CUNY Astro

Project Pitches 2023

Assessment of Stellar Variability for JWST Exoplanet Observations Using Citizen Science Tools

Mentors: Joshua Tan (CUNY/AMNH) and Rob Zellem (Jet Propulsion Laboratory – California Institute of Technology)

Abstract: Exoplanet Watch is a citizen/community science project to monitor transiting exoplanets with small, ground-based telescopes to ensure the efficient use of large observatories, such as Hubble and James Webb. However, stellar activity can also impact one’s transit measurements: if a host star is active due to spots or flares, it could drastically alter one’s interpretation of the planet’s atmosphere. Correcting for this effect is particularly important for JWST’s high precision transit observations. With this project, we will expand Exoplanet Watch’s capability to monitor and quantify stellar activity. The student will work with our community to produce instructions and guides to upload their stellar activity data generated from EXOTIC, our Python3 data analysis code, to the American Association of Variable Star Observers’ stellar database. They will also lead the efforts to bring this data into Exoplanet Watch’s CITISENS pipeline and be displayed on our Results page. The intention with this project is to ultimately end up with a first-author, peer-reviewed paper about assessing stellar variability for known JWST targets and their potential impact on these high-precision observations.

Comparing 1D and 3D models of red supergiant stars to observations of Betelgeuse and its surrounding environment

Mentors: Jared Goldberg (CCA) and Thavisha Dharmawardena (CCA)

Abstract: This project works with both data (Thavisha’s expertise) and stellar models (Jared’s expertise). On the observational end, the concrete goal is to use radio observations to determine the radial density profile of material outside of Betelgeuse (both pre- and post-dimming). If possible, we would try to constrain both its average properties, as well as any 3D asymmetries. On the theory end, the student would then compare this data to modeling in order to constrain the star’s structure and evolution, and/or supernova observables. For that, more of a “pick your own adventure” will proceed depending on interests/speed/what proves to be harder than any of us expect. If the student is interested in stellar pulsations and mass loss, there is plenty of room to build up predictions for how much mass loss *should* occur given pulsational outbursts, as well as where that mass would go, so comparing to the data would constrain and sharpen our theories of stellar pulsations and outbursts. If the student is more interested in stars as supernova progenitors, they could instead (or in addition!) take the density distribution around Betelgeuse as a given, then stitch a Betelgeuse-like circumstellar environment onto stellar evolution models and explode them in order to model a variety of supernova outcomes. This would yield predictions for observables when a star blows up in a *realistic* circumstellar environment rather than an ad-hoc one (which is what’s still being done in current state-of-the-art calculations). Either way, we already have 3D simulations which could inform expected levels of inhomogeneity near the stellar surface, and 1D models (or at least the tools to construct them) which could be used to inform how much mass is lost and when (perhaps aided by information about the 3D velocity and density distribution near the stellar photosphere, both from the observations and the 3D simulations).

For what galaxies can we obtain birth radii? (Galactic archaeology; Cosmological simulation analysis)

Mentor: Lucy Lu (AMNH)

Abstract: Stars move away from their birth radii via a process called radial migration. Without taking into account radial migration, we will not be able to understand the true chemical evolution while looking at mono-age populations. One way to do so is by obtaining the radii the stars are born (birth radii). We know we can do this for the Milky Way (MW), and there are some hints that we can do this for the LMC. However, what is the limit of obtaining birth radii for galaxies other than the MW? Can we obtain birth radii for external galaxies? For this project, we will be looking at cosmological simulations and try to understand what kind of galaxy properties will enable us to obtain birth radii. By testing this, we will be taking the first step in understanding the ability to extend birth radii inference to external galaxies. 

You will learn: How to work with simulation data, understand how to obtain birth radii for stars and why they are important, and understand the chemical evolution of galaxies.

Understanding stellar spin down with Galactic kinematics (Stars; Observational data analysis)

Mentor: Lucy Lu (AMNH)

Abstract: Stars spin-down over time as they lose angular momentum. Thus, we can constrain the magnetic dynamo of stars by looking at their spin-down. In particular, what the stellar dynamos of low-mass stars look like is still unclear. In order to understand stellar spin-down, we need information on stellar ages of these stars. However, ages for low-mass stars are extremely hard. As a result, in this project, we will be exploring stellar spin-down with kinematic information for old, low-mass stars. 

You will learn: How to obtain rotation periods for stars, understand stellar spin-down, explore Gaia data.

Different Areas of Work to Develop a Project

Mentor: Mordecai-Mark Mac Low (AMNH)

Abstract: There are different areas of work that I would be happy to work with a student to develop a project in. The biggest possibilities are:

  • work with the Torch framework to model star cluster formation in clouds drawn from larger-scale galactic models or perform simulated observations of existing models for comparison with the real world. Another option here for someone with a strong programming background would be working to incorporate a new radiative transfer model into the framework.
  • compare magnetic fields generated in supernova-driven dynamo models to observations, again using simulated observations.
  • perform simulated observations of models of dust in protoplanetary disks to compare to ALMA observations of structure in disks.
  • study the formation and behavior of stellar mass black hole binaries in AGN accretion disks, probably using the FARGO code.

Generative Data-driven Models of Stellar Spectra.

Primary Mentor: Adrian Price-Whelan (CCA)

Co-mentors: David W. Hogg (CCA, NYU), Danny Horta (CCA)

Abstract: Our study of the Milky Way and its stellar populations has been revolutionized over the last decade thanks to data from the Gaia Mission and large stellar surveys. Gaia measures stellar positions and velocities, photometry, and low-resolution spectroscopy with incredible precision for almost 2 billion stars. Ground-based spectroscopic surveys like the APOGEE surveys and SDSS-V provide an additional wealth of information by measuring high-resolution spectra for millions of stars, including detailed measurements of stellar composition and evolutionary stage. Combining these datasets will lead to more robust determinations of stellar distances and intrinsic stellar parameters that will push our ability to measure properties of our Galaxy and its dark matter content out to much larger distances than are possible with existing catalogs. The goal of this set of projects will be to (1) develop and implement a data-driven statistical model for stellar parameter inference given combined spectroscopy and Gaia data, (2) use this framework to measure improved distances, ages, and stellar compositions for Gaia and SDSS-V stars, and (3) use the resulting catalog to study the stellar population structure of the Milky Way. These projects will be led by the student, and the goal will be to produce at least two first-author publications: One with a catalog of stellar parameters for Gaia and SDSS-V stars, and one with a study of the Milky Way’s stellar populations using this catalog, but we are open to adapting the project as much as necessary to align with your interests! The CCA is a partner data center for the Gaia Mission, and the CCA+CUNY are member institutions of the SDSS-V.

Open Clusters as Galactic Laboratories

Mentor: Dave Zurek (AMNH)

Abstract: Open Clusters are gravitationally bound groupings of 100s to about 10,000 stars, span ages from essentially zero to about 8 Gyrs, and trace the star formation of the Galactic disk. Stellar clusters larger than Open Clusters (i.e., Globular Clusters) are difficult to model dynamically and thus Open Clusters provide a direct comparison with modern N-body simulations.  Dynamics of the stars in stellar clusters are predicted to produce a plethora of compact binaries such as cataclysmic binaries and low mass X-ray binaries. However, these specific compact binaries are rare in Open Clusters and thus other tracers of dynamical interactions such as blue stragglers and main sequence binaries are needed to test N-body simulations. There are thousands of Open Clusters, so some constraints need to be used and specifically one that makes sense from the perspective of both cluster properties and data availability. There are 73 Open Clusters that are older than a Gyr and are in the Zwicky Transient Factory (ZTF) archive. Membership for the stars in each of these clusters can be found using the GAIA-DR3 data. Given the data that exists for each of these clusters we can determine the stellar membership (GAIA), the photometry for the stars within each cluster (ZTF), identification of photometrically peculiar stars (ZTF; for example blue stragglers but there are others), variability (ZTF and possibly TESS), and cluster ages (and an estimate of metallicity). The expectation is not necessarily that a full analysis of all 73 clusters will be accomplished by the end of the MS program, but even a subset of clusters will result in at least one and possibly multiple papers.

High Energy Astrophysics of Galaxy Clusters

Mentors: Tim Paglione (CUNY/AMNH) and the AMNH Gamma-Rays and Compact Objects Group

Abstract: Gamma-rays probe the most energetic processes in the universe. Most gamma-rays are created by light and matter interacting with cosmic rays, particles accelerated to nearly the speed of light usually by supernova explosions and the pulsars they leave behind. However, any strong shocks can accelerate cosmic rays, ranging from those found between interacting wind binaries to the very large-scale shocks in galaxy clusters launched by starburst super winds or jets from active galactic nuclei. Our group uses 15 years of data from the Fermi Gamma-Ray Space Telescope to stack signals from any and all potential sources of gamma-rays including pulsars, novae, hot stars, interstellar clouds, and a variety of interacting binaries and other exotica (even Jupiter!). Our current study targets galaxy clusters by cross-matching potential targets with the Planck Telescope’s catalog of those detected by the Sunyaev-Zeldovich Effect, which can help us quantify the energy budgets of galaxy clusters, the mass loading of winds and jets, and the impact of cosmic rays on cluster evolution and feedback in general.

Using Kepler and TESS to study rotation and stellar activity in the open cluster NGC 6819

Mentor: Isabel Colman (AMNH)

Abstract: The Kepler mission broke new ground in time domain astronomy, observing the same area of sky for four continuous years and providing data of unprecedented quality for the detection of transiting exoplanets and stellar variability. Now, we have the opportunity to revisit targets of interest from the Kepler mission with overlapping observations from TESS, and search for changes on the scale of years — which we know is feasible, because our own Sun has activity cycles. In this project, we’ll look at the open cluster NGC 6819 through the lens of gyrochronology, the study of the link between stellar rotation and age. As a star ages, it “spins down” — the process of losing angular momentum, which results in a decreasing rotation period. The link between rotation and activity cycles is an ongoing area of study, and both can be detected in time domain light curves from variations in brightness. Open clusters are the perfect testing ground for gyrochronology because their member stars are coeval: knowing the ages of these stars, we can use rotation periods across different types of star to calibrate gyrochronological relations. NGC 6819 is one of four open clusters in Kepler’s field of view which have been extensively studied and characterized, so we can use these stars that have been observed with both Kepler and TESS to paint a thorough picture of the long-term stability of magnetically active stars, as well as creating a benchmark sample that can be monitored and updated as the TESS mission progresses. Practical skills gained through this project will include familiarity with Kepler and TESS data, time series analysis in both the time domain and frequency space, and an introduction to “big data” astronomy. There will be the opportunity to work with machine learning, and to expand the scope of this project to other stars common to Kepler and TESS, including those in NGC 6811 and NGC 6866, and field stars.

Primeval Beads: Connecting galaxies to the cosmic filaments they live in.

Primary Mentor: Charlotte Welker (CUNY/CCA)

Co-mentors: Charlotte Olsen (LSSTC/CUNY), Viraj Pandya (Columbia/CCA)

Abstract: Galaxies like our Milky Way do not live in a smooth, homogeneous, quiet Universe. In fact, on its largest scales, the Universe resembles a spider web, with filaments stretching across near empty regions and connecting to dense knots, a structure dubbed the cosmic web. Galaxies flow, live and evolve in these filamentary rivers of dark matter and gas. How do these cosmic filaments shape the diversity of galaxies? This question is all the more complex that filaments are themselves dynamic structures that zip, stretch or collapse over billions of years. In your project, you will use new high-resolution simulations of large volumes of the Universe and large surveys of thousands of observed galaxies to investigate some of these questions. You may choose to focus on dwarf galaxies, the tiniest galaxies of all and the likely “untouched” remains of  the early Universe, hence potential witnesses of the slow evolution of thin, primeval filaments. Or you may choose to focus on more massive galaxies in clusters, the densest regions of the Universe where the largest cosmic filaments plunge together, potentially protecting galaxies from the hot, buoyant intra-cluster medium. Either way, you will learn topology and methods to identify filaments, and learn about types of galaxies and how to measure their properties. The project will give you insights into the evolution of galaxies, the formation and detection of the Cosmic Web, and the interplay between cosmology (large scales) and astrophysics (galaxy evolution). Additionally, the project will help you develop transferable skills: cutting-edge machine learning techniques, geometry with Python or Fortran, visualization methods, and supercomputing using a high-performance computing machine.

Satellite Galaxies


Abstract: Galaxies can often be grouped into central galaxies and satellite galaxies that orbit them. While observationally it can be challenging to always tell which galaxies are central and which are satellites, in simulations it is considerably easier to classify galaxies into these two categories. In this project we will analyze the satellite population in a number of simulations and seek to understand why they differ. We will seek to understand what causes the number of satellites to vary between different simulations and then try to identify the ‘correct’ model by matching with observations. We can then test our idea of a correct model by incorporating it into a semi-analytic model and comparing our results with observations.

“ScienceGPT” Project at Flatiron Institute

Mentor: Shirley Ho (Flatiron Institute/ NYU/ Princeton) 

Abstract: Are you passionate about merging the frontiers of astronomy and artificial intelligence? Do you envision a future where machine learning plays a pivotal role in unraveling the mysteries of the universe? We invite you to be part of an exciting initiative that combines the power of AI with the insights of astronomy. At the Flatiron Institute, we are embarking on a groundbreaking project: the creation of “ScienceGPT.” This visionary endeavor aims to develop a comprehensive scientific generalist machine learning model that harnesses the collective knowledge and insights from various physical science disciplines. Our goal is to pre-train this model on a vast and diverse dataset, enabling it to provide invaluable insights and solutions across a wide range of scientific challenges.

We are currently seeking enthusiastic and motivated master’s students to join our dynamic team and contribute to two key threads of the project:

  1. Astronomy Integration: Dive into the universe’s mysteries by bridging the gap between astronomy and AI. Work closely with our experts to incorporate astronomical data and theories into the “ScienceGPT” framework. Your contributions will empower the model to understand and interpret astronomical phenomena, ultimately advancing our understanding of the cosmos.
  2. AI Advancement: Push the boundaries of artificial intelligence by developing innovative algorithms and techniques tailored to the unique demands of scientific data. Collaborate with our AI specialists to enhance the model’s capacity to process complex information, recognize patterns, and provide meaningful insights.

As a member of our team, you will have the opportunity to:

  • Collaborate with leading researchers in both astronomy and AI fields.
  • Gain hands-on experience in building and fine-tuning large-scale machine learning models.
  • Contribute to groundbreaking advancements with potential implications for various scientific disciplines.
  • Work in a stimulating and collaborative environment that encourages creativity and critical thinking.

Whether you have a background in astronomy, AI, or both, this project offers a platform for you to apply your skills in a meaningful and transformative way. Together, let’s push the boundaries of knowledge and technology to redefine the future of astrophysics and AI.