We present the cross-correlation between the far-infrared background fluctuations as measured with the Herschel Space Observatory at 250, 350, and 500 {mu}m and the near-infrared background fluctuations with Spitzer Space Telescope at 3.6 {mu}m. The cross-correlation between far and near-IR background anisotropies are detected such that the correlation coefficient at a few to ten arcminute angular scales decreases from 0.3 to 0.1 when the far-IR wavelength increases from 250 {mu}m to 500 {mu}m. We model the cross-correlation using a halo model with three components: (a) far-IR bright or dusty star-forming galaxies below the masking depth in Herschel maps, (b) near-IR faint galaxies below the masking depth at 3.6 {mu}m, and (c) intra-halo light, or diffuse stars in dark matter halos, that likely dominates fluctuations at 3.6 {mu}m. The model is able to reasonably reproduce the auto correlations at each of the far-IR wavelengths and at 3.6 {mu}m and their corresponding cross-correlations. While the far and near-IR auto-correlations are dominated by faint dusty, star-forming galaxies and intra-halo light, respectively, we find that roughly half of the cross-correlation between near and far-IR backgrounds is due to the same galaxies that remain unmasked at 3.6 {mu}m. The remaining signal in the cross-correlation is due to intra-halo light present in the same dark matter halos as those hosting the same faint and unmasked galaxies. In this model, the decrease in the cross-correlation signal from 250 {mu}m to 500 {mu}m comes from the fact that the galaxies that are primarily contributing to 500 {mu}m fluctuations peak at a higher redshift than those at 250 {mu}m.
We discuss the restframe UV emission from the starbursting galaxy HFLS3 at z=6.34, discovered in Herschel/SPIRE data due to its red color in the submm wavelengths from 250-500 um. The apparent inst. SFR of HFLS3 inferred from the total FIR luminosity measured with over 15 photometric data points between 100 to 1000 um is 2900 Msun/yr. Keck/NIRC2 Ks band adaptive optics imaging data showed two potential NIR counterparts near HFLS3. Previously, the northern galaxy was taken to be in the foreground at z=2.1 while the southern galaxy was assumed to HFLS3s NIR counterpart. New HST/WFC3 and ACS imaging data show both optically bright galaxies are in the foreground at z<6. A new lensing model based on HST data and mm-wave continuum emission yields a magnification of 2.2+/-0.3. The lack of multiple imaging constrains the lensing magnification to be lower than either 2.7 or 3.5 at the 95% confidence level for the two scenarios, which attribute one or two components to HFLS3 in the source plane. Correcting for gravitational lensing, the inst. SFR is 1320 Msun/yr with the 95% confidence lower limit around 830 Msun/yr. Using models for the restframe UV to FIR SED, the ave. SFR over the last 100 Myr is around 660 Msun/yr. The dust and stellar masses of HFLS3 from the same SED models are 3x10^8 Msun and ~5x10^10 Msun, respectively, with large systematic uncertainties on assumptions related to the SED model. With HST/WFC3 images we also find diffuse NIR emission about 0.5 (~3 kpc) SW of HFLS3 that remains undetected in the ACS data. The emission has a photometric redshift consistent with either z~6 or a dusty galaxy template at z~2. If at the same redshift as HFLS3 the detected diffuse emission could be part of the complex merger system that could be triggering the starburst. Alternatively, it could be part of the foreground structure at z~2.1 that is responsible for lensing of HFLS3.
Stellar archeology shows that massive elliptical galaxies today formed rapidly about ten billion years ago with star formation rates above several hundreds solar masses per year (M_sun/yr). Their progenitors are likely the sub-millimeter-bright galaxies (SMGs) at redshifts (z) greater than 2. While SMGs mean molecular gas mass of 5x10^10 M_sun can explain the formation of typical elliptical galaxies, it is inadequate to form ellipticals that already have stellar masses above 2x10^11 M_sun at z ~ 2. Here we report multi-wavelength high-resolution observations of a rare merger of two massive SMGs at z = 2.3. The system is currently forming stars at a tremendous rate of 2,000 M_sun/yr. With a star formation efficiency an order-of-magnitude greater than that of normal galaxies, it will quench the star formation by exhausting the gas reservoir in only ~200 million years. At a projected separation of 19 kiloparsecs, the two massive starbursts are about to merge and form a passive elliptical galaxy with a stellar mass of ~4x10^11 M_sun. Our observations show that gas-rich major galaxy mergers, concurrent with intense star formation, can form the most massive elliptical galaxies by z ~ 1.5.
Unresolved near-infrared background anisotropies are expected to have contributions from the earliest galaxies during reionization and faint, dwarf galaxies at intermediate redshifts. Previous measurements were unable to conclusively pinpoint the dominant origin because they did not sample spatial scales that were sufficiently large to distinguish between these two possibilities. Here we report a measurement of the anisotropy power spectrum from sub-arcminute to one degree angular scales and find the clustering amplitude to be larger than the model predictions involving the two existing explanations. As the shot-noise level of the power spectrum is consistent with that expected from faint galaxies, a new source population on the sky is not necessary to explain the observations. A physical mechanism that increases the clustering amplitude, however, is needed. Motivated by recent results related to the extended stellar light profile in dark matter halos, we consider the possibility that the fluctuations originate from diffuse intrahalo stars of all galaxies. We find that the measured power spectrum can be explained by an intrahalo light fraction of 0.07 to 0.2 % relative to the total luminosity in dark matter halos of masses log(M/M_Sun) ~ 9 to 12 at redshifts of ~ 1 to 4.
The cosmic far-infrared background (CFIRB) is expected to be generated by faint, dusty star-forming galaxies during the peak epoch of galaxy formation. The anisotropy power spectrum of the CFIRB captures the spatial distribution of these galaxies in dark matter halos and the spatial distribution of dark matter halos in the large-scale structure. Existing halo models of CFIRB anisotropy power spectrum are either incomplete or lead to halo model parameters that are inconsistent with the galaxy distribution selected at other wavelengths. Here we present a conditional luminosity function approach to describe the far-IR bright galaxies. We model the 250 um luminosity function and its evolution with redshift and model-fit the CFIRB power spectrum at 250 um measured by the Herschel Space Observatory. We introduce a redshift dependent duty-cycle parameter so that we are able to estimate the typical duration of the dusty star formation process in the dark matter halos as a function of redshifts. We find the duty cycle of galaxies contributing to the far-IR background is 0.3 to 0.5 with a dusty star-formation phase lasting for sim0.3-1.6 Gyrs. This result confirms the general expectation that the far-IR background is dominated by star-forming galaxies in an extended phases, not bright starbursts that are driven by galaxy mergers and last sim10-100 Myrs. The halo occupation number for satellite galaxies has a power-law slope that is close to unity over 0<z<4. We find that the minimum halo mass for dusty, star-forming galaxies with L_250>10^{10} L_Sun is 2times10^{11}M_Sun and 3times 10^{10}M_Sun at z=1 and 2, respectively. Integrating over the galaxy population with L_250>10^{9} L_Sun, we find that the cosmic density of dust residing in the dusty, star-forming galaxies responsible for the background anisotropies Omega_{dust}sim3times10^{-6} to 2times10^{-5}.
We present a list of 13 candidate gravitationally lensed submillimeter galaxies (SMGs) from 95 square degrees of the Herschel Multi-tiered Extragalactic Survey, a surface density of 0.14pm0.04deg^{-2}. The selected sources have 500um flux densities (S_500) greater than 100mJy. Gravitational lensing is confirmed by follow-up observations in 9 of the 13 systems (70%), and the lensing status of the four remaining sources is undetermined. We also present a supplementary sample of 29 (0.31pm0.06deg^{-2}) gravitationally lensed SMG candidates with S_500=80--100mJy, which are expected to contain a higher fraction of interlopers than the primary candidates. The number counts of the candidate lensed galaxies are consistent with a simple statistical model of the lensing rate, which uses a foreground matter distribution, the intrinsic SMG number counts, and an assumed SMG redshift distribution. The model predicts that 32--74% of our S_500>100mJy candidates are strongly gravitationally lensed (mu>2), with the brightest sources being the most robust; this is consistent with the observational data. Our statistical model also predicts that, on average, lensed galaxies with S_500=100mJy are magnified by factors of ~9, with apparently brighter galaxies having progressively higher average magnification, due to the shape of the intrinsic number counts. 65% of the sources are expected to have intrinsic 500micron flux densities less than 30mJy. Thus, samples of strongly gravitationally lensed SMGs, such as those presented here, probe below the nominal Herschel detection limit at 500 micron. They are good targets for the detailed study of the physical conditions in distant dusty, star-forming galaxies, due to the lensing magnification, which can lead to spatial resolutions of ~0.01 in the source plane.
We use spitzer-IRAC data to identify near-infrared counterparts to submillimeter galaxies detected with Herschel-SPIRE at 250um in the Herschel Astrophysical Terahertz Large Area Survey (H-ATLAS). Using a likelihood ratio analysis we identify 146 reliable IRAC counterparts to 123 SPIRE sources out of the 159. We find that, compared to the field population, the SPIRE counterparts occupy a distinct region of 3.6 and 4.5um color-magnitude space, and we use this property to identify a further 23 counterparts to 13 SPIRE sources. The IRAC identification rate of 86% is significantly higher than those that have been demonstrated with wide-field ground-based optical and near-IR imaging of Herschel fields. We estimate a false identification rate of 3.6%, corresponding to 4 to 5 sources. Among the 73 counterparts that are undetected in SDSS, 57 have both 3.6 and 4.5um coverage. Of these 43 have [3.6] - [4.5]> 0 indicating that they are likely to be at z > 1.4. Thus, ~ 40% of identified SPIRE galaxies are likely to be high redshift (z > 1.4) sources. We discuss the statistical properties of the IRAC-identified SPIRE galaxy sample including far-IR luminosities, dust temperatures, star-formation rates, and stellar masses. The majority of our detected galaxies have 10^10 to 10^11 L_sun total IR luminosities and are not intense starbursting galaxies as those found at z ~ 2, but they have a factor of 2 to 3 above average specific star-formation rates compared to near-IR selected galaxy samples.
While the arcminute-scale Cosmic Microwave Background (CMB) anisotropies are due to secondary effects, point sources dominate the total anisotropy power spectrum. At high frequencies the point sources are primarily in the form of dusty, star-forming galaxies. Both Herschel and Planck have recently measured the anisotropy power spectrum of cosmic infrared background (CIB) generated by dusty, star-forming galaxies from degree to sub-arcminute angular scales, including the non-linear clustering of these galaxies at multipoles of 3000 to 6000 relevant to CMB secondary anisotropy studies. We scale the CIB angular power spectra to CMB frequencies and interpret the combined WMAP-7 year and arcminute-scale Atacama Cosmology Telescope (ACT) and South Pole Telescope (SPT) CMB power spectra measurements to constrain the Sunyaev-Zeldovich (SZ) effects. Allowing the CIB clustering amplitude to vary, we constrain the amplitudes of thermal and kinetic SZ power spectra at 150 GHz.
The extragalactic background light at far-infrared wavelengths originates from optically-faint, dusty, star-forming galaxies in the universe with star-formation rates at the level of a few hundred solar masses per year. Due to the relatively poor spatial resolution of far-infrared telescopes, the faint sub-millimetre galaxies are challenging to study individually. Instead, their average properties can be studied using statistics such as the angular power spectrum of the background intensity variations. A previous attempt at measuring this power spectrum resulted in the suggestion that the clustering amplitude is below the level computed with a simple ansatz based on a halo model. Here we report a clear detection of the excess clustering over the linear prediction at arcminute angular scales in the power spectrum of brightness fluctuations at 250, 350, and 500 microns. From this excess, we find that sub-millimetre galaxies are located in dark matter halos with a minimum mass of log[M_min/M_sun ]= 11.5^+0.7_-0.2 at 350 microns. This minimum dark matter halo mass corresponds to the most efficient mass scale for star formation in the universe, and is lower than that predicted by semi-analytical models for galaxy formation.
The temperature fluctuations and polarization of the Cosmic Microwave Background (CMB) are now a well-known probe of the Universe at an infant age of 400,000 years. During the transit to us from the surface of last scattering, the CMB photons are expected to undergo modifications induced by the intervening large-scale structure. Among the expected secondary effects is the weak gravitational lensing of the CMB by the foreground dark matter distribution. We derive a quadratic estimator that uses the non-Gaussianities generated by the lensing effect at the four-point function level to extract the power spectrum of lensing potential fluctuations integrated out to z ~ 1100 with peak contributions from potential fluctuations at z of 2 to 3. Using WMAP 7-year temperature maps, we report the first direct constraints of this lensing potential power spectrum and find that it has an amplitude of A_L = 0.96 +/- 0.60, 1.06 +/- 0.69 and 0.97 +/- 0.47 using the W, V and W+V bands, respectively.
