The University of Wisconsin–Madison Department of Physics, in collaboration with PGSC, is pleased to host the first ever Physics Department Special Colloquium for Graduate Student Research on Friday, March 31, 2023!
|2:35 PM||Vedant Basu & Jessie Thwaites||Chasing Ghosts: neutrino astronomy at the South Pole with IceCube|
|3:15 PM||Jimena Gonzalez||Searching for strong gravitational lenses in the Dark Energy Survey (DES)|
|3:35 PM||Roark Habegger||Cosmic Ray Buoyancy in the Interstellar Medium|
|3:55 PM||Break – food available|
|4:10 PM||Yeqing Zhou||Quantum Approximate Optimization Algorithm with Adaptive Bias Fields|
|4:30 PM||Sam Norrell||Quantum Computing and Error Correction with Neutral Cesium and Rubidium Atoms A|
|4:50 PM||Zain Abhari||X-FAST: A New XUV Femtosecond Absorption Spectroscopy Tabletop Instrument|
|5:30 – 6:30 PM||Graduate School Panel||Q&A with current Department of Physics graduate students|
Speakers & Presentations
Vedant Basu, Physics PhD Student, Astroparticle Physics (top)
Jessie Thwaites, Physics PhD Student, High EnergyAstrophysics (bottom)
Chasing Ghosts: neutrino astronomy at the South Pole with IceCube
In this talk, we’ll introduce the IceCube neutrino observatory, and how we can use neutrinos to probe the highest energy astrophysical processes in the universe. We’ll start by discussing IceCube, the currently operational cubic-kilometer detector in the ice, and R&D on planned new hardware for IceCube-Gen2. We’ll talk about IceCube’s measurements of the diffuse astrophysical neutrino spectrum, shedding light on models of particle acceleration and neutrino production in some of the most violent phenomena in the universe. Although IceCube has discovered this diffuse flux, the sources of the vast majority of those neutrinos is still unknown. We’ll discuss how we can use information from multiple messenger particles to search for the sources of astrophysical neutrinos.
Physics PhD Student, Observational Cosmology
Physics PhD Student
Searching for Strong Gravitational Lenses in the Dark Energy Survey (DES)
In strong gravitational lensing (SL), light from a distant source galaxy is deflected by a massive foreground lens galaxy, producing multiple magnified and distorted images. SL systems can be used to study the nature of dark matter and dark energy, and to improve measurements of the Hubble constant as a complementary probe. One of the primary challenges in this field is to develop SL search techniques that achieve sufficiently high completeness and purity for practical application on datasets with billions of galaxies. Present-day methods use machine learning (ML) models to extract an initial sample of candidates that needs to be visually inspected by humans to identify a final set of candidates, which often only contains about 1% of the sample returned by the ML algorithm. In order to reduce the amount of human time required in future searches, we need to design better automatic search techniques that generate samples of post-ML candidates with higher purity rates. Here we present an ML search technique based on the attention mechanism and implemented as a multi-class classification problem. We create specific training classes for various types of common false positives. We apply this technique to the data from the Dark Energy Survey, for which the ML model inspects ~230 million cutout images and returns ~20,500 (0.01%) of the highest-scored images for visual inspection. We successfully recover a large fraction of compelling strong lensing candidates found with other methods while maintaining a manageable false-positive rate.
Cosmic Ray Buoyancy in the Interstellar Medium
In a galaxy, energy gets partitioned into several components: stars, gas, magnetic fields, and electromagnetic radiation. The gas, called the interstellar medium (ISM), is made up of multiple components – there are warm and cold phases, ionized phases, interstellar dust grains, and even particles moving at relativistic speeds. Those relativistic particles are cosmic rays. In our own galaxy, the total energy in cosmic rays is similar to the energy of the galactic magnetic field. How do these cosmic rays change the long-term state of the ISM and the local structure of a galaxy? We explore one way: injection of cosmic rays by supernova explosions. With 3D cosmic ray magnetohydrodynamic simulations, we show these injections can drive the buoyant rise of magnetic flux tubes in a galaxy. The buoyant rise changes the structure of gas and reorients the magnetic field. To measure this buoyant rise, we examine the movement of a magnetic flux tube containing the injection location for a long time after the injection. Additionally, we vary initial parameters like the fraction of energy in cosmic rays and the magnetic field strength to determine how those parameters adjust the buoyant rise of the flux tubes. Overall, our simulations show how cosmic rays are an important component when studying a galaxy’s ISM and they should not be neglected in simulations attempting to create realistic models of the ISM.
Quantum Approximate Optimization Algorithm with Adaptive Bias Fields
QAOA (quantum approximate optimization algorithm) is one of the most promising algorithms for NISQ (noisy intermediate-scale quantum) era, the current state of quantum computing. In this presentation I will first introduce the standard QAOA. I will start with physics behind it (time evolution of quantum state, concept of Hamiltonian, quantum adiabatic theorem), and proceed to how the classical optimization part of it work. Then I will show the performance of abQAOA (quantum approximate optimization algorithm with adaptive bias fields) in comparison to standard QAOA, and offer some speculations of further research directions related to QAOA.
Quantum Computing and Error Correction with Neutral Cesium and Rubidium Atoms A
Quantum computers are theorized to revolutionize information processing. Problems across disciplines such as medicine, chemistry, and finance that are too difficult to be solved on even the largest supercomputers today can be encoded onto a quantum computer and solved quickly. Despite this motivation, no sufficiently large quantum computers have been built due to technical limitations. In this talk, I will focus on one potential platform for quantum computers: neutral atomic qubits. I will address benefits and challenges to the platform, with a focus on quantum error correction.
X-FAST: A New XUV Femtosecond Absorption Spectroscopy Tabletop Instrument
Core-level spectroscopies enable element specific measurements of electronic
and vibrational dynamics on the femtosecond timescale. Advancements
in the generation of XUV light through high harmonic generation
(HHG) has opened the door to doing such measurements in-house; we are
currently in the commissioning phase of the X-FAST instrument at the University
of Wisconsin – Madison. Here, we describe the instrument and show
the initial capabilities and studies carried out using the X-FAST instrument.
The instrument is comprised of two paths, an optical pump in air and
an XUV probe in vacuum. We used this system to carry out two preliminary
experiments: a study of the kinetics and dynamics of transition-metal
perovskite oxides, and a collection of static transient absorption spectra of a
metallic shape-memory alloy, Ni2MnGa.
Current Graduate Student Panel (5:30 – 6:30 PM)
Are you considering going to graduate school? Attend this panel for the opportunity to ask questions and hear from a panel of current UW-Madison Physics graduate students about their research, paths, application process and preparation, and what it’s like to be a graduate student in the Department of Physics. We will hear perspectives from various research areas in Physics, so bring any and all questions!