Subaru Seminars

TBD

TBD

TBD

We are entering an era of extremely large spectroscopic surveys, with Subaru/Prime Focus Spectrograph (PFS) poised to play a leading role. Dense spectroscopy provides a powerful tools of probing the physical properties of statistically large samples of galaxies and their environments. In this talk, I present two key applications of dense spectroscopy to the study of galaxy clusters and their member galaxies. The first application is the stellar velocity dispersion function (VDF) of cluster galaxies, which serves as a probe of the underlying dark matter subhalo mass distribution. Using dense spectroscopic data, we measure the VDF in galaxy clusters and derive corresponding theoretical predictions from cosmological simulations. I will discuss the comparison between the observational measurements and theoretical expectations, along with their implications for cluster substructure and galaxy-halo connections. The second application is weak-lensing spectroscopic tomography ("spectrotomography"), which combines weak lensing with dense spectroscopy. This technique reconstructs the cluster lensing signal using exclusively spectroscopically confirmed background galaxies, thereby minimizing uncertainties associated with photometric redshifts. I will present the methodology and early results from a sample of 10 galaxy clusters. Spectrotomography provides not only an independent measurement of cluster mass but also a promising cosmological probe in the era of large spectroscopic surveys.

The SCExAO team at the Subaru Telescope has launched an intensive survey, OASIS, for substellar companions around accelerating stars. Dynamical masses derived from orbital fitting allow us to compare evolutionary models of age-mass-luminosity relations with observational results. For example, hot (Baraffe et al. 2003) vs. cold start (Marley et al. 2007) are competing hypotheses of initial entropy. Differences between two models are most prominent at young ages (≲ 200 Myr), making high-contrast imaging particularly sand exhibit temperature variations across their surfaces, making apparent magnitudes inclination-dependuitable for characterizing such young and hot substellar companions. However, age estimates from the color-magnitude diagram (CMD) suffer from large uncertainties and may even be misleading when metallicity effects are underestimated. Our survey targets are mainly early-type stars (B, A, and early-F), which typically rotate rapidly. We developed software that applies Bayesian analysis to jointly sample six parameters: mass, age, metallicity, rotation, inclination, and parallax. Our method better accounts for parameters affecting age-dating, providing more reliable age estimates than simple CMD-based results. We validated our software with well-known clusters: Hyades (600-800 Myr), Praesepe (600-800 Myr), and Pleiades (~100 Myr), finding good agreement with previous studies (e.g., Brandt & Huang 2015a,b,c). We further conclude well-constrained extinction and metallicity are essential for reliable ages with small uncertainties. In addition, we will present a newly detected brown dwarf candidate from our survey with a mass of 51 (+12, -16) M_Jup, a semi-major axis of 54.4 (+11, -8.3) AU, and an eccentricity of 0.36 (+0.13, -0.17). Its spectral type is M4-M5, with an effective temperature of 3148 (+175, -142) K. We also derive a Bayesian age of 484 (+248, -235) Myr. We will also compare it with ages from evolutionary models of substellar companions.

Ground-Based direct imaging of exoplanets is limited by atmospheric turbulence, which introduces phase errors resulting in complex "speckle" patterns forming in the Point Spread Function (PSF). Understanding the relationship between the wavefront sensor (WFS) telemetry and the speckles is essential for high-contrast imaging; however traditional analytical reconstructions often struggle with the non-linearities in the WFS response. An alternative is a machine learning-based approach using a U-Net architecture. By treating PSF reconstruction as an image-to-image translation task, the model is trained to map telemetry from the wavefront sensor directly to the corresponding focal plane intensity distribution. This approach aims to leverage the U-Net’s ability to capture spatial features across multiple scales, potentially offering a more robust method for handling atmospheric speckles compared to standard linear models. This presentation outlines the current status of the model architecture, training data, and current model performance.