SFRG ongoing projects

[ back to sfrgs ]


Based on population studies of local galaxies, we expect that a large fraction of high redshift galaxies should lie in the unexplored region of hot dusty systems, where there are currently no sensitive surveys. If we confirm that these galaxies are star formation dominated with slightly hotter dust temperatures than the well studied SMGs, then they could make up a large fraction of the cosmic star formation history at early times. Delineating the full range of SED properties for distant, luminous galaxies is also an important and currently unknown quantity that is essential for modeling the counts and redshift distributions of our radio selected sample.

The distribution of dust temperature, Td, versus total infrared luminosity (8-1100um) for SFRGs with spectroscopic redshifts.

   MERLIN radio selection

All SFRGs of our sample were originally selected from radio-detected (uJy) high redshift galaxies. This population of S_1.4GHz > 30uJy sources has always been fundamental in pinpointing and studying the submillimeter galaxy population (SMGs) who show fantastic amounts of star formation near z~2. These galaxies were found in a similar way, but lack submm detections, suggesting hotter dust temperatures. While VLA coverage of these objects has previously provided flux estimates in 1.4GHz, it is the fantastic high-resolution MERLIN radio imaging which shows the details of the emission regions on sub-kpc scales. We use the MERLIN data to distinguish AGN from starburst dominated SFRGs. In the case of an AGN-dominated galaxy, we investigate the core strength vs. jet and lobe structure, and in the case of starbursts, analyze dust extinction between HST and MERLIN imagery.

MERLIN radio contours shown here over HST i-band for an SFRG which has recently undergone a major merger. Without high-resolution radio, it would be impossible to distinguish two merging galaxies from a single quiessant galaxy with dust obscuration.

   Spitzer IRS

As part of a pilot study containing six galaxies, we have already obtained IRS spectra to understand the mid-IR properties of SFRGs. An approved program for the current term will increase the sample size from six galaxies to sixteen, including galaxies that present a wider range in spectral properties. The original set was intrinsically brighter at 24um than the whole SFRG sample (in order to guarantee detection in IRS with relatively short exposures), and all show clear signs of AGN in their UV-spectra. The ten new galaxies (with redshifts between 1.3 and 2.7) we target this term all exhibit pure starburst UV-spectra and are part of our IRAM-PdBI CO detection program and have 24um fluxes from 0.3-0.8mJy. Whil many 24um selected galaxies have been studied with IRS, many with radio detections, the uniqueness of this set stems from the fact that these galaxies have no submm detection. We expect to detect PAH emission features in individual galaxies, and also plan to make a composite spectrum coadding the spectra in their restframes, and by doing that, will be able to test the physical cause for their apparent hot dust temperatures (and hence lack of submm detections).

Composite of the six SFRGs observed with Spitzer-IRS from the previous GO2 program here shown in the rest frame (Menendez-Delmestre et al., in prep).
   IRAM-PdBI CO detection

In a pilot PdBI project, we detected CO(3-2) in two SFRGs, confirming that these galaxies are similar in most respect to the ultraluminous SMGs. Assuming they simply differ by having hotter dust temperatures, their similarities would thus double the average FIR luminosity density at z~2. To place this conclusion on a better statistical footing, three more detections of CO(3-2) in SFRGs have been gathered (one new, two from Daddi et al 2008) using the IRAM Plateau de Bure Interferometer. More observations are currently planned for a larger sample of 16 SFRGs. Detailed 2-D CO maps will be used to analyze the galaxies' star formation properties in relation to the canonical Schmidt-Kennicutt law and star formation efficiencies in similar high-z galaxies. Using the high-sensitivity and bandwidth of the new PdBI 3mm receivers, we will be able to probe the SFRGs as deep as our extensive PdBI SMG survey had probed CO in SMGs, allowing a direct comparison between the two classes of objects.

Detection of CO(3-2) in five SFRGs.

   Keck follow-up spectroscopy

Following-up longer wavelength observations with the Keck 10m telescope, we are able to classify the galaxies' rest-UV spectra and determine what fraction are AGN dominated versus starburst dominated. Making this distinction is key to understanding the population - if the hotter dust temperatures give rise to more prominant AGN activity than is seen in SMGs. The rest-UV spectra may also be used to constrain metallicities and star formation rates independent of similar radio and infrared measurements.

Moderate starburst activity is revealed in these LRIS rest-UV spectra of two SFRGs.

Managed by Caitlin Casey:
Institute of Astronomy, University of Cambridge
Madingley Rd, Cambridge CB3 0HA, UK   tel: (44)1223 766653