Gregory F. Snyder’s research while affiliated with Space Telescope Science Institute and other places

What is this page?


This page lists works of an author who doesn't have a ResearchGate profile or hasn't added the works to their profile yet. It is automatically generated from public (personal) data to further our legitimate goal of comprehensive and accurate scientific recordkeeping. If you are this author and want this page removed, please let us know.

Publications (125)


Figure 12. Merger rate based on visual classifications (left panel), random forest classifications (middle panel), and neural network classifications (right panel) as compared to the theoretical Illustris merger rate (black line) derived from V. Rodriguez-Gomez et al. (2015) and the major merger rate (gray line: power-law + exponential model; gray stars: binned JWST data) of Q. Duan et al. (2024) derived from the close pair method. The green shaded regions and the observed galaxy error bars indicate the binomial 95% confidence interval.
Comparison of F1 Scores for the Random Forests (RF) and DeepMerge Neural Networks (NN) on the Simulated CEERS Test Sets
CEERS Key Paper. IX. Identifying Galaxy Mergers in CEERS NIRCam Images Using Random Forests and Convolutional Neural Networks
  • Article
  • Full-text available

November 2024

·

13 Reads

·

1 Citation

The Astrophysical Journal Letters

Caitlin Rose

·

Jeyhan S. Kartaltepe

·

Gregory F. Snyder

·

[...]

·

A crucial yet challenging task in galaxy evolution studies is the identification of distant merging galaxies, a task that suffers from a variety of issues ranging from telescope sensitivities and limitations to the inherently chaotic morphologies of young galaxies. In this paper, we use random forests and convolutional neural networks to identify high-redshift JWST Cosmic Evolution Early Release Science Survey (CEERS) galaxy mergers. We train these algorithms on simulated 3 < z < 5 CEERS galaxies created from the IllustrisTNG subhalo morphologies and the Santa Cruz SAM light cone. We apply our models to observed CEERS galaxies at 3 < z < 5. We find that our models correctly classify ∼60%–70% of simulated merging and nonmerging galaxies; better performance on the merger class comes at the expense of misclassifying more nonmergers. We could achieve more accurate classifications, as well as test for a dependency on physical parameters such as gas fraction, mass ratio, and relative orbits, by curating larger training sets. When applied to real CEERS galaxies using visual classifications as ground truth, the random forests correctly classified 40%–60% of mergers and nonmergers at 3 < z < 4 but tended to classify most objects as nonmergers at 4 < z < 5 (misclassifying ∼70% of visually classified mergers). On the other hand, the CNNs tended to classify most objects as mergers across all redshifts (misclassifying 80%–90% of visually classified nonmergers). We investigate what features the models find most useful, as well as the characteristics of false positives and false negatives, and also calculate merger rates derived from the identifications made by the models.

Download

CEERS Key Paper. IX. Identifying Galaxy Mergers in CEERS NIRCam Images Using Random Forests and Convolutional Neural Networks

July 2024

·

42 Reads

A crucial yet challenging task in galaxy evolution studies is the identification of distant merging galaxies, a task which suffers from a variety of issues ranging from telescope sensitivities and limitations to the inherently chaotic morphologies of young galaxies. In this paper, we use random forests and convolutional neural networks to identify high-redshift JWST CEERS galaxy mergers. We train these algorithms on simulated 3<z<53<z<5 CEERS galaxies created from the IllustrisTNG subhalo morphologies and the Santa Cruz SAM lightcone. We apply our models to observed CEERS galaxies at 3<z<53<z<5. We find that our models correctly classify 6070%\sim60-70\% of simulated merging and non-merging galaxies; better performance on the merger class comes at the expense of misclassifying more non-mergers. We could achieve more accurate classifications, as well as test for the dependency on physical parameters such as gas fraction, mass ratio, and relative orbits, by curating larger training sets. When applied to real CEERS galaxies using visual classifications as ground truth, the random forests correctly classified 4060%40-60\% of mergers and non-mergers at 3<z<43<z<4, but tended to classify most objects as non-mergers at 4<z<54<z<5 (misclassifying 70%\sim70\% of visually-classified mergers). On the other hand, the CNNs tended to classify most objects as mergers across all redshifts (misclassifying 8090%80-90\% of visually-classified non-mergers). We investigate what features the models find most useful, as well as characteristics of false positives and false negatives, and also calculate merger rates derived from the identifications made by the models.


The Next Generation Deep Extragalactic Exploratory Public Near-infrared Slitless Survey Epoch 1 (NGDEEP-NISS1): Extragalactic Star-formation and Active Galactic Nuclei at 0.5 < z < 3.6

July 2024

·

17 Reads

·

5 Citations

The Astrophysical Journal

The Next Generation Deep Extragalactic Exploratory Public (NGDEEP) survey program was designed specifically to include Near Infrared Slitless Spectroscopic observations (NGDEEP-NISS) to detect multiple emission lines in as many galaxies as possible and across a wide redshift range using the Near Infrared Imager and Slitless Spectrograph. We present early results obtained from the first set of observations (Epoch 1, 50% of the allocated orbits) of this program (NGDEEP-NISS1). Using a set of independently developed calibration files designed to deal with a complex combination of overlapping spectra, multiple position angles, and multiple cross filters and grisms, in conjunction with a robust and proven algorithm for quantifying contamination from overlapping dispersed spectra, NGDEEP-NISS1 has achieved a 3 σ sensitivity limit of 2 × 10 ⁻¹⁸ erg s ⁻¹ cm ⁻² . We demonstrate the power of deep wide field slitless spectroscopy (WFSS) to characterize the star formation rates, and metallicity ([O iii ]/H β ), and dust content, of galaxies at 1 < z < 3.5. The latter showing intriguing initial results on the applicability and assumptions made regarding the use of Case B recombination. Further, we identify the presence of active galactic nuclei and infer the mass of their supermassive black holes using broadened restframe Mg ii and H β emission lines. The spectroscopic results are then compared with the physical properties of galaxies extrapolated from fitting spectral energy distribution models to photometry alone. The results clearly demonstrate the unique power and efficiency of WFSS at near-infrared wavelengths over other methods to determine the properties of galaxies across a broad range of redshifts.


UVCANDELS: The role of dust on the stellar mass-size relation of disk galaxies at 0.5 z\leq z \leq 3.0

May 2024

·

8 Reads

·

1 Citation

We use the Ultraviolet Imaging of the Cosmic Assembly Near-infrared Deep Extragalactic Legacy Survey fields (UVCANDELS) to measure half-light radii in the rest-frame far-UV for \sim16,000 disk-like galaxies over 0.5z30.5\leq z \leq 3. We compare these results to rest-frame optical sizes that we measure in a self-consistent way and find that the stellar mass-size relation of disk galaxies is steeper in the rest-frame UV than in the optical across our entire redshift range. We show that this is mainly driven by massive galaxies (1010\gtrsim10^{10}M_\odot), which we find to also be among the most dusty. Our results are consistent with the literature and have commonly been interpreted as evidence of inside-out growth wherein galaxies form their central structures first. However, they could also suggest that the centers of massive galaxies are more heavily attenuated than their outskirts. We distinguish between these scenarios by modeling and selecting galaxies at z=2 from the VELA simulations suite in a way that is consistent with UVCANDELS. We show that the effects of dust alone can account for the size differences we measure at z=2. This indicates that, at different wavelengths, size differences and the different slopes of the stellar mass-size relation do not constitute evidence for inside-out growth.


The Next Generation Deep Extragalactic Exploratory Public (NGDEEP) Survey

April 2024

·

34 Reads

·

30 Citations

The Astrophysical Journal Letters

We present the Next Generation Deep Extragalactic Exploratory Public (NGDEEP) Survey, a deep slitless spectroscopic and imaging Cycle 1 JWST treasury survey designed to constrain feedback mechanisms in low-mass galaxies across cosmic time. NGDEEP targets the Hubble Ultra Deep Field (HUDF) with NIRISS slitless spectroscopy ( f lim , line , 5 σ ≈ 1.2 × 10 ⁻¹⁸ erg s ⁻¹ cm ⁻² ) to measure metallicities and star formation rates (SFRs) for low-mass galaxies through the peak of the cosmic SFR density (0.5 < z < 4). In parallel, NGDEEP targets the HUDF-Par2 parallel field with NIRCam ( m lim , 5 σ = 30.6 − 30.9 ) to discover galaxies to z > 12, constraining the slope of the faint end of the rest-ultraviolet luminosity function. NGDEEP overlaps with the deepest HST Advanced Camera for Surveys optical imaging in the sky, F435W in the HUDF ( m lim , F 435 W = 29.6 ) and F814W in HUDF-Par2 ( m lim , F 814 W = 30 ), making this a premier HST+JWST deep field. As a treasury survey, NGDEEP data are public immediately, and we will rapidly release data products and catalogs in the spirit of previous deep-field initiatives. In this paper we present the NGDEEP survey design, summarize the science goals, and detail plans for the public release of NGDEEP reduced data products.


The connection between mergers and AGN activity in simulated and observed massive galaxies

December 2023

·

34 Reads

·

7 Citations

Monthly Notices of the Royal Astronomical Society

We analyze a suite of 29 high-resolution zoom-in cosmological hydrodynamic simulations of massive galaxies with stellar masses Mstar > 1010.9 M⊙, with the goal of better understanding merger activity among AGN, AGN activity in merging systems, SMBH growth during mergers, and the role of gas content in triggering AGN. Using the radiative transfer code Powderday, we generate HST-WFC3 F160W mock observations of central galaxies at redshift 0.5 < z < 3; convolve each image with a CANDELS-like PSF; stitch each image into a real CANDELS image; and identify mergers within the synthetic images using commonly adopted non-parametric statistics. We study the connection between mergers and AGN activity in both the simulations and synthetic images and find reasonable agreement with observations from CANDELS. We find that AGN activity is not primarily driven by major mergers (stellar mass ratio >1:4) except in a select few cases of gas-rich mergers at low redshifts (0.5 < z < 0.9). We also find that major mergers do not significantly grow the central SMBHs, indicating major mergers do not sustain long-term accretion. Moreover, the most luminous AGN in our simulations (Lbol > 1045 erg s−1) are no more likely than inactive galaxies (Lbol < 1043 erg s−1) to be found in merging systems. We conclude that mergers are not the primary drivers of AGN activity in the simulated massive galaxies studied here.


Figure 1. HST/ACS+WFC3 postage stamps for a random subset of the galaxies in the sample are shown. We select edge-on galaxies using a cut on axis ratio. The top row are high-mass galaxies ( M M 10 log 11
Figure 6. The distribution of the galaxy scale heights of our sample is shown in red. The sample spans 0.4 „ z „ 2.5. The measurements are carried out at one effective radius from the center of each galaxy. The distribution of a population-matched sample of disk galaxies in the local Universe (SDSS; Bizyaev et al. 2014) is shown in blue. The histograms for the two samples have been normalized. The median and inclination-corrected median of our high-redshift sample are shown with solid and dashed red lines, respectively. The median of the local sample is indicated with a vertical blue line. The range of estimates for the thick (light blue) and thin (gray) disks of the Milky Way are also shown (Section 5.1.1 of Bland-Hawthorn & Gerhard 2016 and references therein). The inclination-corrected median scale height of the high-redshift galaxies (∼750 pc) is smaller than that of the disk galaxies today (∼1500 pc) but is similar to that of the thick disk of the Milky Way (∼600-1100 pc).
Figure 7. This figure shows the percentage difference between the intrinsic and recovered scale heights for toy models of disk galaxies that are randomly inclined to our LOS. The toy models have ratios (q) of intrinsic scale height to scale length that vary between 0.15 and 0.35. The horizontal line represents the median of the models, 22%. We use the results shown here to correct the population median of the observed scale heights measured in this paper.
The Physical Thickness of Stellar Disks to z ∼ 2

October 2023

·

45 Reads

·

7 Citations

The Astrophysical Journal

In local disk galaxies such as our Milky Way, older stars generally inhabit a thicker disk than their younger counterparts. Two competing models have attempted to explain this result: one in which stars first form in thin disks that gradually thicken with time through dynamical heating, and one in which stars form in thick disks at early times and in progressively thinner disks at later times. We use a direct measure of the thicknesses of stellar disks at high redshift to discriminate between these scenarios. Using legacy Hubble Space Telescope imaging from the CANDELS and GOODS surveys, we measure the rest-optical scale heights of 491 edge-on disk galaxies spanning 0.4 ≤ z ≤ 2.5. We measure a median intrinsic scale height for the full sample of 0.74 ± 0.03 kpc, with little redshift evolution of both the population median and scatter. The median is consistent with the thick disk of the Milky Way today (0.6–1.1 kpc), but it is smaller than the median scale height of local disks (∼1.5 kpc) that are matched to our high-redshift sample by descendant mass. These findings indicate that, while (1) disks as thick as the Milky Way’s thick disk were in place at early times, (2) to explain the full disk galaxy population today, the stellar disks in galaxies need to on average physically thicken after formation.


Effects of feedback on galaxies in the VELA simulations: elongation, clumps and compaction

April 2023

·

4 Reads

·

10 Citations

Monthly Notices of the Royal Astronomical Society

The evolution of star-forming galaxies at high redshifts is very sensitive to the strength and nature of stellar feedback. Using two sets of cosmological, zoom-in simulations from the VELA suite, we compare the effects of two different models of feedback: with and without kinetic feedback from the expansion of supernovae shells and stellar winds. At a fixed halo mass and redshift, the stellar mass is reduced by a factor of ∼1-3 in the models with stronger feedback, so the stellar-mass-halo-mass relation is in better agreement with abundance matching results. On the other hand, the three-dimensional shape of low-mass galaxies is elongated along a major axis in both models. At a fixed stellar mass, M* < 1010 M⊙, galaxies are more elongated in the strong-feedback case. More massive, star-forming discs with high surface densities form giant clumps. However, the population of round, compact, old (agec > 300 Myr), quenched, stellar (or gas-poor) clumps is absent in the model with strong feedback. On the other hand, giant star-forming clumps with intermediate ages (agec = 100 − 300 Myr) can survive for several disc dynamical times, independently of feedback strength. The evolution through compaction followed by quenching in the plane of central surface density and specific star-formation rate is similar under the two feedback models.


The Science Performance of JWST as Characterized in Commissioning

April 2023

·

454 Reads

·

324 Citations

Publications of the Astronomical Society of the Pacific

This paper characterizes the actual science performance of the James Webb Space Telescope (JWST), as determined from the six month commissioning period. We summarize the performance of the spacecraft, telescope, science instruments, and ground system, with an emphasis on differences from pre-launch expectations. Commissioning has made clear that JWST is fully capable of achieving the discoveries for which it was built. Moreover, almost across the board, the science performance of JWST is better than expected; in most cases, JWST will go deeper faster than expected. The telescope and instrument suite have demonstrated the sensitivity, stability, image quality, and spectral range that are necessary to transform our understanding of the cosmos through observations spanning from near-earth asteroids to the most distant galaxies.


CEERS Key Paper. III. The Diversity of Galaxy Structure and Morphology at z = 3–9 with JWST

March 2023

·

52 Reads

·

102 Citations

The Astrophysical Journal Letters

We present a comprehensive analysis of the evolution of the morphological and structural properties of a large sample of galaxies at z = 3–9 using early James Webb Space Telescope (JWST) CEERS NIRCam observations. Our sample consists of 850 galaxies at z > 3 detected in both Hubble Space Telescope (HST)/WFC3 and CEERS JWST/NIRCam images, enabling a comparison of HST and JWST morphologies. We conduct a set of visual classifications, with each galaxy in the sample classified three times. We also measure quantitative morphologies across all NIRCam filters. We find that galaxies at z > 3 have a wide diversity of morphologies. Galaxies with disks make up 60% of galaxies at z = 3, and this fraction drops to ∼30% at z = 6–9, while galaxies with spheroids make up ∼30%–40% across the redshift range, and pure spheroids with no evidence for disks or irregular features make up ∼20%. The fraction of galaxies with irregular features is roughly constant at all redshifts (∼40%–50%), while those that are purely irregular increases from ∼12% to ∼20% at z > 4.5. We note that these are apparent fractions, as many observational effects impact the visibility of morphological features at high redshift. On average, Spheroid-only galaxies have a higher Sérsic index, smaller size, and higher axis ratio than disk or irregular galaxies. Across all redshifts, smaller spheroid and disk galaxies tend to be rounder. Overall, these trends suggest that galaxies with established disks and spheroids exist across the full redshift range of this study, and further work with large samples at higher redshift is needed to quantify when these features first formed.


Citations (55)


... Historically, convolutional neural networks (CNNs) have been the NN of choice for identifying galaxy mergers due to their exceptional performance in image processing and analysis (e.g. Huertas-Company et al. 2015;Pearson et al. 2019a,b;Bickley et al. 2021;Rose et al. 2024). Recently, Ferreira et al. (2024) combined vision transformers (ViTs) with CNNs in a two-stage hierarchical merger identification framework, achieving post-merger purities ≳ 95 per cent even at ∼1 Gyr after the merger event. ...

Reference:

Unions with UNIONS: Using galaxy-galaxy lensing to probe galaxy mergers
CEERS Key Paper. IX. Identifying Galaxy Mergers in CEERS NIRCam Images Using Random Forests and Convolutional Neural Networks

The Astrophysical Journal Letters

... illiams et al. 2023;Y. Harikane et al. 2024; T. Y.-Y. Hsiao et al. 2024). However, one remarkably UV-bright (M UV ; -22) galaxy candidate at z ; 16 turns out to be at z = 4.9 with strong emission lines and red continuum that mimic the expected colors of more distant objects (R. P. Naidu et al. 2022a;S. Fujimoto et al. 2023b;J. McKinney et al. 2023;J. A. Zavala et al. 2023). Recent JWST spectroscopic observations also find a numerous number of faint AGN populations at z ; 4-7 (e.g., Y. Harikane et al. 2023b;R. Maiolino et al. 2024a;J. Matthee et al. 2024), indicative of steeper faint ends in the quasar/AGN luminosity functions (LFs) than suggested in previous studies, and some studies suggest the identific ...

Dusty Starbursts Masquerading as Ultra-high Redshift Galaxies in JWST CEERS Observations

The Astrophysical Journal Letters

... deviate from Case B recombination (e.g. Pirzkal et al. 2024;Scarlata et al. 2024). Due to these uncertainties with using the H /H ratio to determine the attenuation, we do not attempt a reddening correction of the emission line fluxes based on the Balmer decrement for the four galaxies at ≳ 7, where H is beyond the wavelength coverage of NIRSpec. ...

The Next Generation Deep Extragalactic Exploratory Public Near-infrared Slitless Survey Epoch 1 (NGDEEP-NISS1): Extragalactic Star-formation and Active Galactic Nuclei at 0.5 < z < 3.6

The Astrophysical Journal

... Compared to the typical PSF FWHM of the HST images, the sizes of sample LAEs are mostly resolved due to the high spatial-resolution of the HST images. The size is corrected for lensing magnification, and measured as the circularized effective size r eff (that is, r eff = r maj × b/a, where b/a and r maj are the galaxy axis ratio and effective radius along major-axis, respectively); This size definition is commonly adopted in other studies (e.g., van der Wel et al. 2014;Shibuya et al. 2015;Kim et al. 2021;Nedkova et al. 2024), which allows us to compare our size measurements with the literature in a consistent manner. ...

UVCANDELS: The role of dust on the stellar mass-size relation of disk galaxies at 0.5 z\leq z \leq 3.0
  • Citing Preprint
  • May 2024

... Hα, [OIII]). For the remaining source, ID:695, which is not covered in the FRESCO Survey, we search the Next Generation Deep Extragalactic Exploratory Public (NGDEEP, PI: S. Finkelstein, #2079) Survey (Bagley et al. 2024). Again adopting the pro- Notes. ...

The Next Generation Deep Extragalactic Exploratory Public (NGDEEP) Survey

The Astrophysical Journal Letters

... Gas-rich mergers have been found to trigger starburst events (exceptionally high rates of star formation), and merging galaxies exhibit enhanced SFRs compared to isolated ones (Mihos & Hernquist 1994;Patton et al. 2011;Torrey et al. 2012;Patton et al. 2013;Lanz et al. 2013;Moreno et al. 2015;Pearson et al. 2019;Moreno et al. 2019;Patton et al. 2020;Garduño et al. 2021;Ellison et al. 2022; Thorp et al. 2022;Montenegro-Taborda et al. 2023;Duan et al. 2024b;Reeves & Hudson 2024;Yuan et al. 2024). Mergers can also trigger active galactic nuclei (AGN) activity (Silk & Rees 1998;Hopkins et al. 2008;Ellison et al. 2011;Satyapal et al. 2014;Gao et al. 2020;Li et al. 2023;Bickley et al. 2023;Sharma et al. 2024;Duan et al. 2024b), and the most luminous AGN, ultra-luminous infrared galaxies, and extreme emission line galaxies are often associated with mergers (Kartaltepe et al. 2010;Ellison et al. 2013;Perna et al. 2023a;Gupta et al. 2023;Marshall et al. 2023). Moreover, Witten et al. (2024) recently discovered several Ly emitters with close companions at > 7, concluding that mergers may drive Ly emission in these early systems and facilitate the escape of Ly photons from ionized bubbles (see also Saxena et al. 2023;Witstok et al. 2024). ...

The connection between mergers and AGN activity in simulated and observed massive galaxies

Monthly Notices of the Royal Astronomical Society

... Recent investigations with JWST and ALMA have found that galactic disks (Ferreira et al. 2022;Wu et al. 2023;Kartaltepe et al. 2023;Nelson et al. 2023;Hamilton-Campos et al. 2023;Sun et al. 2024;Lian & Luo 2024), some of which are dynamically cold (Fraternali et al. 2021;Xu et al. 2024;Rowland et al. 2024), as well as bars and spiral arms therein (Robertson et al. 2023;Guo et al. 2023;Huang et al. 2023;Le Conte et al. 2024;Kuhn et al. 2024), seem to develop early in the cosmic time. These exciting findings hint at a possible revision to the secularly versus merger-built bulge-like structures in the Universe. ...

The Physical Thickness of Stellar Disks to z ∼ 2

The Astrophysical Journal

... Disk properties, such as gas fraction, contribute to clump longevity (Renaud et al. 2024), but stellar and supernova feedback are likely the primary factors. Thus, clumps hold critical insights into feedback processes in galaxies (Mayer et al. 2016;Ceverino et al. 2023;Dekel et al. 2022;cf. Fensch & Bournaud 2020), helping to constrain them. ...

Effects of feedback on galaxies in the VELA simulations: elongation, clumps and compaction
  • Citing Article
  • April 2023

Monthly Notices of the Royal Astronomical Society

... The time-correlated noise operating on timescales of ∼0.3-5 minutes can be explained by thermal cycling of the heaters (J. Rigby et al. 2023). However, a subset of the lightcurves additionally show lower-frequency flux undulations. ...

The Science Performance of JWST as Characterized in Commissioning

Publications of the Astronomical Society of the Pacific

... A host of observational results is revolutionising this field of research, including the discovery of dynamically cold galactic discs at very high redshift (1 < z < 8) with ALMA and JWST (e.g. Ferreira et al. 2022;Nelson et al. 2023;Ferreira et al. 2023;Kartaltepe et al. 2023;Pope et al. 2023), the existence of already-mature discs containing bars (Costantin et al. 2023a;Guo et al. 2024) and spiral structures (Kuhn et al. 2024) as early as 2 Gyr after the Big Bang. These discoveries concern objects that are expected to be massive, settling to discs earlier than Milky Way-like galaxies (disc downsizing, see Tsukui et al. 2024a), in some cases evolving into ellipticals in the present-day Universe. ...

CEERS Key Paper. III. The Diversity of Galaxy Structure and Morphology at z = 3–9 with JWST

The Astrophysical Journal Letters