An analysis of the Hubble Space Telescope's deepest view of the universe offers compelling evidence that monster black holes in the centres of galaxies were not born big but grew over time through repeated galactic mergers. The Hubble Ultra Deep Field (HUDF) studies also confirm recent computer simulations that predict that that newly merging galaxies are enshrouded in so much dust that astronomers cannot see black holes feasting on stars and gas from the mergers. The computer simulations, as supported by Hubble, suggest that it takes hundreds of millions to a billion years before enough dust clears so that astronomers can see the black holes feasting on stars and gas from the merger.
These postage-stamp-size images reveal 36 young galaxies caught in the act of merging with other galaxies. These galaxies appear as they existed many billions of years ago. Astronomers have dubbed them "tadpole galaxies" because of their distinct knot-and-tail shapes, which suggest that they are engaging in galactic mergers.
Expand (689kb, 3000 x 2400) Credit: NASA Each image is roughly 84,000 light-years on a side, which is about the size of our Milky Way Galaxy today. The tadpole galaxies are shown in the middle of each image and are considerably smaller than today's giant galaxies. The image was taken by Hubble's Advanced Camera for Surveys.
The galaxies were captured in 2004 in the Hubble Space Telescope's Ultra Deep Field (HUDF) survey of thousands of distant galaxies. They are part of more than 165 tadpole galaxies in the HUDF studied recently by a team of astronomers. The team was looking for indications of black hole activity in these young galaxies. A characteristic signature of such activity is a fluctuation in brightness over time, an indication that a black hole is feasting on surrounding stars and gas. The flickering light does not come from the black hole itself but from the area immediately surrounding the black hole. Astronomers did not see brightness fluctuations in any of the tadpole galaxies they surveyed. They did, however, observe the fluctuations in 46 different faint galaxies in the HUDF. These galaxies existed millions of years after the tadpole galaxies. This result suggests that black holes did not begin eating when galaxies merged. Rather, it took several hundred million years for the gas and stars from the merger to arrive on the black hole's dinner plate and become visible as flickering light. This finding agrees with recent computer models, which predicted that the feeding habits of black holes would become visible after galactic mergers.
At visible wavelengths, the light from the galaxy HUDF-JD2 is absorbed by intervening hydrogen gas, and so the galaxy appears faint in the Hubble visible and near-infrared images. The surprise is how bright is appears to Spitzer in the infrared, suggesting a very massive and distant galaxy. Spitzer observations were also independently reported by Dr. Laurence Eyles from the University of Exeter in the United Kingdom and Dr. Haojing Yan of the Spitzer Science Centre, Pasadena. They also revealed evidence for mature stars in more ordinary, less massive galaxies at similar distances, when the universe was less than one billion years old.
Expand(192kb, 900 x 720) Credit: NASA, ESA/JPL-Caltech/B. Mobasher (STScI/ESA)
Expand (87kb, 742 x 742) Position(2000): R.A. 3h 32m 28s.74 Dec -27° 48' 39".9
Composite of visible-light (Hubble) and infrared (Spitzer) images of the distant galaxy HUDF-JD2, UDF033238.74-274839.9 in the Hubble Ultra Deep Field of the constellation Fornax.
One of the most distant galaxies ever studied is more massive and mature than expected, astronomers announced today. The finding suggests some galaxies grew up much more quickly than conventional wisdom held.
The galaxy, named HUDF-JD2, is seen as the universe was only about 800 million years old. The universe today is about 13.6 billion years old. The galaxy HUDF-JD2 was found among approximately 10,000 others in a small area of sky called the Hubble Ultra Deep Field (HUDF). This is the deepest images of the universe ever made at optical and near-infrared wavelengths.
"This galaxy appears to have 'bulked up' amazingly quickly, within a few hundred million years after the Big Bang. It made about eight times more mass in terms of stars than are found in our own Milky Way today, and then, just as suddenly, it stopped forming new stars. It appears to have grown old prematurely" - Bahram Mobasher, European Space Agency and the Space Telescope Science Institute.
The galaxy was spotted in an infrared image made last year as part of the Hubble Space Telescope's Ultra Deep Field survey. Follow-up observations were done with NASA’s Spitzer Space Telescope and the European Southern Observatory's Very Large Telescope. Spitzer records infrared light at wavelengths longer than what Hubble sees, making it better at spotting older, redder stars. These old stars in the galaxy are the clues to its overall maturity. The leading theory of galaxy formation holds that small galaxies merged to gradually form larger ones. But the newfound galaxy is so massive at such an early epoch that astronomers now think that at least some galaxies formed more quickly in a monolithic manner. Other recent observations have begun to reveal similar disparities between theory and reality.
"This would be quite a big galaxy even today. At a time when the Universe was only 800 million years old, it's positively gigantic" - Mark Dickinson, study team member at the National Optical Astronomy Observatory.
The galaxy likely harbours a supermassive black hole at its centre. In recent years, black holes have been seen as a crucial element in the formation and evolution of galaxies, but theorists are still unsure of the exact details of how it all comes together.
It is one of the major goals of observational cosmology to trace the way galaxies formed and evolved and to compare it to predictions from theoretical models. It is therefore essential to know as precisely as possible how many galaxies were present in the Universe at different epochs.
This is easier to say than to do. Indeed, if counting galaxies from deep astronomical images is relatively straightforward, measuring their distance - hence, the epoch in the history of the universe where we see it - is much more difficult. This requires taking a spectrum of the galaxy and measuring its redshift.
However, for the faintest galaxies - that are most likely the farthest and hence the oldest - this requires a lot of observing time on the largest of the telescopes. Until now, astronomers had thus to first carefully select the candidate high-redshift galaxies, in order to minimise the time spent on measuring the distance. But it seems that astronomers were too careful in doing so, and hence had a wrong picture of the population of galaxies.
Expand (448kb, 800 x 847) This image shows a small patch of the sky surveyed by the VVDS team in the Cetus (The Whale) constellation. This colour-composite image based on observations made with Megacam at the CFHT indicates a few of the newly found distant galaxies (encircled ones) based on VIMOS/VLT data. In this field of view of about 3 x 3 arcmin, 13 distant galaxies were discovered. The VVDS team discovered in total about 1,000 of such galaxies. (credit: LAM-OAMP/CFHT)
It would be better to "simply" observe in a given patch of the sky all galaxies brighter than a given limit. But looking at one object at a time would make such a study impossible. To take up the challenge, a team of French and Italian astronomers used the largest possible telescope with a highly specialised, very sensitive instrument that is able to observe a very large number of (faint) objects in the remote universe simultaneously.
The astronomers made use of the VIsible Multi-Object Spectrograph (VIMOS) on Melipal, one of the 8.2-m telescopes of ESO's Very Large Telescope Array. VIMOS can observe the spectra of about 1,000 galaxies in one exposure, from which redshifts, hence distances, can be measured. The possibility to observe two galaxies at once would be equivalent to using two VLT Unit Telescopes simultaneously. VIMOS thus effectively multiplies the efficiency of the VLT hundreds of times. This makes it possible to complete in a few hours observations that would have taken months only a few years ago. With capabilities up to ten times more productive than competing instruments, VIMOS offers the possibility for the first time to conduct an unbiased census of the distant Universe.
Using the high efficiency of the VIMOS instrument, the team of astronomers embarked in the VIMOS VLT Deep Survey (VVDS) whose aim is to measure in some selected patch of the sky the redshift of all galaxies brighter than magnitude 24 in the red, that is, galaxies that are up to 16 million fainter than what the unaided eye can see. In a total sample of about 8,000 galaxies selected only on the basis of their observed brightness in red light, almost 1,000 bright and vigorously star forming galaxies were discovered at an epoch 1,500 to 4,500 million years after the Big Bang (redshift between 1.4 and 5).
"To our surprise this is two to six times higher than had been found by previous works. These galaxies had been missed because previous surveys had selected objects in a much more restrictive manner than we did. And they did so to accommodate the much lower efficiency of the previous generation of instruments" - Olivier Le Fèvre, from the Laboratoire d'Astrophysique de Marseille (France) and co-leader of the VVDS project.
While observations and models have consistently indicated that the Universe had not yet formed many stars in the first billion years of cosmic time, the discovery made by the scientists calls for a significant change in this picture. Combining the spectra of all the galaxies in a given redshift range (i.e. belonging to the same epoch), the astronomers could estimate the amount of star formed in these galaxies. They find that the galaxies in the young Universe transform into stars between 10 and 100 times the mass of our Sun in a year.
"This discovery implies that galaxies formed many more stars early in the life of the Universe than had previously been thought. These observations will demand a profound reassessment of our theories of the formation and evolution of galaxies in a changing Universe" - Gianpaolo Vettolani, the other co-leader of the VVDS project, working at INAF-IRA in Bologna (Italy).
It now remains for astronomers to explain how one can create such a large population of galaxies, producing more stars than previously assumed, at a time when the Universe was about 10-20% of its current age.