Efstathiou, Andreas

We present 7-180μm photometry of a sample of hyperluminous infrared galaxies (HyLIGs) obtained with the photometer and camera mounted on the Infrared Space Observatory (ISO). We have used state-of-the-art' radiative transfer models of obscured starbursts and dusty tori to model their broadband spectral energy distributions (SEDs). We find that IRAS F00235+1024, IRAS F14218+3845 and IRAS F15307+3252 require a combination of starburst and AGN components to explain their mid to far-infrared emission, while for TXS0052+471 a dust torus model alone is sufficient. For IRAS F00235+1024 and IRAS F14218+3845 the starburst component is the predominant contributor whereas for IRAS F15307+3252 the dust torus component dominates. The implied star formation rates (SFR) estimated from the starburst infrared luminosities are dM*,all/dt > 1000M⊙yr-1h50-2 and are amongst the highest SFRs estimated to date. We also demonstrate that the well-known radio-FIR correlation observed for extragalactic sources extends into both higher radio and infrared power than previously investigated. The relation for HyLIGs has a mean q value of 1.94. The results of this study imply that better sampling of the IR SEDs of HyLIGs may reveal that both AGN and starburst components are required to explain their emission from the NIR to the sub-millimetre.

The ISOPHOT Serendipity Survey utilized the slew time between ISO's pointed observations with strip scanning measurements of the sky in the far-infrared (FIR) at 170μm. The integral 170μm fluxes for compact sources derived from the slews are put on an absolute flux level by using a number of galaxies as calibrator sources observed with ISOPHOT's photometric mapping mode, supplemented by Serendipity Survey observations of two planets and two asteroids with available model fluxes. A first group of 115 well-observed sources with a high signal-to-noise ratio in all four detector pixels having a galaxy association were extracted from the slew data with low (I100μm ≤ 15 MJy/sr) cirrus background. For all but a few galaxies, the 170μm fluxes are determined for the first time, which represents a significant increase in the number of galaxies with measured FIR fluxes beyond the IRAS 100μm limit. The large fraction of sources with a high F170μm/F100μm flux ratio indicates that a very cold (T < 20 K) dust component is present in many galaxies. The typical mass of the coldest dust component is MDust = 107.5 ± 0.5 M⊙, a factor 2 - 10 larger than that derived from IRAS fluxes alone. As a consequence, the gas-to-dust ratios are much closer to the canonical value for the Milky Way. A similar Serendipity Survey with FIRST has the prospects of delivering FIR data with a much higher angular resolution (PACS) or at longer wavelengths (SPIRE) than ISOPHOT, thereby providing either crucial information for the identification of compact sources in confused regions or extending the spectral coverage for a large number of sources and finding rare classes of very cold FIR emitters.

The first set of compact sources observed in the ISOPHOT 170 μm Serendipity Survey is presented. From the slew data with low (I100μm ≤ 15 MJy/sr) cirrus background, 115 well-observed sources with a high signal-to-noise ratio in all detector pixels having a galaxy association were extracted. Of the galaxies with known optical morphologies, the vast majority are classified as spirals, barred spirals, or irregulars. The 170 μm fluxes measured from the Serendipity slews have been put on an absolute flux level by using calibration sources observed additionally with the photometric mapping mode of ISOPHOT. For all but a few galaxies, the 170 μm fluxes are determined for the first time, which represents a significant increase in the number of galaxies with measured Far-Infrared (FIR) fluxes beyond the IRAS 100 μm limit. The 170 μm fluxes cover the range 2 ≲ F170μm ≲ 100 Jy. Formulae for the integrated FIR fluxes F40-220μm and the total infrared fluxes F1-1000/μm incorporating the new 170 μm fluxes are provided. The large fraction of sources with a high F170μm/F100μm flux ratio indicates that a cold (T/Dust ≲ 20 K) dust component is present in many galaxies. The detection of such a cold dust component is crucial for the determination of the total dust mass in galaxies, and, in cases with a large F170μm/F100μm flux ratio, increases the dust mass by a significant factor. The typical mass of the coldest dust component is MDust = 107.5 ±0.5 M⊙, a factor 2-10 larger than that derived from IRAS fluxes alone. As a consequence, the majority of the derived gas-to-dust ratios are much closer to the canonical value of ≈ 160 for the Milky Way. By relaxing the selection cri-teria, it is expected that the Serendipity Survey will eventually lead to a catalog of 170 μm fluxes for ≈ 1000 galaxies.

he mid-infrared spectra of galaxies are rich in features such as the Polycyclic Aromatic Hydrocarbon (PAH) and silicate dust features which give valuable information about the physics of galaxies and their evolution. For example they can provide information about the relative contribution of star formation and accretion from a supermassive black hole to the power output of galaxies. However, the mid-infrared spectra are currently available for a very small fraction of galaxies that have been detected in deep multi-wavelength surveys of the sky. In this paper we explore whether Deep Generative Network methods can be used to reconstruct mid-infrared spectra in the 5–35μm range using the limited multi-wavelength photometry in ∼20 bands from the ultraviolet to the submillimeter which is typically available in extragalactic surveys. For this purpose we use simulated spectra computed with a combination of radiative transfer models for starbursts, active galactic nucleus (AGN) tori and host galaxies. We find that our method using Deep Generative Networks, namely Generative Adversarial Networks and Generative Latent Optimization models, can efficiently produce high quality reconstructions of mid-infrared spectra in ∼ 60% of the cases. We discuss how our method can be improved by using more training data, photometric bands, model parameters or by employing other generative networks.