A Search for the Host Galaxy of FRB 171020

The progenitors and local environments responsible for the bright (F_ν ∼ 0.55–380 Jy ms; Petroff et al. 2016), extragalactic millisecond-duration pulses known as fast radio bursts (FRBs) are open questions. Only a single known repeating FRB (121102; Spitler et al. 2016) has been localized with sufficient accuracy to unambiguously identify its host galaxy (Chatterjee et al. 2017; Tendulkar et al. 2017) and permit examination of the burst environment. Its host galaxy, a low-metallicity dwarf galaxy at redshift z = 0.193, harbors a "persistent" compact radio source (Chatterjee et al. 2017) co-located within 40 pc of the FRB (Marcote et al. 2017). The radio source, and therefore the FRB, likely resides in a bright star-forming region in the outskirts of the galaxy (Bassa et al. 2017). The persistent radio source has inspired suggestions that detections of bright radio emission may help to identify hosts of other FRBs (Eftekhari et al. 2018).

However, the unusual properties of FRB 121102 make it difficult to apply these findings to other FRBs. So far, no other FRBs have been observed to repeat, despite extensive follow-up campaigns (e.g., Bhandari et al. 2018). Radio diagnostics of the repeater's host galaxy and circumburst medium reveal that FRB 121102 resides in a highly magnetized medium (Michilli et al. 2018) whose rotation measure exceeds other FRB measurements by 3–4 orders of magnitude (Masui et al. 2015; Petroff et al. 2017; Caleb et al. 2018).

Here we examine FRB 171020 which has the lowest dispersion measure (DM) measured to date (114 pc cm⁻³; Shannon et al. 2018). No repeat bursts above a signal-to-noise ratio (S/N) of 9 were found in 32.7d of observations. Despite the large localization region of this FRB, 50 × 34 arcmin at a position angle of 296 (95% containment), its low DM demands a sufficiently close proximity to attempt identification of its host galaxy.

In Section 2 we report the properties of FRB 171020 and estimate the maximum redshift, followed by a search for its host galaxy in existing catalogs (Section 3). We discuss follow-up optical and radio observations of our best candidate host galaxy ESO 601–G036 in Section 4, and compare the properties of this galaxy with the host galaxy of FRB 121102 in Section 5, before concluding in Section 6. The cosmological parameters assumed in this Letter are from the Planck 2015 results (Planck Collaboration et al. 2016).

We use the DM of FRB 171020 to estimate an upper limit to its distance. We assume that the total DM is given by

$$\begin{eqnarray*}&&\mathrm{DM}={\mathrm{DM}}_{\mathrm{MW}(\mathrm{disk})}+{\mathrm{DM}}_{\mathrm{MW}(\mathrm{halo})}+{\mathrm{DM}}_{\mathrm{IGM}}+{\mathrm{DM}}_{\mathrm{Host}}\end{eqnarray*}$$

where ${\mathrm{DM}}_{\mathrm{MW}(\mathrm{disk})}$ and ${\mathrm{DM}}_{\mathrm{MW}(\mathrm{halo})}$ are the contributions, respectively, of the Milky Way disk interstellar medium (ISM) and halo, DM_IGM is the contribution of the intergalactic medium along the line of sight, and DM_Host is the contribution from both the host galaxy itself and its circumburst environment. Given that the DM of FRB 171020 is so low, the contribution from the Milky Way represents a significant fraction of the total. As such, the assumptions used to derive these quantities can lead to large fractional uncertainty in the maximum redshift of the FRB.

At the Galactic coordinates (l, b) = (36.4, −53.6) of the burst, DM_MW(disk) is 38 pc cm⁻³ according to the NE2001 model (Cordes & Lazio 2002), or 26 pc cm⁻³ following the YMW16 model (Yao et al. 2017).

The Milky Way halo contribution is much more uncertain; we assume DM_MW(halo) = 12 pc cm⁻³ calculated by using the excess DM of pulsars detected on the near side of the Large Magellanic Cloud (LMC; Manchester et al. 2005). By selecting the closest LMC pulsars it is assumed that there is negligible DM contribution from the LMC itself, and subtracting the Milky Way contribution leads to a halo contribution of 8 < DM_MW(halo) < 13 pc cm⁻³ out to 50 kpc. Extragalactic FRBs travel farther, however, implying a higher ${\mathrm{DM}}_{\mathrm{MW}(\mathrm{halo})}$ that depends on its electron density distribution. Physically plausible models predict ${\mathrm{DM}}_{\mathrm{MW}(\mathrm{halo})}$ with an additional 2–21 pc cm⁻³ at distances >50 kpc (e.g., Miller & Bregman 2013, J. X. Prochaska & Y. Zheng, 2018, in preparation). This is consistent with the estimate of ${\mathrm{DM}}_{\mathrm{MW}(\mathrm{halo})}=30$ pc cm⁻³ (Dolag et al. 2015), but we use a lower estimate here to determine the maximum redshift.

For ${\mathrm{DM}}_{\mathrm{Host}}$, we adopt two different assumptions. One assumes ${\mathrm{DM}}_{\mathrm{Host}}=0\,\mathrm{pc}\,{\mathrm{cm}}^{-3}$, and the other assumes DM_Host = 45 pc cm⁻³ typical for a dwarf galaxy (Xu & Han 2015). Note that ${\mathrm{DM}}_{\mathrm{Host}}$ includes the halo, disk, and circumburst environment components of the host galaxy.

The two assumptions imply a range for DM_IGM of 0–76 pc cm⁻³ (Table 1). Adopting the relation between redshift and DM_IGM of z ≈ DM_IGM/1000 (Ioka 2003; Inoue 2004) yields an upper limit on the redshift of 0.02 ≲ z_max ≲ 0.08. Given the low observed DM of FRB 171020, the contribution from the intergalactic medium (IGM) could vary significantly depending on the number and properties of intervening galaxy halos (McQuinn 2014). However, given that the maximum redshift of z = 0.08 assumes conservative estimates of both the Milky Way halo and host galaxy contributions, any scatter in the DM-z relation is compensated for by the large range in maximum redshifts obtained by using the different models in Table 1.

The area of the localization region of FRB 171020 (0.38 deg²) and its maximum distance of 350 Mpc (using z_max = 0.08 from Model (a) in Table 1) together confine its host galaxy to a maximum comoving volume of 1620 Mpc³. Taking the lower value of z_max = 0.03 from Model (b), the search volume is only 90 Mpc³. These volumes are small enough to search for an optical host galaxy counterpart to FRB 171020 in spite of its poor localization. We searched the NASA Extragalactic Database for cataloged galaxies within the FRB localization ellipse, yielding only two galaxies with a published redshift at z < 0.08.

3.1. ESO 601–G036

3.2. 2MASX J22150112-1925373

3.3. Candidates from the WISE × SCOSPZ Catalog

Given the low redshift and location close to the center of the localization ellipse, ESO 601–G036 is the most likely host galaxy of FRB 171020. In low-resolution images of this galaxy (i.e., DSS) there is a clear stellar tail just south of the galaxy (separately cataloged by Lauberts (1982) as ESO 601−G037). Bright UV emission of both objects, detected in GALEX images, indicates significant star formation that likely originated from tidal interactions or accretion of a dwarf galaxy companion. In the infrared, ESO 601–G036 is detected in the VISTA Hemisphere Survey with Y = 14.53, J = 14.31, Ks = 13.73 mag (Irwin et al. 2004). We estimate a total star formation rate (SFR) of ∼0.13 M_⊙ yr⁻¹ from the GALEX data,¹⁰ and a total stellar mass of ∼9 × 10⁸ M_⊙ from the VISTA Ks-band magnitude.

The HI Parkes All Sky Survey (HIPASS; Barnes et al. 2001) shows H i emission in ESO 601–G036, with an H i-integrated flux density of ≈7 Jy km s⁻¹, corresponding to an H i mass of 2.3 × 10⁹ M_⊙.¹¹ For a rotational velocity of ∼80 km s⁻¹ and maximum radius of 11 kpc, we derive a dynamical mass of ∼1.6 × 10¹⁰ M_⊙.

ESO 601–G036 is a member of a loose galaxy group, with four other gas-rich members detected in HIPASS at similar velocities. These other four galaxies lie outside the FRB 171020 2σ error region.

4.1. Optical Follow-up of ESO 601–G036

We observed ESO 601–G036 using the Gemini Multi-Object Spectrograph (GMOS; Hook et al. 2004; Gimeno et al. 2016) on Gemini-South on 2018 July 11 UT. We used the B600 grating with a 075 wide slit oriented along its major axis for a total of 5 minutes exposure time with wavelength coverage of 3700–6850 Å and dispersion of ≈1 Å/pixel.

The GMOS spectrum gives a redshift measurement of z ≈ 0.0082, consistent with the H i observations. We observe strong Hα, [O iii] emission lines but little [N ii] emission, indicating that this is a low-metallicity galaxy. The flux ratios of log ([N ii] λ6583/Hα) ≈ −1.09 and log{([O iii] λ5007/Hβ)/([N ii] λ6583/Hα)} ≈ 1.41 imply an oxygen abundance 12 + log(O/H) ≈ 8.3 ± 0.2 (Pettini & Pagel 2004). This is consistent with the upper limit of <8.4 found for the host galaxy of FRB 121102 (Tendulkar et al. 2017). The flux line ratios log([O iii] λ5007/Hβ) ≈ 0.32 and log([N ii] λ6583/Hα) ≈ −1.09 place this galaxy in the star-forming region of the Baldwin, Phillips, and Terlevich (BPT) diagram (Baldwin et al. 1981).

In addition, we obtained narrow-band imaging of ESO 601–G036 with GMOS on Gemini-South on 2018 July 15. We observed for an exposure time of 3 × 180 s using the HaC filter (6590–6650 Å), ensuring full coverage of the Hα emission from this galaxy at redshift z ∼ 0.008 (Figure 1).

Zoom In Zoom Out Reset image size

4.2. Radio Continuum Follow-up of ESO 601–G036

To search for a "persistent" radio source, we carried out radio continuum observations using the Australia Telescope Compact Array (ATCA). The initial observations were carried out on 2018 May 27 UT across a wide frequency range to search for a compact radio source in ESO 601−G036 and to compare its spectral energy distribution (SED) with the persistent source of FRB 121102.

Using an extended ATCA array (6D configuration) and Briggs weighting with robust = 0.5 resulted in synthesized beam sizes ranging from 174 × 40 at 2.1 GHz to 18 × 03 at 21.2 GHz. This gives rms noise levels of 40.5, 13.7, 13.2, 21.1, and 31.8 μJy at frequencies of 2.1, 5.5, 9.0, 16.7, and 21.2 GHz, respectively.¹² No radio continuum emission associated with ESO 601−G036 was detected at any frequency.

Figure 2 shows the 5σ flux density limits reached at each frequency. If ESO 601–G036 hosts this FRB, there is no coincident radio source above a radio luminosity of ${L}_{2.1\mathrm{GHz}}=3.6\times {10}^{19}$ W Hz⁻¹ (5σ), i.e., 600 times fainter than that seen in the repeating FRB.

Zoom In Zoom Out Reset image size

Subsequent observations were carried out over the entire localization area at 1–3 GHz using a more compact array configuration (1.5D) on 2018 June 28 UT. The increased sensitivity (rms = 13.0 μJy) and lower resolution of these observations revealed a faint continuum source at the center of ESO 601−G036 with a flux density of ${S}_{2.1\mathrm{GHz}}=240\,\mu \mathrm{Jy}$ (with a beam size of 371 × 50). Re-imaging using natural weighting gives a beam size of 838 × 123 and a flux density of ${S}_{2.1\mathrm{GHz}}=315\,\mu \mathrm{Jy}$ and indicates that the source is resolved. Using the naturally weighted flux density gives a radio luminosity of ${L}_{2.1\mathrm{GHz}}=4.3\times {10}^{19}$ W Hz⁻¹. As this radio detection is resolved we discount it as being similar to the persistent radio source detected in FRB 121102, which is compact on mas-scales. Figure 1 shows the 2.1 GHz data of the field. None of the candidates selected from the WISE × SCOSPZ catalog have associated radio emission (5σ limits are listed in Table 2).

Table 3 compares key properties of ESO 601–G036 with those of the host galaxy of the repeating FRB 121102. DM_host was estimated by assuming that all excess DM is attributed to the host galaxy.

ESO 601–G036 is about a magnitude more luminous than the host galaxy of FRB 121102 (Tendulkar et al. 2017), and the measured projected size of the galaxy (9.2 × 3.7 kpc) is slightly larger. The current star formation rate in ESO 601–G036 is two to three times lower than in the FRB 121102 host, but the two galaxies are qualitatively similar.

The most striking difference is that ESO 601–G036 does not contain a luminous persistent radio source like that seen in the FRB 121102 host galaxy. If ESO 601–G036 is indeed the host galaxy of FRB 171020, this would imply that not all FRBs are associated with bright, compact, and persistent radio emission.

5.1. Searching for FRB 121102 Analogs

We have searched for a potential host galaxy of FRB 171020 and found ESO 601–G036 to be the most likely candidate given its low redshift and position close to the center of the error ellipse. UV imaging from GALEX and follow-up spectroscopic observations reveal that ESO 601–G036 is a low-metallicity galaxy with an SFR of 0.13 M_⊙ yr⁻¹, similar to the host galaxy of FRB 121102. However, no compact persistent radio continuum source is detected above L_2.1GHz = 3.6 × 10¹⁹ W Hz⁻¹, which is 600 times fainter than the persistent source associated with FRB 121102. There is no galaxy within the localization uncertainty region that has similar properties to the host galaxy of FRB 121102 at redshifts z ≲ 0.06. This suggests that not all FRBs have an associated "persistent" radio source. As such, identifying host galaxies based on the presence of a compact, luminous radio continuum source may not necessarily help in identifying host galaxies of FRBs.

The authors wish to thank S. Chatterjee, C. Bassa, and the anonymous referee for useful discussions and suggestions that helped to improve this Letter. This research is based on observations collected at the European Southern Observatory under ESO programme 0101.A-0455(B) (PI: J.-P. Macquart), as well as on observations obtained as part of program GS-2018A-Q-205 (PI: N. Tejos) at the Gemini Observatory and C3211 (PI: R.M. Shannon) on the Australia Telescope Compact Array. J.P.M., R.M.S., and K.B. acknowledge Australian Research Council grant DP180100857 and R.M.S. acknowledges support through Australian Research Council (ARC) grants FL150100148 and CE17010000.

The Australian SKA Pathfinder is part of the Australia Telescope National Facility which is managed by CSIRO. We acknowledge the Wajarri Yamatji as the traditional owners of the Observatory site. The Australia Telescope Compact Array is part of the Australia Telescope National Facility which is funded by the Australian Government for operation as a National Facility managed by CSIRO. The Gemini Observatory is operated by the Association of Universities for Research in Astronomy, Inc., under a cooperative agreement with the NSF on behalf of the Gemini partnership: the National Science Foundation (United States), the National Research Council (Canada), CONICYT (Chile), Ministerio de Ciencia, Tecnología e Innovación Productiva (Argentina), and Ministério da Ciência, Tecnologia e Inovação (Brazil).

Facilities: ASKAP - , ATCA - , WISE - Wide-field Infrared Survey Explorer, VISTA - , GALEX - Galaxy Evolution Explorer satellite, Gemini-S - , VLT - .

Software: astropy (Astropy Collaboration et al. 2013), PyRAF.