Frank McNally

Carleton College
mcnallyf (at) carleton (dot) edu


Introduction -- Research -- Cosmic Rays -- IceCube and IceTop -- Simulating Tank Response to Particle Showers -- Results -- Summary



Simulating Light Production in IceTop

Simulating Light Production in IceTop



Well, you've found your way to my webpage. I'm a rising senior physics major at Carleton College in Minnesota. Want more information? Too bad.


Introduction

About Me

This should probably be an introduction about myself...

Research

This summer I worked under Stefan Westerhoff and the IceCube Neutrino Observatory. Stefan's work centers around the changing composition of the cosmic ray spectrum at its two critical points. By working with both the Pierre Auger Observatory in Argentina and IceCube, Stefan is able to look at both of these sections in detail. My work was somewhat more concrete -- using a program called TankTop I simulated the production of light by cosmic ray byproducts in IceTop tanks.


Cosmic Ray Basics


What Are Cosmic Rays?
[balloon]

Don't judge a book by its cover: "cosmic rays" are not rays at all. The phrase is actually kind of a catch-all term for high-energy charged particles striking our atmosphere. They're mostly protons, but can be heavier nuclei as well.

Cosmic rays were discovered in 1912 by Victor Hess. He was trying to test the theory that most of the Earth's radioactivity came from the ground, so he took three detectors up to 5300 meters in a hot air balloon, expecting a decrease in the ionization rate. Instead, he found an increase, indicating the primary source of radiation was coming not from below, but from above. Hess thought that he was observing high energy photons, hence the name "cosmic rays".


Cosmic Ray Showers
[Particle Shower Breakdown]

When a cosmic ray strikes a molecule in the atmosphere, it produces a cascade of lighter particles. Initial collisions produce exotic particles like pions, but by the time the shower has reached the ground, these particles have decayed, and we're generally left with several common byproducts -- muons, electrons, neutrons, neutrinos, and gamma rays.

Since cosmic rays interact in our atmosphere, it seems that all of our cosmic ray detectors should be in balloons so we can detect them before they create particle showers. So why build detectors on the ground? Statistics: as the energy of the incoming cosmic rays increases, their frequency decreases. There are two important points of study on the spectrum where the slope of flux as a function of energy changes -- the "knee" and the "ankle". These points of interest have fluxes of roughly 1 particle per square meter per year and 1 particle per square kilometer per year, respectively.

Detecting one particle per year obviously is not going to provide enough data, so we need to take an alternate approach. To maximize the flux detected, we use very large ground detectors to indirectly detect cosmic rays that interact in the atmosphere. Instead of detecting the incoming particle itself, cosmic ray detectors look for the byproducts produced in a particle shower and use them to infer information about the original. IceTop is one such detector.

[Cosmic Ray Spectrum]
Cosmic Ray Spectrum


IceCube and IceTop

The Project

IceCube was not built as a cosmic ray detector. It was designed to detect neutrinos, and, when finished, will have 4200 detectors buried deep in the ice (anywhere from 1450 to 2450 meters) dedicated to this task. These detectors (Digital Optical Modules, or DOMs) use photomultiplier tubes to detect light produced by collisions between neutrinos and particles in the ice. As you can see in the picture to the right, the DOMs are located on strings arranged over a roughly hexagonal area. When IceCube is finished, this area will be about 1 square kilometer.

What you can't see is that at the top of every string are two tanks built to detect cosmic rays. These tanks combined form IceTop, a partner project with IceCube. IceTop takes advantage of IceCube's large surface area and grid-like pattern to detect cosmic ray showers. IceTop's arrangement allows for the observation of cosmic rays with energies ranging from 10^14 to 10^17 eV, which is right around the "knee".

[IceCube]
Inside the Tanks

Below is a crude illustration of what happens in an IceTop tank. In this side-on view, you can see that each tank has two Digital Optical Modules (DOMs). These are essentially photomultiplier tubes (PMTs); they output a pulse of electrons dependent on the intensity of detected light. Tanks are ~1 meter tall, with 90cm of it being filled with ice, and the remaining 10 with air. In this case, a muon enters from above. First it penetrates the snow layer, then into the tank itself. Cosmic ray shower particles are capable of producing light in many ways, from bremsstrahlung ("braking radiation") to Cerenkov radiation. Light emitted in the tank will be reflected around the tank until it reaches a DOM.

[IceTop Tank]


Since this is the way in which we detect cosmic rays, it becomes very important to know how much light is produced by different particles of different energies entering these tanks, and also at what point they stop producing light. This is where I come in.


Simulating Tank Response to Particle Showers

The Goal

The goal of my research was to see at what energy the tanks stop detecting particles from a cosmic ray shower. This is a very complex problem because of the variables involved. First, we're detecting muons, electrons, neutrons and gamma rays, all of which stop producing light at different energies. Second, each of these particles can enter at a different angle or location in the tank, also affecting light production. Finally -- the focus of my work -- the depth of the snow layer on top of the tanks can strip energy from the particles. So, the primary focus of my work was to look at how the light cutoff changes with snow layer depth.

Tanktop

In order to accomplish this, I relied on three programs: Geant4, Tanktop, and ROOT. Geant4 (GEometry ANd Tracking) is a platform developed by CERN to simulate the passage of particles through matter using Monte Carlo methods. Tanktop is built on top of Geant4 and developed specifically for IceTop. ROOT is a data analysis tool also developed by CERN that allowed me to look at my data. So, the standard process for a run looked like this.

1. Write a steering file for Tanktop.
        Steering files dictate the tank setup (dimensions, reflectivity, etc), incoming particle information(particle type, energy, incoming angle...), what information to record, and where to output it

2. Run Tanktop for given steering file.
        Here I would choose the number of incoming particles to simulate

3. Look at output data using ROOT

Results
Summary