Science Nugget—Understanding the Dynamics of Galaxy Formation in the Early Universe
  David Law, California Institute of Technology

In the local universe, we see a familiar population of galaxies: dwarfs, spirals and ellipticals whose structures have shaped our understanding of morphological diversity. In contrast to our knowledge of the present-day structure and distribution of these galaxies however, our knowledge of their formation mechanisms is rudimentary. Observations of galaxies at higher redshifts (i.e. at earlier times in the history of the universe) indicate that these galaxies began as blobby, highly irregular clumps of star formation (Fig. 1) quite dissimilar to their present-day descendants. Understanding the formation of these early, rapidly star-forming galaxies and their subsequent evolution into the local galaxy population is one of the most significant challenges to the current generation of astronomers in the effort to determine how, when, and why the galactic building blocks of the visible universe formed.

FIGURE 1: Representative HST-ACS morphologies of z ~ 2 galaxies in the GOODS-N field are shown sorted into quartiles according to their Gini (G) and multiplicity (W) parameters. The vertical axis corresponds to increasing emission nucleation, the horizontal axis to increasing number of components.

According to one popular theoretical description of early galaxy formation, the process may have occurred somewhat as follows. Early in the history of the universe, small regions of space slightly denser than their neighbors began to decouple from the cosmic expansion, collapsing and accreting into an intricate web of dark matter. Into the deepest of these gravitational valleys fell baryonic hydrogen gas, whose collisional nature permitted it to radiate away its dynamical energy and collapse to a rotating disk supported by residual angular momentum. In isolation, these gas disks could slowly form stars via local gravitational instabilities (perhaps producing spiral-type galaxies), and later collide with other such galaxies to produce more massive, dispersion-supported elliptical galaxies. As compelling as such a tidy picture may be, numerous recent observations of galaxies at the peak epoch of star formation (redshift z ~ 2) have begun to challenge some of the underlying assumptions of these theoretical models.

Using advanced integral-field spectroscopy and adaptive optics (AO) technologies, it has recently become possible to probe the spatially resolved kinematic structure of star forming galaxies in the early universe through observation of nebular line emission redshifted into the near-IR (1-2.5 microns). The 3D data "cubes" (which contain full spectral information for each spatial pixel across the target galaxies) provided by such instruments have revealed that these early galaxies are not as dynamically cold as originally expected, and in many cases are dominated more by kinematic dispersion than by coherent rotation in flattened gaseous disks. Indeed, it may be that the bursts of star formation observed at this epoch may intrinsically occur as a result of strong instabilities and dynamical friction which completely destroy the original disk and result in large-scale gas flows and a turbulent nuclear starburst. Alternatively, the patchy emission morphologies may represent star formation occurring in gas as it is accreted directly from filamentary 'highways' rather than gradual spherical accretion onto a well-defined cold disk. It is unknown however whether these kinematic characteristics are typical of early star forming galaxies, or are a feature of only the rarest and brightest galaxies to which current technology is sensitive. Indeed, by pushing current technology to the limits of its sensitivity, it is currently only possible to detect the very brightest regions of a single bright galaxy for every few hours of observing on a 10m class telescope.

In contrast TMT, with a factor of ~ 10 greater collecting area, optimized AO system, and advanced multiplexing technology will be able to surpass in a single night the tentative studies that have taken current telescopes years to accomplish. Since the power of an AO-equipped telescope to detect emission from individual star forming regions grows as the fourth power of the primary mirror diameter (by both increasing the collecting area and concentrating the diffraction-limited PSF), simulations indicate that a single hour of observing with TMT will be sufficient to trace star formation throughout galaxies whose brightest knots are only barely detected with current technology (Fig. 2). The unparalleled power of TMT will therefore permit the efficient study of large and statistically representative populations of early galaxies, representing a ground-breaking advance in our understanding of the nature and kinematics of galaxy formation in the universe.

FIGURE 2: Simulated maps of the S/N ratio expected for detection of nebular line emission from IFU observations of a typical z ~ 2 star forming galaxy observed for 1 hour with Keck (top) versus TMT (bottom).