From a search of a ~ 2400 square degree region covered by both the SDSS and UKIDSS databases, we have attempted to identify galaxies at z ~ 0.5 that are consistent with their being essentially unmodified examples of the luminous passive compact galax
ies found at z ~ 2.5. After isolating good candidates via deeper imaging, we further refine the sample with Keck moderate-resolution spectroscopy and laser-guide-star adaptive-optics imaging. For 4 of the 5 galaxies that so far remain after passing through this sieve, we analyze plausible star-formation histories based on our spectra in order to identify galaxies that may have survived with little modification from the population formed at high redshift. We find 2 galaxies that are consistent with having formed > 95% of their mass at z > 5. We attempt to estimate masses both from our stellar population determinations and from velocity dispersions. Given the high frequency of small axial ratios, both in our small sample and among samples found at high redshifts, we tentatively suggest that some of the more extreme examples of passive compact galaxies may have prolate morphologies.
Although mergers and starbursts are often invoked in the discussion of QSO activity in the context of galaxy evolution, several studies have questioned their importance or even their presence in QSO host galaxies. Accordingly, we are conducting a stu
dy of z~0.2 QSO host galaxies previously classified as passively evolving elliptical galaxies. We present deep Keck LRIS spectroscopy of a sample of 15 hosts and model their stellar absorption spectra using stellar synthesis models. The high S/N of our spectra allow us to break various degeneracies that arise from different combinations of models, varying metallicities, and contamination from QSO light. We find that none of the host spectra can be modeled by purely old stellar populations and that the majority of the hosts (14/15) have a substantial contribution from intermediate-age populations with ages ranging from 0.7 to 2.4 Gyr. An average host spectrum is strikingly well fit by a combination of an old population and a 2.1 (+0.5, -0.7) Gyr population. The morphologies of the host galaxies suggest that these aging starbursts were induced during the early stages of the mergers that resulted in the elliptical-shaped galaxies that we observe. The current AGN activity likely corresponds to the late episodes of accretion predicted by numerical simulations, which occur near the end of the mergers, whereas earlier episodes may be more difficult to observe due to obscuration. Our off-axis observations prevent us from detecting any current star formation or young stellar populations that may be present in the central few kiloparsecs.
From a search of a portion of the sky covered by the SDSS and UKIDSS databases, we have located 2 galaxies at z~0.5 that have properties similar to those of the luminous passive compact galaxies found at z~2.5. From Keck moderate-resolution spectrosc
opy and laser-guided adaptive-optics imaging of these galaxies, we can begin to put together a more detailed picture of what their high-redshift counterparts might be like. Spectral-synthesis models that fit the u to K photometry also seem to give good fits to the spectral features. From these models, we estimate masses in the range of 3-4 10^11 M_sun for both galaxies. Under the assumption that these are spheroidal galaxies, our velocity dispersions give estimated masses about a factor of 3 smaller. However, our high-resolution imaging data indicate that these galaxies are not normal spheroids, and the interpretation of the kinematic data depends critically on the actual morphologies and the nature of the stellar orbits. While recent suggestions that the population of high-redshift compact galaxies is present locally as the inner regions of local massive elliptical galaxies are quite plausible, the peak mass surface densities of the two galaxies we discuss here appear to be up to a factor of 10 higher than those of the highest density local ellipticals, assuming that our photometric masses are roughly correct. It thus seems possible that some dynamical puffing-up of the high-redshift galaxies might still be required in this scenario.
We present Spitzer IRS spectra and MIPS photometry of 12 radio-loud QSOs with FR II morphologies at z ~ 0.3. Six of the sources are surrounded by luminous extended emission-line regions (EELRs), while the other six do not have such extended nebulae.
The two subsamples are indistinguishable in their mid-infrared spectra and overall infrared spectral energy distributions (SEDs). For both subsamples, the mid-infrared aromatic features are undetected in either individual sources or their stacked spectra, and the SEDs are consistent with pure quasar emission without significant star formation. The upper limits to the star formation rate are sufficiently low that starburst-driven superwinds can be ruled out as a mechanism for producing the EELRs, which are instead likely the result of the ejection of most of the gas from the system by blast waves accompanying the launching of the radio jets. The FR II quasars deviate systematically from the correlation between host galaxy star formation rate and black hole accretion rate apparently followed by radio-quiet QSOs, implying little or no bulge growth coeval with the current intensive black hole growth. We also present a new Spitzer estimate of the star formation rate for the starburst in the host galaxy of the compact steep-spectrum radio quasar 3C 48.
We present extensive ground-based spectroscopy and HST imaging of 3C79, an FR II radio galaxy associated with a luminous extended emission-line region (EELR). Surface brightness modeling of an emission-line-free HST R-band image reveals that the host
galaxy is a massive elliptical with a compact companion 0.8 away and 4 magnitudes fainter. The host galaxy spectrum is best described by an intermediate-age (1.3 Gyr) stellar population (4% by mass), superimposed on a 10 Gyr old population and a power law (alpha_{lambda} = -1.8); the stellar populations are consistent with super-solar metallicities, with the best fit given by the 2.5 Z_sun models. We derive a dynamical mass of 4E11 M_sun within the effective radius from the velocity dispersion. The EELR spectra clearly indicate that the EELR is photoionized by the hidden central engine. Photoionization modeling shows evidence that the gas metallicity in both the EELR and the nuclear narrow-line region is mildly sub-solar (0.3 - 0.7 Z_sun) -- significantly lower than the super-solar metallicities deduced from typical active galactic nuclei in the SDSS. The more luminous filaments in the EELR exhibit a velocity field consistent with a common disk rotation. Fainter clouds, however, show high approaching velocities that are uncoupled with this apparent disk rotation. The striking similarities between this EELR and the EELRs around steep-spectrum radio-loud quasars provide further evidence for the orientation-dependent unification schemes. The metal-poor gas is almost certainly not native to the massive host galaxy. We suggest that the close companion galaxy could be the tidally stripped bulge of a late-type galaxy that is merging with the host galaxy. The interstellar medium of such a galaxy is probably the source for the low-metallicity gas in 3C79.
We give an overview of our recent integral-field-unit spectroscopy of luminous extended emission-line regions (EELRs) around low-redshift quasars, including new observations of 5 fields. Previous work has shown that the most luminous EELRs are found
almost exclusively around steep-spectrum radio-loud quasars, with apparently disordered global velocity fields, and little, if any, morphological correlation with either the host-galaxy or the radio structure. Our new observations confirm and expand these results. The EELRs often show some clouds with velocities exceeding 500 km/s, ranging up to 1100 km/s, but the velocity dispersions, with few exceptions, are in the 30-100 km/s range. Emission-line ratios show that the EELRs are clearly photoionized by the quasars. Masses of the EELRs range up to >10^10 Msun. Essentially all of the EELRs show relatively low metallicities, and they are associated with quasars that, in contrast to most, show similarly low metallicities in their broad-line regions. The two objects in our sample that do not have classical double-lobed radio morphologies (3C48, with a compact-steep-spectrum source; Mrk1014, radio-quiet, but with a weak compact-steep-spectrum source) are the only ones that appear to have recent star formation. While some of the less-luminous EELRs may have other origins, the most likely explanation for the ones in our sample is that they are examples of gas swept out of the host galaxy by a large-solid-angle blast wave accompanying the production of the radio jets. The triggering of the quasar activity is almost certainly the result of the merger of a gas-rich galaxy with a massive, gas-poor galaxy hosting the supermassive black hole.
We have examined in detail the morphologies of seven z~1.5 passively evolving luminous red galaxies using high resolution HST NICMOS and ACS imaging data. Almost all of these galaxies appear to be relaxed systems, with smooth morphologies at both res
t-frame UV and visible wavelengths. Previous results from spectral synthesis modeling favor a single burst of star formation more than 1 Gyr before the observed epoch. The prevalence of old stellar populations, however, does not correlate exclusively with early-type morphologies as it does in the local universe; the light profiles for some of these galaxies appear to be dominated by massive exponential disks. This evidence for massive old disks, along with the apparent uniformity of stellar age across the disk, suggests formation by a mechanism better described as a form of monolithic collapse than as a hierarchical merger. These galaxies could not have undergone a single major merging event since the bulk of their stars were formed, more than 1 Gyr earlier. There is at least one case, however, that appears to be undergoing a dry merger, which may be an example of the process that converts these unusual galaxies into the familiar spheroids that dominate galaxies comprising old stellar populations at the present epoch.
We present the results of NICMOS imaging of two massive galaxies photometrically selected to have old stellar populations at z ~ 2.5. Both galaxies are dominated by apparent disks of old stars, although one of them also has a small bulge comprising a
bout 1/3 of the light at rest-frame 4800 A. The presence of massive disks of old stars at high redshift means that at least some massive galaxies in the early universe have formed directly from the dissipative collapse of a large mass of gas. The stars formed in disks like these may have made significant contributions to the stellar populations of massive spheroids at the present epoch.
We present a correlation between the presence of luminous extended emission-line regions (EELRs) and the metallicity of the broad-line regions (BLRs) of low-redshift quasars. The result is based on ground-based [O III] 5007 narrow-band imaging and Hu
bble Space Telescope UV spectra of 12 quasars at 0.20 < z < 0.45. Quasars showing luminous EELRs have low-metallicity BLRs (Z < 0.6 Z_Solar), while the remaining quasars show typical metal-rich gas (Z > Z_Solar). Previous studies have shown that EELRs themselves also have low metallicities (Z < 0.5 Z_Solar). The correlation between the occurrence of EELRs and the metallicity of the BLRs, strengthened by the sub-Solar metallicity in both regions, indicates a common external origin for the gas, almost certainly from the merger of a gas-rich galaxy. Our results provide the first direct observational evidence that the gas from a merger can indeed be driven down to the immediate vicinity (< 1 pc) of the central black hole.
We present Gemini integral field spectroscopy and Keck II longslit spectroscopy of the extended emission-line region (EELR) around the quasar 4C 37.43. The velocity structure of the ionized gas is complex and cannot be explained globally by a simple
dynamical model. The spectra from the clouds are inconsistent with shock or ``shock + precursor ionization models, but they are consistent with photoionization by the quasar nucleus. The best-fit photoionization model requires a low-metallicity (12+log(O/H) < 8.7) two-phase medium, consisting of a matter-bounded diffuse component with a unity filling-factor (N ~ 1 cc, T ~ 15000 K), in which are embedded small, dense clouds (N ~ 400 cc, T ~ 10^4 K). The high-density clouds are transient and can be re-generated through compressing the diffuse medium by low-speed shocks (V_S lesssim 100 kms). Our photoionization model gives a total mass for the ionized gas of about 3x10^{10} M_sun, and the total kinetic energy implied by this mass and the observed velocity field is ~2x10^{58} ergs. The fact that luminous EELRs are confined to steep-spectrum radio-loud quasars, yet show no morphological correspondence to the radio jets, suggests that the driving force producing the 4C 37.43 EELR was a roughly spherical blast wave initiated by the production of the jet. That such a mechanism seems capable of ejecting a mass comparable to that of the total interstellar medium of the Milky Way suggests that ``quasar-mode feedback may indeed be an efficient means of regulating star formation in the early universe.
