Probing the IGM with High Resolution Spectroscopy of Quasars

Between Galaxies and Intergalactic Medium exists a strong interaction that can significantly influence the processes that led to the formation of our Universe as we can see it today. The Intergalactic Medium represents the reservoir of raw material from which stars and galaxies were originated. These ones in turn are responsible for the production of heavy elements commonly referred to as "metals", these metals can be expelled in order to "enrich" the Intergalactic Medium. The search for metals in the Intergalactic Medium allows to obtain valuable information on the mechanisms of expulsion from Galaxies, like for example galactic winds or supernovae explosions, which are commonly referred to as "feedbacks". One of the methods used for this kind of study is represented by the measurement of spectra of galaxies called "Quasars". Each time that the light of the Quasar is absorbed by a specific chemical element along the line of sight, an absorption line is produced in the observed spectrum to a characteristic frequency that allows to identify the element responsible for the absorption. Given the particular nature of these absorptions, in order to measure their corresponding absorption lines in the spectrum high spectral resolution (R ~ 50000) and high sensitivity or "signal-to-noise ratio" (SNR> 50) are basic requirements for a successful observation. State of the art measurements in this field are currently limited by the available sensitivity which rarely reaches SNR>50. The main objective of this work was to obtain a sample of spectra with SNR>100 in order to repeat those measurements with an improved detection of weak CIV metal lines and understand the contribution that those very weak lines could give to the total CIV mass density in the Universe. In doing this it has been shown that the measurements made with the best currently available resources are still far from the instrumental limits imposed by systematic errors, this means that the sensitivity of such measures can still be increased as the square root of the observation time. In this context, it is crucial the type of approach used for the data reduction which, if not suitable for the high quality of the measurements, could introduce systematics much more larger than the instrumental ones compromising the quality of the finished product. This type of problems has been studied in the thesis by analyzing the practical case with the development of a dedicated software, written in IDL language, for the data reduction of UVES and HIRES echelle spectrographs. Major attention was given to the handling of errors in order to obtain an estimation of the uncertainties able to reproduce in a realistic way the actual quality of the data. This software represents a first step in the development of pipelines dedicated to data reduction for the next generation of telescopes, which will require new approaches able to minimize those systematic errors generally introduced by the data reduction processes themselves. As a case study in the thesis have been presented the measurements of the Column Density Distribution Function, the cosmic evolution of the total CIV mass density and the two points correlation function. These measurements were carried out for the first time with a sample of spectra at very high signal-to-noise ratio highlighting how the sensitivity of these spectra allows to significantly improve the observations for column densities logN(CIV)<12 cm^-2 corresponding to the deepest areas, and thus more representative of the Intergalactic Medium being located at great distance from the galaxies in which the metals are produced.