EDP Sciences logo
Subscriber Authentication Point
Search
Menu
Home All issues Volume 638 (June 2020) A&A, 638 (2020) L14 Full HTML
Free Access
Issue
A&A
Volume 638, June 2020
Article Number L14
Number of page(s) 7
Section Letters to the Editor
DOI https://doi.org/10.1051/0004-6361/202038483
Published online 29 June 2020

© ESO 2020

1. Introduction

Atomic neutral hydrogen gas, H I, is found to be one of the most extended baryon components of galaxies (Yun et al. 1994). The width of the H I emission line has been historically used to estimate the dynamical mass via the large-scale rotation velocity derived from the double horn shape of the emission line (Roberts 1978). These estimates provide a proxy for estimating the dark matter content in galaxies (Salucci 2019). On the other hand, the optically thin nature of the H I emission leads to the possibility of probing possible gas inflows via the asymmetry of the line profile (Bournaud et al. 2005; Deg et al. 2020). The H I fluxes are found to very closely follow scaling relations such as the star formation rate (SFR) surface density ΣSFR and the combined surface density of molecular (H2) and H I gas, ΣHI+H2 (Schmidt 1959; Kennicutt 1998), H I mass versus stellar mass (Huang et al. 2012; Maddox et al. 2015; Romeo 2020), H I mass versus H I size (Wang et al. 2016; Stevens et al. 2019), and the H I-to-H2 ratio as a function of stellar or gas surface density (Leroy et al. 2008). These relations enable us to reveal important information even when we are unable to perform spatially resolved H I observations (Giovanelli & Haynes 2015).

Given the fundamental importance of the H I component, several wide-field H I surveys have been carried out to probe the neutral atomic gas at low redshift. Previous blind H I surveys such as the H I Parkes All-Sky Survey (HIPASS, Barnes et al. 2001; Meyer et al. 2004; Wong et al. 2006, 13 beams), the H I Jodrell All Sky Survey (Lang et al. 2003, 4 beams), the Effelsberg-Bonn H I Survey (EBHIS, Winkel et al. 2010; Kerp et al. 2011, 7 beams), and the Arecibo Legacy Fast ALFA Survey (ALFALFA, Giovanelli et al. 2005, 7 beams) detected large numbers of gas-rich galaxies in both northern and southern skies. The ongoing Widefield ASKAP L-band Legacy All-sky Blind surveY (WALLABY, Koribalski et al. 2020), and Deep Investigation of Neutral Gas Origins (DINGO, Meyer 2009) are planning to perform an H I survey over much larger areas using the Australian Square Kilometer Array Pathfinder (ASKAP), expecting to detect a million new H I sources up to z ∼ 0.4.

Current facilities have been able to detect the H I emission line up to z = 0.1 (e.g. Giovanelli & Haynes 2015). Major factors limiting the detectability of higher redshift sources include the sensitivity, the detector frequency range, and radio frequency interference (RFI). One way to overcome the sensitivity issues is by using gravitational magnification. For example, HI ultra-deep survey projects such as Blind Ultra Deep H I Environmental Survey (BUDHIES, Jaffé et al. 2013), or the COSMOS H I Large Extragalactic Survey (CHILES) have spent hundreds hours extending the H I detection up to z ≃ 0.3 (z = 0.37 in Fernández et al. 2016, and z = 0.32 in Rhee et al. 2018). Higher redshift H I studies have been available from damped Lyα absorbers (Neeleman et al. 2016), or intensity mapping (e.g. Hu et al. 2020). To detect the H I emission line beyond z ∼ 0.1 is still a major challenge.

The new fully operational Five-hundred-meter Aperture Spherical radio Telescope (FAST, Nan 2006; Nan et al. 2011) with the tracking active reflector design (Qiu 1998) provides us with the opportunity to probe H I in the higher redshift Universe with unprecedented sensitivity. For example, the upcoming Commensal Radio Astronomy FasT Survey (CRAFTS) project is expected to detect H I out to z = 0.35 (Zhang et al. 2019). As the largest filled-aperture radio telescope, FAST has been designed to achieve many challenging scientific goals, including hunting for pulsar, H I map of local galaxies, mapping the Milky Way central region, and so on (see the review of Jiang et al. 2019, 2020, and references therein).

In this Letter, we report the results of our pilot H I survey with FAST on a sample of four z ∼ 0.05 star-forming galaxies taken from the Valparaíso ALMA/APEX Line Emission Survey (Villanueva et al. 2017; Cheng et al. 2018). These are among the first extragalactic H I detection cases observed by FAST in its commissioning phase. Throughout this Letter, we assume a standard ΛCDM cosmology with H0 = 70 km s−1 Mpc−1, ΩM = 0.3, and ΩΛ = 0.7. All magnitudes are provided in the AB magnitude system (Oke & Gunn 1983).

2. FAST observations

2.1. Sample selection

The ongoing VALES project targets the low-J CO transitions in 91 low-z dusty star-forming galaxies taken from the Herschel Astrophysical Terahertz Large Area Survey (H-ATLAS, Eales et al. 2010). The CO(J = 1−0) and CO(J = 2−1) emission lines were observed using the Atacama Large Millimeter/sub-millimiter Array (ALMA, Villanueva et al. 2017) and the Atacama Pathfinder EXperiment (APEX, Cheng et al. 2018), respectively. The VALES galaxies are selected from the equatorial GAMA fields, which provides extensive multi-wavelength coverage (Driver et al. 2009).

In this pilot work, we extract four galaxies previously detected in CO(J = 1−0) to target their H I emission line with FAST. To mitigate sensitivity effects, we focus on those at the lower redshift end of the VALES distribution. This is particularly useful to avoid the strong RFI seen at lower frequencies. The targeted sources have: (1) an expected H I flux higher than 0.5 Jy km s−1 as estimated using the Zhang et al. (2009) prescription, (2) they do not have bright nearby galaxies at similar redshift within one beam (2.9 arcmin diameter), and (3) their Declinations are in the range of 0 to 40 deg, so that we can have a high collecting area (full 300 m dish). Most of the VALES galaxies are massive star forming galaxies as shown in Fig. 1. The four galaxies observed by FAST, shown as red dots, are local main sequence galaxies.

thumbnail Fig. 1.

Main sequence of the low-z galaxies (black contour and greyscale) and our ALMA detected VALES sample (blue dots). The FAST observed targets are shown as red dots.

Aided by the previous coverage provided by ALFALFA (Haynes et al. 2011, 2018) in part of the GAMA fields, we include one source with ALFALFA measurements (HATLASJ083601.5+002617) to check consistency and the relative performance of the FAST telescope.

2.2. Observation setup and calibration

We were allocated 10 h in L-band with the 19-Beam receiver during the FAST Commissioning Phase (Shared-risk Observing Proposal1 2019A-012-S; PI: Cheng Cheng). The focal plane is covered by 19 beams, each one with a 2.9 arcmin diameter and separated by 5.8 arcmin from each other. Our observations were carried out in the tracking mode, using only the central beam (M01). Details for the FAST technical performance can be found in Jiang et al. (2020).

To mitigate the errors due to an unsteady baseline at the GHz frequency, we set the observations as 5 min ON-target +5 min OFF-target per iteration, and made from three to eight iterations per target. The OFF-target pointing position was chosen at 5.8 arcmin distance, with no sources in the field of view (FoV) with a similar redshift (from SDSS spec-z catalogue) to that of our targets. We use the flux from the OFF pointing to estimate the background, and thus the ON-OFF spectra is the flux from our target.

We calibrate the flux by the noise diode with known antenna temperature that injected into the receiver during the observation (Jiang et al. 2020). Details of the data reduction and calibration processes are briefly described in Appendix A.

3. Results

The noise levels reached for each target are listed in Table 1, while the H I spectra are shown in Fig. 2. For a 5 min on-target integration, we obtained a noise level of about 2.6 mJy beam−1 at 1.7 km s−1 velocity resolution bin. The rms declines with the on-target integration time, but not according to ∝ 1 / t $ {\propto}1/\sqrt{t} $, which might be caused by the pointing jitter in the commissioning stage, or weak interference we were not aware of. We detect the H I emission line in three galaxies. For the H I undetected galaxy, we estimate the flux upper limit as 5 × rms × 300 W300 mJy km s−1, where W300 = Width/300 km s−1. All three detected spectra show the double-horn pattern with a typical full width at half maximum (FWHM) of about 300 km s−1. We re-sample the H I spectrum into ΔV = 20 km s−1 channels, and derive the integrated H I flux by summing over the velocity channels within the full width at zero intensity (FWZI), which includes the channels at both line wings with the channel flux higher than the rms. We highlight the velocity channels that were used to derive the H I flux in grey in Fig. 2).

thumbnail Fig. 2.

H I and CO spectra of our targets. The black thick lines show the H I spectra obtained from FAST. The H I spectrum are rebinned into 20 km s−1 resolution. The red dot lines are the CO(1−0) line from our previous ALMA survey. The scales of the CO emission lines are shown in the right y-axis. We adopt the optical spectroscopy redshift to derive the line velocity. We show the ALFALFA H I spectrum of the galaxy HATLASJ083601.5+002617 in blue in the upper left panel. Effective on-target integration time of Arecibo telescope is about 48 s. The edges of the emission lines are very sharp, and we obtain a line width consistent with the ALFALFA results. The grey regions highlight the velocity channels we used to derive the H I flux.

Table 1.

FAST observation results.

We derive the H I mass following Giovanelli & Haynes (2015):

M H I M ⊙ = 2.35 × 10 5 D Mpc 2 1 + z ∫ FWZI S ( V ) d V , $$ \begin{aligned} \frac{M_{\mathrm{H}\small{\rm I} }}{M_\odot } = \frac{2.35\times 10^5 D_{\rm Mpc}^{2}}{1+z} \int _{\rm FWZI} S(V) \mathrm{d}V, \end{aligned} $$(1)

where the S(V) is the flux density in units of Jy beam−1. The uncertainty is estimated using σ 2 = ( N channel Δ V × rms ) 2 + ( 10 % × flux ) 2 $ \sigma^2 = (\sqrt{N_{\mathrm{channel}}} \Delta V \times \mathrm{rms})^2 + (10\% \times \mathrm{flux})^2 $, where Nchannel is the number of integration channels, and the 10%×flux would account for the uncertainty of the flux calibration. The H I mass or mass upper limit of our targets are listed in Table 2.

Table 2.

Target properties.

4. Discussion

4.1. Comparison with a previous ALFALFA result

For HATLASJ083601.5+002617, the upper left panel of Fig. 2 shows the results from both FAST and Arecibo telescopes. The ALFALFA survey is in the drift scan mode, and the effective on the target integration time for the ALFALFA spectrum is about 48 s (Haynes et al. 2011). The observed spectra show consistent line profiles. The integrated flux of the Arecibo spectrum given in Haynes et al. (2018) is 1.68 ± 0.14 mJy km s−1, and the FWHM is 390 ± 19 km s−1, both values consistent with our results (see Table 1). This indicates a reasonable flux calibration for the FAST 19-beam receiver in this commissioning phase experiment. In terms of noise levels, at 20 km s−1, we reach 0.8 mJy beam−1 for 5 min integration time, corresponding to about 2 mJy beam−1 for 48 s integration, while the rms given in ALFALFA H I catalogue is 2.89 mJy beam−1 at 10 km s−1 resolution, which is about 2 mJy beam−1 at 20 km s−1 resolution. Thus it appears that, in the commissioning phase, FAST has already reached a sensitivity similar to that of ALFALFA.

4.2. The HI gas mass

Previous studies have shown empirical correlations to derive the H I gas mass from the global optical properties such as the colour, surface brightness, and the stellar mass densities (e.g. Zhang et al. 2009; Catinella et al. 2013; Parkash et al. 2018) with typical scatters about 0.5 dex. In Fig. 3, we show the observed and estimated H I mass of our VALES sample based on the relations given by Catinella et al. (2013):

log ( M H I / M ∗ ) = − 0.338 log Σ ∗ − 0.235 ( NUV − r ) + 2.908 , $$ \begin{aligned} \log (M_{\mathrm{H}\small{\rm I} }/M_*) = -0.338 \log \Sigma _* - 0.235 (\mathrm{NUV}-r) + 2.908, \end{aligned} $$(2)

thumbnail Fig. 3.

Predicted H I masses of our VALES sample based on the empirical relations in Zhang et al. (2009, in purple circles) and in Catinella et al. (2013, in blue circles). The black dots are the galaxies from HIPASS (Parkash et al. 2018). The red dots show the results of the FAST observation. We link the predicted and the observed H I mass by dotted lines. We show the FAST non-detected target by upper limits.

where Σ * = M * /π R opt 2 [ M ⊙ kpc −2 ] $ \Sigma_* = M_*/\pi R_{\rm opt}^2\,[M_\odot\,{\rm kpc}^{-2}] $ is the stellar mass surface density, Ropt is the r band half light radius, and the relation from Zhang et al. (2009):

log ( M H I / M ∗ ) = − 1.73238 ( g − r ) + 0.215182 μ i − 4.08451 , $$ \begin{aligned} \log (M_{\mathrm{H}\small{\rm I} }/M_*) = -1.73238 (g-r) + 0.215182 \mu _i - 4.08451, \end{aligned} $$(3)

where μi [mag arcsec−2] is the SDSS i band average surface brightness. We adopt the stellar mass, radius and the photometry from GAMA galaxy structure catalogue (Kelvin et al. 2012) and H-ATLAS catalogue (Valiante et al. 2016). We also show the observed results from H I Parkes All-Sky Survey catalogue (HIPASS, Parkash et al. 2018) in Fig. 3 as a comparison.

In Villanueva et al. (2017) we show that the VALES sample stands for the galaxy population from star-forming (for the z ∼ 0.05 galaxies) to starburst (mainly at z >  0.1, see the Fig. 1 of Villanueva et al. 2017). We find that the VALES sample follows the Kennicutt-Schmidt relation based on the H2 gas mass from CO(J = 1−0) observations, and the H I gas mass derived from the empirical relation (Zhang et al. 2009). Our pilot survey results reveal a consistency between estimated and observed H I mass, suggesting that the star-forming galaxies at redshift 0.05 in VALES sample may still follow the Kennicutt-Schmidt relation. More H I observations of the VALES galaxies by FAST are still ongoing.

4.3. Dynamical masses

In Molina et al. (2019), we presented a dynamical analysis of 39 VALES galaxies using the spatial extension of the CO(J = 1−0) emission. For the three H I detected galaxies in this latter work, the dynamical study shows that the CO rotation curves are still increasing or are just about to turn to flat rotation curve with the radius (see the Appendix A of Molina et al. 2019). While comparing the H I and CO(J = 1−0) emission line spectra, we find that the CO(J = 1−0) line profiles in Fig. 2 show a sharp decline at larger velocities, and a similar line width and profile to those of the H I spectra (see the CO FWHM in Table 2). The agreement of the CO and H I line width suggests that the observed CO(J = 1−0) emission might extend up to the region where the rotation curve is flat (Dickey & Kazes 1992; Sofue 1992; Schoniger & Sofue 1994; de Blok et al. 2016; Tiley et al. 2016).

We can roughly estimate the dynamical mass based on the H I FWHM, and H I radius (RH I) by M dyn H I = ( F W H M / 2 / sin θ ) 2 R H I / G $ M_{\mathrm{dyn}}^{\mathrm{H}\textsc{\rm I}} = ({FWHM/2/\sin \theta})^2 R_{\mathrm{H}\textsc{\rm I}}/G $, where θ is the H I inclination angle and RH I is the H I radius. Since for galaxies the H I size is tightly correlated with the H I mass (Wang et al. 2016; Stevens et al. 2019), the RH I can be estimated from MH I using the relation derived by Wang et al. (2016). For the H I mass range of 9.5 < log(MH I/M⊙) < 10, the H I radius is in the range of 16.5 to 30 kpc (see Eq. (2) or Fig. 1 in Wang et al. 2016), which is much larger than the CO radius (see the red contours in Fig. B.1). We do not have inclination information of the H I gas. But if we assume the Ks band image and H I have a similar inclination angle (Molina et al. 2019), we can roughly estimate the M dyn H I $ M_{\mathrm{dyn}}^{\mathrm{H}\textsc{\rm I}} $ (see Table 2). The M dyn H I $ M_{\mathrm{dyn}}^{\mathrm{H}\textsc{\rm I}} $ is about one order of magnitude larger than the baryon mass ( M bary = M H I + M H 2 + M ∗ $ M_{\mathrm{bary}} = M_{\mathrm{H}\textsc{\rm I}} + M_{\mathrm{H_2}} + M_* $), also listed in Table 2. For HATLASJ085112.9+010342, which has no H I detection, we estimate the lower limit of the baryon mass to be M bary lower limit = M H 2 + M ∗ $ M_{\mathrm{bary}}^{\mathrm{lower\,limit}} = M_{\mathrm{H_2}}+M_* $. The H I radii of our targets are much larger than the optical and CO radii, yielding a larger M dyn H I $ M_{\mathrm{dyn}}^{\mathrm{H}\textsc{\rm I}} $. The larger M dyn H I $ M_{\mathrm{dyn}}^{\mathrm{H}\textsc{\rm I}} $ may suggest that H I mainly traces the dynamical mass within a larger radius where the dark matter starts to dominate the gravitational potential.

On the other hand, our previous work also derived the dynamical mass within 2 × r1/2CO from CO velocity map (Molina et al. 2019), where the half light CO radius (r1/2CO) is about 4 kpc. We list the M dyn CO $ M_{\mathrm{dyn}}^{\mathrm{CO}} $ in Table 2 and we can see good consistency between the M dyn CO $ M_{\mathrm{dyn}}^{\mathrm{CO}} $ and the baryon mass. Considering the higher dynamical masses derived from the H I data, this implies that CO dynamics is restricted, as expected, to the gravitational potential of the central regions in galaxies. Therefore, it might be possible to trace dark matter halos using both CO and H I observations. We show the optical image, CO contours, and the H I spectra of our targets in Appendix B. The target HATLASJ085340.7+013348 has a clumpy CO morphology, which might be the reason of the relatively higher CO-to-H I flux ratio in Fig. 2.

5. Conclusion

We report on some of the first extragalactic H I line observations of four star-forming galaxies at z ≃ 0.05 taken from the VALES survey which were made during the commissioning phase of the FAST 19-beam receiver. These are among the first extragalactic H I detection results during the FAST commissioning phase stage. Using 5 min/5 min ON/OFF pointing observations, we reached an rms of 2.6 mJy beam−1 at a spectral resolution of 1.7 km s−1. We detected three out of the four observed galaxies. One of our targets was detected previously by the Arecibo ALFALFA survey, with results consistent with ours. The observed H I masses are consistent with values estimated using previously determined empirical relations (Zhang et al. 2009; Catinella et al. 2013), with a scatter of about 0.5 dex. We find the width of the H I emission is similar to the CO(J = 1−0) width revealed by ALMA, suggesting that ALMA observations already observed the flat rotation curve from the galaxy outskirts. The dynamical mass estimated from H I ( M dyn H I ) $ \left(M_{\mathrm{dyn}}^{\mathrm{H}\textsc{\rm I}}\right) $ is an order of magnitude higher than the baryon mass ( M ∗ + M H 2 + M H I ) $ \left(M_*+M_{\mathrm{H_2}}+M_{\mathrm{H}\textsc{\rm I}}\right) $ and the dynamical masses derived from CO observations ( M dyn CO ) $ \left(M_{\mathrm{dyn}}^{\mathrm{CO}}\right) $, implying that the dynamical mass traced by H I may be more dominated by the dark matter halo.

Acknowledgments

We thank the referee for carefully reading and for patiently providing constructive comments that helped us to improve the quality of this Letter. C.C. appreciates the kindness help from FAST, especially to Lei Qian, Ningyu Tang, Jing Tang and Zheng Zheng, and helpful discussions about FAST observation configuration and data reduction with Pei Zuo, Niankun Yu and Lizhi Xie. This work made use of the data from FAST (Five-hundred-meter Aperture Spherical radio Telescope). FAST is a Chinese national mega-science facility, operated by National Astronomical Observatories, Chinese Academy of Sciences. C.C. is supported by the National Natural Science Foundation of China (NSFC), No. 11803044, 11673028. B.Z. is supported by NSFC, No. 11903056. J.H. is supported by NSFC, No. 11933003. This work is sponsored (in part) by the Chinese Academy of Sciences (CAS), through a grant to the CAS South America Center for Astronomy (CASSACA). E.I. acknowledges partial support from FONDECYT through grant N° 1171710. D.W. is supported by NSFC, No. U1931109, 11733006. T.M.H. acknowledges the support from the Chinese Academy of Sciences (CAS) and the National Commission for Scientific and Technological Research of Chile (CONICYT) through a CAS-CONICYT Joint Postdoctoral Fellowship administered by the CAS South America Center for Astronomy (CASSACA) in Santiago, Chile. This work was supported by the National Science Foundation of China (11721303, 11991052) and the National Key R&D Program of China (2016YFA0400702). CKX acknowledges support from the National Key R&D Program of China No. 2017YFA0402704 and National Natural Science Foundation of China No. 11873055 and No. 11733006. This Letter makes use of the following ALMA data: ADS/JAO.ALMA#2013.1.00530.S. ALMA is a partnership of ESO (representing its member states), NSF (USA) and NINS (Japan), together with NRC (Canada), MOST and ASIAA (Taiwan), and KASI (Republic of Korea), in cooperation with the Republic of Chile. The Joint ALMA Observatory is operated by ESO, AUI/NRAO and NAOJ. In addition, publications from NA authors must include the standard NRAO acknowledgement: The National Radio Astronomy Observatory is a facility of the National Science Foundation operated under cooperative agreement by Associated Universities, Inc.

References

  1. Aihara, H., AlSayyad, Y., Ando, M., et al. 2019, PASJ, 71, 114 [CrossRef] [Google Scholar]
  2. Barnes, D. G., Staveley-Smith, L., de Blok, W. J. G., et al. 2001, MNRAS, 322, 486 [NASA ADS] [CrossRef] [Google Scholar]
  3. Bournaud, F., Combes, F., Jog, C. J., & Puerari, I. 2005, A&A, 438, 507 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  4. Catinella, B., Schiminovich, D., Cortese, L., et al. 2013, MNRAS, 436, 34 [NASA ADS] [CrossRef] [Google Scholar]
  5. Cheng, C., Ibar, E., Hughes, T. M., et al. 2018, MNRAS, 475, 248 [CrossRef] [Google Scholar]
  6. de Blok, W. J. G., Walter, F., Smith, J. D. T., et al. 2016, AJ, 152, 51 [NASA ADS] [CrossRef] [Google Scholar]
  7. Deg, N., Blyth, S. L., Hank, N., Kruger, S., & Carignan, C. 2020, MNRAS, 495, 1984 [CrossRef] [Google Scholar]
  8. Dickey, J. M., & Kazes, I. 1992, ApJ, 393, 530 [CrossRef] [Google Scholar]
  9. Driver, S. P., Norberg, P., Baldry, I. K., et al. 2009, Astron. Geophys., 50, 5.12 [NASA ADS] [CrossRef] [Google Scholar]
  10. Eales, S., Dunne, L., Clements, D., et al. 2010, PASP, 122, 499 [NASA ADS] [CrossRef] [Google Scholar]
  11. Fernández, X., Gim, H. B., van Gorkom, J. H., et al. 2016, ApJ, 824, L1 [NASA ADS] [CrossRef] [Google Scholar]
  12. Giovanelli, R., & Haynes, M. P. 2015, A&ARv, 24, 1 [NASA ADS] [CrossRef] [Google Scholar]
  13. Giovanelli, R., Haynes, M. P., Kent, B. R., et al. 2005, AJ, 130, 2598 [NASA ADS] [CrossRef] [Google Scholar]
  14. Haynes, M. P., Giovanelli, R., Martin, A. M., et al. 2011, AJ, 142, 170 [NASA ADS] [CrossRef] [Google Scholar]
  15. Haynes, M. P., Giovanelli, R., Kent, B. R., et al. 2018, ApJ, 861, 49 [Google Scholar]
  16. Hu, W., Wang, X., Wu, F., et al. 2020, MNRAS, 493, 5854 [CrossRef] [Google Scholar]
  17. Huang, S., Haynes, M. P., Giovanelli, R., & Brinchmann, J. 2012, ApJ, 756, 113 [NASA ADS] [CrossRef] [Google Scholar]
  18. Jaffé, Y. L., Poggianti, B. M., Verheijen, M. A. W., Deshev, B. Z., & van Gorkom, J. H. 2013, MNRAS, 431, 2111 [CrossRef] [Google Scholar]
  19. Jiang, P., Yue, Y., Gan, H., et al. 2019, Sci. China Phys. Mech. Astron., 62, 959502 [CrossRef] [Google Scholar]
  20. Jiang, P., Tang, N. Y., Hou, L. G., et al. 2020, Res. Astron. Astrophys., 20, 064 [CrossRef] [Google Scholar]
  21. Kelvin, L. S., Driver, S. P., Robotham, A. S. G., et al. 2012, MNRAS, 421, 1007 [NASA ADS] [CrossRef] [Google Scholar]
  22. Kennicutt, R. C., Jr. 1998, ARA&A, 36, 189 [Google Scholar]
  23. Kerp, J., Winkel, B., Ben Bekhti, N., Flöer, L., & Kalberla, P. M. W. 2011, Astron. Nachr., 332, 637 [NASA ADS] [CrossRef] [Google Scholar]
  24. Koribalski, B. S., Staveley-Smith, L., Westmeier, T., et al. 2020, MNRAS, submitted [arXiv:2002.07311] [Google Scholar]
  25. Lang, R. H., Boyce, P. J., Kilborn, V. A., et al. 2003, MNRAS, 342, 738 [NASA ADS] [CrossRef] [Google Scholar]
  26. Leroy, A. K., Walter, F., Brinks, E., et al. 2008, AJ, 136, 2782 [NASA ADS] [CrossRef] [Google Scholar]
  27. Maddox, N., Hess, K. M., Obreschkow, D., Jarvis, M. J., & Blyth, S. L. 2015, MNRAS, 447, 1610 [NASA ADS] [CrossRef] [Google Scholar]
  28. Meyer, M. 2009, Panoramic Radio Astronomy: Wide-field 1-2 GHz Research on Galaxy Evolution, 15 [Google Scholar]
  29. Meyer, M. J., Zwaan, M. A., Webster, R. L., et al. 2004, MNRAS, 350, 1195 [NASA ADS] [CrossRef] [Google Scholar]
  30. Molina, J., Ibar, E., Villanueva, V., et al. 2019, MNRAS, 482, 1499 [CrossRef] [Google Scholar]
  31. Nan, R. 2006, Sci. China Ser. G, 49, 129 [CrossRef] [Google Scholar]
  32. Nan, R., Li, D., Jin, C., et al. 2011, Int. J. Mod. Phys. D, 20, 989 [NASA ADS] [CrossRef] [Google Scholar]
  33. Neeleman, M., Prochaska, J. X., Ribaudo, J., et al. 2016, ApJ, 818, 113 [NASA ADS] [CrossRef] [Google Scholar]
  34. Oke, J. B., & Gunn, J. E. 1983, ApJ, 266, 713 [NASA ADS] [CrossRef] [Google Scholar]
  35. Parkash, V., Brown, M. J. I., Jarrett, T. H., & Bonne, N. J. 2018, ApJ, 864, 40 [CrossRef] [Google Scholar]
  36. Qiu, Y. H. 1998, MNRAS, 301, 827 [CrossRef] [Google Scholar]
  37. Rhee, J., Lah, P., Briggs, F. H., et al. 2018, MNRAS, 473, 1879 [NASA ADS] [CrossRef] [Google Scholar]
  38. Roberts, M. S. 1978, AJ, 83, 1026 [NASA ADS] [CrossRef] [Google Scholar]
  39. Romeo, A. B. 2020, MNRAS, 491, 4843 [NASA ADS] [CrossRef] [Google Scholar]
  40. Salucci, P. 2019, A&ARv, 27, 2 [CrossRef] [Google Scholar]
  41. Schmidt, M. 1959, ApJ, 129, 243 [NASA ADS] [CrossRef] [Google Scholar]
  42. Schoniger, F., & Sofue, Y. 1994, A&A, 283, 21 [Google Scholar]
  43. Sofue, Y. 1992, PASJ, 44, L231 [Google Scholar]
  44. Stevens, A. R. H., Diemer, B., Lagos, C. D. P., et al. 2019, MNRAS, 490, 96 [CrossRef] [Google Scholar]
  45. Tiley, A. L., Bureau, M., Saintonge, A., et al. 2016, MNRAS, 461, 3494 [CrossRef] [Google Scholar]
  46. Valiante, E., Smith, M. W. L., Eales, S., et al. 2016, MNRAS, 462, 3146 [NASA ADS] [CrossRef] [Google Scholar]
  47. Villanueva, V., Ibar, E., Hughes, T. M., et al. 2017, MNRAS, 470, 3775 [NASA ADS] [CrossRef] [Google Scholar]
  48. Wang, J., Koribalski, B. S., Serra, P., et al. 2016, MNRAS, 460, 2143 [NASA ADS] [CrossRef] [Google Scholar]
  49. Winkel, B., Kalberla, P. M. W., Kerp, J., & Flöer, L. 2010, ApJS, 188, 488 [NASA ADS] [CrossRef] [Google Scholar]
  50. Wong, O. I., Ryan-Weber, E. V., Garcia-Appadoo, D. A., et al. 2006, MNRAS, 371, 1855 [NASA ADS] [CrossRef] [Google Scholar]
  51. Yun, M. S., Ho, P. T. P., & Lo, K. Y. 1994, Nature, 372, 530 [NASA ADS] [CrossRef] [PubMed] [Google Scholar]
  52. Zhang, W., Li, C., Kauffmann, G., et al. 2009, MNRAS, 397, 1243 [NASA ADS] [CrossRef] [Google Scholar]
  53. Zhang, K., Wu, J., Li, D., et al. 2019, Sci. China Phys. Mech. Astron., 62, 959506 [CrossRef] [Google Scholar]

Appendix A: Data reduction and flux calibration

Figure A.1 shows the data reduction flowchart. The data is reduced in batches of 10 min. Each batch includes 300 spectra for the ON target and 300 spectra red for the OFF target (background). If the redshifted H I line suffers from temporal RFI, we simply ignore these 10 min data. The left two panels in Fig. A.1 are the spectra we obtained in the 5 min ON (upper panel) and 5 min OFF the target (bottom panel).

thumbnail Fig. A.1.

Data reduction pipeline. Left two panels: each panel shows the raw spectra for the 5 min ON- or OFF-target positions (upper and lower panel respectively). The x-axis is the observed frequency and the y-axis shows the sampling time of the spectra, with a sampling rate of once every 1.006632926 s (0.993411 Hz). The signal intensity in individual spectra (300 of them in each panel) is represented by colours, with the scale illustrated by the colour bar on the right side of each panel. We note that those spectra in red stripes were sampled when calibration noise was on, and those in blue stripes were sampled when the calibration noise was off (see text). The units of the colour bar are digital number/1012 from ADU, which needs to be calibrated into temperature units. Middle left panels: we split the raw spectra into four cases: ON/OFF the target with calibration noise on/off. Middle right panel: average spectra of the four classes. We show the total background in blue colour, and the background from the first and remaining 74 s (green and purple, respectively) in the lower two panels. The green line shows a lower temperature than the purple line, implying an increasing system temperature. Right panel: spectra ON and OFF the target.

A noise diode is used for FAST signal calibration (Jiang et al. 2020). When it is ‘on’, it injects a noise of known temperature to the receiver. The noise diode is switched between ‘on’ and ‘off’ repeatedly during the whole FAST observation. Noise ‘on’ and ‘off’ are in periods that are twice the signal sampling period. We extract one spectrum per 1.006632926 s. Half of the spectra contains noise with known temperature from the noise diode. For a given target, comparing the spectra with noise diode ‘on’ and ‘off’, we can derive the intensity (i.e. antenna temperature) of these spectra (Jiang et al. 2020).

Our spectra can be separated into four classes, ON target with noise diode ‘on’ and ‘off’ ( f ON cal on $ f_{\mathrm{ON}}^{\mathrm{cal\,on}} $ and f ON cal off $ f_{\mathrm{ON}}^{\mathrm{cal\,off}} $), and OFF target with noise diode ‘on’ or ‘off’ ( f OFF cal on $ f_{\mathrm{OFF}}^{\mathrm{cal\,on}} $ and f OFF cal off $ f_{\mathrm{OFF}}^{\mathrm{cal\,off}} $). These four classes are shown in the middle-left panels of Fig. A.1. We note that these have not yet been calibrated in flux density.

The flux calibration is done via the following two formulas: T ON cal off / f ON cal off = T noise / ( f ON cal on − f ON cal off ) $ T_{\mathrm{ON}}^{\mathrm{cal\,off}}/f_{\mathrm{ON}}^{\mathrm{cal\,off}} = T_{\mathrm{noise}}/(f_{\mathrm{ON}}^{\mathrm{cal\,on}}-f_{\mathrm{ON}}^{\mathrm{cal\,off}}) $, and ( T ON cal on + T noise ) / f ON cal on = T noise / ( f ON cal on − f ON cal off ) $ (T_{\mathrm{ON}}^{\mathrm{cal\,on}}+T_{\mathrm{noise}})/f_{\mathrm{ON}}^{\mathrm{cal\,on}} = T_{\mathrm{noise}}/(f_{\mathrm{ON}}^{\mathrm{cal\,on}}-f_{\mathrm{ON}}^{\mathrm{cal\,off}}) $, where Tnoise is the known noise temperature in Kelvin, the T ON cal on $ T_{\mathrm{ON}}^{\mathrm{cal\,on}} $ and T ON cal off $ T_{\mathrm{ON}}^{\mathrm{cal\,off}} $ are the calibrated antenna temperatures of the signal data, with noise diode ‘on’ and ‘off’ (see Sect. 3.1 of Jiang et al. 2020). The average calibrated spectra of four classes are shown in the middle right panels of Fig. A.1. These are inverse-variance weighted means of individual spectra, and the rms error is estimated as σ ∝ T sys / Δ V t $ \sigma \propto T_{\mathrm{sys}}/\sqrt{\Delta V t} $, where the system temperature Tsys ∼ 20 K, the channel width ΔV = 1.7 km s−1, and the t here is the integration time. The right panel of Fig. A.1 shows the weighted average spectra temperature (Ta) of the ON and OFF targets in the units of Kelvin.

We find a clear increment of the noise as we move to higher frequencies, starting at ∼1373.5 MHz to 1374.5 MHz (see white boxes in Fig. A.1), indicating an unstable background noise. In order to assess this quantitatively, we derive additional mean spectra for the OFF position in two time intervals: the first 0 ∼ 74 s and the second 75 ∼ 147 s separately. We compare the mean OFF spectra of different time intervals and find that the background temperature increases by ∼0.1 K during the 5 min integration, with a larger increase of about 0.3 K at 1373.5 MHz (purple and green curves in the middle-right panel of Fig. A.1). In the right panel of Fig. A.1, we can also see that the target OFF spectrum (blue line) is about 0.15 K higher than the target ON spectrum (red).

We adopt a 14.86 K Jy−1 conversion factor, which has an accuracy of the order of 10% (Jiang et al. 2020).

We subtract the baseline using model-fit based on a sinusoidal plus a linear function, where the sine function accounts for the standing wave, and the linear function represents the trend of the baseline between 1372 MHz and 1376 MHz. As an example, the final steps of the data reduction of a target are illustrated in Fig. A.2. The generation of the final spectrum can be seen in the left panel of Fig. A.2. The minus temperature is caused by the increasing background noise. We correct the Doppler velocity and convert the spectrum velocity into kinematical local standard of rest, and show the final spectrum in the right panel of Fig. A.2.

thumbnail Fig. A.2.

Flux calibration. Left panel: we fit the baseline from the data shown in green dots with the function A * sin(k x + b)+m x + n to account for the stand wave. Right panel: we correct the Doppler velocity and convert the antenna temperatures to flux intensities in mJy. H I emission from galaxies should have no polarisation. As a consistency check, we also reduce the spectrum including only one polarisation, and show the results in blue and red.

Each FAST spectrum includes the spectra with two polarisations individually. As H I emission from galaxies should not be polarised, we inspected both polarisations separately to check whether or not the rms or detections are consistent. The right panel of Fig. A.2 shows the two polarisations in blue and red. We do not find a significant differences in polarisation, and so we simply combined the spectra of both polarisations together during the data reduction.

Appendix B: Fake colour image of our targets

We show the CO contour and the optical images of our targets in Fig. B.1. We can see that the CO morphology is compact in the galaxy centre for HATLASJ085346.4+001252, HATLASJ083601.5+002617, and HATLASJ085112.9+010342. For the galaxy HATLASJ085340.7+013348, the CO morphology is clumpy. We also see that the H2 to H I flux ratio for HATLASJ085340.7+013348 is about 15, while the other two targets have CO-to-H I flux ratios of about ten. This might suggest a more efficient formation of H2 molecules in a nuclear region, and a possible connection between CO morphology and the CO-to-H I flux ratio.

thumbnail Fig. B.1.

Colour images composed of g, r, i band images from the HSC survey wide fields (Aihara et al. 2019), as well as the red contour of ALMA CO (1−0) moment-zero map. The contours are shown in [2, 3, 5] × rms. We also show the H I spectrum in the 20 km s−1 velocity bin in the lower right corner of each panel. The FAST beam size (2.9 arcmin) corresponds to about 130 kpc at a redshfit of 0.05, which is much larger than the scale of the images.

All Tables

Table 1.

FAST observation results.

In the text
Table 2.

Target properties.

In the text

All Figures

thumbnail Fig. 1.

Main sequence of the low-z galaxies (black contour and greyscale) and our ALMA detected VALES sample (blue dots). The FAST observed targets are shown as red dots.
In the text
thumbnail Fig. 2.

H I and CO spectra of our targets. The black thick lines show the H I spectra obtained from FAST. The H I spectrum are rebinned into 20 km s−1 resolution. The red dot lines are the CO(1−0) line from our previous ALMA survey. The scales of the CO emission lines are shown in the right y-axis. We adopt the optical spectroscopy redshift to derive the line velocity. We show the ALFALFA H I spectrum of the galaxy HATLASJ083601.5+002617 in blue in the upper left panel. Effective on-target integration time of Arecibo telescope is about 48 s. The edges of the emission lines are very sharp, and we obtain a line width consistent with the ALFALFA results. The grey regions highlight the velocity channels we used to derive the H I flux.
In the text
thumbnail Fig. 3.

Predicted H I masses of our VALES sample based on the empirical relations in Zhang et al. (2009, in purple circles) and in Catinella et al. (2013, in blue circles). The black dots are the galaxies from HIPASS (Parkash et al. 2018). The red dots show the results of the FAST observation. We link the predicted and the observed H I mass by dotted lines. We show the FAST non-detected target by upper limits.
In the text
thumbnail Fig. A.1.

Data reduction pipeline. Left two panels: each panel shows the raw spectra for the 5 min ON- or OFF-target positions (upper and lower panel respectively). The x-axis is the observed frequency and the y-axis shows the sampling time of the spectra, with a sampling rate of once every 1.006632926 s (0.993411 Hz). The signal intensity in individual spectra (300 of them in each panel) is represented by colours, with the scale illustrated by the colour bar on the right side of each panel. We note that those spectra in red stripes were sampled when calibration noise was on, and those in blue stripes were sampled when the calibration noise was off (see text). The units of the colour bar are digital number/1012 from ADU, which needs to be calibrated into temperature units. Middle left panels: we split the raw spectra into four cases: ON/OFF the target with calibration noise on/off. Middle right panel: average spectra of the four classes. We show the total background in blue colour, and the background from the first and remaining 74 s (green and purple, respectively) in the lower two panels. The green line shows a lower temperature than the purple line, implying an increasing system temperature. Right panel: spectra ON and OFF the target.
In the text
thumbnail Fig. A.2.

Flux calibration. Left panel: we fit the baseline from the data shown in green dots with the function A * sin(k x + b)+m x + n to account for the stand wave. Right panel: we correct the Doppler velocity and convert the antenna temperatures to flux intensities in mJy. H I emission from galaxies should have no polarisation. As a consistency check, we also reduce the spectrum including only one polarisation, and show the results in blue and red.
In the text
thumbnail Fig. B.1.

Colour images composed of g, r, i band images from the HSC survey wide fields (Aihara et al. 2019), as well as the red contour of ALMA CO (1−0) moment-zero map. The contours are shown in [2, 3, 5] × rms. We also show the H I spectrum in the 20 km s−1 velocity bin in the lower right corner of each panel. The FAST beam size (2.9 arcmin) corresponds to about 130 kpc at a redshfit of 0.05, which is much larger than the scale of the images.
In the text

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.