skip to main content


Title: The MOSDEF-LRIS Survey: The Interplay Between Massive Stars and Ionized Gas in High-Redshift Star-Forming Galaxies1
Abstract We present a joint analysis of rest-UV and rest-optical spectra obtained using Keck/LRIS and Keck/MOSFIRE for a sample of 62 star-forming galaxies at z ∼ 2.3. We divide our sample into two bins based on their location in the [OIII]5007/Hβ vs. [NII]6584/Hα BPT diagram, and perform the first differential study of the rest-UV properties of massive ionizing stars as a function of rest-optical emission-line ratios. Fitting BPASS stellar population synthesis models, including nebular continuum emission, to our rest-UV composite spectra, we find that high-redshift galaxies offset towards higher [OIII]λ5007/Hβ and [NII]λ6584/Hα have younger ages ($\log (\textrm {~Age/yr})=7.20^{+0.57}_{-0.20}$) and lower stellar metallicities ($Z_*=0.0010^{+0.0011}_{-0.0003}$) resulting in a harder ionizing spectrum, compared to the galaxies in our sample that lie on the local BPT star-forming sequence ($\log (\textrm {Age/yr})=8.57^{+0.88}_{-0.84}$, $Z_*=0.0019^{+0.0006}_{-0.0006}$). Additionally, we find that the offset galaxies have an ionization parameter of $\log (U)=-3.04^{+0.06}_{-0.11}$ and nebular metallicity of ($12+\log (\textrm {~O/H})=8.40^{+0.06}_{-0.07}$), and the non-offset galaxies have an ionization parameter of $\log (U)=-3.11^{+0.08}_{-0.08}$ and nebular metallicity of $12+\log (\textrm {~O/H})=8.30^{+0.05}_{-0.06}$. The stellar and nebular metallicities derived for our sample imply that the galaxies offset from the local BPT relation are more α-enhanced ($7.28^{+2.52}_{-2.82}\textrm {~O/Fe}_{\odot }$) compared to those consistent with the local sequence ($3.04^{+0.95}_{-0.54}\textrm {~O/Fe}_{\odot }$). However, even galaxies that are entirely consistent with the local nebular excitation sequence appear to be α-enhanced – in contrast with typical local systems. Such differences must be considered when estimating gas-phase oxygen abundances at high redshift based on strong emission-line ratios. Specifically, a similarity in the location of high-redshift and local galaxies in the BPT diagram may not be indicative of a similarity in their physical properties.  more » « less
Award ID(s):
2009313 2009085
NSF-PAR ID:
10249682
Author(s) / Creator(s):
; ; ; ; ; ; ;
Date Published:
Journal Name:
Monthly Notices of the Royal Astronomical Society
Volume:
495
Issue:
4
ISSN:
0035-8711
Page Range / eLocation ID:
4430-4444
Format(s):
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. null (Ed.)
    ABSTRACT We analyse the rest-optical emission-line spectra of z ∼ 2.3 star-forming galaxies in the complete MOSFIRE Deep Evolution Field (MOSDEF) survey. In investigating the origin of the well-known offset between the sequences of high-redshift and local galaxies in the [O iii]λ5008/Hβ versus [N ii]λ6585/Hα (‘[N ii] BPT’) diagram, we define two populations of z ∼ 2.3 MOSDEF galaxies. These include the high population that is offset towards higher [O iii]λ5008/Hβ and/or [N ii]λ6585/Hα with respect to the local SDSS sequence and the low population that overlaps the SDSS sequence. These two groups are also segregated within the [O  iii]λ5008/Hβ versus [S ii]λλ6718,6733/Hα and the [O iii]λλ4960,5008/[O ii ]λλ3727,3730 (O32) versus ([O  iii]λλ4960,5008+[O ii]λλ3727,3730)/Hβ (R23) diagrams, which suggests qualitatively that star-forming regions in the more offset galaxies are characterized by harder ionizing spectra at fixed nebular oxygen abundance. We also investigate many galaxy properties of the split sample and find that the high sample is on average smaller in size and less massive, but has higher specific star formation rate (SFR) and SFR surface density values and is slightly younger compared to the low population. From Cloudy+BPASS photoionization models, we estimate that the high population has a lower stellar metallicity (i.e. harder ionizing spectrum) but slightly higher nebular metallicity and higher ionization parameter compared to the low population. While the high population is more α-enhanced (i.e. higher α/Fe) than the low population, both samples are significantly more α-enhanced compared to local star-forming galaxies with similar rest-optical line ratios. These differences must be accounted for in all high-redshift star-forming galaxies – not only those ‘offset’ from local excitation sequences. 
    more » « less
  2. null (Ed.)
    ABSTRACT We present constraints on the massive star and ionized gas properties for a sample of 62 star-forming galaxies at z ∼ 2.3. Using BPASS stellar population models, we fit the rest-UV spectra of galaxies in our sample to estimate age and stellar metallicity which, in turn, determine the ionizing spectrum. In addition to the median properties of well-defined subsets of our sample, we derive the ages and stellar metallicities for 30 high-SNR individual galaxies – the largest sample of individual galaxies at high redshift with such measurements. Most galaxies in this high-SNR subsample have stellar metallicities of 0.001 < Z* < 0.004. We then use Cloudy + BPASS photoionization models to match observed rest-optical line ratios and infer nebular properties. Our high-SNR subsample is characterized by a median ionization parameter and oxygen abundance, respectively, of log (U)med = −2.98 ± 0.25 and 12 + log (O/H)med = 8.48 ± 0.11. Accordingly, we find that all galaxies in our sample show evidence for α-enhancement. In addition, based on inferred log (U) and 12 + log (O/H) values, we find that the local relationship between ionization parameter and metallicity applies at z ∼ 2. Finally, we find that the high-redshift galaxies most offset from the local excitation sequence in the BPT diagram are the most α-enhanced. This trend suggests that α-enhancement resulting in a harder ionizing spectrum at fixed oxygen abundance is a significant driver of the high-redshift galaxy offset on the BPT diagram relative to local systems. The ubiquity of α-enhancement among z ∼ 2.3 star-forming galaxies indicates important differences between high-redshift and local galaxies that must be accounted for in order to derive physical properties at high redshift. 
    more » « less
  3. ABSTRACT

    The ionizing photon escape fraction [Lyman continuum (LyC) fesc] of star-forming galaxies is the single greatest unknown in the reionization budget. Stochastic sightline effects prohibit the direct separation of LyC leakers from non-leakers at significant redshifts. Here we circumvent this uncertainty by inferring fesc using resolved (R > 4000) Lyman α (Lyα) profiles from the X-SHOOTER Lyα survey at z = 2 (XLS-z2). With empirically motivated criteria, we use Lyα profiles to select leakers ($f_{\mathrm{ esc}} > 20{{\ \rm per\ cent}}$) and non-leakers ($f_{\mathrm{ esc}} < 5{{\ \rm per\ cent}}$) from a representative sample of >0.2L* Lyman α emitters (LAEs). We use median stacked spectra of these subsets over λrest ≈ 1000–8000 Å to investigate the conditions for LyC fesc. Our stacks show similar mass, metallicity, MUV, and βUV. We find the following differences between leakers versus non-leakers: (i) strong nebular C iv and He ii emission versus non-detections; (ii) [O iii]/[O ii] ≈ 8.5 versus ≈3; (iii) Hα/Hβ indicating no dust versus E(B − V) ≈ 0.3; (iv) Mg ii emission close to the systemic velocity versus redshifted, optically thick Mg ii; and (v) Lyα fesc of ${\approx} 50{{\ \rm per\ cent}}$ versus ${\approx} 10{{\ \rm per\ cent}}$. The extreme equivalent widths (EWs) in leakers ([O iii]+$\mathrm{ H}\beta \approx 1100$ Å rest frame) constrain the characteristic time-scale of LyC escape to ≈3–10 Myr bursts when short-lived stars with the hardest ionizing spectra shine. The defining traits of leakers – extremely ionizing stellar populations, low column densities, a dust-free, high-ionization state interstellar medium (ISM) – occur simultaneously in the $f_{\rm esc} > 20{{\ \rm per\ cent}}$ stack, suggesting they are causally connected, and motivating why indicators like [O iii]/[O ii] may suffice to constrain fesc at z > 6 with the James Webb Space Telescope (JWST). The leakers comprise half of our sample, have a median LyC$f_{\rm esc} \approx 50{{\ \rm per\ cent}}$ (conservative range: $20\!-\!55{{\ \rm per\ cent}}$), and an ionizing production efficiency $\log ({\xi _{\rm {ion}}/\rm {Hz\ erg^{-1}}})\approx 25.9$ (conservative range: 25.7–25.9). These results show LAEs – the type of galaxies rare at z ≈ 2, but that become the norm at higher redshift – are highly efficient ionizers, with extreme ξion and prolific fesc occurring in sync.

     
    more » « less
  4. Abstract

    The Baldwin, Philips, & Terlevich diagram of [Oiii]/Hβversus [Nii]/Hα(hereafter N2-BPT) has long been used as a tool for classifying galaxies based on the dominant source of ionizing radiation. Recent observations have demonstrated that galaxies atz∼ 2 reside offset from local galaxies in the N2-BPT space. In this paper, we conduct a series of controlled numerical experiments to understand the potential physical processes driving this offset. We model nebular line emission in a large sample of galaxies, taken from thesimbacosmological hydrodynamic galaxy formation simulation, using thecloudyphotoionization code to compute the nebular line luminosities from Hiiregions. We find that the observed shift toward higher [Oiii]/Hβand [Nii]/Hαvalues at high redshift arises from sample selection: when we consider only the most massive galaxiesM*∼ 1010–11M, the offset naturally appears, due to their high metallicities. We predict that deeper observations that probe lower-mass galaxies will reveal galaxies that lie on a locus comparable toz∼ 0 observations. Even when accounting for samples-selection effects, we find that there is a subtle mismatch between simulations and observations. To resolve this discrepancy, we investigate the impact of varying ionization parameters, Hiiregion densities, gas-phase abundance patterns, and increasing radiation field hardness on N2-BPT diagrams. We find that either decreasing the ionization parameter or increasing the N/O ratio of galaxies at fixed O/H can move galaxies along a self-similar arc in N2-BPT space that is occupied by high-redshift galaxies.

     
    more » « less
  5. ABSTRACT

    We present detections of [O iii] λ4363 and direct-method metallicities for star-forming galaxies at z = 1.7–3.6. We combine new measurements from the MOSFIRE Deep Evolution Field (MOSDEF) survey with literature sources to construct a sample of 18 galaxies with direct-method metallicities at z > 1, spanning 7.5 < 12+log(O/H) < 8.2 and log(M*/M⊙) = 7–10. We find that strong-line calibrations based on local analogues of high-redshift galaxies reliably reproduce the metallicity of the z > 1 sample on average. We construct the first mass–metallicity relation at z > 1 based purely on direct-method O/H, finding a slope that is consistent with strong-line results. Direct-method O/H evolves by ≲0.1 dex at fixed M* and star formation rate from z ∼ 0 to 2.2. We employ photoionization models to constrain the ionization parameter and ionizing spectrum in the high-redshift sample. Stellar models with supersolar O/Fe and binary evolution of massive stars are required to reproduce the observed strong-line ratios. We find that the z > 1 sample falls on the z ∼ 0 relation between ionization parameter and O/H, suggesting no evolution of this relation from z ∼ 0 to z ∼ 2. These results suggest that the offset of the strong-line ratios of this sample from local excitation sequences is driven primarily by a harder ionizing spectrum at fixed nebular metallicity compared to what is typical at z ∼ 0, naturally explained by supersolar O/Fe at high redshift caused by rapid formation time-scales. Given the extreme nature of our z > 1 sample, the implications for representative z ∼ 2 galaxy samples at ∼1010 M⊙ are unclear, but similarities to z > 6 galaxies suggest that these conclusions can be extended to galaxies in the epoch of reionization.

     
    more » « less