My story ends here. It is a fairly trivial story, and I can only hope that
it has been interesting in the same way as a travel diary is interesting.
George Orwell, from "Down and Out in Paris and London", 1933
This Thesis tackled the issues related to space infrared extragalactic surveys from the eclectic point of view characterizing any space science project. Its main focus was on the development of software for automated and interactive processing of infrared imaging data and on its application to a massive dataset obtained with the ISOCAM instrument on board ESA's ISO satellite, namely the 15 m observations obtained within the framework of the ELAIS project. In so doing, not only scientific results of relevance for galaxy formation and evolution and observational cosmology studies were obtained, but also technical tools which will be of help in future space missions exploiting the infrared wavelength range have been developed.
Trying to follow some logical order, heterogeneous subjects such as infrared astronomy, galaxy formation and evolution, observational cosmology, space mission design and operations, detector physics, software development and engineering, data reduction and analysis, catalogue cross-correlation and finally multi-wavelength astronomy have been in turn touched upon. Although the Thesis focused on data obtained by the ISO satellite, the first space infrared "observatory" ever, throughout an attempt was made at putting the work into an "historical" perspective, with a keen eye both for the efforts of the "pioneers" of infrared astronomy and for the exciting prospects that space missions will offer to this discipline in the years to come.
Original results presented in this Thesis broadly belong to three areas, namely: development of techniques and tools for the reduction and analysis of infrared imaging data (The LARI Method for ISO-CAM/PHOT Data Reduction and Analysis); reduction and analysis of infrared imaging observations using such tools (Final Analysis of ELAIS 15 m Observations); identification of detected infrared sources detected in images at different wavelengths and early scientific results from such analysis (Multi-Wavelength Studies of ELAIS 15 m Final Analysis Sources). Future prospects of the present work were outlined at several stages, with special emphasis on its possible application to the ongoing Spitzer satellite mission.
The LARI Method for ISO-CAM/PHOT Data Reduction and Analysis
Techniques and tools for the reduction and analysis of data obtained with the ISOCAM and ISOPHOT cameras on board ISO were developed and integrated. As a result, the LARI Method and related software now provide a complete and efficient framework for the exploitation of ISO imaging data and a useful toolbox for the challenges in data reduction posed by future infrared missions. On the basis of a mathematical model of detectors' behaviour, it is possible to successfully deal with glitches introduced in data by cosmic rays hitting the detectors. The method has been substantially refined, extended so as to make it possible to apply it to datasets taken at different wavelengths and to observations of different depths, and turned into a powerful but easy-to-use software tool for the detection of faint sources in ISO-CAM/PHOT observations. Most importantly, and differently from other methods, the method is fully interactive by its very nature, allowing to easily check at will the reliability of detected sources. Finally, the engineering and testing process the software was submitted to led to the production of a prototype software package, the LARI Package, and related draft user's manual. As a result, the employed techniques and the developed software were extensively validated and can now be confidently applied to ISO data of particular interest. Reduction of similar data from other present and future infrared missions may undoubtedly benefit from these as well.
Final Analysis of ELAIS 15 m Observations
The LARI Method was applied to the 15 m observations carried out within the framework of the most ambitious extragalactic survey realized with ISO, the ELAIS project. Reduction of the four fields making up the ELAIS 15 m dataset has produced a catalogue of 1923 sources of extreme reliability spanning the 0.5 - 100 mJy range, which are detected with a greater than 5 over a total area of . This is the largest extragalactic catalogue produced to date by any single ISO project.
Identification of 15 m sources has been carried out on heterogeneous optical and near-infrared imaging material, allowing to determine a robust association for about 85% of the sources and to identify 22% of them as bona fide stars, thus furtherly demonstrating the reliability of our data reduction process.
The evaluation of the catalogue's quality has been carried out through both accurate simulations and the aforementioned identification process. Using the same mathematical model the LARI Method is based on, additional sources were simulated on top of true data so as to reproduce the signal and noise features appearing in the data as faithfully as practicable. Simulated data were then reduced exactly as done for true data, and comparison between input and output parameters for simulated sources provided a way to evaluate the performance of our data reduction method.
Estimated astrometric accuracy is of order 1 arcsec in both RA and Dec for , while it increases up to about 2 arcsec in both RA and Dec for , and marginally better than this for sources detected in higher-redundancy regions. Photometric accuracy is estimated to be better than 25% over our whole range of fluxes and redundancy levels, and better than 15% for sources.
The comparison of measured stellar fluxes with fluxes estimated on the basis of stellar atmosphere models calibrated on IRAS data and on near-infrared photometry allowed to achieve an IRAS/ISO photometric calibration. An IRAS/ISO relative calibration factor of was determined, shedding doubts on the goodness of the two independently determined calibrations at the 10% level. For lack of a simple way to identify error sources in IRAS and/or ISO calibration process, it was decided to put our catalogue on the more commonly used IRAS flux scale. However, while this choice was taken for the sake of compatibility of our results with studies adopting the IRAS flux scale (and particularly IRAS-based source counts and luminosity functions), this is not to indicate that IRAS calibration is more secure that ISO's.
Multi-Wavelength Studies of ELAIS 15 m Final Analysis Sources
ELAIS 15 m Final Analysis sources were identified on a variety of multi-wavelength observational material, and the physical properties of the extragalactic populations appearing in the catalogue were discussed.
Optical identification and star-galaxy discrimination, whose reliability is critical to later carry out studies of well-defined samples of extragalactic sources, were carried out in the N1 and N2 fields, i.e. the only ELAIS fields where deep multi-colour imaging from the Isaac Newton Telescope Wide Field Survey was available. Identification of optical counterparts to 15 m sources made use of the likelihood ratio method, quantitatively evaluating the likelihood of each counterpart to a given source being the correct one and thus allowing automated decisions in this respect. Through this sophisticated process, of 15 m sources is thus assigned a reliable counterpart, while of them are classified as blank fields.
Building on the results of the optical identification process, which was extended to ELAIS radio sources, a multi-wavelength catalogue resulting from the combination of ELAIS various follow-up observations into a coherent ensemble was produced, providing a precious legacy from the ELAIS project to the astronomical community.
The ELAIS Band-Merged Catalogue in bands and at 6.7, 15, 90, 175 m and 20 cm contains 3762 sources, a wealth of information which lends itself to further investigation through a variety of statistical analyses. As a preliminary step in this process, the extragalactic populations detected by ELAIS were briefly discussed through colour-colour diagrams, redshift distributions, spectral energy distributions and number of rare objects.
The process of band-merging and optical association of the sources has given considerable insight into the extragalactic infrared populations present. Although the largest single population is relatively nearby () cirrus galaxies, there is a surprisingly large population of Ultraluminous Infrared Galaxies ( 10% of 15 m sources), many of them highly obscured starbursts like Arp220. There appears to be a small proportion of genuine optically blank fields (8% at 15 m, 3% at 6.7 m, up to 20% at 90 m, and 1% at 175 m) which must have high infrared-to-optical ratios, be at higher redshift (), and have high infrared luminosities. They are therefore dusty luminous starbursts or Type 2 AGN.
9 Hyperluminous Infrared Galaxies are found in the survey. The large number of Ultraluminous Infrared Galaxies imply a very strong evolution in the star formation rate between and 1. 9 Extremely Red Objects are also detected, whose mid-infrared and optical fluxes are consistent with all 9 objects being highly extinguished starbursts like Arp220 at . Finally, 9 objects characterized by extremely high mid-to-near infrared flux ratio ( ) are detected. These newly-dubbed Extreme Mid-to-Near Infrared Objects (EMNOs) may belong to a new extragalactic population of obscured AGN and very strong dusty starbursts at which was not detected in deeper ISO surveys because of their small covered area. If this tentative discovery were confirmed, Spitzer surveys such as SWIRE would be bound to uncover large numbers of these objects, which would provide a useful window into the high-redshift obscured nuclear activity and star formation phenomena in a similar way the discovery of Extremely Red Objects has been used to define samples of high-redshift early-type galaxies.
A forthcoming paper (Lari et al., 2004) will present completeness estimates and extragalactic source counts from ELAIS 15 m Final Analysis Catalogue, covering the crucial flux range 0.5 - 100 mJy between ISOCAM 15 m Deep Surveys and IRAS All Sky Survey, where a very strong evolution appears to be taking place. A further paper (Vaccari et al., 2004a) will discuss the clustering of ELAIS 15 m Final Analysis Sources, tracing its evolution up to for the first time. These works will give us further insight into the properties of detected sources.
The ISO mission and its extragalactic survey program find natural successors in the Spitzer mission and in the GOODS and SWIRE extragalactic surveys included within its Legacy Science Program. Release of results from Spitzer's early observations have already offered a glimpse of the satellite's remarkable performance, and a vast portion of the astronomical community is eagerly waiting for the upcoming opening of the data archive to inspect the first public data and for the awarding of the first open time to get new scientific projects started.
In this exciting context, aforementioned techniques and tools developed within the framework of the LARI Method for application on ISO data will no doubt prove extremely useful to projects to be carried out by Spitzer and its successors. In particular, sophisticated techniques for:
"The Space Telescope will help to solve many outstanding
astronomical puzzles. The greatest excitement, however,
will come when the pictures returned from the satellite
reveal things no one in this generation of astronomers
has dreamed of, phenomena that only the next generation
will be privileged to understand."
John Bahcall and Lyman Spitzer