Host Galaxy Properties and Offset Distributions of Fast Radio Bursts: Implications for their Progenitors
Summary (6 min read)
2. Sample and Observations
- In collaboration with the CRAFT (Macquart et al. 2010) and realfast (Law et al. 2018) surveys, the authors have as part of the Fast and Fortunate for FRB Follow-up (F4)18 collaboration endeavored to obtain dedicated photometric and spectroscopic follow-up observations of all ∼arcsecond-localized FRBs.
- These provide a secure identification of the associated host galaxies and allow us to derive their main physical properties.
- All the observational data products are available on the FRB GitHub repository,19 in addition to a large suite of FRB-related scripts.
- The authors then present the new observations of five FRB-host galaxies and compile all previously known FRB hosts reported in the literature, all considered in their meta analysis.
- At the end, the authors summarize the overall sample properties.
2.1. Host-galaxy Associations
- An FRB signal alone cannot directly establish the redshift of the source, and one relies on an association with a host galaxy for a precise measurement.
- To date and in this work, the association of the FRB with a host galaxy is primarily based on probabilistic arguments given their position relative to coincident or nearby galaxies.
- Further work may adopt additional properties and priors for establishing associations.
- Same as Sample C, except that only a photometric redshift zphot has been measured, also known as 4. Sample D.
2.3. Literature Compilation
- In addition to the five new FRB hosts presented here, the authors include all other known FRB-host galaxies in their analysis.
- The authors separate them primarily by FRB survey.
- For the hosts previously reported by Bhandari et al. (2020b), the authors simply include their reported measurements here.
- Finally, the authors used the photometry reported by Bassa et al. (2017) for the host galaxy of FRB 121102 to model their own spectral energy distribution (SED; see Section 3.1) for consistency with the rest of the sample.
2.4. Overall Sample Properties
- The authors have placed the host of FRB 190523 discovered by Ravi et al. (2019) into SampleC because the poor FRB localization makes the host-galaxy association less secure.
- To be conservative and consistent with the other sample classifications, the authors rely only on the statistical properties of the FRB-host associations here to avoid biasing the host identifications.
- For FRB 190614, Law et al. (2020) identified two potential hostgalaxy candidates, for which only photometric redshifts have been obtained, placing it in SampleD.
- This more than doubles the number of ASKAP-detected FRB hosts studied in their previous work (Bhandari et al. 2020b).
2.5. Repeating and Nonrepeating FRBs
- Throughout the paper, the authors distinguish the hosts of the three FRBs that are currently known to repeat (FRBs 121102, 180916, and 190711) from the hosts of the other apparently nonrepeating one-off FRBs.
- This could imply a different emission mechanism for repeating and one-off sources.
- The authors caution that it is possible that FRBs that are currently classified as nonrepeating may exhibit repeat pulses in 6 The Astrophysical Journal, 903:152 (22pp), 2020 November 10 Heintz et al.
- The future, which would change their classification here.
- This may provide additional clues on whether two populations of FRBs exist.
3.1. Stellar Population Modeling
- Following their previous studies of FRB-host galaxies (Bannister et al.
- In all cases, the authors have input observations corrected for Galactic extinction using the E(B−V )Gal values derived from Schlafly & Finkbeiner (2011) and the Fitzpatrick & Massa (2007) extinction law with RV=3.1 implemented in the extinction 21 software package.
- The precise input parameter file for CIGALE is available in the cigale.py module of the FRB repository on GitHub.
3.2. Star Formation Rate
- The authors derive the SFR for each FRB host by first computing the dust-corrected Hα line fluxes using the AV derived from the Balmer decrement to obtain the intrinsic.
- The authors report the uncertainties on the SFR estimates including the scatter in the SFR-LHα relation (≈30%).
- For the three FRB hosts where the Hα line flux has not been measured, the authors derive the SFR from Hβ assuming the nominal relative strength compared to Hα (FRB190523 in Sample C and FRB190711 in Sample A) or from the best-fit SED model from CIGALE (for FRB 200430, Sample A).
- The authors find that the overall sample of FRB hosts are characterized by a large range in SFR, spanning 0.05–10 Me yr −1.
- For the host of FRB 190614, no constraints could be placed on the SFR because the nature of the host galaxy and redshift are uncertain (Sample D; Law et al. 2020).
3.3. Gas-phase Metallicity
- To infer the gas-phase metallicities of the FRB hosts, the authors rely on commonly used diagnostic ratios of strong nebular emission lines (see Maiolino & Mannucci 2019, for a recent review).
- This calibration has been shown to be consistent with more direct Te-based methods, and has an rms uncertainty of 0.111 dex.
- The authors caution that because the oxygen abundances derived using the O3N2 calibration are specifically calibrated to SF galaxies, the actual metallicities might be slightly different if the emission-line ratios do not represent typical SF galaxies (as reported for FRB hosts by Bhandari et al. 2020b, see also Section 4.2).
- 22 Assuming a conversion from the Salpeter-determined SFR of SFRChab= SFRSalp×0.63.
4. Physical Properties of the FRB-host Population
- The physical properties of all the FRB hosts in their sample are summarized in Table 3.
- In the following analysis, the authors examine the FRB-host galaxy environments and place them into context of field-selected galaxies.
- Throughout, the authors separate the hosts of repeating and seemingly nonrepeating, one-off FRBs.
- The authors only consider the 10 FRB hosts in SampleA for the statistical analyses.
4.1. Luminosity and Color
- The authors compare the values of mr to the characteristic luminosity L * across redshift, using available galaxy luminosity functions (Brown et al.
- The authors then consider the color–magnitude properties of the FRB hosts, which is a useful indicator of the overall stellar population in these galaxies.
- These galaxies may be transitioning from SF to the quiescent galaxies of the “red sequence” (Martin et al. 2007).
- Figure 5 further reveals that the FRB hosts do not populate either of the main loci of the blue or red sequences.
4.2. FRB Hosts in the BPT Diagram
- This allows us to assess the dominant source of ionization and distinguish between typical SF galaxies, low-ionization nuclear emissionline region galaxies, and AGNs (see Kewley et al. 2019, for a recent review).
- The authors also include the standard demarcation lines between SF, AGN, and LINER galaxies (Kauffmann et al. 2003; Cid Fernandes et al. 2010).
- Thomas et al. (2018) argue that the excess in line flux ratios can be described by an increased “mixing” of AGN emission with the H II region emission.
- Alternatively, the same ionization effect can be produced by a dominating population of post-asymptotic giant branch stars (Yan & Blanton 2012; Singh et al. 2013).
4.3. Star Formation Rates and Stellar Masses
- For the control sample, the authors again show the galaxies from the PRIMUS survey (Moustakas et al. 2013).
- The hosts of the repeating FRBs are all relatively low-mass galaxies (Må<2×10 9Me) compared to the overall FRB-host population (as described before in Section 4.1).
- Specifically, the authors compare the observed distribution fFRB (Må) with the stellar mass function of low-z galaxies f(Må) weighted by stellar mass, i.e., fFRB (Må)∝Måf(Må).
- The authors first consider all the hosts (top panel) and then only the hosts of the one-off FRBs (bottom panel).
- The uncertainty regions on the cumulative distribution functions (CDFs) are estimated by combining the two sources of uncertainty: the errors on the individual data points, and the error from the sample size.
4.4. Mass–Metallicity Relation
- In addition to the stellar mass and SFR, the gas-phase metallicity is a strong indicator of the present stellar populations and can thus also provide constraints on the most likely progenitor channels.
- Indeed, the typical low-metallicity environments of LGRB host galaxies were vital in the conception of the “collapsar” progenitor model for LGRBs (e.g., Yoon et al. 2006).
- A more direct, quantitative comparison between FRB and LGRB hosts (in addition to the hosts of other types of transients) is provided in Section 5.1.
- For the control sample, the authors show the SF galaxies from the SDSS emission-line sample, with metallicities calibrated using the same strong-line diagnostics as for the FRB hosts (see Section 3.3).
4.5. Locations: Projected Physical and Host-normalized Offsets
- Last the authors consider the projected physical offsets (ρ) of the expanded sample of FRBs, in addition to the projected offsets normalized by the half-light radii of the hosts (ρ/Reff).
- When operating in the ICS mode, ASKAP/CRAFT can now deliver subarcsecond localizations of FRBs upon detection, without requiring the use of follow-up facilities on repeat bursts.
- Indeed, the offset distributions of other transients have provided a key diagnostic for understanding their origins (discussed further in Section 5).
- Overall, the authors find that FRBs have significant offsets relative to the centers of their host galaxies, with median and mean values of 3.3 and 4.8 kpc, respectively.
- Reff with median and mean values of 1.4 Reff and 1.7 Reff, respectively.
5. Implications for FRB Progenitors
- The authors have here shown that FRB hosts exhibit very diverse environments: in particular, they observe a large variety in terms of their morphologies, ranging from early- to late-type galaxies, and found that FRB hosts are characterized by a broad, continuous range of rest-frame colors, luminosities, stellar masses, SFRs and ages.
- The authors now explore the implications for the nature of FRB progenitors through further comparisons of their host-galaxy properties to the hosts of other astronomical transients.
5.1. Comparisons to the Host Properties and Offset Distributions of Other Transients
- The host galaxies of other known transients such as LGRBs, SGRBs, CC, and SNe Ia provide a natural baseline for comparison to the FRB-host population because these have been intensively studied, and have known or likely known progenitors.
- Investigating the connection between their hosts and galaxies hosting FRBs can therefore provide important (though indirect) clues to the most likely FRB progenitor channels.
- Based on the first small samples of FRB-associated hosts (Bhandari et al. 2020b; Li & Zhang 2020), it was already evident that the majority had generally high masses and low SFRs (excluding FRB 121102).
- The authors work has further cemented this picture based on a sample of 10 secure host galaxies.
5.1.1. Luminosity, SFR, and Stellar Mass
- Here, the authors further discuss the connection between FRB hosts and those of other astronomical transients and compare them quantitatively.
- These physical properties are in stark contrast to the typically elevated specific SFRs (sSFR=SFR/Må) observed for the hosts of LGRBs.
- In Figure 10 the authors compare the stellar mass distribution of the FRB hosts to the host galaxies of these other transients, namely SGRBs (Nugent et al. 2020), LGRBs (Vergani et al. 2015), and CC-SNe (Schulze et al. 2020).
- Neither of the SGRB or CC-SNe host populations, however, has statistically inconsistent CDFs.
5.1.2. No Evidence for Metal Aversion in FRB Hosts
- As was demonstrated in Section 4.4, the stellar masses and metallicities of FRB hosts are generally consistent with the mass–metallicity relations observed for field galaxies at z=0.07–0.7.
- This is again consistent with the host galaxies of SGRBs at z<1 (Berger 2009) and CC (Type II)/SNe Ia (Prieto et al. 2008), which are also found to closely track the mass–metallicity or luminosity-metallicity relations of field galaxies at similar redshifts.
- In contrast, the production of LGRBs appears to be heavily suppressed in more metal-rich environments (at least at z<1; Perley et al. 2016) compared to field galaxies at similar masses.
- FRB progenitors show no such metallicity bias in their host galaxies.
5.1.3. Physical and Host-normalized Offsets
- The offset distribution of the full sample of FRBs is found to closely follow the observed distribution of CC and SNe Ia, with onesided KS tests yielding PKS=0.96 and PKS=0.95, respectively.
- The FRB and SGRB offset distributions are also consistent, but a larger fraction of SGRBs are observed to occur at even greater distances from their host-galaxy centers (Fong & Berger 2013), which is more consistent with theoretical expectations of binary neutron star mergers (Fryer & Kalogera 1997; Bloom et al. 1999; Belczynski et al. 2006).
- The gray shaded region represents the 1σuncertainty on the CDF, combining the error on the measurements and due to the sample size (see text for details).
- Figure 9. Mass–metallicity relation of FRB hosts.
5.1.4. Implications for FRB Progenitor Channels
- Because the progenitors of FRBs (and astronomical transients in general) are linked to specific stellar populations and galaxy environments, determining these host-galaxy properties also allows us to place constraints on the likely progenitor channels of FRBs.
- Finally, a metal deficit and high SFR per unit stellar mass seem to be crucial for the probability of a nearby (z<1) galaxy to host LGRBs or SLSNe.
- Indeed, FRBs have been proposed to originate from either young or long-lived stable magnetars produced by a variety of channels, including binary neutron star mergers (at least some of which produce SGRBs) or in the accretion-induced collapse (AIC) of white dwarfs (Moriya 2016; Margalit et al. 2019).
- The FRB-host properties presented here, which span the full range of properties occupied by field galaxies at similar redshifts, are currently consistent with a single progenitor that can accommodate a diverse set of host properties, although contributions from multiple progenitors cannot be ruled out.
- Coupled with the similarities found when compared to both SGRB and SNe.
6. Summary and Outlook
- Here, the authors have presented new observations of five host galaxies of arcsecond-localized FRBs, together with a comprehensive analysis of their stellar population properties (colors, metalliticies, luminosities, stellar masses, massweighted ages, and SFRs) and locations with respect to their host-galaxy centers.
- Moreover, the FRB host-burst offset distribution is consistent with those observed for SGRBs, CC, and SNe Ia, and further disfavors LGRBs (>99% c.l.).
- Establishment of ASKAP, the Murchison Radio-astronomy Observatory, and the Pawsey Supercomputing Centre are initiatives of the Australian Government, with support from the Government of Western Australia and the Science and Industry Endowment Fund.
Did you find this useful? Give us your feedback
Citations
118 citations
Cites background from "Host Galaxy Properties and Offset D..."
...This would be at odds with the offset distribution of the progenitors of long gamma-ray bursts and superluminous supernovae, which are found close to the centers of their respective hosts (Lunnan et al. 2015; Blanchard et al. 2016; Heintz et al. 2020; Mannings et al. 2020)....
[...]
...Heintz et al. (2020) noted that the hosts of apparently nonrepeating FRBs are typically more massive than those of repeating FRBs....
[...]
...However, FRBs, both repeating and apparently nonrepeating, are found in a variety of types of host galaxies, and it is not yet clear if the two populations are intrinsically different (Bhandari et al. 2020; Heintz et al. 2020)....
[...]
...This same issue has been noted for other localized FRBs (Heintz et al. 2020)....
[...]
72 citations
63 citations
52 citations
50 citations
References
15,988 citations
14,295 citations
10,728 citations
10,384 citations
9,745 citations