I would like to thank our sponsors and those who helped over the course of the internship for allowing us high schoolers to have this overall fantastic experience.”

We would like to thank our mentors at WIPAC—Professor Justin Vandenbroucke, Matt Meehan, Peter Karn, Paolo Desiati, Silvia Bravo, and Matthew Plewa—along with QuarkNet, Hilldale professorships, and APS for supporting the grants that made this internship possible.