Empowering Collaborative Data Management for Astronomers
Mentor: Kelle Cruz (CUNY/AMNH/Flatiron)
Astronomy is a data-intensive field, with astronomers collecting vast amounts of data from telescopes and other instruments. This data is used to study a wide range of astronomical phenomena, from the formation of stars and galaxies to the evolution of the universe. Many astronomers use compilations of very wide tables to keep track of their data instead of using databases. Databases are organized collections of data that can be easily searched and queried but they can be difficult for a typical astronomer to build. The goal of this project is to develop new database infrastructure, dubbed the AstroDB Toolkit, to make it easier for astronomers to build, share, and collaboratively maintain databases of astronomical data. This project involves: designing and implementing new database schemas, developing tools and techniques for ingesting data, working with astronomers to identify their needs and requirements for databases. This project is ideal for a student with a strong interest in software development and data management and has experience using Github. Prior experience with databases is not necessary.
Simulating Gravitational Recoil in Galaxies and Other Black Hole Simulations
Mentor: Jillian Bellovary (CUNY/AMNH/Flatiron)
Project 1: Simulating gravitational recoil in galaxies
When black holes merge, there is an effect called gravitational recoil which gives the remnant black hole a “kick.” This kick can be strong enough to eject a black hole out of its host galaxy, especially if the galaxy is small. When people simulate black hole/galaxy co-evolution, no one includes this, but it’s probably very important. I have included recoil in the ChaNGa galaxy simulation code, but it’s pretty simple and needs some modifications and testing. A student working on this project would make modifications to ChaNGa (in C++, don’t worry it’s not too gnarly), run test simulations of isolated galaxies to see what the effects of different recoil scenarios are, and eventually implement and run the recoil models in a cosmological simulation. Simulations are analyzed in python using the pynbody library.
Project 2: Something about simulations and black holes
Jillian works on many aspects of black holes in galaxies, but most of them involve intermediate mass black holes, black holes at high redshift, gravitational waves from merging black holes, and black holes in dwarf galaxies. There are numerous projects that could be done in these realms, and Jillian has some half-baked ideas for a few, but nothing fully fleshed out (unlike project #1). Jillian is interested in talking to people about what they would like to work on and developing a project together. This could include making observational comparisons, studying star formation rates/histories, analyzing black hole merger activity, or other totally different things. Let’s talk!
Cluster Lenses in the Strong Lensing Database
Mentor: Giorgos Vernardos (CUNY/AMNH)
Gravitational lensing is a rare phenomenon that even Einstein did not think would be possible to observe. Yet, with our current advanced observatories, both in space and on the ground, we are able to obtain pristine and impressive images of such lenses. These new discoveries that are happening almost on a daily basis now, will increase the sample of approx. 1000 known lenses by about 100 times. This ‘zoo’ of lenses, apart from including very rare systems, for example quasars lensing other quasars, lensed supernovae, etc, will constitute a treasure trove of information regarding cosmology, dark matter, and galaxy evolution in general. The Strong Lensing Database (SLED, https://sled.amnh.org) is an invaluable resource that can catalogue and enrich the information of all these lenses – it has been called the ‘facebook’ of lenses and it is continuously being extended and improved. The project will consist of extending SLED to incorporate lens clusters and visualization tools, adding results from the literature but also with particular emphasis on recent discoveries by Euclid.
Modeling the AGN Channel for Gravitational Wave Sources
Mentors: K. Saavik Ford (CUNY/AMNH/CCA) & Barry McKernan (CUNY/AMNH/CCA) and the TIDYNYC group
Active galactic nuclei (AGN) are powered by the accretion of disks of gas onto supermassive black holes in the centers of galaxies. AGN disks are expected to contain a dense population of embedded objects, including stars and stellar mass black holes. This embedded population is a promising source of binary black hole mergers, detectable in gravitational waves with LIGO-Virgo-Kagra (the so-called AGN channel). AGN should be the most efficient channel to merge binary black holes to intermediate black hole (IMBH>100Msun) mass.
The recent public, open-source code McFACTS is the only complete AGN channel simulation code; with development led by Profs Ford & McKernan. We are interested in working with a MS student to develop modules within McFACTS to simulate and study encounters between IMBH and other objects (small mass black holes and the supermassive black hole). Our goal is to predict a population of IMBH mergers detectable with the future mission LISA, a European space-based gravitational wave detector scheduled to launch in ~10 yrs.
Solar Flares!
Mentors: Megan Bedell (CCA) and Ryan Rubenzahl (CCA)
The Sun, like most stars, exhibits magnetic activity in many forms, including dramatic flare events. This magnetic activity is an interesting astrophysical signal that has been increasingly well-studied in the era of intensive spectroscopic observations of stars for the purpose of finding exoplanets. Most work to date using spectroscopic data has focused on starspots, long-term activity cycles, and other aspects of stellar magnetic activity, leaving flares relatively underexplored. Recent projects like the Keck Planet Finder’s Solar Calibrator (KPF SoCal) offer the opportunity to address this gap in knowledge thanks to fast cadence (1 min) daily observations of the disk-integrated Sun across the full optical spectrum and at extreme Doppler stability (20 cm/s) and signal-to-noise (1500).
We aim to compile a catalog of known solar flare and/or coronal mass ejection events that have available observations from extreme precision radial velocity (EPRV) spectrographs including KPF SoCal. We’ll use these spectra to (1) diagnose how flare activity manifests in different features of the solar spectrum, (2) develop general flare/CME indicators that may be applied to observations of other stars, and (3) explore the physics of flare effects in the solar photosphere. Along the way, the student will gain experience in high-resolution spectroscopy, astronomical data analysis, and python programming.
Stacking Gamma Rays To Discover New Source Populations
Mentors: Joshua Tan (CUNY/AMNH), Tim Paglione (CUNY/AMNH), Dave Zurek (AMNH) and the AMNH Gamma-Rays and Compact Objects Group
Our group uses 17 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, dwarf galaxies, blazars, galaxy clusters, and a variety of interacting binaries and other exotica (even Jupiter!). Current ongoing projects of the group include CUNY Astro PhD student Owen Henry studying the signal from galactic novae and CUNY Astro MS student Thyra Essyelinck studying galaxy clusters. There are projects available for students interested in stacking the astrophysical phenomena that could plausibly be producing gamma rays with a keen eye towards those potential sources that are yet to be detected. In particular, there are projects waiting in the wings about hot stars, dwarf galaxies, and AGN. Let’s see if your favorite astronomical object is sending gamma rays our way!
Reconstructing the Initial Conditions of the Universe with Diffusion Models
Mentor: Matthew Ho (Columbia)
Initial condition reconstruction involves “rewinding” cosmic evolution to determine the universe’s primordial density fluctuations from observations of its present-day structure. This process is notoriously difficult due to the complexity of structure formation, but recent work has shown that deep learning approaches such as diffusion models can rapidly and accurately infer large-scale initial conditions. This project will apply this technique to the highly non-linear, sub-Mpc scales for the first time, incorporating the effects of baryonic physics. This will involve training a 3D diffusion model on the CAMELS suite of hydrodynamical simulations to reconstruct initial conditions from late-time snapshots. We will then validate the accuracy of these reconstructions by running new, high-resolution ‘Digital Twin’ simulations from the predicted initial conditions and tracking the consistency of subsequent structure formation. An optional stretch goal is to incorporate realistic mock observables, demonstrating the method’s potential for application to real observations.
Local Group Archaeology
Mentors: Allyson Sheffield (CUNY/CCA), Carrie Filion (CCA), and Adrian Price-Whelan (CCA)
Local Group archaeology provides insight into how galaxies form and evolve by leveraging the detail, breadth, and depth of observations that are only possible for galaxies within the Local Group. Stellar spectroscopy is an essential tool in the Local Group archaeology toolbox, as it allows us to constrain both the chemical makeup of a star and its line-of-sight velocity. The elemental abundances that we measure in stellar atmospheres tell us about their birth environment, giving us a window into the past. We have three, relatively open-ended Local Group archaeology + spectroscopy-themed projects, all built around new and incoming data.
Project 1: Globular Clusters
Globular clusters are dense collections of stars that often showcase specific, unexplained elemental abundance patterns. For example, some subpopulations of stars have anomalously high nitrogen abundances. By identifying stars with high nitrogen abundances, we can be pretty sure that those stars came from a cluster. For this project, we will use low-resolution spectra and data-driven methods to identify new candidate nitrogen-rich stars. This project will result in a catalog of candidate stars that observers can follow-up with higher resolution spectroscopy to get detailed abundances, along with a new technique for identifying these stars.
Project 2: Gaia Observations
The Gaia mission has observed over a billion stars, and will soon be releasing the data for each individual observation of each star (December 2026). A number of stars will have multiple spectra taken over multiple years, which will make it possible to track how these stars have changed over this time interval in a very detailed way. There are a number of possible ways to prepare for and use these data, with applications in exoplanets, variable stars, and multiple-star systems. For this project, we would first make predictions for what Gaia will see, identify a specific application that we want to pursue, and we will then develop a pipeline to apply to the real data when it is public.
Project 3: Milky Way Disk
The disk of the Milky Way goes through a transition near the solar radius. Interior to the solar orbit, the gravitational field is dominated by the self-gravity of stars and other baryons, but exterior to the solar circle the orbits of stars act more like tracers of the combined stellar and dark matter distribution. This means that the outer disk of the Milky Way is much more sensitive to perturbations from orbiting dwarf galaxies like the Sagittarius dwarf galaxy and the MCs. Evidence of these perturbations has been seen in a large-scale “warp” of the stellar and gas disks, in a “flaring” of the scale height, and in “wiggles” in the stellar midplane. The goal of this project is to characterize the age, chemical, kinematical, and/or spatial structure of the outer disk in new survey data
Exploring NGC 6791: A Fascinating Open Cluster
Mentor: Isabel Colman (NYU, AMNH)
My research revolves around taking a data science approach to solving detection and classification problems in stellar astrophysics, using data from NASA’s TESS and Kepler space telescopes. The Kepler mission ran from 2009-2013, staring at one region of the sky for four continuous years. In this field is the open cluster NGC 6791, an exciting target due to its unusually high age (~8Gyr) and metallicity (+0.3dex). I have access to a newly-available dataset of detrended light curves from NGC 6791 which are yet to be studied, providing several avenues for a two-year masters project depending on student interests. The cluster is host to oscillating red giants, binaries, blue stragglers, and at least one gamma Doradus star and one pulsating subdwarf, all of which are of scientific interest in the time domain. For example, studying the effects of high metallicity on stellar oscillations. There is also the potential to find signals of stellar rotation here, which is a long shot, but it’s never been done in an open cluster this old and it would be exciting to try. The wide variety of stellar variability in this cluster (and field) also opens up an avenue for a machine learning classification and cataloging project. Finally, I am not an exoplanet person, but a colleague once told me he found three potential planets in NGC 6791, and that was with access to only a fraction of the data we have now; there are very few (if any) known exoplanet hosts in open clusters, so this presents an opportunity to perform a systematic search.
Cross the Misty Mountains through Moria: New frontiers of the cosmic web
Mentor: Charlotte Welker (CUNY/CCA)
On scales much larger than stars, solar systems and even galaxies, the Universe is shaped in a highly structured network resembling a spider web: the cosmic web. This network of filaments serve as highways for dark matter, cosmic gas and galaxies and constrain their properties, such as their spin, shape or ability to form stars. However, the diversity of cosmic filaments and their ability to evolve over time remain a mostly uncharted territory.
In the Gotham Web Lab, we produce and use a variety of datasets ( high-resolution cosmological simulations, telescope observations ) and techniques (hydrodynamic modeling, HPC computing, AI) to better understand the diversity of cosmic filaments, develop models of their evolution, understand how they shape populations of galaxies….and how galaxies shape them back, for instance through feedback associated with central black hole activity.
In our group, you will be able to choose between a variety of projects including for instance modeling the merger of two large filaments connecting to a cluster of galaxies and hunting for signatures in observations of galaxies, using Graph Neural Networks to reconstruct the evolutionary stage of filament from the population of galaxies within and around it, understanding how the faintest, most elusive filaments affect the tiniest galaxies (dwarf galaxies) detected by millions in new surveys like EUCLID and LSST, exploring the impact of AGN feedback in galaxies hosted by a cosmic filament on its long-term evolution or building a high resolution simulation of a merger of two faint filaments hosting dwarf galaxies.
You will join a highly collaborative team including a postdoc, two PhD students and a number of undergraduate students. You will also be given the opportunity to contribute to cross-disciplinary engineering/astronomy projects we plan to offer to undergraduate at City Tech in the coming two years, including among others building a mini radio-telescope and building a mechanical simulation of a galaxy merger.
Mentor: Jared Goldberg (CCA)
Mentor: Rachel Sommerville (CCA)
Mentor: Mordecai-Mark Mac Low (AMNH)
This entry is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International license.
