Research
Welcome to my research website! 👋
During my undergraduate career, I have been involved with research projects in a variety of astronomical subfields, spanning from the characterizing Hot Jupiters’ atmosphere to mining some of the faintest galaxies.
Thanks to my research experience, I have been working on various datasets, like DELVE, WOCS, and JWST Cycle 2 data. I also learned some useful statistical tools for astrophysics along the way, like unsupervised clustering, rejection/MCMC sampling, and atmospheric retrievals with nested sampling.
I am interested in continuing my research career in graduate school. I enjoy using modern statistical tools to answer astrophysical problems. I have been working on ultra-faint dwarf galaxies (UFD), blue straggler stars (BSS), and hot Jupiters, but I would love to explore more astronomical fields, like AGN, ISM, quasars, etc.
If you are interested in learning about my research or simply chatting about science in general, please don’t hesitate to email me (panpi@umich.edu).
Kinematics of ultra-faint dwarf galaxies 🌌
Advisor: Marla Geha
05/2024 - Present
Ultra-faint dwarf (UFD) satellite galaxies are believed to have the oldest, most dark-matter dominated, and least chemically evolved stellar populations (Simon 2019). Therefore, UFDs serve as effective cosmological probes for the early Universe and are pristine laboratories for dark matter physics. In this study, we analyze resolved stars of two Milky Way satellite galaxies, Pegasus 3 and Pegasus 4, to provide more precise and accurate measurements of their velocity dispersion, dynamical masses, and mass to light ratios compared to previous studies (Cerny et al. 2023a; Kim et al. 2016). We first select member stars within these two galaxies using radial velocity, proper motion, equivalent width, and color & magnitude measurements. With these selected member stars’ radial velocity measurements, we apply a customized MCMC procedure to fit for their systematic velocity and velocity dispersion, which gives us their dynamical mass estimates (Walker et al. 2006; Wolf et al. 2010; Foreman-Mackey et al. 2013). We reconfirm and conclude these two systems are dark matter dominated dwarf galaxies by comparing them with other similarly faint stellar systems. This work offers us more accurate dynamical mass estimates of Peg 3 & 4 with more spectroscopically selected member stars.
Hot Jupiter multi-dimensional atmospheric modeling 🪐
Advisor: Ryan MacDonald
01/2024 - Present
We observed the hot Jupiter WASP-52b with JWST as part of a Cycle 2 Program aiming to measure differences between the east and west limbs of the planet. We analyzed the chemical composition, temperature, and cloud properties on each side of the planet. The science goals of this project are:
- Measure the chemical composition of WASP-52b’s atmosphere using a Bayesian atmospheric retrieval code. (ongoing)
- Identify whether a 2D atmospheric model is preferred over a 1D model, and if so, quantify differences between the morning and evening terminators (west and east). (ongoing)
- Investigate and correct for the influence of unocculted starspots on WASP-52b’s JWST transmission spectrum. (ongoing)
Stellar inhomogeneities (unocculted starspots or faculae) are believed to influence exoplanet transmission spectroscopy. Therefore, modeling and retrieving stellar features is crucial for understanding the exoplanet’s atmospheric properties. In this study, we apply atmospheric retrievals on the optical to infrared (0.6–2.7 μm) transmission spectrum of the hot Jupiter WASP-52b observed by JWST NIRISS/SOSS instruments. Our objective is to measure its atmospheric chemical composition and study the influence of unocculted stellar features on the transmission spectrum using POSEIDON (a Bayesian atmospheric retrieval algorithm). Since WASP-52b presents significant stellar features and starspot crossing events during the transit, we introduce unocculted starspot or facula (fraction, temperature, and surface gravity as free parameters) in the retrieval process. This work allows us to establish constraints on WASP-52b’s planetary atmosphere and the stellar feature of its host star. We are close to understand WASP-52’s stellar features and results are coming soon. Limb asymmetry analysis results coming soon
Alternative evolutionary pathway for stellar population by orbit modeling ⭐
Advisor: Robert Mathieu
Started 05/2023 - Present
Astrobites
Blue Straggler Stars (BSS) are believed to have originated from binary star interactions. Therefore, modeling binary orbits is crucial for understanding their formation and evolution. In this study, we applied The Joker (a Monte Carlo rejection sampler) to model binary orbits in the WIYN Open Cluster Survey (WOCS). Our objective was to evaluate The Joker’s efficacy on binaries with ample radial-velocity data (RVD). Exploring the potential usage of The Joker, designed for sparsely measured radial velocities, we applied it to known single-lined binary orbits in M67 and NGC 188. With 70+ binary systems, we compared our Joker orbital parameters among binaries having a diverse array of period-eccentricity combinations (Geller et al. 2021). We found that 5-7 RVD start to meaningfully constrain orbital parameters, and 8-10 RVD are required for unimodal period solutions with eccentricities converging to previous studies’ solutions. Additionally, we explored the possibility of devising observing plans having limited prior measurements with The Joker predicting possible periastron approach times to constrain eccentricity. This work enables us to estimate binary orbital solutions with fewer RVD and to enhance our understanding of the alternative evolutionary tracks of binary systems.
For our future work, we will fit more binary solutions for WD-MS binaries, while improving the pipeline to better suit observation plans and reduce bias our priors might introduce. This study is also used to design future observation plans to break the degeneracies between possible orbits.
image credit: Aaron Geller
Mining Ultra-Faint Galaxies in the Local Group 🌌
Advisor: Eric Bell
01/2023 - 05/2024
Ultra-faint dwarf (UFD) satellite galaxies are believed to have the oldest, most dark-matter-dominated, and least chemically evolved stellar populations. Therefore, UFDs serve as effective cosmological probes for the early stage of the Universe. In this study, we applied unsupervised machine learning (DBSCAN & OPTICS) to search for potential UFDs by analyzing spatial overdensities of the tip of red giant branch (TRGB) stars in the DELVE survey.
Our objective is to complement the current catalog of UFDs in the local volume while exploring the potential of using modern statistical density searching tools in UFD studies. We have focused on searching for M31 UFDs and recovered all 10 M31 satellite galaxies in the DELVE DR2 coverage. This work provides tools to identify UFDs in large sky surveys, which helps to test the cosmological constraints of the Lambda Cold Dark Matter (ΛCDM ) model.
To further this project, I hope to introduce other types of stars (MS, BSS) into the selection criteria. Also, data from larger areas of the sky will be explored for those “missing satellites”.
image credit: DELVE Collaboration