2025 Research Projects
Combining JWST Spectroscopy and HST Imaging to Understand the Efficiency of Dwarf Galaxies in Producing Ionizing Photons at Cosmic Noon
Advisors: Anahita Alavi
JWST continues to amaze us by gazing at faint galaxies from the epoch of reionization (EoR). Nevertheless, with many hours of observations, the fundamental question that remains is understanding the key sources of ionizing radiation. To address this, our focus must shift to lower redshifts, where ionizing radiation can traverse IGM.
Dwarf galaxies are expected to play a substantial role in ionizing the IGM. As part of an approved JWST AR program, a prospective graduate student is expected to incorporate the JWST NIRISS grism data of Abell 2744 galaxy cluster, with existing deep HST/WFC3 UV data. The goal is to investigate how effective are dwarf galaxies in producing ionizing photons at cosmic noon. Abell 2744 is a massive galaxy cluster which provides large gravitational lensing magnification. This enables the study of faintest galaxies that would otherwise be challenging to detect. This cluster is part of the Hubble Frontiers Fields program with rich ancillary data from HST and many other ground-based observatories.
A successful visiting student is expected to use H-alpha emission line measurements from the JWST/NIRISS grism spectra and UV luminosity from the HST/WFC3 UVIS imaging to measure the ionizing photon production efficiency. This project may involve working with NIRISS grism spectra and spectral fitting, measuring galaxy properties via SED fitting, and using gravitational lensing tools to estimate lensing magnifications.
The prospective student will collaborate with a group of experts at IPAC and UC Riverside. Students with experience in working with spectra, especially JWST grism spectroscopy, are encouraged to apply.
Extreme Galaxies in the Census of the Local Universe (CLU) Survey
Advisor: David Cook
The Census of the Local Universe (CLU) is a large-area survey that aims to find new galaxies in the local Universe out to a distance of 200 Mpc (z~0.05). We utilize four narrow-band filters to search for H-alpha emission-line sources across 26,000 deg^2 of the sky; similar in footprint to Pan-STARRS. The new galaxies found in this survey will be used to search for counterparts to Gravitational Wave events and enable the discovery of extreme galaxies (BCDs, metal-poor galaxies, green peas, etc.) that are rare in the local Universe. The source catalogs for the entire survey have recently been published, and are ripe for data mining galaxies with extreme emission-line properties. We anticipate discovering several thousand new BCDs and green peas during this project.
We are seeking a student to develop and implement methodologies to separate extreme galaxy candidates from normal star-forming galaxies and nebular regions in the Milky Way. The resulting candidates can then be combined with data from other large surveys to derive physical properties (stellar mass, star formation rate, metallicities, etc.) to enable studies of galaxy evolution and star formation.
We seek a student with a background in galaxy evolution, star formation, or analysis of large data sets to discover thousands of exciting and rare objects.
Harnessing KPF for Exoplanet Atmosphere Exploration
Advisor: Aurora Kesseli
High-resolution transmission and emission spectroscopy has proven to be a powerful tool for characterizing exoplanet atmospheres and has led to the detection of a range of atoms and molecules, precise constraints of bulk abundances, and the detection of atmospheric winds. Keck Planet Finder (KPF), the newly commissioned high-resolution (R~100000) optical (445-870nm) spectrograph on the W.M. Keck Observatory, will lead to unprecedented signal-to-noise ratios due to the improved throughput and stability of the instrument and large collecting area of Keck. With this improvement, a larger sample of planets that includes dimmer host stars and smaller planets will be accessible for atmospheric studies for the first time using high-resolution spectroscopy. During the six-month fellowship, the graduate student will work with Dr. Kesseli and key members of the KPF science team to analyze data of a sample of hot Jupiters that have already been observed with KPF, as well as have an opportunity to join on observing runs for other approved targets. The student will lead the scientific analysis (and expected publication) for one of the planets that has already been observed. As KPF is a new instrument and is still undergoing pipeline development and testing, the graduate student would also be able to contribute to testing and development of the part of the KPF pipeline we are actively updating to better support atmospheric studies.
Students do not need to have prior experience with high-resolution spectroscopy of exoplanet atmosphere, but coding knowledge in Python is recommended.
Joint Photometric and Astrometric Analysis of Roman Microlensing Events
Advisors: Etienne Bachelet, Sebastiano Calchi Novato
It is expected that the Galactic Bulge Time Domain Survey made by NASA's Nancy Grace Roman Space Telescope will revolutionize our knowledge of exoplanets. Indeed, it is expected that thousands of cold-planets, i.e., orbiting at large distances from their host, will be detected via the microlensing effect. Moreover, microlensing has the capabilities to discover stellar remnant hosts. The importance of these lens systems has been demonstrated recently with the discovery of a Jovian planet orbiting a white dwarf as well as with the detection of the first stellar mass black hole in the Milky Way. Generally, the microlensing analysis relies on the modeling of the photometric lightcurve only. But the precision of Roman offers the possibility to study the astrometric time series of microlensing events as well and, ultimately, constrain the mass and distance of the lenses with high precision.
We propose to study the potential of joint astrometry and photometry analysis of the Roman mission.
Studying the Faint Components of the Milky Way with Microlensing
Advisor: Etienne Bachelet
Thanks to a new generation of all-sky surveys, it is now possible to detect microlensing in the entire Milky Way. This is a unique opportunity to study the faint components of the Milky Way, from free-floating planets to stellar remnants. However, the cadence of these survey (~weekly) is insufficient to properly characterize the lensing systems. The OMEGA collaboration intends to conduct a photometric and spectroscopic follow-up of all microlensing candidates in the entire sky. Started in 2020, several planets have already been detected, along stellar binaries and stellar remnants candidates.
The student will join the OMEGA collaboration to plan and perform real time follow-up as well as the data analysis of microlensing events (photometry, spectroscopy and modeling).
Machine Learning for Time Series Imputation
Advisors: Jessica Krick, Shooby Hemmati
Using machine learning (ML) techniques on time series data is ubiquitous in the world of data science, from electrical grids to financial markets. Despite the large amounts of work in this field and available models, astronomy datasets are significantly different from most other datasets. Astronomical datasets can be characterized as unequal length, multivariate time series with missing values (an unenviable triumvirate of challenges). We propose a project to address these challenges in astronomical time series data. To that end, we have already built a tool that can easily, automatically generate multi-band training sets up to a few hundred thousand samples from archival data (ZTF, PanSTARRS, WISE, Gaia, etc.).
We propose here for a collaborator interested in time domain with a background in ML to design, optimize, test, and write a paper on imputation models appropriate for the wider astronomical community.
AGN Analytics
Advisors: Yu-Heng (Ian) Lin, Andreas Faisst, Shooby Hemmati, Jessica Krick
This project intersects machine learning, big data analytics, and archival research, utilizing a cloud-based science platform developed by our NASA-affiliated team. We are engaged in ongoing research into the diverse classifications of Active Galactic Nuclei (AGN), focusing more on those with variable multiband light-curve such as the Changing Look AGN (CLAGNs). Our team has developed tools and machine learning models that effectively differentiate various AGN samples and enable the isolation of distinct groups such as turn-on/turn-off CLAGNs from their multiband lightcurves. Additionally, we have implemented an automated system to collect all available, reduced archival spectroscopy. The proposed role involves conducting an examination of spectroscopic data of optimally targeted variable AGNs to analyze differences and uncover the underlying physical mechanisms leading to the observed variations.
The successful candidate will join us in leveraging this extensive dataset using advanced analytical techniques, providing new insights into the dynamic nature of AGNs and enhancing our understanding of the universe in a research-intensive environment.
Deciphering Galaxy Evolution with Emission Line Galaxies Using JWST Grism Spectroscopy
Advisor: Vihang Mehta
JWST slitless spectroscopy mode is revolutionizing our understanding of galaxy evolution at high redshifts, especially in the important low-mass regime, by providing unprecedented access to study the physical properties the stellar and ISM content of galaxies in great detail at high-redshifts.
The JWST Cycle 1 program, PASSAGE (GO1571, PI Malkan), has acquired up to 400 hours of NIRISS grism data over several tens of independent fields in up to 3 filters, spanning 1-2.3 micron and covering a total area up to 280 sq. deg. This program provides an unbiased spatio-spectroscopic dataset for several thousands of galaxies, including key optical emission lines (such as Balmer lines, [OIII], [OII]), across a wide redshift range (z=1-3.5). This dataset enables a wide range of galaxy evolution science including the time evolution as well as spatially resolved analysis of galaxy physical properties (such as metallicities, emission line strengths, ionization states, dust content, etc.), studying galaxies at the epoch of reionization, and discovery of rare, bright objects.
We seek students to leverage this rich JWST grism dataset to infer the stellar and nebular properties of galaxies and ultimately, study how galaxies grow and evolve over cosmic time. The student will have access to the reduced PASSAGE data and would be welcome to tackle any of the listed or other topics related to galaxy evolution enabled with these data.
Resolving the Dynamic Interstellar Medium with Infrared Echoes around Cas A
Advisor: Jacob Jencson
The content and structure of the interstellar medium (ISM) plays an influential role in the formation of stars and overall evolution of galaxies. In turn, the ISM is also shaped by the massive stars formed out of its gas and dust and the energetic supernovae (SNe) they produce. The infrared echoes around the young SN remnant Cassiopeia A provide a unique laboratory to explore the effects of a high-energy burst of SN radiation as it interacts with ISM dust in real time. Using upcoming JWST time-series observations, this program will exploit the motion of these echoes in NIRCam imaging to map the 3D structure of the echoing dust clouds and precisely measure the changes in their compositions induced by the passing SN radiation in MIRI/MRS spectroscopy.
The graduate student will primarily work with Dr. Jencson at IPAC to reduce and analyze the MIRI data and lead a publication on detailed modeling of the time-resolved spectra. As a member of the team, the student will also collaborate with JWST instrument scientists at the Space Telescope Science Institute and dust-modeling experts at NASA/Goddard.
Galaxies in the Early Growth Phase with JWST and ALMA
Advisor: Andreas Faisst
We seek a student to work on combining the extensive multi-wavelength imaging and spectroscopy from ALMA and JWST to conduct a range of studies as part of the ALPINE ALMA large program on the COSMOS field. The JWST data includes cycle 1 NIRCam imaging data from 255h COSMOS-Web as well as data from a 60h cycle 2 NIRSpec/IFU program. The sample includes main-sequence galaxies at z = 4 – 6 and provides the currently largest sample with observations from UV to far-IR including JWST optical lines. Other archival data can be added. A possible project could focus on the study of internal dust attenuation of the galaxies or their stellar population distribution via resolved SED fitting and spectroscopy. The findings can then be related to the evolution of galaxies on the z ~ 5 main sequence and their far-IR and [CII] properties and dynamics. The student would have access to all ALPINE (including JWST follow-up) data. Ideas for other projects related to the above are very welcome.
Deciphering the Growth of Massive Galaxies with Euclid
Advisors: Andreas Faisst, Kyle Finner
The dominant mass growth mechanisms of very massive (mostly quiescent) galaxies are still under debate. Accretion of mass in the form of minor satellites could be an avenue next to internal star formation or major mergers. Such cases will also alter the sizes of the galaxies. The first images from JWST already showed the abundance of globular clusters and dwarf galaxies around massive galaxies in the low-redshift universe (z~0.5). Euclid covers a large area for statistical studies and benefits from superior surface brightness sensitivity.
We seek a student who is interested in studying the circum-galactic medium around massive galaxies at z < 1 to understand better the contribution of satellite populations to their mass and size growth. A possible avenue is to stack light profiles derived from (public) Euclid images (+ any ancillary data) of selected massive galaxies. We also encourage students with theoretical background to apply—specifically to compare and explain the abundance (or absence) of satellite populations in the context of cosmological simulations.
Exploring the Dusty Universe with CHAMPS
Advisor: Andreas Faisst
CHAMPS is a 144h ALMA Large program in cycle 10 covering the 0.2 square-degrees of the COSMOS-Web JWST field with 1.2mm observations. The program aims at detecting the most dusty sources in the high-redshift universe. We seek a student to work on projects related to the CHAMPS ALMA data as well as JWST (and other) ancillary data. Possible projects include the study of the first heavily dust-obscured galaxies, constraining the ISM masses of quiescent galaxies, or studying the relation between stellar mass, star formation, and ISM mass across cosmic time from individual detections and from stacking.
The student would have access to all ALMA and JWST data in the COSMOS field. Other ideas for projects related to these data are very welcome.
Testing the Method of Cluster Redshifts for Future Surveys
Advisor: Andreas Faisst
Large area surveys often lack the photometric coverage for accurate photometric redshifts. Or, some sources, such as dust-obscured star forming galaxies detected with Herschel or ALMA lack optical counterparts, making it difficult to derive redshifts. Cluster redshifts (redshifts based on the clustering of sources) provide a method to derive a redshift distribution for a population of galaxies via their large-scale clustering on the sky. The only input are positions and a selection function for the training sample. This method could be applied to large area surveys (such as with Euclid or Roman) to provide alternative redshift distributions to classic photometric redshifts or redshifts derived with machine learning techniques. We seek a student for a project which aims at (i) characterizing the biases of such a technique by comparisons to dark-matter-only simulations (readily available), and (ii) applying this method to various fields including the COSMOS field where accurate photometric and spectroscopic redshifts are available for training and testing.
Resolved PAH Emission in z=1 Galaxies
Advisor: Andreas Faisst
We seek a student to work on a project related to an accepted JWST/MIRI MRS cycle 3 program to study the resolved PAH emission in z=1 galaxies. The galaxies are selected to be face-on, thus allow the examination of changes in the PAH emission lines in a spatially resolved manner across the galaxies. Possible projects could involve: (i) optimization of pixel-by-pixel mid-IR SED fitting for this project to derive kpc-resolved PAH line ratios; (ii) the measurement of grain sizes across the galaxies; (iii) the characterization of ionization properties in the galaxies from Ne lines. Other ideas for projects along these lines with these data are welcome.
Hidden AGN in Quiescent Galaxies
Advisor: Andreas Faisst
The occurrence of AGN might be correlated with the emergence of quiescent galaxies, either via starburst or feedback mechanisms. We seek a student to work on a project with the aim of characterizing the occurrence of dust-obscured AGN in quiescent galaxies. For this, we would select quiescent galaxies in multiple public JWST fields with multi-band MIRI coverage to measure the rest-frame 3-20um mid-IR slope from which the contribution of dust-obscured AGN can be computed. The results may be put into context of the large-scale structure in which the quiescent galaxies are residing for addition information on the quenching process.
Unlocking the Early Universe for Weak Lensing with JWST: High-Precision Analysis of z=2 Galaxy Cluster XLSSC122
Advisor: Kyle Finner
XLSSC122 is a galaxy cluster at z=2 with a well-established red sequence. The cluster has a significant XMM-Newton X-ray detection of the intracluster medium and a 7.6 sigma detection of the Sunyaev Zel'dovich decrement. However, these two measurements result in contrasting mass estimates. The SZ mass (M200~3e14 solar masses) is extreme for a galaxy cluster at redshift 2. Furthermore, the cluster has a signature of a recent merger with the X-ray peak and SZ centroid being significantly offset. We have been granted JWST NIRCam observations to perform a weak-lensing analysis. By mapping the dark matter distribution of the cluster with weak lensing, we will investigate its merging nature. We will also fit dark matter models to the radial mass profile to quantify the mass of the galaxy cluster. This will be the most distant cluster analyzed with weak lensing.
The successful candidate will work with Dr. Finner on JWST NIRCam data reduction and analysis.
Galaxy Survey Enhancements with Deep Learning
Advisor: Shoubaneh Hemmati
The success of many of the upcoming large cosmology missions (e.g., LSST, Roman, Euclid, SPHEREx) relies on the precise measurements of the redshifts, shapes, and other physical properties of galaxies. These measurements often require a joint processing of the observations and the higher the spatial and spectral resolution, wavelength coverage, and depth of these observations, the more information they entail. Given the wealth of the galaxy data today and advancements in image processing with deep learning, huge improvements for future surveys and their combinations can come about using data-driven approaches.
Through this project, we will design and optimize multi-band image enhancement deep learning structures and explore the extent of spatial/spectral resolution boosting, denoising, and in-filling of hyper-spectral images for future large galaxy surveys by training on the deepest existing multi-band observations in the COSMOS and CANDELS fields. We will quantify the gain in the estimation of various physical properties given the enhanced data products.
Using Machine Learning to Generate the Palomar Spectroscopic Survey of Young Stars
Advisor: Jessie Christiansen
There are an increasing number of exoplanet demographics studies that rely on stellar ages, with the goal of mapping out how planets form, migrate, and evolve with time (e.g., Christiansen et al. 2023; Vach et al. 2024). However, for young stars, there are few catalogs of stellar parameters of the size and uniformity needed for robust demographics analyses. Over two semesters we acquired spectra of >200 young FGK stars with DoubleSpec (an optical spectrograph on the 5-m Hale Telescope at Palomar) and 25 young M dwarfs with TripleSpec (a near-infrared spectrograph on the same telescope). Our goal is to use these spectra as a machine-learning training set and ultimately build a large, uniform, well understood atlas of young star parameters.
A visiting graduate student would reduce the spectra, derive stellar properties, and ultimately use these to build the machine-learning training set for the much larger sample of young stars observed by Kepler, K2, and TESS. The student would also be encouraged to come up with a better acronym or name for the final atlas.
