While the dark energy program drives the design of the high-latitude imaging and spectroscopic surveys, the scientific motivation for them is much broader. These surveys will provide an extraordinary resource for a wide range of investigations.
A candidate for a galaxy at z=9.6, magnified by a factor of ~15 by the foreground cluster MACS J1149+2223 (z = 0.54). The object was found in an HST survey using the WFC3IR camera (Zheng et al. 2012). This young object is seen when the universe is only about 500 million years old.
Dark matter and cluster gas distribution around the merging cluster Abell 2744. Such maps require Wide Field imaging. To map the entire area above with HST WFC3 or ACS would require 100 — 200 separate pointings, due to their small field of view. The same field is mapped with a single Roman Space Telescope pointing (red rectangle).
The First Billion Years of Cosmic History
Deep, multi-band, near-IR imaging over 2000 deg2 will be a revolutionary data set for understanding the first generations of galaxies and quasars. Extrapolations of high-redshift luminosity functions are uncertain, but they suggest that the high latitude imaging survey will detect ~ 30,000 z > 8 galaxies above the 10σ, mag ≈ 26 limit, and ~ 100 at z > 10 (rising to ~ 1000 at the 5σ limit). Some of these galaxies will be highly magnified by the gravitational lensing of foreground clusters, making them ideal targets for follow-up observations with JWST. The synergy of a Wide Field telescope that can discover the most luminous or highly magnified systems and a large aperture telescope that can characterize them in detail is essential for understanding the earliest galaxies.
Mapping Dark Matter on Intermediate and Large Scales
The high latitude imaging survey will produce weak lensing catalogs with surface densities of ~75 galaxies arcmin-2 in co-added J and H bands. These can be used to map the distributions of dark matter in clusters and superclusters with much higher fidelity than achievable from the ground (or from Euclid, which has less than half this source density).
Weak+strong lensing measurements of cluster mass profiles and substructure, compared to predictions from numerical simulations, will yield new insights into the mechanisms of structure formation and the interactions between baryons and dark matter. Merging/colliding clusters, such as the famed Bullet Cluster, are especially interesting systems, allowing novel tests of the properties of dark matter and the impact of cluster assembly on galaxy-scale halos and intracluster gas. Deeper observations in the General Observer program could achieve surface densities of 200 — 300 arcmin-2 for detailed study of interesting systems, with a survey speed fast enough to yield statistically informative samples.
Kinematics of Stellar Streams in Our Local Group of Galaxies
Using the high density of slightly resolved, medium-brightness galaxies as references, the Roman Space Telescope will measure accurate proper motions to a limit several magnitudes fainter than the GAIA mission. For G stars at V ≈ 20, we estimate accuracy of about 125 μas/year from the 2-year baseline of the high latitude imaging survey, which could be improved to 50 μas/year in GO programs that extend the baseline to 5 years. This depth probes the main sequence turnoff to distances of ~ 10 kpc and red giants throughout the Galactic halo. Tidal streams from disrupted dwarfs or star clusters have velocity dispersions of ~ 10 km s-1, so with proper motions of this accuracy they will stand out at high contrast from the background halo stars, which have a dispersion of ~ 150 km s-1. Measurements of the kinematics and structure of cold tidal streams can determine the structure of the Milky Way's gravitational potential and detect or rule out the small scale perturbations from dark matter subhalos that are a critical prediction of the cold dark matter scenario.
Discovering the Most Extreme Star Forming Galaxies and Quasars
The high latitude spectroscopic survey will enable the largest census yet performed of powerful emission-line galaxies and quasars up to (and possibly beyond) lookback times of 90% of the current age of the universe. Samples of Hα, OIII, and OII emitters will measure the contribution of luminous galaxies to the global star formation rate at 1 < z < 4, when most stars in the universe formed. At redshifts 8 < z < 15 (roughly 200 — 600 million years after the Big Bang), Lyman-alpha emitting galaxies will be detectable if they have attenuated star formation rates of at least 100 — 200 solar masses per year. The survey will detect ~ 2600 z > 7 quasars, with an estimated 20% of them at z > 8, probing the assembly of billion solar mass black holes in the first Gyr of cosmic history and finding the rare, bright backlights that JWST and ground-based telescopes can use to trace the earliest enrichment of the intergalactic medium with heavy elements (e.g., via CIV absorption).
