Current theory predicts that structure develops quite differently under different cosmological models3 (see Figure 2). Because these are best differentiated at high redshift, the relative merits of different models can be determined from observational studies at z > 1. Although existing (e.g., CfA) and ongoing (e.g., Sloan) galaxy surveys will map out the structure in galaxies at low redshifts (z < 0.5), the challenge of the coming decades will be to carry out definitive tests of theories of structure formation by comparing available theoretical predictions with future observations of the CMBR fluctuation spectrum and observations of structure evolution in the redshift range 1 < z < 3.5. The redshift range 1 < z < 3.5 encompasses roughly half the star formation history of the universe and, in this redshift range, structures seen at z < 1 will be in their initial stages of assembly. Indeed, current work over small angular fields already indicates the existence of significant structure at z > 3 (e.g., Steidel et al. 1998; see Figure 3)4. Although space missions such as MAP (Microwave Anisotropy Probe) will measure the CMBR fluctuation spectrum, extensive future ground-based programs will be needed to measure the evolution of structure as traced by the galaxy and intergalactic gas distributions.
Although galaxies are the traditional tracers of large scale structure, additional emphasis has recently been given to the intergalactic gas distribution, as typically observed in absorption, because the gas that gives rise to the Ly forest is expected to be a nearly direct (i.e., nearly unbiased) tracer of the total matter distribution (see Figure 4). Moreover, the bulk of the baryonic matter may be in a hot phase (e.g., Davé et al.; 106 K) 5 and difficult to detect except in absorption against background sources. Measuring the structure traced by both galaxies and intergalactic gas would be very useful, because it would allow us to construct a high dynamic range, three-dimensional tomographic map of the universe that can be used to test theories of structure formation. In this case, the high-density sampling in the angular dimensions as traced by galaxies would be complemented by high-density sampling in the redshift dimension, as traced by the intergalactic gas.
With this set of observations, we will be able to test not only theories of structure formation, but also fundamental assumptions about how the underlying mass distribution can be measured. For example, we will be able to make a detailed test of the current belief that the Ly forest traces mass fluctuations; we will also be able to measure the degree to which galaxies are biased tracers of the underlying mass distribution. The data obtained will also be useful beyond these fundamental tests. For example, the same data can be used to measure the detailed spatial distribution of metals in the IGM (Intergalactic Medium) compared to galaxies, to help understand the history of IGM enrichment. With these data, we can explore, for example, whether known populations of galaxies can account for the measured enrichment of the IGM, or whether new sources of enrichment are required. More generally, the same data will also allow the study of how galaxies form and evolve in the context of large scale structure .
In order to measure the structure traced by galaxies and intergalactic gas, we will need to study volumes large enough to measure accurate clustering statistics and to characterize the largest structures. These large volumes (~100 Mpc on a side; or 5 x 5 degrees for M = 0.3 and = 0.7) are needed even to measure the clustering of galaxies on fairly small scales, with accuracy required for precision cosmology (see Figure 5).
We will also need to go faint in order to study structure at high-z and to sample densely. Dense sampling is needed to construct the kind of high dynamic range structure map that will elucidate the complex structures (e.g., filaments) that are predicted, accurately measure the magnitude of large overdensities or underdensities, and probe structure on small angular scales (e.g., Figure 6).
Because these measurements require a significant amount of spectroscopy, completing the measurements in a reasonable amount of time requires a high throughput GSMT that has a wide field of view, a large telescope aperture, and the capability of highly multi-plexed spectroscopy, primarily at optical wavelengths.
Probing Structure in Galaxies: To effectively probe the galaxy distribution from z = 0.5 to the epoch of formation of the first galaxies (zf 10) over angular scales 100 Mpc, we will need to obtain spectra of hundreds of thousands of faint galaxies spread over a range of mass and over large areas (> 5º x 5º). Spectroscopy at R = 1000-5000 is required in order to measure redshifts and to characterize the tracer population. (The typical uncertainties in photometric redshifts are much too large to provide a detailed description of the three-dimensional galaxy distribution.) Faint galaxies must be studied in order to probe both the high redshift population and to obtain the dense sampling and the consequent high dynamic range that will reveal the complex structures that are predicted.
With the high sensitivity of a 30-m GSMT, we will be able to obtain moderate resolution spectroscopy (R = 1000-5000) of large numbers of galaxies (> 50,000/sq. deg.) to faint magnitudes (R ~ 26.5). We will thereby be able to probe, at high redshift, the equivalent of L* densities in the present-day universe. By probing comparable number densities at each epoch, we will be able to make robust evolutionary connections between present-day galaxies and their progenitors. Given a large FOV (field of view) (~ 20'), we will also be able to survey large areas (several degrees on a side), covering volumes large enough to provide an accurate measure of clustering statistics and to characterize the largest structures (several 106 Mpc3).
Probing Structure in the IGM: To effectively probe the three-dimensional structure of the intergalactic gas distribution, we will require dense sampling of the gas distribution on 100 Mpc size scales and over a large, well-sampled redshift range (i.e., 106 absorbers over a 5º x 5º degree field per z = 1). The object density and the relevant redshift ranges to be probed require the use of faint, distant objects (~ 25 AB mag) as background beacons. With the high sensitivity of the GSMT, it will be possible to probe Mpc scales in three dimensions. That is, because the surface density of background beacons (quasars and galaxies) depends very sensitively on apparent magnitude, we will have access to an unprecedentedly high surface density (> 5000/sq. deg.) of faint (R ~ 24) background beacons, and consequently be able to produce a map of the IGM on fine angular scales. The high sensitivity of the GSMT will also enable spectroscopy at a high enough resolution (R = 5,000-20,000) to probe Mpc scales along the line of sight.
For the estimates given below, the seeing is assumed to be 0.5", and the total efficiency of the spectrograph and telescope system is assumed to be 0.2". The sky background is assumed to be V = 22 from the ground and V = 24 in space. For JWST (James Webb Space Telescope), the s/n given is for the diffraction-limited point source case. This is overly optimistic, because galaxies will be resolved.
As described above, the combination of sensitivity and FOV is a powerful combination for studies of large scale structure. To illustrate the scientific potential of a GSMT equipped with a large (20') FOV, we can compare the total observing time it would take to carry out a seeing-limited spectroscopic survey of large scale structure probed by galaxies and gas using the GSMT, in comparison with other observing facilities. As described in the previous section, we need to probe structure on ~100 Mpc scales, which implies survey areas ~25 sq. deg.
The characteristic target, spectral resolution, and s/n for the faint galaxy portion of the survey is ~27.5 AB magnitude, R = 1000, and s/n = 5 per resolution element. To reach s/n = 5 per resolution element at R = 1000 at V on a 27.5 AB mag object, we would require integration times of approximately 1 hr, 14 hrs, and 1 hr with the GSMT (seeing-limited), 8m Gemini (seeing-limited), and 6.5m JWST (diffraction-limited), respectively. In order to survey a total area of 25 sq. deg. with a sampling density of 100,000 galaxies per sq. deg., we would need to obtain spectra for a total of 2.5 million objects. If we can obtain spectra for 1000 targets at a time (cf. MOMFOS (multi-object multi-fiber optical spectrograph)), we would require 2500 hrs (or 312 nights) of observing time to finish the survey. In comparison, a seeing-limited 8m ground-based telescope with the same FOV and instrumentation (currently nonexistent) would require ~12 yrs of observing time to complete the same project. JWST with a 5' FOV and the ability to measure spectra for 100 objects at a time (and assuming a 20 hr observing "day") would require 3.4 yrs of observing time to complete the project.
For the IGM portion of the survey, the corresponding parameters are ~25 AB magnitude, R = 5000; and s/n = 20 per resolution element. To reach s/n = 20 per resolution element at R = 5000 at V on a 25 AB mag object, we would require integration times of approximately 4 hrs, 55 hrs, and 20 hrs with the GSMT (seeing-limited), 8m Gemini (seeing-limited), and 6.5m JWST (diffraction-limited), respectively. In order to survey a total area of 25 sq. deg. with a sampling density of 10,000 galaxies per sq. deg., we would need to obtain spectra for a total of 250,000 objects. If we can obtain spectra for 1000 targets at a time (cf. MOMFOS), we would require 1000 hrs (or 125 nights) of observing time to finish the survey. In comparison, a seeing-limited 8m ground-based telescope with the same FOV and instrumentation (currently nonexistent) would require 5 yrs of observing time to complete the same project. JWST with a 5' FOV and the ability to measure spectra for 100 objects at a time (and assuming a 20 hr observing "day") would require 7 yrs to complete the project.
Thus, the total observing times (not including weather delays, instrument and telescope failures, etc.) required for the large scale structure project are 1.2 yrs with a 30m GSMT, 17 yrs with an 8m Gemini equipped with similar instrumentation, and 10.4 yrs with JWST (i.e., twice the nominal 5 yr mission lifetime). Note that the comparison with JWST is actually moot, because JWST is not currently envisioned to have optical spectroscopic capability.
All aspects of this program require multi-object spectroscopy over very large fields of view(>10'). To meet these requirements, we have explored the concept of a multi-fiber prime focus spectrograph.
- Bennett, C. et al. "Cosmic temperature fluctuations from two years of COBE differential microwave radiometers observations." ApJ, 436, 423, 1994.
- Eisenhardt, P.R.M., et al. 2002, in preparation
- White, S.D.M. The Early universe with the VLT. [ed. J. Bergeron Berlin]: Springer, p. 219 (1997).
- Steidel, C.C., Adelberger, K.L., Dickinson, M., Giavalisco, M., Pettini, M., & Kellogg, M. "A Large Structure of Galaxies at Redshift Z approximately 3 and Its Cosmological Implications." ApJ, 492, 428, 1998.
- Davé, R., Cen, R., Ostriker, J.P., Bryan, G.L., Hernquist, L., Katz, N., Weinberg, D.H., Norman, M.L., & O'Shea, B. "Baryons in the Warm-Hot Intergalactic Medium." ApJ, 552, 473, 2001.
- Katz, N., Weinberg, D.H., Hernquist, L., & Miralda-Escude, J. "Damped Lyman-Alpha and Lyman Limit Absorbers in the Cold Dark Matter Model." ApJ, 457, L57, 1996.