2023 Program

Who are we?

Program Staff:

• Michael Stroh – Co-Director, Computational Lead
• James Schottelkotte – Co-Director
• Fulya Kiroglu – Brinson Mentor, Computation Lesson Development and Instruction
• Elena González – Brinson Mentor, Reach Further Mentor Lead
• Nycole Wenner – Programs Coordinator
• Kari Frank – CIERA Director of Operations

Research Mentors: Darsan Swaroop Bellie, Anna Childs, Shinjan Dutta, Saarah Hall, Ben Hyatt, Emma Kaufman, Fulya Kiroglu, Chang Liu, Miguel Martinez, Liam O’Connor, Jillian Rastinejad, Jonathan Roberts, Kyle Rocha, Huei Sears, Philipp Srivastava, Elizabeth Teng, Rachel Zhang

Additional Instructors and Helpers: Daniel Campos, Erin Cox, Alexa Gordon, Nathalie Jones, Charlie Kilpatrick, Dean Kousiounelos, Shane Larson, Dennis Lee, Caroline von Raesfeld, Ethan Rengifo, Pedro Rodríguez, Imran Sultan, Jason Wang, Caitlin Witt

What are the research topics?

Potential research projects for both sessions during Summer 2023 currently include, but are not limited to: 

• • The Velocities of Stars in the Milky Way 
  • The Habitable Zones of Other Worlds in the Cosmos 
  • The Effects of Stellar X-Ray and UV Flares on Exoplanetary Atmospheres 
  • Climate Models of the Earth
  • Short Gamma Ray Burst Afterglow Fitting 
  • How Bright are Accreting Black Holes in Binary Systems? 
  • Stellar Evolution – The Dynamic Lives of Stars 
  • Binary Stars with COSMIC 
  • How Can You Extract Energy from Black Holes? 

New projects are added each year, though, so more may be available during the program! 

Research projects for the extended portion of the program are drawn from ongoing research at CIERA, an overview of which can be found here. 

Reach Further Topics

The 2023 cohort of High School students worked on the following research projects, created by the CIERA scientists below:

Accelerating binary stellar evolution simulations with machine learning

Project Design: Elizabeth Teng

Student: Vivian Nguyen

Simulating the evolution of binary star systems is very computationally challenging because of the level of detailed physics required. When simulating all of the millions of binaries in a galaxy, it would take way too long to actually perform all of the physics calculations. In this project we speed up the simulation process by training machine learning models on the evolutionary outcomes.

 

Sonification of Waves in Massive Stars

Project Design: Emma Kaufman & Benjamin Hyatt

Student: Chloe Kim

 In this project, we introduced the student to tools for somifying data in Python. The student learned to use the stellar evolution code MESA to generate models of massive stars native to different galactic environments. Utilizing pre-run 3D fluid dynamics simulations of the waves generated by core convection in massive stars, the student was able to sonify the resulting simulated wave data. They then used these auditory representations of the waves in massive stars to listen to how the star’s environment changes the sound.

 

Fitting Isochrones to Stellar Cluster Photometry Data

Project Design: Saarah Hall

Student: Zeb O’Hara

 In this project, we connected theory to observations by fitting MIST isochrone models to Gaia photometry data describing a star cluster of the student’s choosing. With the best fit model, the student was able to report their measurements of the cluster’s age, metallicity, and extinction. The student gained practice querying, managing, and “cleaning” large data sets, as well as writing Python code to quantify the “best-fitting” model to their data. The student gained an understanding of several isochrone-fitting methods from the literature, such as geometric fitting, maximum likelihood, and Hess diagram fitting.

 

A Star in a Box: Modeling Thermal Convection in Dedalus

Project Design: Liam O’Connor

Student: Brandon Rogers

Our project studied convective heat transfer in sun-like stars by performing simulations of Rayleigh-Benard convection using the Dedalus Python package. We found that the number of convective cells (and their respective aspect ratios) followed two distinct regimes: at low Rayleigh number, more convective cells formed while at high Rayleigh number, fewer cells formed. Between the regimes we observed a rapid shift in the cell counts which is likely  accompanied by other behavioral changes. The student learned how to solve Partial Differential Equations (PDEs) in Dedalus, along with processing and interpreting the simulation data.

 

A Radius-Magnitude relation for GRB Host Galaxies

Project Design: Huei Sears

Student: Carissa Chen

Our project studied a sample of GRB host galaxies at high-redshift.  These galaxies were imaged with the Hubble Space Telescope in rest-frame IR.  We measured the half-light radius and apparent magnitude of these galaxies using aperture photometry in Python.  We sky subtracted, PSF corrected, and aperture corrected our measurements to find the intrinsic radius and the absolute UV magnitude of these galaxies.  We compared our data set to that of star-forming galaxies and found that the two galaxy samples had similar properties.

 

Predicting the Long Term Stability of TRAPPIST-1-like Systems

Project Design: Anna Childs

Student: Alexa Hua

We extended previous N-body simulations from Childs et al. (2023) that modeled the formation of  the TRAPPIST-1 (T1) planets through pebble accretion in a dynamically evolving gas disk.  These simulations produced 24 systems in resonance chains with at least six planets.  We extended the integration time of these systems from 3 Myr to 100 Myr in the absence of gas to test for the long term stability of such systems and to identify two and three-body resonance configurations that contribute to long term stability.  We found that the majority of our simulations went unstable (lost planets via merging or ejections).  This finding suggests that T1-like systems may be uncommon in our Galaxy and systems of long resonance chains that form in the presence of gas are more likely to merge into lower multiplicity systems of super-Earths after the gas disk dissipates.

 

Exploring the spectral diversity in Type Ia supernovae

Project Design: Chang Liu

Student: Isabella Chen

Using a spectroscopic sample of ~100 Type Ia supernovae (SNe Ia) from Carnegie Supernova Project we explored the diversity in the kinematics of certain species in the SN ejecta, which is essential in understanding the progenitors of these splendid explosions. We focused on the velocity of the prominent Si II λ6355 line in these SNe when they reached their maximum brightness. We developed a pipeline which automatically fits the Si II line profile to Gaussian profiles for a large number of SNe, and estimated the corresponding expansion velocities and the uncertainties using Markov Chain Monte Carlo. We reconfirm the existence of two subclasses of SNe Ia in the maximum brightness – Si II velocity phase space, suggesting multiple formation channels in the SN Ia population.

 

Galaxy classification on image data from GalaxyZoo using deep learning

Project Design: Shinjan Dutta

Student: Jayden Freelove

Galaxy zoo is a project for capturing images of different galaxies and being able to classify them into classes depending on their shape, for eg: spiral, spherical, etc. We used a subset of the galaxy zoo training data, approximately 10,000 images and trained a convolutional neural network to be able to identify the type of galaxy. This was more of a computer vision focused project, to highlight the state of the art in current deep learning models and their applications in astronomy. 

 

Gamma Ray Bursts and Afterglows

Project Design: Jillian Rastinejad 

Student: Jake Meltzer

The sources of the heaviest elements in the Universe remain unknown, but may be tied to the extremely bright supernovae following long gamma-ray bursts. In this projects, we analyzed the properties of the observed jets accompanying long gamma-ray bursts and compared them to predictions for the explosions that might synthesize heavy elements.

 

Machine Learning in Binary Population Synthesis 

Project Design: Philipp Moura Srivastava 

Student: Leah Shaman

Until recently, the evolution of large numbers of binary star systems either took too long or gave inaccurate results. Through this project, the aim is to learn about and implement the Machine Learning techniques that have alleviated both issues as well as to learn about crucial components of theoretical astrophysics such as binary star system simulation.

 

Machine Learning and WD Binaries 

Project Design: Jonathan Roberts 

Student: Katie Butterworth and Rachel Durango-Cohen

During the summer of 2023, a project focused on employing machine learning techniques to identify low signal-to-noise ratio (SNR) white dwarf binaries. The primary objective was to analyze gravitational wave strain data associated with these binaries, leveraging ML algorithms to discern patterns and extract information that could reveal the mass of the systems. The initiative aimed to enhance our understanding of these astronomical phenomena through the application of advanced computational methods.

 

High-Mass X-Ray Binaries

Project Design: Kyle Akira Rocha

Student: Mia Resendiz

High-mass X-ray binaries (XRBs) contain a massive star and either a neutron star or black hole accreting stellar material from the donor producing high energy X-ray emission. Modeling XRBs allows us to investigate the formation of compact objects through supernovae, binary stellar evolution physics, and even the reionization of the Universe. In this project we use state-of-the-art binary evolution models to compare our predictions to an observed sample of XRBs in the Milky Way.

 

Studying the Formation of Black Hole X-Ray Binaries Through Hydrodynamical Models

Project Design: Fulya Kiroglu

Student: Venus Aradhya

The goal of this project is to explore formation channels of low-mass X-ray binaries in globular clusters. The abundance of bright X-ray binaries in Galactic globular clusters exceeds that in the field by many orders of magnitude, indicating that these binaries are formed through dynamical processes. Motivated by this, we performed hydrodynamic simulations of close encounters between black holes and giant stars using the smoothed particle hydrodynamice code StarSmasher. Venus explored the outcomes of these simulations using Python tools to identify the region of  parameter space leading to formation of low-mass X-ray binaries through stellar collisions.

 

Modeling Binary Evolution To Produce Compact Objects

Project Design: Rachel Zhang

Student: Michael Bae

The goal of this project is to understand the evolutionary steps of how binaries evolve from stars to become white dwarfs, neutron stars, and black holes depending on varying initial conditions. Through this project, my student learned how to use a population binary synthesis code (COSMIC) and gain an understanding of how stars can form compact objects, focusing in particular on binary black hole formation.

 

Gravitational Wave Theory and Sonification

Project Design: Darsan Swaroop Bellie

Student: Ariane Akayezu

The goal of this project was to use both literature and gravitational-wave modeling tools to understand the core theoretical aspects underlying gravitational waves from compact binary mergers. A secondary goal of the project was to sonify the simulated gravitational-wave signals as a first step towards developing an open-source gravitational-wave sonification package.

 

Simulating Planets in Clusters Using REBOUND

Project Design: Miguel Martinez

Student: Jayden Vargas

It is now well known that planets are born in star clusters. In this project, Jayden looked at what happens to planets as a result of their birth environment. Using the planetary integration Python package Rebound, he studied how close a star could pass to a planet orbiting a star before the planet is ejected from the system.