Prof. Dimitra Rigopoulou

Abstract:

We present a multi-wavelength, UV-to-radio analysis for a sample of massive (M * ∼ 1010 M ⊙) IRAC- and MIPS 24 μm detected Lyman break galaxies (LBGs) with spectroscopic redshifts z 3 in the GOODS-North field. For LBGs without individual 24 μm detections, we employ stacking techniques at 24 μm, 1.1mm, and 1.4GHz to construct the average UV-to-radio spectral energy distribution and find it to be consistent with that of a luminous infrared galaxy with L IR = 4.5 +1.1-2.3 × 1011 L ⊙ and a specific star formation rate of 4.3 Gyr-1 that corresponds to a mass doubling time 230 Myr. On the other hand, when considering the 24 μm detected LBGs we find among them galaxies with L IR>10 12 L ⊙, indicating that the space density of z 3 UV-selected ultra-luminous infrared galaxies (ULIRGs) is (1.5 0.5) × 10-5 Mpc-3. We compare measurements of star formation rates from data at different wavelengths and find that there is tight correlation (Kendall's τ>99.7%) and excellent agreement between the values derived from dust-corrected UV, mid-IR, millimeter, and radio data for the whole range of L IR up to L IR 1013 L ⊙. This range is greater than that for which the correlation is known to hold at z ∼ 2, possibly due to the lack of significant contribution from polycyclic aromatic hydrocarbons to the 24 μm flux at z ∼ 3. The fact that this agreement is observed for galaxies with L IR> 1012 L ⊙ suggests that star formation in UV-selected ULIRGs, as well as the bulk of star formation activity at this redshift, is not embedded in optically thick regions as seen in local ULIRGs and submillimeter-selected galaxies at z = 2. © 2010 The American Astronomical Society. All rights reserved.

Abstract:

We present spectral energy distributions (SEDs) for 68 Herschel sources detected at 5σ at 250, 350 and 500 μm in the HerMES SWIRE-Lockman field. We explore whether existing models for starbursts, quiescent star-forming galaxies and active galactic nucleus dust tori are able to model the full range of SEDs measured with Herschel. We find that while many galaxies ( 56 per cent) are well fitted with the templates used to fit IRAS, Infrared Space Observatory (ISO) and Spitzer sources, for about half the galaxies two new templates are required: quiescent ('cirrus') models with colder (10-20 K) dust and a young starburst model with higher optical depth than Arp 220. Predictions of submillimetre fluxes based on model fits to 4.5-24 μm data agree rather poorly with the observed fluxes, but the agreement is better for fits to 4.5-70 μm data. Herschel galaxies detected at 500 μm tend to be those with the highest dust masses. © 2010 The Authors. Journal compilation © 2010 RAS.

Abstract:

We study the dust properties of galaxies in the redshift range 0.1 ≲z≲ 2.8 observed by the Herschel Space Observatory in the field of the Great Observatories Origins Deep Survey-North as part of the PACS Extragalactic Probe (PEP) and Herschel Multi-tiered Extragalactic Survey (HerMES) key programmes. Infrared (IR) luminosity (LIR) and dust temperature (Tdust) of galaxies are derived from the spectral energy distribution fit of the far-IR (FIR) flux densities obtained with the PACS and SPIRE instruments onboard Herschel. As a reference sample, we also obtain IR luminosities and dust temperatures of local galaxies at z < 0.1 using AKARI and IRAS data in the field of the Sloan Digital Sky Survey. We compare the LIR-Tdust relation between the two samples and find that the median Tdust of Herschel-selected galaxies at z< 0.5 with LIR≲ 5 × 1010L· appears to be 2-5 K colder than that of AKARI-selected local galaxies with similar luminosities, and the dispersion in Tdust for high-z galaxies increases with LIR due to the existence of cold galaxies that are not seen among local galaxies. We show that this large dispersion of the LIR-Tdust relation can bridge the gap between local star-forming galaxies and high-z submillimetre galaxies (SMGs). We also find that three SMGs with very low Tdust (≲20 K) covered in this study have close neighbouring sources with similar 24-μm brightness, which could lead to an overestimation of FIR/(sub)millimetre fluxes of the SMGs. © 2010 The Authors. Journal compilation © 2010 RAS.

Abstract:

We have carried out two extremely deep surveys with SPIRE, one of the two cameras on Herschel, at 250 μm, close to the peak of the far-infrared background. We have used the results to investigate the evolution of the rest-frame 250-μm luminosity function out to z = 2. We find evidence for strong evolution out to z ≃ 1 but evidence for at most weak evolution beyond this redshift. Our results suggest that a significant part of the stars and metals in the universe today were formed at z ≤ 1.4 in spiral galaxies. © 2010 ESO.

Abstract:

Dusty, star-forming galaxies contribute to a bright, currently unresolved cosmic far-infrared background. Deep Herschel-Spectral and Photometric Imaging Receiver (SPIRE) images designed to detect and characterize the galaxies that comprise this background are highly confused, such that the bulk lies below the classical confusion limit. We analyse three fields from the Herschel Multi-tiered Extragalactic Survey (HerMES) programme in all three SPIRE bands (250, 350 and 500 μm); parametrized galaxy number count models are derived to a depth of ~2 mJy beam-1, approximately four times the depth of previous analyses at these wavelengths, using a probability of deflection [P(D)] approach for comparison to theoretical number count models. Our fits account for 64, 60 and 43 per cent of the far-infrared background in the three bands. The number counts are consistent with those based on individually detected SPIRE sources, but generally inconsistent with most galaxy number count models, which generically overpredict the number of bright galaxies and are not as steep as the P(D)-derived number counts. Clear evidence is found for a break in the slope of the differential number counts at low flux densities. Systematic effects in the P(D) analysis are explored. We find that the effects of clustering have a small impact on the data, and the largest identified systematic error arises from uncertainties in the SPIRE beam. © 2010 The Authors. Journal compilation © 2010 RAS.