Browsing by Author "Ananna, Tonima Tasnim"

Now showing 1 - 3 of 3
  • No Thumbnail Available
    Item
    GaMPEN: A Machine-learning Framework for Estimating Bayesian Posteriors of Galaxy Morphological Parameters
    (2022) Ghosh, Aritra; Urry, C. Megan; Rau, Amrit; Perreault-Levasseur, Laurence; Cranmer, Miles; Schawinski, Kevin; Stark, Dominic; Tian, Chuan; Ofman, Ryan; Ananna, Tonima Tasnim; Auge, Connor; Cappelluti, Nico; Sanders, David B.; Treister, Ezequiel
    We introduce a novel machine-learning framework for estimating the Bayesian posteriors of morphological parameters for arbitrarily large numbers of galaxies. The Galaxy Morphology Posterior Estimation Network (GaMPEN) estimates values and uncertainties for a galaxy's bulge-to-total-light ratio (L ( B )/L ( T )), effective radius (R ( e )), and flux (F). To estimate posteriors, GaMPEN uses the Monte Carlo Dropout technique and incorporates the full covariance matrix between the output parameters in its loss function. GaMPEN also uses a spatial transformer network (STN) to automatically crop input galaxy frames to an optimal size before determining their morphology. This will allow it to be applied to new data without prior knowledge of galaxy size. Training and testing GaMPEN on galaxies simulated to match z < 0.25 galaxies in Hyper Suprime-Cam Wide g-band images, we demonstrate that GaMPEN achieves typical errors of 0.1 in L ( B )/L ( T ), 0.'' 17 (similar to 7%) in R ( e ), and 6.3 x 10(4) nJy (similar to 1%) in F. GaMPEN's predicted uncertainties are well calibrated and accurate (<5% deviation)-for regions of the parameter space with high residuals, GaMPEN correctly predicts correspondingly large uncertainties. We also demonstrate that we can apply categorical labels (i.e., classifications such as highly bulge dominated) to predictions in regions with high residuals and verify that those labels are greater than or similar to 97% accurate. To the best of our knowledge, GaMPEN is the first machine-learning framework for determining joint posterior distributions of multiple morphological parameters and is also the first application of an STN to optical imaging in astronomy.
  • No Thumbnail Available
    Item
    Probing the Structure and Evolution of BASS Active Galactic Nuclei through Eddington Ratios
    (2022) Ananna, Tonima Tasnim; Urry, C. Megan; Ricci, Claudio; Natarajan, Priyamvada; Hickox, Ryan C.; Trakhtenbrot, Benny; Treister, Ezequiel; Weigel, Anna K.; Ueda, Yoshihiro; Koss, Michael J.; Bauer, F. E.; Temple, Matthew J.; Balokovic, Mislav; Mushotzky, Richard; Auge, Connor; Sanders, David B.; Kakkad, Darshan; Sartori, Lia F.; Marchesi, Stefano; Harrison, Fiona; Stern, Daniel; Oh, Kyuseok; Caglar, Turgay; Powell, Meredith C.; Podjed, Stephanie A.; Mejia-Restrepo, Julian E.
    We constrain the intrinsic Eddington ratio (lambda(Edd)) distribution function for local active galactic nuclei (AGN) in bins of low and high obscuration [log(N-H/cm(-2)) <= 22 and 22 < log(N-H/cm(-2)) < 25], using the Swift Burst Alert Telescope 70 month/BASS DR2 survey. We interpret the fraction of obscured AGN in terms of circumnuclear geometry and temporal evolution. Specifically, at low Eddington ratios (log lambda(Edd) < -2), obscured AGN outnumber unobscured ones by a factor of similar to 4, reflecting the covering factor of the circumnuclear material (0.8, or a torus opening angle of similar to 34 degrees). At high Eddington ratios (log lambda(Edd) > - 1), the trend is reversed, with <30% of AGN having log(N-H/cm(-2)) > 22, which we suggest is mainly due to the small fraction of time spent in a highly obscured state. Considering the Eddington ratio distribution function of narrow-line and broad-line AGN from our prior work, we see a qualitatively similar picture. To disentangle temporal and geometric effects at high lambda(Edd), we explore plausible clearing scenarios such that the time-weighted covering factors agree with the observed population ratio. We find that the low fraction of obscured AGN at high lambda(Edd) is primarily due to the fact that the covering factor drops very rapidly, with more than half the time spent with <10% covering factor. We also find that nearly all obscured AGN at high-lambda(Edd) exhibit some broad lines. We suggest that this is because the height of the depleted torus falls below the height of the broad-line region, making the latter visible from all lines of sight.
  • No Thumbnail Available
    Item
    Stripe 82-XL: The ∼54.8 deg2 and ∼18.8 Ms Chandra and XMM-Newton Point-source Catalog and Number of Counts
    (2024) Peca, Alessandro; Cappelluti, Nico; Lamassa, Stephanie; Urry, C. Megan; Moscetti, Massimo; Marchesi, Stefano; Sanders, David; Auge, Connor; Ghosh, Aritra; Ananna, Tonima Tasnim; Torres-Alba, Nuria; Treister, Ezequiel
    We present an enhanced version of the publicly available Stripe 82X catalog (S82-XL), featuring a comprehensive set of 22,737 unique X-ray point sources identified with a significance greater than or similar to 4 sigma. This catalog is four times larger than the original Stripe 82X catalog, by including additional archival data from the Chandra and XMM-Newton telescopes. Now covering similar to 54.8 deg(2 )of nonoverlapping sky area, the S82-XL catalog roughly doubles the area and depth of the original catalog, with limiting fluxes (half-area fluxes) of 3.4 x 10(-16) (2.4 x 10(-15)), 2.9 x 10(-15) (1.5 x 10(-14)), and 1.4 x 10(-15) (9.5 x 10(-15)) erg s(-1) cm(-2) across the soft (0.5-2 keV), hard (2-10 keV), and full (0.5-10 keV) bands, respectively. S82-XL occupies a unique region of flux-area parameter space compared to other X-ray surveys, identifying sources with rest-frame luminosities from 1.2 x 1038 to 1.6 x 1047 erg s(-1) in the 2-10 keV band (median X-ray luminosity, 7.2 x 1043 erg s(-1)), and spectroscopic redshifts up to z similar to 6. By using hardness ratios, we derived the obscuration of active galactic nuclei (AGNs), obtaining a median value of log(NH/cm(-2))=21.6-1.6+1.0 and an overall, obscured fraction ( log(NH/cm-2)>22 ) of similar to 36.9%. S82-XL serves as a benchmark in X-ray surveys and, with its extensive multiwavelength data, is especially valuable for comprehensive studies of luminous AGNs.