ALMA band 8 observations of DLA 2233+131 at $z$ = 3.150

Quasar absorption line systems in ultraviolet (UV) spectra of high-redshift quasars have been used to investigate the nature of the intergalactic medium (IGM) and intervening objects such as gas-rich galaxies along lines of sight of quasars (Wolfe et al., 1986, 2005). Among such quasar absorption line systems, damped Ly$\alpha$ absorption (DLA) systems [$N$(HI) $\geq 2\times 10^{20}$ cm^-2 (Wolfe et al., 1986)] as well as Lyman limit systems (LLSs) [$10^{17}$ cm${}^{-2}\leq N$(HI) $<2\times 10^{20}$ cm^-2 (Péroux et al., 2003)], and metal-line absorbers like Mg ii absorbers (Bergeron, & Boissé, 1991; Steidel, 1995b; Rao et al., 2006; Monier et al., 2009; Rao et al., 2011) are useful in studying the evolution of gas-rich galaxies from high through intermediate to low redshifts. However, one serious problem is that their optical counterparts are rarely identified (Wolfe et al., 2005; Krogager et al., 2017). Moreover, even when they are identified, their impact parameters are sometimes large, several tens kpc to more than 100 kpc; e.g., Rao et al. (2011); Fynbo et al. (2018); Mackenzie et al. (2019). Christensen et al. (2014) examined the mass-luminosity (MZ) relation for DLAs (e.g., Møller et al. (2013)) to find that the stellar masses of optical counterparts of DLAs show a good agreement with those predicted by the MZ relation, at least for the metal-rich DLAs. Krogager et al. (2017) showed that the low detection rate of optical DLA counterparts and the distribution of impact parameters between DLAs and background quasars are explained by a simple scenario where the “metallicity-luminosity relation” of a DLA is given by $M_{\rm UV}=-5\times[{\rm M/H}+0.3]-20.8$ and the cross-section of DLA is proportional to the luminosity following a relation $\sigma_{\rm DLA}\propto L^{0.8}$.

An open question for the nature of high-$z$ DLAs is how the counterpart makes the absorption feature. The most probable candidates are disks of gas-rich galaxies including dwarf and low-surface brightness ones (e.g., Prochaska, & Wolfe (1997, 1998); Jimenez et al. (1999)). On the other hand, in these two decades, the following new ideas have been proposed; galactic outflows (e.g., Taniguchi & Shioya (2000); Nagamine et al. (2004)), tidal tails in the galaxy interaction (e.g., Kacprzak et al. (2010)), and cosmic filaments (e.g., Fumagalli et al. (2017)).

In the case of DLAs with a larger impact parameter, outflowing gas is a good candidate of the origin of absorption feature. For example, Taniguchi & Shioya (2000) proposed that an initial starburst in a galaxy with mass of $\sim$10¹¹ $M_{\odot}$ makes a large-scale shocked gaseous shell around the galaxy because of the superwind outflow. Since its typical radius reaches to $\sim$100 kpc, this can explain the observed large impact parameters of DLAs. The galaxy itself could become fainter during the course of its passive evolution making it difficult to identify optical counterparts in many cases which would then explain the observed low detection rate of DLA counterparts in the optical. The metal enrichment by Type II supernovae gives a typical metallicity of $Z\sim 0.02Z_{\odot}$ or [M/H] $\sim-1.7$ in the shocked shell, consistent with the observed metallicity of DLAs. According to Taniguchi, & Shioya (2001), the shock velocity ranges from $\sim$0 km s^-1 to a few thousands km s^-1, depending on the energy of a superwind, the evolutionally phase of a superwind, and the viewing angle toward a part of the shocked shell. They also found a positive correlation between the metallicity and velocity width, being consistent with the observed velocity-metallicity relation of high-$z$ DLAs (e.g., Ledoux et al. (2006)). Such superwind models can also explain the observed H i column density distribution function (Nagamine et al., 2004). Although it is unclear whether or not the galactic outflow is a major origin of high-$z$ DLAs, some DLAs at high redshift with evidence for outflows have been observed so far (e.g., Christensen et al. (2004); Krogager et al. (2013); Kashikawa et al. (2014)). For understanding the nature of high-$z$ DLAs, it is important to confirm the plausible origin of each DLA.

Optical observations even with the largest and most sensitive telescopes are extremely challenging because these systems are extremely faint in the optical due to cosmological surface brightness dimming whereas millimeter/submillimeter windows offer a relatively constant flux due to the negative K-correction. Therefore, instead of the optical window, we are now able to choose the millimeter/submillimeter windows using the Atacama Large Millimeter and Submillimeter Array (ALMA). In fact, recently, ALMA has been used to investigate the nature of high-redshift DLA counterparts by using dust continuum emission and radio atomic and molecular emission lines such as [C ii] 158 $\mu$m and CO emission lines (Neeleman et al., 2016, 2017; Fynbo et al., 2018; Kanekar et al., 2018; Møller et al., 2018; Neeleman et al., 2018, 2019). Encouraged by these pioneering ALMA observations of high-redshift DLAs, in this study, we focus on a $z=3.150$ DLA system in the line of sight of Q2233+131 whose redshift is $z=3.295$ (Christensen et al., 2004). Because the Ly$\alpha$ emission from this optical counterpart shows a double-peaked profile, a possible origin of this DLA is thought to be outflowing gas. Since this DLA has been well studied by optical observations (Djorgovski et al., 1996; Christensen et al., 2004, 2007; Kashikawa et al., 2014), it provides a good laboratory to examine the scenario for the origin of the DLA, by combining optical and submm properties.

The structure of this paper is as follows. In Section 2, we summarize general properties of our target, DLA2233+131. Section 3 describes the observations and data reduction. The results of our ALMA observations are shown in Section 4. In Section 5, we discuss possible scenarios to explain the origin of DLA2233+131 and serendipitously detected submm galaxies (SMGs). Finally, we present our concluding remark in Section 6. Throughout this paper, we adopt a flat cosmology with $\Omega_{\rm m}$ = 0.3, $\Omega_{\Lambda}$ = 0.7, and $H_{0}$ = 70 km s^-1 Mpc^-1.

Our target is DLA2233+131 found toward the quasar Q2233+131 at $z=3.295$ (see Table 2). This is the first intervening DLA whose optical counterpart was identified in both Ly$\alpha$ emission and stellar continuum by Djorgovski et al. (1996). The H i column density of DLA2233+131 has been measured as follows; log$N$(H i) (cm^-2) = 20.00 (Lu et al., 1993), 20.2 (Curran et al., 2002), and 19.95 [SDSS DR5; see Kashikawa et al. (2014)]. Since all these values are smaller than log$N$(H i) (cm^-2) = 20.3 required to define the DLA, strictly speaking, this absorber is not a DLA but a sub-DLA or a Lyman limit system (LLS); see Péroux et al. (2003). However, following Djorgovski et al. (1996), we refer this absorber as DLA2233+131 in this paper. Its optical image obtained with the Hubble Space Telescope is shown in the right panel of Figure 2.

This DLA is located 2^′′.3 away from the quasar at PA = 159^∘. The projected separation from the line of sight to Q2333+133 is $\sim$ 18 kpc. A summary of the position of DLA2233+131 measured in the literature is given in Table 2. Note that the presence of this DLA feature was already found by Sargent et al. (1989) and by Lu et al. (1993) and the quasar Q2233+131 was found in the CFHT Blue Grens Quasar Survey Plates by Crampton et al. (1985).

After the optical identification of DLA2233+131 by Djorgovski et al. (1996), a number of follow up studies have been made to date. First, its metallicity is found as [Fe/H] = $-$1.4 (Lu et al., 1997). Møller, & Christensen (2019) re-derived the metallicity to be [Si/H] = $-0.97\pm 0.13$. Since Fe tends to be depleted by dust (e.g., Wolfe et al. (2005); Rafelski et al. (2014)), we use [Si/H] as the metallicity of DLA2233+131 throughout this paper. The Ly$\alpha$ emission-line region is further investigated by Christensen et al. (2004) and Kashikawa et al. (2014). Christensen et al. (2004) performed integral field spectroscopic (IFS) observations of this DLA2233+131 to find a spatially extended Ly$\alpha$ blob with a size of $\sim$23 kpc $\times$ 38 kpc. However, the extended morphology of Ly$\alpha$ emission was not confirmed by higher spectral resolution IFS observations (Christensen et al., 2007). Christensen et al. (2007) claimed that the IFS data obtained by Christensen et al. (2004) is likely affected by the systematic noise. Furthermore, Kashikawa et al. (2014) found not an extended but a compact Ly$\alpha$ emission-line region at DLA2233+131. Their analysis shows that its effective radius is 1.35 arcsec, corresponding to 10.24 kpc, and the Sersic index is $n$ = 1.80. Christensen et al. (2014) estimated the stellar mass of the counterpart of DLA2233+131 to be $\sim 10^{9.8}$ $M_{\odot}$ based on their SED fitting. The observational properties of DLA2233+131 are summarized in Table 2. It is noted that another absorber at $z=2.543$ is also identified in the UV spectrum of Q2233+131 (Steidel et al., 1995a). However, we call the absorber at $z=3.150$ as DLA2233+133 in this paper.

*4c Previously reported properties of the optical counterpart of DLA2233+131 Redshift ($z_{\rm DLA,em}$) of the counterpart measured from its emission line
$z_{\rm DLA,em}$ References Note
\endfirsthead\endhead\endfoot^∗*∗*footnotemark: $*$ Christensen et al. (2007).
\endlastfoot 3.1476 Christensen et al. (2004)
Impact parameter ($b$)
$b$
[arcsec] [kpc] References Note
2.3 17.5 Steidel et al. (1995a) broad-band imaging
2.3 17.5 Djorgovski et al. (1996) long-slit spectroscopy
2.37 18.0 Christensen et al. (2004) IFS
2.41 18.3 Kashikawa et al. (2014) narrow-band imaging
Photometric measurements
Measurements References Note
$V=25.1$, $R=24.8$ Steidel et al. (1995a)
$H=25.34$ Warren et al. (2001) HST/NICMOS
$V_{50}=25.75$ Møller et al. (2002) $V_{50}$ is a $V$-band filter of HST/STIS
$V>24.40$, $R>24.44$, $N502=24.30$ Kashikawa et al. (2014) N502 is a narrow-band filter of Subaru/FOCAS
($\lambda_{\rm eff}=5025$ Å, $FWHM=60$ Å)
Ly$\alpha$ flux measurements
Measurements References Note
$(6.4\pm 1.2)\times 10^{-17}$ [erg s^-1] Djorgovski et al. (1996) broad-band imaging
$(2.8\pm 0.3)\times 10^{-16}$ [erg s^-1] Christensen et al. (2004) IFS, overestimated^∗*∗*footnotemark: $*$
24.30 [AB mag] Kashikawa et al. (2014) narrow-band imaging

We observed a sky area with a radius of 11^′′ centered at the position of Q2233+131 in ALMA Cycle 5 (Proposal ID 2017.1.00345.S, PI: Y. Taniguchi). The observations were conducted on 24 May 2018 using forty-six 12-m antennas with baseline lengths of 15.0 m – 313.7 m. We used the ALMA Band-8 receivers to detect redshifted [C ii] 158 $\mu$m and continuum emissions from the counterpart of DLA at $z=3.150$ along the sightline to Q2233+131. We used 4 bands with a width of 1.875 GHz which are divided into 128 channels. We set the central frequency of the band to be 457.329 GHz and 459.121 GHz for the [C ii] line, while 469.329 GHz and 471.121 GHz for the continuum emission. The total on-source integration time was 31.7 minutes.

All the data were reduced by using the Common Astronomy Software Application (CASA) ver 5.4.0 (McMullin et al., 2007). To search for the emission line, we create a dirty cube per 100 km s^-1 bin. We then applied uvcontsub task for each side band to make the final cube with the natural weighting. As for the continuum emission, we first created a dirty map with natural weighting using all the 4 spectral windows. We then performed tCLEAN down to 2$\sigma$, resulting that we detected 3 continuum sources, one of which is Q2233+131. We detected no other objects with $>3\sigma$. Here, the 1$\sigma$ noise level is $\sim$67 $\mu$Jy/beam at the phase center. The synthesized beam size is 0^′′.84 $\times$ 0^′′.61 with a position angle of 52 degree. The maximum recoverable scale is 5^′′.15. Flux measurements were performed via Gaussian fitting with the CASA task, imfit. We extracted spectra of these 3 continuum sources by adopting 1^′′.0 diameter aperture.

We have conducted ALMA Band-8 observations of DLA2233+131 at $z=3.150$ found in the UV spectrum of quasar Q2233+131 at $z=3.295$. This DLA is the first intervening DLA at high redshift whose optical counterpart was identified (Djorgovski et al., 1996). The Ly$\alpha$ emission associated with this DLA shows double peak profile whose separation corresponds to 600 – 750 km s^-1(Christensen et al., 2004). High spectral resolution IFS observations by Christensen et al. (2007) and an excellent quality image obtained by Kashikawa et al. (2014) showed the optical counterpart of DLA2233+131 is a compact galaxy, although some faint extended Ly$\alpha$ emitting regions appear to be associated.

To investigate the origin of this DLA, we have conducted ALMA Band 8 observations to detect [C ii] 158 $\mu$m emission. However, we have not detected significant [C ii] emission in our ALMA observations. A possible reason of this non-detection is that the line is too faint to be detected in the relatively short integration carried out here. We estimate the expected [C ii] from DLA2233+131 based on its SFR derived by optical observations, $S$([C ii]) $\delta v$ = 24 – 3200 mJy km s^-1. The large spread of estimated flux comes from the scatter of observed [C ii]/FIR ratio. In the case that the [C ii]/FIR ratio is significantly low, the expected [C ii] flux is fainter than the detection limit of our ALMA observations. Although such a lower [C ii]/FIR value suggests higher dust temperature (Malhotra et al., 2017), optical observations imply that DLA2233+131 may have only little dust content. This is consistent with the non-detection of dust continuum from DLA2233+131 in this work. Therefore, we consider another possible scenario in which the counterpart of DLA2233+131 is optically thick enough to activate the self-shielding effect. In this scenario, the PDR cannot be formed and thus [C ii] emission is not radiated. The observed double peak Ly$\alpha$ emission implies the existence of the outflowing gas. The origin of faint emissions around DLA2233+131 may be explained by the resonant scattering.

The previous ALMA [C ii] 158 $\mu$m or CO observations of low- to intermediate- to high-$z$ DLAs are successful (Neeleman et al., 2016, 2017; Fynbo et al., 2018; Kanekar et al., 2018; Møller et al., 2018; Neeleman et al., 2018, 2019). Although we cannot detect [C ii] 158 $\mu$m emission from DLA2233+131, we would like to recommend that one carry out a systematic search for [C ii] 158 $\mu$m emission for a large sample of DLA from low to high redshifts. One problem is that it is difficult to estimate reasonable integration time because we do not have any reliable line flux estimators. Therefore, the first step is to carry out this survey with one hour integration for any DLAs. Such a large survey will tell us useful information on the origin of DLAs although there may be a variety of DLA counterparts, e.g., outflowing gas, disks of gas-rich galaxies from dwarf to giant ones.

Despite the non-detection of [C ii] 158 $\mu$m emission from DLA2233+131, we have serendipitously found the two SMGs in the observed field; SMG1 and SMG2. Their angular separation from Q2233+131 are 4^′′.7 and 8^′′.1, respectively. They are not detected in the optical imaging with HST and the Subaru Telescope. Since most of previously observed low-$z$ SMGs are also detected in optical observations, they may not be low-$z$ galaxies although we cannot rule out the possibility that they are optically very faint low-$z$ objects.

Because it is hard to drive the redshifts of SMG1 and SMG2, we consider the possibility that these two SMGs and Q2233+131 are associated with each other; i.e., SMG1 and SMG2 are at $z=3.3$ as well as Q2233+131. If this is the case, the linear projected separations from the quasar to SMG1 and SMG2 are 35 and 60 kpc, respectively. These separations are similar to that of gas-rich merging systems observed as quasar-SMG pairs at high redshift (10 – 100 kpc). If this is the case, the Q2233+131 provides an interesting target to study gas-rich merging systems at high redshift. Because we cannot confine the redshifts of two SMGs, new ALMA and/or deep optical observations will be necessary to obtain their redshifts.

We are very grateful to the referee, Lise Chiristensen, for careful reading and many helpful comments. This paper uses the ALMA data ADS/JAO.ALMA#2017.1.00345.S. KO is grateful to Masahiro Nagasima for kindly giving many useful comments. This work is supported from JSPS KAKENHI No. 16H02166 (PI: Y. Taniguchi). YT would like to thank Jason X. Prochaska for his kind communication for these years. Actually, he showed me ALMA observations made by his group prior to publication. We would also like to thank Kotaro Kohno for useful discussion on the metallicity effect for the detection of [C ii] 158 $\mu$m emission from DLAs.