Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2022 Feb;602(7895):58-62.
doi: 10.1038/s41586-021-04294-5. Epub 2022 Feb 2.

Microwave background temperature at a redshift of 6.34 from H2O absorption

Dominik A Riechers  1 , Axel Weiss  2 , Fabian Walter  3   4 , Christopher L Carilli  4 , Pierre Cox  5 , Roberto Decarli  6 , Roberto Neri  7
Affiliations

Microwave background temperature at a redshift of 6.34 from H2O absorption

Dominik A Riechers et al. Nature. 2022 Feb.

Abstract

Distortions of the observed cosmic microwave background provide a direct measurement of the microwave background temperature at redshifts from 0 to 1 (refs. 1,2). Some additional background temperature estimates exist at redshifts from 1.8 to 3.3 based on molecular and atomic line-excitation temperatures in quasar absorption-line systems, but are model dependent3. No deviations from the expected (1 + z) scaling behaviour of the microwave background temperature have been seen4, but the measurements have not extended deeply into the matter-dominated era of the Universe at redshifts z > 3.3. Here we report observations of submillimetre line absorption from the water molecule against the cosmic microwave background at z = 6.34 in a massive starburst galaxy, corresponding to a lookback time of 12.8 billion years (ref. 5). Radiative pumping of the upper level of the ground-state ortho-H2O(110-101) line due to starburst activity in the dusty galaxy HFLS3 results in a cooling to below the redshifted microwave background temperature, after the transition is initially excited by the microwave background. This implies a microwave background temperature of 16.4-30.2 K (1σ range) at z = 6.34, which is consistent with a background temperature increase with redshift as expected from the standard ΛCDM cosmology4.

PubMed Disclaimer

Conflict of interest statement

The authors declare no competing interests.

Figures

Fig. 1
Fig. 1. Broad-band, 3-mm spectroscopy of the starburst galaxy HFLS3 at a redshift of 6.34 with NOEMA.
Black/yellow histogram, NOEMA spectroscopy data, binned to 40-MHz (158 km s−1 at 75.9 GHz) spectral resolution. Expected frequencies of molecular and atomic lines at the redshift of HFLS3 are indicated, with the dominant species labelled in black. The dashed red box indicates the frequency range of the ortho-H2O(110−101) line, which is detected in absorption against the CMB.
Fig. 2
Fig. 2. H2O line and continuum properties and modelling of HFLS3.
Ortho-H2O energy-level structure (a, red solid arrows are detected transitions, grey dotted lines are upper limits and blue dashed arrows are pumping transitions, with observed and model-predicted absorption/emission lines indicated as upward/downward arrows, respectively; percentages are the level populations in the model) and zoom-in on the H2O line at the same spectral resolution as in Fig. 1 to show that the line absorbs into the CMB (b, blue shading added for emphasis). The black curve is a fit to the spectrum. The red dashed curve is the best-fit radiative transfer model. Source data
Fig. 3
Fig. 3. Radiative transfer models for HFLS3 and constraints on the CMB temperature.
a, Model grid for the predicted line-absorption strength for TCMB(z = 6.34) = 20.0 K (greyscale) as a function of H2O column density (y axis) and radius of the dust-emission region at 108 μm (x axis). The white curves show the parameter space allowed by the measurement (solid line) and the −1σ r.m.s. uncertainty region (dotted line). The dashed black lines show the measured continuum size (left) and +1σ r.m.s. uncertainty region (right). The overlapping region between the white boundary (that is, the minimum allowed absorption strength) and the size measurement (that is, the minimum required emitting area at 100% covering fraction) is the allowed parameter space for the absorption strength within 1σ r.m.s. The minimum required radius at N(H2O)  ~ 1017 cm−2 is due to a minimum in Tex in the models. b, Constraints on TCMB for the observed absorption strength (green line and shaded region) at the minimum size compatible with the observations (a), based on the same models (red/blue shaded regions are the allowed ranges within the source radius +1σ/+2σ r.m.s.). The source radius at face value (black line), as well as the −1σ and −2σ r.m.s. regions (not shown), are ruled out by the observations. The minimum filling factor of the dust emission region CFmin is indicated for the +1σ and +2σ r.m.s. regions. The grey dashed line shows a model assuming a continuum radius of 5 kpc, which provides a conservative lower limit on TCMB. c, Observability of the H2O absorption as a function of redshift for three solutions allowed by the data without and with collisional excitation. The effect becomes observable at z ~ 4.5 and remains visible at similar strength to z > 12. The lower-redshift limit is higher in cases where collisional excitation is important, but the impact is minor below n(H2) = 105 cm−3.
Fig. 4
Fig. 4. Measurements of the CMB temperature as a function of redshift,,–.
a, 1σ (black) and 2σ r.m.s. (grey) uncertainties are shown for HFLS3 and 1σ r.m.s. uncertainties elsewhere. b, Same data but dividing out the (1 + z) redshift scaling of the CMB expected from ΛCDM. Previous direct measurements are from CMB mapping at z = 0 and SZ effect measurements of galaxy clusters in contrast to the CMB out to z ~ 1 (left zoom-in panel in b). Additional measurements are from ultraviolet absorption systems along the lines of sights to quasars out to z ~ 3. The downward (upward) triangles are not corrected (corrected) for the contribution of collisional excitation in the diffuse interstellar medium to the excitation temperature Tex of the tracer (right zoom-in panel in b; green dots show an alternative proposed correction),. The separation of these pairs of points for the same sources is an indication of the systematic uncertainties on top of the statistical uncertainties indicated by the error bars. The H2O-based measurement of HFLS3, like those up to z ~ 1, is in contrast to the CMB, but—as a line measurement—it is more precise in redshift. It is not subject to the same uncertainties in Tex as the intermediate redshift measurements, because collisions can only decrease (rather than boost) the resulting absorption strength into the CMB for the H2O-based measurement. They are also unlikely to play an important role due to the high density required to collisionally excite the relevant H2O lines. Ignoring collisions results in the most conservative estimate of TCMB for HFLS3. The orange shaded region shows a TCMB = TCMB(z = 0)*(1 + z)1 − β fit to the data in Extended Data Table 1 and its uncertainty (where TCMB(z = 0) = 2.72548 ± 0.00057 K (ref. ) and β = (3.4−7.3+8.1) × 10−3), the orange dashed line indicates the β = 0 case corresponding to the standard cosmology and the dotted lines indicate a ±10% deviation in 1 − β.
Extended Data Fig. 1
Extended Data Fig. 1. Combined effect of CMB absorption and starburst radiation field on the strength of the H2O(110–101) line in HFLS3.
Top, the cold H2O vapour is exposed to the CMB radiation field, which has the shape of a black-body function (TCMB = 20.0 K at z = 6.34), and the starburst infrared radiation field, which has the shape of a grey-body function (Tdust = 63 K). NOEMA observed the signal in contrast to the CMB and therefore detects only the dust emission from the starburst and the H2O line, but not the CMB itself (which therefore fills the region below zero flux density, as seen by the telescope). Bottom left, as the energy-level difference for the H2O(110–101) line is only 26.7 K, there are sufficient CMB photons at z = 6.34 to thermalize the level population between both levels, such that Tex is the same as TCMB in equilibrium. Therefore no H2O emission or absorption will be observed, despite the presence of a ‘seed’ population in the upper level. Bottom right, the radiation field of the starburst alters the level populations towards increased higher-level populations. Owing to the grey-body shape of its spectral energy distribution, more photons are available at 108 μm to increase the 221 level population from the 110 state than there are 538-μm photons available to increase the 110 level population from the 101 state, relative to the ‘seed’ population provided by the absorption of CMB photons. Therefore the relative population of the 110 and 101 levels is lower than in thermal equilibrium, such that the resulting Tex is lower than TCMB. As a result, the H2O(110–101) line is observed in absorption towards the CMB due to the negative temperature contrast—as observed towards HFLS3.
Extended Data Fig. 2
Extended Data Fig. 2. H2O line emission integrated moment 0 and continuum maps of HFLS3.
a–c, H2O contour maps (blue) before (b) and after (c) continuum subtraction, and local continuum (a, green contours) at the wavelength of the H2O line, overlaid on the 158-μm continuum (intensity scale). H2O emission is integrated over the central 395 km s−1 (100 MHz). d–e, Rest-frame 122-μm continuum emission (orange contours and intensity scale) as a proxy for the 108-μm continuum size, showing the full emission (d), and the compact nuclear region that accounts for two-thirds of the emission at higher resolution (e), overlaid with 158-μm contours (red) for orientation. f, Radially averaged visibility amplitude as a function of interferometer baseline length for the data in d and e. The radial profile of the visibility amplitude (binned to 50-m steps, with 1σ error bars) shows that the 122-μm dust emission is clearly resolved. Observed-frame 538-μm continuum contours (a) are shown in steps of 1σ = 22.5 μJy beam−1, starting at ±3σ. H2O contours (b, c) are shown in steps of 1σ = 0.0375 Jy kms−1 beam−1, starting at ±2σ. Contours of 122 μm (d, e) are shown in steps of +/−10σ and +/−5σ, where 1σ = 229 and 374 μJy beam−1, respectively. Contours of 158 μm (d, e) are shown in steps of 3σ, starting at ±5σ, where 1σ = 400 μJy beam−1 (all uncertainties are r.m.s.). Negative intensity contours are dashed.
See this image and copyright information in PMC

Similar articles

  • A dust-obscured massive maximum-starburst galaxy at a redshift of 6.34.
    Riechers DA, Bradford CM, Clements DL, Dowell CD, Pérez-Fournon I, Ivison RJ, Bridge C, Conley A, Fu H, Vieira JD, Wardlow J, Calanog J, Cooray A, Hurley P, Neri R, Kamenetzky J, Aguirre JE, Altieri B, Arumugam V, Benford DJ, Béthermin M, Bock J, Burgarella D, Cabrera-Lavers A, Chapman SC, Cox P, Dunlop JS, Earle L, Farrah D, Ferrero P, Franceschini A, Gavazzi R, Glenn J, Solares EA, Gurwell MA, Halpern M, Hatziminaoglou E, Hyde A, Ibar E, Kovács A, Krips M, Lupu RE, Maloney PR, Martinez-Navajas P, Matsuhara H, Murphy EJ, Naylor BJ, Nguyen HT, Oliver SJ, Omont A, Page MJ, Petitpas G, Rangwala N, Roseboom IG, Scott D, Smith AJ, Staguhn JG, Streblyanska A, Thomson AP, Valtchanov I, Viero M, Wang L, Zemcov M, Zmuidzinas J. Riechers DA, et al. Nature. 2013 Apr 18;496(7445):329-33. doi: 10.1038/nature12050. Nature. 2013. PMID: 23598341
  • A massive, quiescent galaxy at a redshift of 3.717.
    Glazebrook K, Schreiber C, Labbé I, Nanayakkara T, Kacprzak GG, Oesch PA, Papovich C, Spitler LR, Straatman CM, Tran KH, Yuan T. Glazebrook K, et al. Nature. 2017 Apr 5;544(7648):71-74. doi: 10.1038/nature21680. Nature. 2017. PMID: 28382981
  • A galaxy at a redshift z = 6.96.
    Iye M, Ota K, Kashikawa N, Furusawa H, Hashimoto T, Hattori T, Matsuda Y, Morokuma T, Ouchi M, Shimasaku K. Iye M, et al. Nature. 2006 Sep 14;443(7108):186-8. doi: 10.1038/nature05104. Nature. 2006. PMID: 16971942
  • Molecular gas in the host galaxy of a quasar at redshift z = 6.42.
    Walter F, Bertoldi F, Carilli C, Cox P, Lo KY, Neri R, Fan X, Omont A, Strauss MA, Menten KM. Walter F, et al. Nature. 2003 Jul 24;424(6947):406-8. doi: 10.1038/nature01821. Nature. 2003. PMID: 12879063
  • Measuring our Universe from Galaxy Redshift Surveys.
    Lahav O, Suto Y. Lahav O, et al. Living Rev Relativ. 2004;7(1):8. doi: 10.12942/lrr-2004-8. Epub 2004 Jul 16. Living Rev Relativ. 2004. PMID: 28163643 Free PMC article. Review.
See all similar articles

Cited by

  • Cosmology and fundamental physics with the ELT-ANDES spectrograph.
    Martins CJAP, Cooke R, Liske J, Murphy MT, Noterdaeme P, Schmidt TM, Alcaniz JS, Alves CS, Balashev S, Cristiani S, Di Marcantonio P, Génova Santos R, Gonçalves RS, González Hernández JI, Maiolino R, Marconi A, Marques CMJ, Melo E Sousa MAF, Nunes NJ, Origlia L, Péroux C, Vinzl S, Zanutta A. Martins CJAP, et al. Exp Astron (Dordr). 2024;57(1):5. doi: 10.1007/s10686-024-09928-w. Epub 2024 Feb 26. Exp Astron (Dordr). 2024. PMID: 39308933 Free PMC article.

References

    1. Sunyaev RA, Zel’dovich IAB. Microwave background radiation as a probe of the contemporary structure and history of the universe. Annu. Rev. Astron. Astrophys. 1980;18:537–560.
    1. Rephaeli Y. Comptonization of the cosmic microwave background: the Sunyaev-Zeldovich effect. Annu. Rev. Astron. Astrophys. 1995;33:541–580.
    1. Klimenko VV, et al. Estimation of the cosmic microwave background temperature from atomic C I and molecular CO lines in the interstellar medium of early galaxies. Astron. Lett. 2020;46:715–725.
    1. Planck Collaboration Planck 2018 results. I. Overview and the cosmological legacy of Planck. Astron. Astrophys. 2020;641:A1.
    1. Riechers DA, et al. A dust-obscured massive maximum-starburst galaxy at a redshift of 6.34. Nature. 2013;496:329–333. - PubMed

Publication types