Local galaxies harbor about a third more neutral hydrogen gas (HI) than was previously thought, a re-analysis of radio telescope observations from Australia, the USA and the Netherlands has found.
And the same study provides evidence for a dramatic drop-off in the prevalence of galaxy HI halos in the universe during the last 12 billion years. These halos are the galaxy refueling reservoirs that are necessary for future generations of stars to be born.
Such a decline in HI was thought likely, as gas accreted onto galaxies and was turned into stars, and the previous lack of evidence for it had been puzzling, says Dr. Robert Braun, Chief Scientist at CSIRO Astronomy and Space Science in Sydney, Australia, who conducted the analysis.
Braun's interest was piqued by looking at images of the Andromeda galaxy (M31) made by observing the 21-cm radio emission from HI. "There was a pattern, but it was not a pattern of positive emission features -- it was a pattern of negative emission features, local minima in the brightness distribution," he said.
It reminded him of the dark dust features seen at optical wavelengths in the Milky Way -- areas where light is being absorbed.
In the radio images, absorption was also going on in the darker areas. But, rather than being absorbed by dust, gas glowing in HI was being hidden -- its radiation being absorbed -- by more cool gas in front of it. Paradoxically, these regions with more gas look dimmer, not brighter. The phenomenon is called "self-absorption".
Radio astronomers have known since the 1950s that the 21-cm hydrogen line can be subject to self-absorption. But, Braun says, in recent decades researchers have often ignored this "inconvenient truth", basing their estimates of HI mass solely on the strength of the spectral line.
Self-absorption, however, is characterized by flattening of the spectral line peak, and broadening of the overall profile. Measuring the detailed shape of the entire profile allows you to better estimate the total amount of HI present.
Temperature variations and physical disruption of the gas clouds can also affect the shape of the HI line, but these changes can be distinguished from the effects of "hidden" HI mass when the measurements have enough precision, Braun says.
Taking self-absorption into account, Braun determined the amount of mass that the spectrum indicated was present at each point observed. The result? Tiny but massive HI clouds popped out in the galaxy images.
The extra HI in such small but massive clouds appears to add 30-36% to the estimate of total HI in the galaxies studied: M31, M33 and the LMC. These galaxies were the only ones for which observations had been made at sufficiently high resolution and sensitivity for the current study to be carried out: they are, however, representative in terms of their morphology and gas content of 90% of the galaxies in the local universe and so are a reasonable guide to the HI content of the local universe.
From this result, Braun has derived how much HI in the local universe (roughly within a 100 million light-year radius) is present as diffuse gas, and how much as dense gas. Comparing this to similar estimates made at higher redshifts (derived from quasar absorption line studies), he suggests that "faint" 21 cm absorbers (i.e., diffuse HI) have declined by a factor of 2.5 since z=1 (8 Gyr ago) and a factor of 5 since z=3 (12 Gyr ago), probably as a result of diffuse HI gas in galaxy haloes and satellites having been "mopped up" by galaxy growth and star formation.
Perhaps surprisingly, the number of "deep" absorbers (dense HI regions) has not changed significantly over the past 12 billion years, Braun says. It is the more extended "refueling" reservoir that has disappeared as the universe has evolved. Today's galaxies are using their last "tankful" of gas to form today's stars. Once that has been used up over the next few billion years, the lights will start to go out.
Braun's study was possible only with data that had high linear resolution (about 100 light-years, the narrowest dimension of the HI clouds). The best result was obtained from the CSIRO image of the Large Magellanic Cloud, the resolution of which was 50 light-years. Making similarly high-resolution images of more distant galaxies can't be done with current telescopes, Braun says, but will be possible with the future Square Kilometer Array radio telescope.
Meanwhile, however, the result of the present study will be useful for a project called FLASH (the First Large Absorption Survey in HI), which is to be carried out with CSIRO's new Australian SKA Pathfinder Telescope in Western Australia. FLASH is a blind HI absorption-line survey that uses background radio continuum sources to identify and characterize foreground neutral hydrogen, and will increase the total number of HI absorption line systems in the redshift range 0.5 < z < 1.0 by up to a hundred-fold.
Ms. Helen Sim
CSIRO Astronomy and Space Science
+61 419 635 905 (UT + 10 hours)
Dr. Robert Braun
+61 2 9372 4271 (UT + 10 hours)
Braun, Robert. "Cosmological Evolution of Atomic Gas and Implications for 21 cm HI Absorption". ApJ 749 87 (10 April 2012).
ApJ website (limited access):
ArXiv preprint server:
Braun's work, now published in The Astrophysical Journal, used archival data from the Parkes and Australia Telescope Compact Array telescopes in eastern Australia, operated by CSIRO (the Commonwealth Scientific and Industrial Research Organization); the Robert C. Byrd Green Bank Telescope and the Karl G. Jansky Very Large Array, both operated by NRAO in the USA; and the Westerbork Synthesis Radio Telescope, operated by ASTRON in the Netherlands.