arXiv subjects

M. S. Oey

Publications and source records attributed to M. S. Oey.

At least 55 records · Page 3Linked to original sources

arXiv2022

Haro 11 -- Untying the knots of the nuclear starburst

Star formation is a clustered process that regulates the structure and evolution of galaxies. We investigate this process in the dwarf galaxy Haro 11, forming stars in three knots (A, B, C). The exquisite resolution of HST imaging allows us to resolve the starburst into tens of bright star clusters. We derive masses between $10^5$ and $10^7\,\rm M_{\odot}$ and ages younger than 20 Myr, using photometric modeling. We observe that the clustered star formation has propagated from knot C (the oldest) through knot A (in between) towards knot B (the youngest). We use aperture-matched ultraviolet and optical spectroscopy (HST + MUSE) to independently study the stellar populations of Haro 11 and determine the physical properties of the stellar populations and their feedback in 1 kpc diameter regions. We discuss these results in light of the properties of the ionised gas within the knots. We interpret the broad blue-shifted components of the optical emission lines as outflowing gas ($v_{max} \sim 400$ km/s). The strongest outflow is detected in knot A with a mass-rate of $\dot{M}_{out}\sim 10\,\rm M_{\odot}/yr$, ten times higher than the star-formation in the same region. Knot B hosts a young and not fully developed outflow, whereas knot C has likely been already evacuated. Because Haro 11 has properties similar to high-redshift unresolved galaxies, our work can additionally aid the understanding of star formation at high redshift, a window that will be opened by upcoming facilities.

astro-ph.GA↗

arXiv2021

Catastrophic Cooling in Superwinds. II. Exploring the Parameter Space

Superwinds and superbubbles driven by mechanical feedback from super star clusters (SSCs) are common features in many star-forming galaxies. While the adiabatic fluid model can well describe the dynamics of superwinds, several observations of starburst galaxies revealed the presence of compact regions with suppressed superwinds and strongly radiative cooling, i.e., catastrophic cooling. In the present study, we employ the non-equilibrium atomic chemistry and cooling package MAIHEM, built on the FLASH hydrodynamics code, to generate a grid of models investigating the dependence of cooling modes on the metallicity, SSC outflow parameters, and ambient density. While gas metallicity plays a substantial role, catastrophic cooling is more sensitive to high mass-loading and reduced kinetic heating efficiency. Our hydrodynamic simulations indicate that the presence of a hot superbubble does not necessarily imply an adiabatic outflow, and vice versa. Using CLOUDY photoionization models, we predict UV and optical line emission for both adiabatic and catastrophic cooling outflows, for radiation-bounded and partially density-bounded models. Although the line ratios predicted by our radiation-bounded models agree well with observations of star-forming galaxies, they do not provide diagnostics that unambiguously distinguish the parameter space of catastrophically cooling flows. Comparison with observations suggests the possibility of minor density bounding, non-equilibrium ionization, and/or observational bias toward the central outflow regions.

astro-ph.GA↗

arXiv2021

Non-equilibrium Photoionization and Hydrodynamic Simulations of Starburst-driven Outflows

Starburst-driven galactic outflows in star-forming galaxies have been observed to contain complex thermal structures and emission line features that are difficult to explain by adiabatic fluid models and plasmas in photoionization equilibrium (PIE) and collisional ionization equilibrium (CIE). We previously performed hydrodynamic simulations of starburst-driven outflows using the MAIHEM module for non-equilibrium atomic chemistry and radiative cooling functions in the hydrodynamics code FLASH, and calculated emission lines in combined CIE and PIE conditions. In the present study, we consider time-dependent non-equilibrium ionization (NEI) states produced by our MAIHEM simulations. Through extensive CLOUDY calculations made with the NEI states from our hydrodynamic simulations, we predict the UV and optical emission line profiles for starburst-driven outflows in time-evolving non-equilibrium photoionization conditions. Our hydrodynamic results demonstrate applications of non-equilibrium radiative cooling for H II regions in starburst galaxies hosting cool outflows.

astro-ph.GA↗

arXiv2021

Emission-line Wings Driven by Lyman Continuum in the Green Pea Analog Mrk 71

We propose that the origin of faint, broad emission-line wings in the Green Pea (GP) analog Mrk 71 is a clumpy, LyC and/or Ly$α$-driven superwind. Our spatially-resolved analysis of Gemini-N/GMOS-IFU observations shows that these line wings with terminal velocity $>3000~\rm{km~s^{-1}}$ originate from the super star cluster (SSC) Knot A, and propagate to large radii. The object's observed ionization parameter and stellar surface density are close to their theoretical maxima, and radiation pressure dominates over gas pressure. Together with a lack of evidence for supernova feedback, these imply a radiation-dominated environment. We demonstrate that a clumpy, radiation-driven superwind from Knot A is a viable model for generating the extreme velocities, and in particular, that Lyman continuum and/or Ly$α$ opacity must be responsible. We find that the Mrk 71 broad wings are best fitted with power laws, as are those of a representative extreme GP and a luminous blue variable star, albeit with different slopes. This suggests that they may share a common wind-acceleration mechanism. We propose that high-velocity, power-law wings may be a distinctive signature of radiation feedback, and of radiatively-driven winds, in particular.

astro-ph.GA↗

arXiv2021

Limits to Ionization-Parameter Mapping as a Diagnostic of HII Region Optical Depth

We employ ionization-parameter mapping (IPM) to infer the optical depth of HII regions in the northern half of M33. We construct [OIII]$λ5007$/[OII]$λ3727$ and [OIII]$λ5007$/[SII]$λ6724$ ratio maps from narrow-band images continuum-subtracted in this way, from which we classify the HII regions by optical depth to ionizing radiation, based on their ionization structure. This method works relatively well in the low metallicity regime, $12 + \log(\rm O/H) \leq 8.4$, where [OIII]$λ\lambda4949,5007$ is strong. However, at higher metallicities, the method breaks down due to the strong dependence of the [OIII]$λ\lambda4959,5007$ emission lines on the nebular temperature. Thus, although O$^{++}$ may be present in metal-rich HII regions, these commonly used emission lines do not serve as a useful indicator of its presence, and hence, the O ionization state. In addition, IPM as a diagnostic of optical depth is limited by spatial resolution. We also report a region of highly excited [OIII] extending over an area $\sim$ 1 kpc across and [OIII]$\lambda5007$ luminosity of $4.9\pm 1.5\times10^{38}$ erg/s, which is several times higher than the ionizing budget of any potential sources in this portion of the galaxy. Finally, this work introduces a new method for continuum subtraction of narrow-band images based on the dispersion of pixels around the mode of the diffuse-light flux distribution. In addition to M33, we demonstrate the method on C III]$λ$1909 imaging of Haro~11, ESO 338-IG004, and Mrk~71.

astro-ph.GA↗

arXiv2021

The Low-redshift Lyman-continuum Survey: [S II]-deficiency and the leakage of ionizing radiation

The relationship between galaxy characteristics and the reionization of the universe remains elusive, mainly due to the observational difficulty in accessing the Lyman continuum (LyC) at these redshifts. It is thus important to identify low-redshift LyC-leaking galaxies that can be used as laboratories to investigate the physical processes that allow LyC photons to escape. The weakness of the [S II] nebular emission lines relative to typical star-forming galaxies has been proposed as a LyC predictor. In this paper, we show that the [S II]-deficiency is an effective method to select LyC-leaking candidates using data from the Low-redshift LyC Survey, which has detected flux below the Lyman edge in 35 out of 66 star-forming galaxies with the Cosmic Origins Spectrograph onboard the Hubble Space Telescope. We show that LyC leakers tend to be more [S II]-deficient and that the fraction of their detections increases as [S II]-deficiency becomes more prominent. Correlational studies suggest that [S II]-deficiency complements other LyC diagnostics (such as strong Lyman-$α$ emission and high [O III]/[O II]). Our results verify an additional technique by which reionization-era galaxies could be studied.

astro-ph.GA↗

arXiv2021

The source of leaking ionizing photons from Haro11 -- Clues from HST/COS spectroscopy of knots A, B and C

Understanding the escape of ionizing (Lyman continuum) photons from galaxies is vital for determining how galaxies contributed to reionization in the early universe. While directly detecting Lyman continuum from high redshift galaxies is impossible due to the intergalactic medium, low redshift galaxies in principle offer this possibility, but requirie observations from space. The first local galaxy for which Lyman continuum escape was found is Haro11 , a luminous blue compact galaxy at z=0.02, where observations with the FUSE satellite revealed an escape fraction of 3.3 %. However the FUSE aperture covers the entire galaxy, and it is not clear from where the Lyman continuum is leaking out. Here we utilize HST/COS spectroscopy in the wavelength range 1100-1700 A of the three knots (A, B, and C) of Haro11 to study the presence of Ly-$α$ emission and the properties of intervening gas. We find that all knots have bright Ly-$α$ emission. UV absorption lines, originating in the neutral interstellar medium, as well as lines probing the ionized medium, are seen extending to blue shifted velocities of 500 km/s in all three knots, demonstrating the presence of an outflowing multiphase medium. We find that knots A and B have large covering fractions of neutral gas, making LyC escape along these sightlines improbable, while knot C has a much lower covering fraction ($\lesssim50$%). Knot C also has the the highest Ly-$α$ escape fraction and we conclude that it is the most likely source of the escaping Lyman continuum detected in Haro11.

astro-ph.GA↗

arXiv2020

Spatially resolved CIII]$λ$1909 emission in Haro 11

The CIII]1909 (hereafter, C III]) line is the strongest ultraviolet emission line after Ly$α$ and is therefore of interest to high redshift studies of star-forming (SF) galaxies near the epoch of reionization. It is thought that C III] emission is strongest in galaxies with subsolar metallicity and low mass, however, spectral observations of numerous such galaxies at high and low redshift produce inconclusive or even contradictory results. We present the first ever C III ] imaging, obtained with HST/STIS for the low-redshift SF galaxy Haro 11. Cluster parameters like stellar mass, dust fraction and dust attenuation, and ionization parameter, obtained through spectral energy distribution fitting, show no correlation with the CIII] equivalent width (EW), which may be due to a combination of the limitation of the models and the age-homogeneity of the cluster population. Comparing the ratio of C III] emission line flux from individual clusters to that of H$α$, [O III], and [O II] we find that the clusters with the highest EW(C III]) can be reconciled only with Cloudy models with an extremely high C/O ratio of $\ge$ 1.4(C/O)$_\odot$ for an ionizing population of single stars, binary stars, or a mixture of binary stars and active galactic nuclei. Given the point-like nature of strong C III], the integrated total strength of EW(C III]) becomes dependent on the morphology of the galaxy, which would explain the large scatter in EW(C III]) strengths, observed in galaxies with otherwise similar SF properties, and of similarly low metallicity and stellar mass.

astro-ph.GA↗

arXiv2020

Runaway OB Stars in the Small Magellanic Cloud: Dynamical Versus Supernova Ejections

Runaway OB stars are ejected from their parent clusters via two mechanisms, both involving multiple stars: the dynamical ejection scenario (DES) and the binary supernova scenario (BSS). We constrain the relative contributions from these two ejection mechanisms in the Small Magellanic Cloud (SMC) using data for 304 field OB stars from the spatially complete, Runaways and Isolated O-Type Star Spectroscopic Survey of the SMC (RIOTS4). We obtain stellar masses and projected rotational velocities $v_r\sin i $ for the sample using RIOTS4 spectra, and use transverse velocities $v_{\rm loc}$ from $\it{Gaia}$ DR2 proper motions. Kinematic analyses of the masses, $v_r\sin i $, non-compact binaries, high-mass X-ray binaries, and Oe/Be stars largely support predictions for the statistical properties of the DES and BSS populations. We find that dynamical ejections dominate over supernova ejections by a factor of $\sim 2-3$ in the SMC, and our results suggest a high frequency of DES runaways and binary ejections. Objects seen as BSS runaways also include two-step ejections of binaries that are reaccelerated by SN kicks. We find that two-step runaways likely dominate the BSS runaway population. Our results further imply that any contribution from $\it{in-situ}$ field OB star formation is small. Finally, our data strongly support the post-mass-transfer model for the origin of classical Oe/Be stars, providing a simple explanation for the bimodality in the $v_r\sin i $ distribution and high, near-critical, Oe/Be rotation velocities. The close correspondence of Oe/Be stars with BSS predictions implies that the emission-line disks are long-lived.

astro-ph.SR↗

arXiv2020

A Search for In-Situ Field OB Star Formation in the Small Magellanic Cloud

Whether any OB stars form in isolation is a question central to theories of massive star formation. To address this, we search for tiny, sparse clusters around 210 field OB stars from the Runaways and Isolated O-Type Star Spectroscopic Survey of the SMC (RIOTS4), using friends-of-friends (FOF) and nearest neighbors (NN) algorithms. We also stack the target fields to evaluate the presence of an aggregate density enhancement. Using several statistical tests, we compare these observations with three random-field datasets, and we also compare the known runaways to non-runaways. We find that the local environments of non-runaways show higher aggregate central densities than for runaways, implying the presence of some "tips-of-iceberg" (TIB) clusters. We find that the frequency of these tiny clusters is low, $\sim 4-5\%$ of our sample. This fraction is much lower than some previous estimates, but is consistent with field OB stars being almost entirely runaway and walkaway stars. The lack of TIB clusters implies that such objects either evaporate on short timescales, or do not form, implying a higher cluster lower-mass limit and consistent with a relationship between maximum stellar mass ($m_{\rm max}$) and the mass of the cluster ($M_{\rm cl}$). On the other hand, we also cannot rule out that some OB stars may form in highly isolated conditions. Our results set strong constraints on the formation of massive stars in relative isolation.

astro-ph.SR↗

arXiv2020

Size distribution of superbubbles

We consider the size distribution of superbubbles in a star forming galaxy. Previous studies have tried to explain the distribution by using adiabatic self-similar evolution of wind driven bubbles, assuming that bubbles stall when pressure equilibrium is reached. We show, with the help of hydrodynamical numerical simulations, that this assumption is not valid. We also include radiative cooling of shells. In order to take into account non-thermal pressure in the ambient medium, we assume an equivalent higher temperature than implied by thermal pressure alone. Assuming that bubbles stall when the outer shock speed becomes comparable to the ambient sound speed (which includes non-thermal components), we recover the size distribution with a slope of $\sim -2.7$ for typical values of ISM pressure in Milky Way, which is consistent with observations. Our simulations also allow us to follow the evolution of size distribution in the case of different values of non-thermal pressure, and we show that the size distribution steepens with lower pressure, to slopes intermediate between only-growing and only-stalled cases.

astro-ph.GA↗

arXiv2020

Accretion-Driven Sources in Spatially Resolved Ly$α$ Emitters

Ly$α$ emission is a standard tracer of starburst galaxies at high redshift. However, a number of local Ly$α$ emitters (LAEs) are X-ray sources, suggesting a possible origin of Ly$α$ photons other than young, hot stars, and which may be active at much later ages relative to the parent starburst. Resolved, nearby LAEs offer the opportunity to discriminate between diffuse X-ray emission arising from supernova-heated gas, high-mass X-ray binaries (HMXBs), or low-luminosity active galactic nuclei (LLAGN). We examine archival X-ray imaging from Chandra and XMM-Newton for 11 galaxies with spatially resolved Ly$α$ imaging to determine the luminosity, morphology, and spectral hardness of the X-ray sources. The data are consistent with 9 of the 12, bright Ly$α$ sources being driven by luminous, $10^{40}$ erg s$^{-1}$ X-ray sources. Half of the 8 Chandra sources are unresolved. The data suggest that nuclear activity, whether from LLAGN or nuclear starbursts, may play an important role in Ly$α$ emission. Our results also suggest a significant link between Ly$α$ emission and HMXBs, ULXs, and/or LLAGN, which would imply that Ly$α$ may be generated over timescales 1 - 2 orders of magnitude longer than produced by photoionization from OB stars. This highlights a critical need to quantify the relative contributions of different sources across cosmic time, to interpret Ly$α$ observations and the resulting properties of distant galaxies.

astro-ph.GA↗

arXiv2019

Catastrophic Cooling in Superwinds: Line Emission and Non-equilibrium Ionization

Outflows are a pervasive feature of mechanical feedback from super star clusters (SSC) in starburst galaxies, playing a fundamental role in galaxy evolution. Observations are now starting to confirm that outflows can undergo catastrophic cooling, suppressing adiabatic superwinds. Here we present a suite of one-dimensional, hydrodynamic simulations that study the ionization structure of these outflows and the resulting line emission generated by the cooling gas. We use the non-equilibrium atomic chemistry package within MAIHEM, our modified version of FLASH, which evolves the ionization state of the gas and computes the total cooling rate on an ion-by-ion basis. We find that catastrophically cooling models produce strong nebular line emission compared to adiabatic outflows. We also show that such models exhibit non-equilibrium conditions, thereby generating more highly ionized states than equivalent equilibrium models. When including photoionization from the parent SSC, catastrophically cooling models show strong C IV λ1549 and O VI λ1037 emission. For density bounded photoionization, He II λ1640, λ4686, C III] λ1908, Si IV λ1206, and Si III λ1400 are also strongly enhanced. These lines are seen in extreme starbursts where catastrophic cooling is likely to occur, suggesting that they may serve as diagnostics of such conditions. The higher ionization generated by these flows may help to explain line emission that cannot be attributed to SSC photoionization alone.

astro-ph.GA↗

arXiv2019

New Insights on Ly-alpha and Lyman Continuum Radiative Transfer in the Greenest Peas

As some of the only Lyman continuum (LyC) emitters at z~0, Green Pea (GP) galaxies are possible analogs of the sources that reionized the universe. We present HST COS spectra of 13 of the most highly ionized GPs, with [O III]/[O II]=6-35, and investigate correlations between Ly-alpha, galaxy properties, and low-ionization UV lines. Galaxies with high [O III]/[O II] have higher H-alpha equivalent widths (EWs), and high intrinsic Ly-alpha production may explain the prevalence of high Ly-alpha EWs among GPs. While Ly-alpha escape fraction is closely linked to low gas covering fractions, implying a clumpy gas geometry, narrow Ly-alpha velocity peak separation (delta_v,LyA) correlates with the ionization state, suggesting a density-bounded geometry. We therefore suggest that delta_v,LyA may trace the residual transparency of low-column-density pathways. Metallicity is associated with both [O III]/[O II] and delta_v,LyA. This trend may result from catastrophic cooling around low-metallicity star clusters, which generates a compact geometry of dense clouds within a low-density inter-clump medium. We find that the relative strength of low-ionization UV emission to absorption correlates with Ly-alpha emission strength and is related to Ly-alpha profile shape. However, as expected for optically thin objects, the GPs with the lowest delta_v,LyA show both weak low-ionization emission and weak absorption. The strengths of the low-ionization absorption and emission lines in a stacked spectrum do not correspond to any individual spectrum. Galaxies with high [O III]/[O II] contain a high fraction of LyC emitter candidates, but [O III]/[O II] alone is an insufficient diagnostic of LyC escape.

astro-ph.GA↗

arXiv2019

Radial abundance gradients in the outer Galactic disk as traced by main-sequence OB stars

Using a sample of 31 main-sequence OB stars located between galactocentric distances 8.4 - 15.6 kpc, we aim to probe the present-day radial abundance gradients of the Galactic disk. The analysis is based on high-resolution spectra obtained with the MIKE spectrograph on the Magellan Clay 6.5-m telescope on Las Campanas. We used a non-NLTE analysis in a self-consistent semi-automatic routine based on TLUSTY and SYNSPEC to determine atmospheric parameters and chemical abundances. Stellar parameters (effective temperature, surface gravity, projected rotational velocity, microturbulence, and macroturbulence) and silicon and oxygen abundances are presented for 28 stars located beyond 9 kpc from the Galactic centre plus three stars in the solar neighborhood. The stars of our sample are mostly on the main-sequence, with effective temperatures between 20800 - 31300 K, and surface gravities between 3.23 - 4.45 dex. The radial oxygen and silicon abundance gradients are negative and have slopes of -0.07 dex/kpc and -0.09 dex/kpc, respectively, in the region $8.4 \leq R_G \leq 15.6$\,kpc. The obtained gradients are compatible with the present-day oxygen and silicon abundances measured in the solar neighborhood and are consistent with radial metallicity gradients predicted by chemodynamical models of Galaxy Evolution for a subsample of young stars located close to the Galactic plane.

astro-ph.SR↗

arXiv2019

Neutral Gas Properties and Ly$α$ Escape in Extreme Green Pea Galaxies

Mechanisms regulating the escape of Ly$α$ photons and ionizing radiation remain poorly understood. To study these processes we analyze VLA 21cm observations of one Green Pea (GP), J160810+352809 (hereafter J1608), and HST COS spectra of 17 GP galaxies at $z<0.2$. All are highly ionized: J1608 has the highest [O III] $\lambda5007$/[O II] $\lambda3727$ for star-forming galaxies in SDSS, and the 17 GPs have [O III]/[O II] $\geq6.6$. We set an upper limit on J1608's HI mass of $\log M_{HI}/M_\odot=8.4$, near or below average compared to similar mass dwarf galaxies. In the COS sample, eight GPs show Ly$α$ absorption components, six of which also have Ly$α$ emission. The HI column densities derived from Ly$α$ absorption are high, $\log N_{HI}/$cm$^{-2}=19-21$, well above the LyC optically thick limit. Using low-ionization absorption lines, we measure covering fractions ($f_{\mbox{cov}}$) of $0.1-1$, and find that $f_{\mbox{cov}}$ strongly anti-correlates with Ly$α$ escape fraction. Low covering fractions may facilitate Ly$α$ and LyC escape through dense neutral regions. GPs with $f_{\mbox{cov}}\sim1$ all have low neutral gas velocities, while GPs with lower $f_{\mbox{cov}}=0.2-0.6$ have a larger range of velocities. Conventional mechanical feedback may help establish low $f_{\mbox{cov}}$ in some cases, whereas other processes may be important for GPs with low velocities. Finally, we compare $f_{\mbox{cov}}$ with proposed indicators of LyC escape. Ionizing photon escape likely depends on a combination of neutral gas geometry and kinematics, complicating the use of emission-line diagnostics for identifying LyC emitters.

astro-ph.GA↗

arXiv2018

Resolved Kinematics of Runaway and Field OB Stars in the Small Magellanic Cloud

We use GAIA DR2 proper motions of the RIOTS4 field OB stars in the Small Magellanic Cloud (SMC) to study the kinematics of runaway stars. The data reveal that the SMC Wing has a systemic peculiar motion relative to the SMC Bar of (v_RA, v_Dec) = (62 +/-7, -18+/-5) km/s and relative radial velocity +4.5 +/- 5.0 km/s. This unambiguously demonstrates that these two regions are kinematically distinct: the Wing is moving away from the Bar, and towards the Large Magellanic Cloud with a 3-D velocity of 64 +/- 10 km/s. This is consistent with models for a recent, direct collision between the Clouds. We present transverse velocity distributions for our field OB stars, confirming that unbound runaways comprise on the order of half our sample, possibly more. Using eclipsing binaries and double-lined spectroscopic binaries as tracers of dynamically ejected runaways, and high-mass X-ray binaries (HMXBs) as tracers of runaways accelerated by supernova kicks, we find significant contributions from both populations. The data suggest that HMXBs have lower velocity dispersion relative to dynamically ejected binaries, consistent with the former corresponding to less energetic supernova kicks that failed to unbind the components. Evidence suggests that our fast runaways are dominated by dynamical, rather than supernova, ejections.

astro-ph.GA↗

arXiv2018

The spectroscopic Hertzsprung-Russell diagram of hot massive stars in the SMC

We present a comprehensive stellar atmosphere analysis of 329 O- and B-type stars in the Small Magellanic Cloud (SMC) from the RIOTS4 survey. Using spectroscopically derived effective temperature (Teff) and surface gravities, we find that classical Be stars appear misplaced to low Teff and high luminosity in the spectroscopic Hertzsprung-Russell diagram (sHRD). Together with the most luminous stars in our sample, the stellar masses derived from the sHRD for these objects are systematically larger than those obtained from the conventional HRD. This suggests that the well-known, spectroscopic mass-discrepancy problem may be linked to the fact that both groups of stars have outer envelopes that are nearly gravitationally unbound. The non-emission-line stars in our sample mainly appear on the main-sequence, allowing a first estimate of the terminal-age main-sequence (TAMS) in the SMC, which matches the predicted TAMS between 12 and 40$\,$M$_{\odot}$ at SMC metallicity. We further find a large underabundance of stars above $\sim 25\,$M$_{\odot}$ near the ZAMS, reminiscent of such earlier findings in the Milky Way and LMC.

astro-ph.SR↗