When the Universe was Young
The current basic model of the Universe is the Big Bang in which the Universe had an initial extremely hot phase and has expanded and cooled since its initial ``Bang". This paradigm has had remarkable success in predicting the existence of the Cosmic Microwave Background (CMB, the ubiquitous radiation field of the Universe), the abundances of the light elements in the Universe and the expansion. The physical nature of the observable Universe is usually described by a number of parameters from which its past and future history may be determined. Among these are the present average density of the Universe, the Hubble Constant (which measures the rate of expansion of the Universe) the age of the Universe, and the smoothness of the CMB. Unfortunately, at the current time most of these parameters are not very well known. By 2010, the second generation CMB satellites will have completed their missions and large ground-based millimeter array surveys of the CMB will be in progress. These programs will either determine most of the cosmological parameters or discover a major flaw in our ideas about the Universe.
In the earliest moments of the Big Bang, the Universe is predicted to have been remarkably isotropic and homogeneous. One of the great attractions of the inflationary model (which predicts that the Universe went through an epoch of extremely rapid expansion very early on in its evolution) is that such homogeneity arises naturally from a wide range of initial conditions. And yet the Universe that we see today is very inhomogeneous on all except very large scales. This inhomogeneity is of course profoundly important, it led directly or indirectly to the emergence of complexity in the Universe on different scales and to the existence of essentially all phenomena that are studied in astrophysics and science generally, including life.
The generally successful theoretical paradigm for the development of density inhomogeneities in the expanding Universe is that small fluctuations in the density initially present grew through the process of gravitational instability. These initial density inhomogeneities may have been the result of quantum fluctuations that became imprinted on the Universe during the inflationary phase, although other possibilities have also been investigated. All structure in the Universe on galactic and larger scales is thought to have arisen through the processes outlined above. The nature of this structure can be studied from a variety of approaches. Locally this can be done from studies of our own and other nearby galaxies which give information on the distribution of masses of galactic systems while studies of the fluctuations on the last scattering surface of the CMB provides information on the largest scales.
The 1990's have seen great progress in observational cosmology, leading in many areas to a first overview of the observational situation and to a resulting sharpening of the observational questions. The major unanswered questions that are the focus of current research in this area are as follows.
On smaller scales astronomers are concerned with how these fluctuations actually produced the galaxies we see around us. Some of the unanswered questions in this field are enumerated below.
The scientific questions in observational cosmology outlined above represent the focus of activity for a substantial segment of the astronomical community and, correspondingly, are the motivation for much of the present investment in new observing facilities in space and on the ground. Therefore it is reasonable to expect substantial progress in addressing them over the next ten years and the state of the field in 2008 is thus quite hard to predict.
The study of large scale structure is primarily statistical, since it is thought to be the product of random processes occuring in the early Universe. Several new facilities will come on line in the next ten years that will extract essentially all of the available information from large scale structure and our knowledge will be to a certain degree limited by cosmic sampling - we have only one Universe available for study!
Several new CMB observational experiments, including Boomerang and CBI in the near future and the satellite observatories MAP and ultimately Planck (2007) will determine the CMB temperature fluctuations on all angular scales down to where the finite thickness of the last scattering surface erases information. While there may well be surprises, it is also possible that the spectrum will be a perfect match to the expectations of the currently popular model (cold dark matter, CDM) and a particular set of cosmological parameters. The latter would be a stunning achievement and mean that much of observational cosmology on the largest scales had been correctly understood.
The Sloan Digital Sky Survey (SDSS) project will measure the redshifts for the 106 brightest galaxies in the northern Galactic Cap, determing the 3-dimensional location of all the more massive galaxies out to a redshift of about z 2#2 0.1, or a tenth of the horizon scale. At greater distance, evolutionary effects would in any case become important, and so SDSS will again be essentially limited by cosmic sampling. The SDSS will be preceded by a smaller but still very impressive survey (105 galaxies) carried out on the AAT with the 2dF facility. The SDSS and other local surveys will be supplemented by very large surveys of order 105 galaxies at redshifts z 2#2 1 (e.g. the VIRMOS survey on the VLT). These will clearly show the development of large scale structure in the Universe.
Major progress in the field of weak gravitational lensing which is used to map the large scale mass distribution in the Universe, will come about through the implementation of wide field imagers (e.g. CFHT12k and MegaCam) with exquisite image quality on telescopes such as the CFHT.
The hectic progress in the area of high redshift galaxies during the last few years has been driven primarily by three new observational facilities that came on-line in the 1990's; the first years of operation of the Keck 10m telescope with its superior light gathering power, the introduction of the highly multiplexed MOS/SIS spectrograph on the CFHT, and the repair of the Hubble Space Telescope (HST) yielding kpc-scale resolution on high redshift galaxies. It should be noted that these facilities all operate in the optical waveband (from 0.3 to 1 3#3). Future progress is likely to be dominated by progress at longer wavelengths.
The near-infrared (1 - 5 3#3) will be critical for two reasons. First, the most distant galaxies presently known have z 2#2 6, and the regime of the earliest stars and galaxies in the Universe is almost certainly at z > 7. Galaxies at these redshifts are completely invisible at wavelengths less than 1 3#3 because of neutral Hydrogen absorption in the intervening Universe. Second, the 1 < z < 5 regime is likely where most of the stars seen today in the Universe were formed and where the different morphological components of galaxies that produce the characteristic Hubble sequence of galaxies today were put in place. The spectral range between 0.3 and 0.7 3#3 is rich in diagnostic features for stellar population ages, metallicities and kinematics and the spectrum at > 0.4 3#3 is also essential for studying the older components of galaxies. This rich spectral range is redshifted into the 1 - 5 3#3 waveband in this crucial 1 < z < 5 redshift range.
The far-infrared, > 10 3#3, will also be critical. Locally, about 35% of stellar luminosity emitted in the optical and ultraviolet is absorbed by dust and is reradiated in the far-IR at a temperature of 30 - 60 K. Globally, the far-IR/sub-mm background that has been recently detected by the COBE satellite has the same or larger energy content as the optical background that is obtained by adding up the light from detected galaxies, indicating that dust absorption continues to be important at earlier epochs. Indeed the picture of galaxy evolution that has emerged over the last few years from optical studies is, literally, only half the story, and very little has been known until now about the nature and redshifts of the sources producing the cosmic far-IR/sub-mm background.
While the requirements are demanding, several developments in the next ten years will almost certainly combine to produce great progress in this area. A much wider community will have had access to 8 - 10m class telescopes. By 2008 there will have been approximately 30,000 clear nights of time on such telescopes, accessed by astronomers in every major astronomical community. This can be compared to about 1000 nights available hitherto to a very restricted user community (Caltech and University of California/Hawaii). The sub-mm waveband around 1 mm is for the first time being opened up with instruments such as SCUBA on the 15m UK/Canada/NL JCMT on Mauna Kea and at IRAM where the potential of these instruments was recently demonstrated with the spectacular detection of CO in a galaxy at redshift 4.7. The sub-mm will continue to be developed with first and second generation interferometers such as OVRO, BIMA and in the future LSA/MMA. In space, SIRTF and FIRST will build on the pioneering ISO mission in the 10 - 100 3#3 waveband.
Multi-object spectrographs on 8 - 10m telescopes will dramatically extend the first exploratory redshift surveys carried out on CFHT - the 103 - 104 galaxies of the CFRS and CNOC surveys carried out with MOS on CFHT will be extended to samples of 105 galaxies in the VIRMOS project on VLT. Many of the new generation of 8 - 10m telescopes coming on line in the next five years are optimized for the near-IR (e.g. Gemini) and are achieving image quality (recently spectacularly demonstrated with the VLT) that is comparable to that achieved by HST in space. Scheduled for launch in 2007, NGST will be an 8m class telescope placed far from Earth where it will passively cool to about 30 K. With diffraction-limited optics, the sensitivity will be limited by the natural zodiacal light. The science mission of the $1B NGST is squarely targeted at solving the science areas of galaxy formation and evolution outlined above.
Stars, Star Formation and Stellar Clusters
We are currently witnessing a resurgence in the field of stellar astronomy. Many of the processes which occur in stellar interiors and play a major role in the way stars function and evolve, are extremely complex and essentially are not understood. These phenomena include convection, turbulence, mixing, transport of angular momentum, the origin of the stellar dynamo and are either totally ignored or grossly parameterized in current modelling of stellar evolution. Other processes occurring at the surface of stars or in their immediate environment also play a decisive role in several stages of stellar evolution, and are basically unknown. Among these are disk accretion, jet collimation, driving of winds from stars and from accretion disks, magnetic phenomena in stellar winds and circumstellar disks.
The attention of the stellar astrophysics community in the coming decade will be directed toward the detailed understanding of these processes. This will require a substantial theoretical effort which must be accompanied by serious observational progress in order to provide the adequate constraints that our models are lacking today.
The physical processes at work in stellar interiors are usually studied by the observation of their manifestation at the stellar surface, most often by high resolution, high S/N spectroscopy. Convection controls the transfer of energy in a significant fraction of stellar interiors. Granulation is one of its surface manifestations, and can be studied by observing the shapes of spectral lines in various types of stars. Turbulence in stellar interiors results in mixing of chemicals, and affects the surface abundances. The measurement of these abundances provides us with an indirect tool for studying turbulence inside stars. Transfer of angular momentum in the internal layers is also an important pending problem. Measurements of surface rotation and surface differential rotation constitute the only conventional tool to study this problem. This can be achieved either statistically on ensembles of stars by measuring the projected rotation velocity, or more directly by measuring the rotational modulation of stellar light (photometric monitoring) and of stellar line profiles (spectroscopic monitoring). Dynamo generation of magnetic fields in stellar interiors is another unsolved major problem in stellar physics. Observations of surface magnetic field geometries are used to constrain dynamo theories, and can be obtained using the methods of Doppler and Zeeman-Doppler imaging. These methods require high resolution (R > 30,000) spectroscopy and spectropolarimetry at very high S/N ratios, from several hundreds for Doppler imaging to several thousands for Zeeman-Doppler imaging, but multi-line techniques reduce considerably these requirements. Monitoring of the stars during several rotational cycles is necessary to apply these imaging methods.
All of these observational tools have been applied to bright stars in the past few years, and have yielded important results concerning the study of the processes listed above. The next important step will be to study how these processes evolve during stellar evolution, and therefore to apply the above tools to ensembles of stars offering a complete age sequence, including pre-main sequence stars and members of stellar clusters.
In addition to these indirect tools for studying the processes at work in stellar interiors, asteroseismology is developing quickly, and is likely to provide us in the near future with direct probes of stellar interiors. Asteroseismology consists of the detection of stellar oscillation modes, and the measurement of their frequencies, amplitudes, and lifetimes. These measurements require very high precision photometry or spectroscopy as well as monitoring for a long enough period of time to yield sufficient frequency precision, typically two weeks or more.
While we have basic ideas on the structure and evolution of stars, their formation remains a key unsolved problem in astrophysics. Without a good theory of the star formation process, no complete understanding of the origin of planetary systems or galaxy evolution can be achieved. Tremendous progress has been made in this field in the last 20 years following the development of infrared astronomy, both from space - (e.g. IRAS, HST) and ground-based facilities. For instance, the circumstellar environment around protostars are better known, important results have been established on the global star formation activity in galaxies, and some knowledge on the level of star formation in high-redshift objects has been gathered. However, many crucial questions remain to be answered.
It is actually rare to have a star form in isolation. They tend to form in pairs, triple or higher order systems and very often in clusters. Star clusters are gravitationally bound groups of stars that all formed at the same time, likely from the same gas cloud. These objects are the laboratories for studying the ages of the stars, how stars evolve, the effects of differing metal abundances on this evolution, and the dynamical evolution of hundreds or thousands of gravitationally bound stars. One example of the great utility of star clusters is that the most precise determination of the age of the Universe comes from old globular star clusters in our Galaxy.
The study of star clusters and large samples of individual stars in the Milky Way Galaxy are fields in which explosive growth is expected during the decade ahead. Enormous progress is anticipated on several fronts, as a wide variety of new telescopes, both large and small, are poised to begin operation. All-sky, near-IR surveys such as DENIS and 2MASS will provide magnitudes and colors for several hundred million Galactic stars. The SDSS will revolutionize studies of Galactic structure by measuring magnitudes for 4#4 50 million stars; low-resolution spectroscopy for several million of these will be obtained during the second, spectroscopic, portion of the survey. Meanwhile, 15 telescopes having apertures in the range 6 - 10m are set to begin operation before 2003 (see Tables 3 and 4) including seven in the south, a hemisphere in which only one telescope larger than 4m is currently in operation. Yet this hemisphere contains the best and brightest Galactic globular clusters, the Galactic bulge and several of the nearest Local Group galaxies, including the Sagittarius Dwarf Galaxy (which is currently being absorbed into our own Galaxy) and the Magellanic Clouds.
Important questions which may remain unanswered by 2010 include the following.
1. What is the number and distribution in mass of the oldest stars?Microlensing surveys such as MACHO and OGLE have provided evidence that 2#2 50% of the mass of the halo of our Galaxy is in the form of dark, 0.5 solar-mass objects (presumably white dwarfs). Since globular clusters are prototypical old objects and should contain large numbers of white dwarfs, it should be possible to detect the dynamical signature of such white dwarfs in these old clusters. At present, such studies are possible with relatively nearby globular clusters, but a thorough understanding of their contribution to the outer halo and bulge of our Galaxy is probably beyond the capabilities of planned 8m-class telescopes.
2. What is the age of the Galaxy from the cooling of white dwarfsAs white dwarfs age they cool, so that the temperature of the white dwarf can be viewed as a cosmic clock. The search for very cool (and hence very old) white dwarfs in the halo of our Galaxy is certain to be a high-priority goal during the decade ahead. In principal, this is a feasible, though challenging, project for a 4m-class telescope equipped with a wide-field, mosaic camera and a suitable set of narrow-band filters. Nevertheless, sample contamination by white dwarfs in the disk of our Galaxy is expected to be a concern and, ideally, the halo sample should not be drawn locally but should instead be selected in situ (at distances greater than 2#2 3 kpc above the Galactic plane). Identifying such a sample would require not only a very large telescope, but also the ability to perform accurate astrometry on very faint stars (5#5).
3. What are the element ratios among stars in the halo and the bulge of our Galaxy?The chemical composition of old, metal-poor halo and bulge stars provides information on the early chemical homogeneity and star formation history of the Milky Way Galaxy as well as constraining nucleosynthesis in the Big Bang. High-resolution spectroscopy from forthcoming 8m-class telescopes will provide much new information on the element ratios in bulge and halo stars (perhaps telling us if they shared a common origin), yet high-resolution, high-S/N spectroscopy of many of the most interesting targets (i.e., distant halo dwarfs and highly reddened bulge dwarfs) will be impossible even with the new large telescopes.
4. How dark are dwarf spheroidal galaxies?The large number of faint dwarf galaxies and their apparently high mass-to-light ratios indicate that they may be a major contributor to the mass budget of the Universe. At present, however, the evidence for dark matter in these faint galaxies is limited to the anomalously high central velocity dispersions of the Draco, Ursa Minor and Sagittarius dwarfs: the three nearest galaxies, and three dwarfs which are most susceptible to Galactic tidal effects. Several additional dwarfs will soon become available in the south (e.g. Phoenix, Tucanae) but even with 10m-class telescopes, measuring internal velocity dispersions for these extremely low-surface-brightness galaxies becomes very difficult at the distance of the Andromeda Galaxy, the nearest galaxy to us that is similar to the Milky Way. Thus, the question of the dark matter content of dwarf spheroidal galaxies is likely to remain an open one until a much larger sample of dwarfs, spanning a wide range in environment, have had their internal kinematics studied by larger telescopes.
After 2010, NGST will completely dominate research on star clusters and Galactic structure in the near IR. Nevertheless, NGST's modest aperture, small field-of-view, low spectral resolution and limited (and, possibly, non-existent) performance in the optical regime suggest that a large optical telescope (e.g. diameter 4#4 25m) having superb image quality over a large field-of-view and the capacity for high-resolution, multi-object spectroscopy would be an extremely competitive instrument for studies of Galactic structure, Local Group globular cluster systems, nearby dwarf galaxies and stellar structure.
The Search for Life on Extrasolar PlanetsIn a mini-survey of members of the committee in which they were asked what single observation they would like to do in the next few decades, the almost unanimous answer was the detection of a terrestrial planet around a nearby star and confirmation that life could be present on that planet. Such an observation would galvanize the general public, alter opinion about our purpose in the Universe and have a profound effect on political, philosophical and religious thought. There is of course no guarantee that such objects exist, but surely we must search for them.
There are several routes to a firm detection of a life-bearing terrestrial planet in orbit around a solar-type star. But, to be convincing, the planet must be imaged and identified from its gross photometric properties as a likely terrestrial planet. Eventually, a spectrum must be obtained of the planet and shown to contain evidence of life-supporting gases (H2O, CO2, O2, O3).
A first step toward such detection might be indirect. Radial velocity variations of the star in the planetary system could be evidence of the presence of planets. Such work is currently being actively pursued at a number of observatories including CFHT, Keck and ESO. A Jovian planet will cause velocity variations in a solar type star on the order of 10 - 100 m/s. Such measurements are well within current technology and periodic velocity variations have been found in about a dozen systems to date. By 2010 surveys for in excess of several thousand stars will have been completed and there will likely be a sample of more than 100 known systems containing giant planets. A terrestrial planet will produce orbital motion in the central star amounting to <1 - 10 m/s. The low end of this range is not currently achievable and in any case may be dominated by velocity noise from the surface of the star itself and be difficult to detect. The interpretation of all such observations suffers from uncertainty in the inclination of the planet's orbit which makes the mass determination of the planet ambiguous. This technique also naturally selects out those orbits which are the smallest. Further, real situations may be very complex with a number of Jovian and terrestrial planets co-existing in the same system causing difficulty in the interpretation of the velocity variations.
Other possible indirect approaches are astrometry, occultations and microlensing. In astrometric observations of a solar-type star a Jovian planet would produce an orbit of the star with an amplitude amounting to 10 - 100 3#3arcsec and terrestrials about one-tenth these values. Such measurements will have to await the interferometers current being planned for space. The presence of an orbiting planet could also be inferred from observed periodic occultations of its parent star. The brightness change is caused by the planet blocking out a small fraction of the star light and can yield the planetary period, distance from the star and the inclination of the orbit. This is one of the few methods capable of detecting a terrestrial planet in the habitable zone (the temperature of the planet being such that water is largely in the liquid state) of their parent star. These observations require photometric accuracy to about 1 part in 105 and is a good project for a small dedicated telescope in space. COROT, a 27cm telescope funded largely by the French Space Agency and scheduled for launch in May 2002 will attempt such observations but they are very challenging. As regards microlensing, the presence of a planet orbiting a star that is being gravitationally lensed produces a lensing signature that is unique and distinct. While a planet may be found with this technique, it is pure serendipity and will be difficult to follow up.
The direct approach to terrestrial planet detection will be the most convincing: obtain an image of the planet and a spectrum to illustrate its atmospheric composition. This project is extremely difficult, the main problem being that the planet is quite faint and close to a star that is much brighter. This brightness ratio is about a factor of 106 at 10 3#3 increasing to 109 at 1 3#3. The main limitation in detecting an object with such a large contrast ratio is scattered light caused by diffraction of the telescope aperture, the Earth's atmosphere, and the quality of the telescope optics. The Earth's atmosphere is the overwhelming problem here causing image spread, speckles and absorption bands in just the wrong places. The scattered light from each of these can be significantly reduced with, respectively, a coronagraph, adaptive optics and super smooth mirror optics. A planet in the habitable zone of a star 10 pc away from Earth appears separated from it by 0.1 arcsec. The diffraction limit of a 25m telescope at 1 3#3 is 0.01 arcsec.