We present a detailed study of the infrared spectral energy distribution of the high-redshift radio galaxy MRC 1138-26 at z = 2.156, also known as the Spiderweb Galaxy. By combining photometry from Spitzer, Herschel and LABOCA we fit the rest-frame 5
-300 um emission using a two component, starburst and active galactic nucleus (AGN), model. The total infrared (8 - 1000 um) luminosity of this galaxy is (1.97+/-0.28)x10^13 Lsun with (1.17+/-0.27) and (0.79+/-0.09)x10^13 Lsun due to the AGN and starburst components respectively. The high derived AGN accretion rate of sim20% Eddington, and the measured star formation rate (SFR) of 1390pm150 Msun/yr, suggest that this massive system is in a special phase of rapid central black hole and host galaxy growth, likely caused by a gas rich merger in a dense environment. The accretion rate is sufficient to power both the jets and the previously observed large outflow. The high SFR and strong outflow suggest this galaxy could potentially exhaust its fuel for stellar growth in a few tens of Myr, although the likely merger of the radio galaxy with nearby satellites suggest bursts of star formation may recur again on time scales of several hundreds of Myr. The age of the radio lobes implies the jet started after the current burst of star formation, and therefore we are possibly witnessing the transition from a merger-induced starburst phase to a radio-loud AGN phase. We also note tentative evidence for [CII]158um emission. This paper marks the first results from the Herschel Galaxy Evolution Project (Project HeRGE), a systematic study of the evolutionary state of 71 high redshift, 1 < z < 5.2, radio galaxies.
We examine the rest-frame far-infrared emission from powerful radio sources with 1.4GHz luminosity densities of 25<=log(L_1.4/WHz^-1)<=26.5 in the extragalactic Spitzer First Look Survey field. We combine Herschel/SPIRE flux densities with Spitzer/IR
AC and MIPS infrared data to obtain total (8-1000um) infrared luminosities for these radio sources. We separate our sources into a moderate, 0.4<z<0.9, and a high, 1.2<z<3.0, redshift sub-sample and we use Spitzer observations of a z<0.1 3CRR sample as a local comparison. By comparison to numbers from the SKA Simulated Skies we find that our moderate redshift sample is complete and our high redshift sample is 14per cent complete. We constrain the ranges of mean star formation rates (SFRs) to be 3.4-4.2, 18-41 and 80-581Msun/yr for the local, moderate and high redshift samples respectively. Hence, we observe an increase in the mean SFR with increasing redshift which we can parameterise as ~(1+z)^Q, where Q=4.2+/-0.8. However we observe no trends of mean SFR with radio luminosity within the moderate or high redshift bins. We estimate that radio-loud AGN in the high redshift sample contribute 0.1-0.5per cent to the total SFR density at that epoch. Hence, if all luminous starbursts host radio-loud AGN we infer a radio-loud phase duty cycle of 0.001-0.005.
In this paper we examine the contribution of galaxies with different infrared (IR) spectral energy distributions (SEDs) to the comoving infrared luminosity density, a proxy for the comoving star formation rate (SFR) density. We characterise galaxies
as having either a cold or hot IR SED depending upon whether the rest-frame wavelength of their peak IR energy output is above or below 90um. Our work is based on a far-IR selected sample both in the local Universe and at high redshift, the former consisting of IRAS 60um-selected galaxies at z<0.07 and the latter of Spitzer 70um selected galaxies across 0.1<z<1. We find that the total IR luminosity densities for each redshift/luminosity bin agree well with results derived from other deep mid/far-IR surveys. At z<0.07 we observe the previously known results: that moderate luminosity galaxies (L_IR<10^11 Lsun) dominate the total luminosity density and that the fraction of cold galaxies decreases with increasing luminosity, becoming negligible at the highest luminosities. Conversely, above z=0.1 we find that luminous IR galaxies (L_IR>10^11 Lsun), the majority of which are cold, dominate the IR luminosity density. We therefore infer that cold galaxies dominate the IR luminosity density across the whole 0<z<1 range, hence appear to be the main driver behind the increase in SFR density up to z~1 whereas local luminous galaxies are not, on the whole, representative of the high redshift population.
We present the first mid-infrared Spitzer/Infrared Spectrograph (IRS) observations of powerful radio galaxies at z>2. These radio galaxies, 4C +23.56 (z=2.48) and 6C J1908+7220 (z=3.53), both show strong mid-infrared continua, but with 6C J1908+7220
also showing strong PAH emission at rest-frame 6.2 and 7.7um. In 4C+23.56 we see no obvious PAH features above the continuum. The PAH emission in 6C J1908+7220 is the amongst the most distant observed to date and implies that there is a large instantaneous star formation rate (SFR). This is consistent with the strong detection of 6C J1908+7220 at far-IR and sub-mm wavelengths, indicative of large amounts of cold dust, ~10^9Msun. Powerful radio galaxies at lower redshifts tend to have weak or undetectable PAH features and typically have lower far-IR luminosities. In addition, 4C 23.56 shows moderate silicate absorption as seen in less luminous radio galaxies, indicating tau_{9.7um}=0.3+/-0.05. This feature is shifted out of the observed wavelength range for 6C J1908+7220. The correlation of strong PAH features with large amounts of cold dust, despite the presence of a powerful AGN, is in agreement with other recent results and implies that star formation at high redshift is, in some cases at least, associated with powerful, obscured AGN.
We have used several different methods (radio morphology, radio spectral index, mid-IR to radio and near-IR to radio flux density ratios) to discriminate between AGN and SFGs in faint, sub-mJy radio surveys. We find that the latter two methods are th
e most powerful with current multi-wavelength data, but that future radio surveys with eMERLIN, LOFAR etc. (and ultimately the SKA) will greatly increase the power of the morphology and spectral index methods. As an example of the science possible we derive the IR luminosity density from the radio-selected SFGs using the radio/IR luminosity correlation. We also examine the contribution by luminosity to the total IR luminosity density and find evidence that fraction of LIRGs remains constant or decreases above z=1 while the relative fraction of ULIRGs continues to increase up to z=2.5.
Discerning the exact nature of the sub-mJy radio population has been historically difficult due to the low luminosity of these sources at most wavelengths. Using deep ground based optical follow-up and observations from the Spitzer Space Telescope we
are able to disentangle the radio-selected Active Galactic Nuclei (AGN) and Star Forming Galaxy (SFG) populations for the first time in a deep multi-frequency VLA/MERLIN Survey of the 13^H XMM-Newton/Chandra Deep Field. The discrimination diagnostics include radio morphology, radio spectral index, radio/near-IR and mid-IR/radio flux density ratios. We are now able to calculate the extragalactic Euclidean normalised source counts separately for AGN and SFGs. We find that while SFGs dominate at the faintest flux densities and account for the majority of the up-turn in the counts, AGN still make up around one quarter of the counts at ~5 uJy (1.4 GHz). Using radio luminosity as an unobscured star formation rate (SFR) measure we are then able to examine the comoving SFR density of the Universe up to z=3 which agrees well with measures at other wavelengths. We find a rough correlation of SFR with stellar mass for both the sample presented here and a sample of local radio-selected SFGs from the 6df-NVSS survey. This work also confirms the existence of, and provides alternative evidence for, the evolution of distribution of star formation by galaxy mass: ``downsizing. As both these samples are SFR-selected, this result suggests that there is a maximum SFR for a given galaxy that depends linearly on its stellar mass. The low ``characteristic times (inverse specific SFR) of the SFGs in our sample are similar to those of the 6dF-NVSS sample, implying that most of these sources are in a current phase of enhanced star formation.
