BBL 9/21/2020 — 3 Talks! Sidney Lower (University Of Florida), Kaze Wong (Johns Hopkins University), And Kameswara Mantha (University Of Missouri – Kansas City)
Abstract: The ability to accurately infer the physical properties of galaxies is critical for our understanding of galaxy formation and evolution. Modeling the broadband spectral energy distributions (SED) of galaxies is one of the primary diagnostics used to derive major physical properties such as the stellar mass, star formation rate (SFR), and stellar age. Yet the technique rests on numerous assumptions to relate the observed spectrum to the physical properties of that galaxy; these assumptions include modeling the relatively complex physical processes in a galaxy related to star formation, stellar evolution, and dust attenuation. The robustness of an SED model and our ability to accurately recover physical properties of a galaxy depend on our confidence in each SED model component to accurately capture the complexity of the physical processes and growth history of that galaxy. It has been shown in the literature that the assumptions made in SED modeling have severe impacts on the derived physical properties, including galaxy stellar masses and star formation rates. In an attempt to understand these impacts, we use the Simba hydrodynamical cosmological galaxy formation simulation to ground truth and improve upon SED model assumptions. Our aim is to explore the baseline accuracies achievable by current SED modeling methods by testing on the wealth of data available from the Simba cosmological simulation. We find that the commonly used prescriptions for the star formation history (SFH) of a galaxy are not flexible enough to recover the stellar mass, stellar age, and SFR of a galaxy simultaneously, independent of galaxy type. We also find that commonly used models for dust attenuation and emission compound this issue. However, we find that 1) using a more flexible nonparametric SFH model can result in more accurate galaxy properties and 2) problems with model simplicity or inflexibility can be mitigated by considering more complex models alongside thoughtful priors to avoid degeneracies between model parameters.
Simulation-based gravitational-wave population inference with normalizing flow
Abstract: Running population synthesis simulations can be time-consuming. To constrain the physical parameters characterizing the simulations we must compare them to the data at numerous sample points in the physical parameter space. This comparison requires a large number of simulations, and it is often computationally impractical. In this talk, I will present a deep learning technique (normalizing flow) to emulate population synthesis simulations at a much faster speed. The emulator can be used in the population inference process, opening up the possibility of constraining astrophysics directly using the observed gravitational-wave population.
email: kazewong@jhu.edu
Constraining the Major Merger Rate History of the Universe since z~3: Comprehensive Analysis of Close Pairs and Tidal Features using Observations and Simulations
Abstract: The major merger rate between similar mass galaxies (stellar-mass ratio <4:1) is theoretically predicted to increase towards earlier cosmic times and is thought to play a vital role in buildup of high-mass galaxies. Established empirical techniques (close pairs and tidal features) test the theory by measuring the rate as pair or tidal fraction divided by the time during which mergers show empirically detectable signatures (observability timescale). Yet, astronomers face two important challenges that hinder precise constraints on merger rate and feature-based major merger identification: 1) lack of rigorous calibrations quantifying the role of observational systematics when measuring the merger fractions; 2) the close-pair and tidal feature observability timescales have yet to be systematically quantified in a cosmological context. I will summarize my coordinated attempts to solve these challenges using comprehensive Hubble Space Telescope (HST) observations and forefront simulated datasets. I will discuss: i) the redshift evolution of close-pair fraction at during 0.5<z<3 using the CANDELS survey; ii) ongoing efforts to quantify the role of observational systematics (photometric redshift, stellar-mass errors, and detection incompleteness) on measurement of close-pair fractions using mock lightcones from Semi-Analytical Models; iii) a recently developed pipeline to extract and quantify the strength of plausible tidal features within parametric model-subtracted residual images; iv) future efforts towards quantifying close-pair and tidal feature timescales using a large sample of 1E5 synthetic HST H-band images of 6500 major mergers (~20 Myr timesteps) from the Horizon-AGN simulation.
Bio: I am an international doctoral student from India working with Dr. Daniel McIntosh on galaxy mergers and galaxy evolution at University of Missouri Kansas City. I am in my final stages of the degree and am actively applying for post-doc positions.
Email: km4n6@mail.umkc.edu
Host: Mike McDonald
Speakers
- Sidney Lower, University of Florida
- Kaze Wong, Johns Hopkins University
- Kameswara Mantha , University of Missouri - Kansas City