We present a simple theoretical framework for massive galaxies at high redshift, where the main assembly and star formation occurred, and report on the first cosmological simulations that reveal clumpy disks consistent with our analysis. The evolution is governed by the interplay between smooth and clumpy cold streams, disk instability, and bulge formation. Intense, relatively smooth streams maintain an unstable dense gas-rich disk. Instability with high turbulence and giant clumps, each a few percent of the disk mass, is self-regulated by gravitational interactions within the disk. The clumps migrate into a bulge in < ∼ 10 dynamical times, or < ∼ 0.5 Gyr. The cosmological streams replenish the draining disk and prolong the clumpy phase to several Gigayears in a steady state, with comparable masses in disk, bulge and dark matter within the disk radius. The clumps form stars in dense subclumps following the overall accretion rate, ∼ 100 M ⊙ yr −1 , and each clump converts into stars in ∼ 0.5 Gyr. While the clumps coalesce dissipatively to a compact bulge, the star-forming disk is extended because the incoming streams keep the outer disk dense and susceptible to instability and because of angular momentum transport. Passive spheroid-dominated galaxies form when the streams are more clumpy: the external clumps merge into a massive bulge and stir up disk turbulence that stabilize the disk and suppress in situ clump and star formation. We predict a bimodality in galaxy type by z ∼ 3, involving giant-clump star-forming disks and spheroid-dominated galaxies of suppressed star formation. After z ∼ 1, the disks tend to be stabilized by the dominant stellar disks and bulges. Most of the high-z massive disks are likely to end up as today's early-type galaxies.

We have studied the properties of giant star forming clumps in five z~2 starforming disks with deep SINFONI AO spectroscopy at the ESO VLT 1 . The clumps reside in disk regions where the Toomre Q-parameter is below unity, consistent with their being bound and having formed from gravitational instability. Broad Hα/ [NII] line wings demonstrate that the clumps are launching sites of powerful outflows. The inferred outflow rates are comparable to or exceed the star formation rates, in one case by a factor of eight. Typical clumps may lose a fraction of their original gas by feedback in a few hundred million years, allowing them to migrate into the center. inferred gas phase oxygen abundance are broadly consistent with inside-out growing disks, and/or with inward migration of the clumps..

We analyze the first cosmological simulations that recover the fragmentation of high-redshift galactic discs driven by cold streams. The fragmentation is recovered owing to an AMR resolution better than 70 pc with cooling below 10^4 K. We study three typical star-forming galaxies in haloes of approx. 5 10^11 Msun at z=2.3, when they were not undergoing a major merger. The steady gas supply by cold streams leads to gravitationally unstable, turbulent discs, which fragment into giant clumps and transient features on a dynamical timescale. The disc clumps are not associated with dark-matter haloes. The clumpy discs are self-regulated by gravity in a marginaly unstable state. Clump migration and angular-momentum transfer on an orbital timescale help the growth of a central bulge with a mass comparable to the disc. The continuous gas input keeps the system of clumpy disc and bulge in a near "steady state", for several Gyr. The average star-formation rate, much of which occurs in the clumps, follows the gas accretion rate of approx. 45 Msun/yr. The simulated galaxies resemble in many ways the observed star-forming galaxies at high redshift. Their properties are consistent with the simple theoretical framework presented in Dekel, Sari & Ceverino (2009). In particular, a two-component analysis reveals that the simulated discs are indeed marginally unstable, and the time evolution confirms the robustness of the clumpy configuration in a cosmological steady state. By z=1 the simulated systems are stabilized by a dominant stellar spheroid, demonstrating the process of "morphological quenching" of star formation (Martig et al. 2009) . We demonstrate that the disc fragmentation is not a numerical artifact once the Jeans length is kept larger than 7 resolution elements, i.e. beyond the standard Truelove criterion.Comment: 20 pages, 12 figures, accepted in MNRAS

We use cosmological simulations to study a characteristic evolution pattern of high redshift galaxies. Early, stream-fed, highly perturbed, gas-rich discs undergo phases of dissipative contraction into compact, star-forming systems ("blue" nuggets) at z ∼ 4 − 2. The peak of gas compaction marks the onset of central gas depletion and inside-out quenching into compact ellipticals (red nuggets) by z ∼ 2. These are sometimes surrounded by gas rings or grow extended dry stellar envelopes. The compaction occurs at a roughly constant specific starformation rate (SFR), and the quenching occurs at a constant stellar surface density within the inner kpc (Σ 1 ). Massive galaxies quench earlier, faster, and at a higher Σ 1 than lower-mass galaxies, which compactify and attempt to quench more than once. This evolution pattern is consistent with the way galaxies populate the SFR-size-mass space, and with gradients and scatter across the main sequence. The compaction is triggered by an intense inflow episode, involving (mostly minor) mergers, counter-rotating streams or recycled gas, and is commonly associated with violent disc instability. The contraction is dissipative, with the inflow rate >SFR, and the maximum Σ 1 anti-correlated with the initial spin parameter . The central quenching is triggered by the high SFR and stellar/supernova feedback (maybe also AGN feedback) due to the high central gas density, while the central inflow weakens as the disc vanishes. Suppression of fresh gas supply by a hot halo allows the longterm maintenance of quenching once above a threshold halo mass, inducing the quenching downsizing.

We combine high-resolution Hubble Space Telescope/WFC3 images with multi-wavelength photometry to track the evolution of structure and activity of massive (M > 10 10 M ) galaxies at redshifts z = 1.4-3 in two fields of the Cosmic Assembly Near-infrared Deep Extragalactic Legacy Survey. We detect compact, star-forming galaxies (cSFGs) whose number densities, masses, sizes, and star formation rates (SFRs) qualify them as likely progenitors of compact, quiescent, massive galaxies (cQGs) at z = 1.5-3. At z 2, cSFGs present SFR = 100-200 M yr −1 , yet their specific star formation rates (sSFR ∼ 10 −9 yr −1 ) are typically half that of other massive SFGs at the same epoch, and host X-ray luminous active galactic nuclei (AGNs) 30 times (∼30%) more frequently. These properties suggest that cSFGs are formed by gas-rich processes (mergers or disk-instabilities) that induce a compact starburst and feed an AGN, which, in turn, quench the star formation on dynamical timescales (few 10 8 yr). The cSFGs are continuously being formed at z = 2-3 and fade to cQGs down to z ∼ 1.5. After this epoch, cSFGs are rare, thereby truncating the formation of new cQGs. Meanwhile, down to z = 1, existing cQGs continue to enlarge to match local QGs in size, while less-gas-rich mergers and other secular mechanisms shepherd (larger) SFGs as later arrivals to the red sequence. In summary, we propose two evolutionary tracks of QG formation: an early (z 2), formation path of rapidly quenched cSFGs fading into cQGs that later enlarge within the quiescent phase, and a late-arrival (z 2) path in which larger SFGs form extended QGs without passing through a compact state.