Introduction

Observational cosmology made great strides without precedent recently. Principal parameters of the theory of the Big-bang: the growth rate of the universe (the constant of Hubble), densities of its principal components (matter, radiation, cosmological constant), i.e. the value of the cosmological parameters (respectively $\Omega _ \Lambda$) and the geometry of the universe were, 15 years ago, only known with an order of magnitude.

The use of measurement of distance by means of objects whose intrinsic luminosity is known (standard candles) was applied for the first in the Thirties per E Hubble. The measurement of the distances from galaxies had then enabled him to highlight the expansion of the universe and to make the first measurement of the constant door today its name.

The study and the classification of the supernovæa allowed to define a particular class: supernovæ of the Ia type. The characteristics most interesting, for the cosmologists, are the remarkable homogeneity of their luminosity (one observes variations from approximately 50% to the maximum of luminosity) and their brightness equivalent to that of a whole galaxy (the supernovæde standard Ia are visible at cosmological distances) by making the standard candles most precise known. Despite everything, the variations of luminosity were too significant to allow precise measurements of cosmology.

The introduction at the beginning of years 1990 of corrections by considering the correlations between the maximum luminosity and the rate of decrease or the color allowed, in a completely empirical way, to reduce the dispersion of luminosity to approximately 15% in the filter B (corresponding to the blue part of the spectrum) of the reference frame of the supernova.

It is as from the same time as two groups, High-Z TEAM and the Supernova Cosmology Project (SCP) began research and the systematic study of supernovæ of the Ia. type the analysis of several tens of supernovæ with great redshift allowed the spectacular description of the acceleration of the expansion of the universe and the presence of a new energy component: black energy . These results were confirmed thereafter by the observation of the great structures within the framework of great surveys of galaxies (2dF, SLOAN DIGITAL Sky Survey) and of the cosmological diffuse bottom (WMAP), then, by the more recent observation of two new batches of supernovæpar each of the two groups.

The work of this thesis, realized within the framework of SCP, falls under the continuity of this research.

The strategy of research of SCP was directed since towards the search for supernovæà very great redshifts (around 1). These objects make it possible on the one hand to partly raise the degeneration in the measurement of $\Omega _ \Lambda$, but also to test certain alternative models with the cosmological constant (intergalactic dust, gray dust...).

It is within this framework that it is registered the research undertaken by our FrOGS group (for French Observing Group of Supernovæ) of the LPNHE (Laboratory of Nuclear physics and High Energies). The privileged access of which we gained, to the imagor with large field of sight CFH12K assembled on telescope CFHT (Canada France Hawaii Telescope), us made it possible to begin the research of such objects in a systematic way.

The research campaigns of springs 2000 and 2001 thus brought to the discovery, using the dedicated software of reduction developed to the LPNHE, ToADS (Tools for Analysis and Detection of Supernovæ), of 5 supernovæà of the shifts towards the red ranging between 0.5 and 1.2.

Four of them were followed thereafter by the space telescope Hubble. At the time of a similar research undertaken to the CTIO (Cerro Tololo Interamerican Telescope of Chile) our collaborators pnt discovered ten supernovædont 2 who were followed by Hubble. This thesis presents the analysis from these 6 supernovæ.

This manuscript is composed of three parts and 12 chapters.

In the first part, we recall the formalism of the Big-bang, then, the principal results of observational cosmology by concentrating us mainly to the measure of the cosmological parameters. Chapter 2 shows how the measurement of the distance from standard candles allows the measurement of the cosmological parameters, we also present the principal observational difficulties related to measurement.

In chapter 5 , we describe, from an observational point of view then from a theoretical point of view by detailing the mechanisms of explosion, the comprehension which we particularly have concentrating us of the supernovæen on the supernovæ of the Ia type.

In chapter 6 , we present the methods which are used for `` to standardize '' the luminosity of the supernovæ and the various systematic effects related to measurement.

The supernovæ are phenomena évanescents, their lifespan is approximately a hundred days, it is thus possible to detect them by comparison of images taken at a few weeks of variation. In chapter 7 , we describe how starting from subtraction of images, it is possible to detect in a systematic way ten supernovæpar night of observation with a imagor with large field. We describe, in particular, the observations which we carried out at the time of spring 2001 with the CFHT and the chain of ToADS detection.

The measurement of the distance from the supernovæ rests on the comparison of the apparent luminosity of close objects (which make it possible to calibrate the scale of distance) and of more remote objects. For that, measurements of luminosity are made with the same phase: the maximum of luminosity. To estimate them, one builds a lightcurve, i.e., the evolution of the apparent luminosity according to time. One adjusts then this curve of light by a model. This adjustment also enables us to estimate the rate of decrease of the lightcurve.

The chapter 8 presents how starting from the analysis of the images of follow-up, we built the lightcurves of our supernovæ by differential photometry. Chapter 9 details the procedures of adjustment of the lightcurves.

We analyzed a batch of a hundred supernovæ close exits of the literature while following our procedures of curve fitting to light. From this batch, we re-studied the relations of standardization for the filter B and we studied these relations for the filter U (corresponding to the part near ultra purple of the spectrum). These two studies enabled us to calibrate our estimator of distance for respectively our supernovæavec of the redshifts around 0.5 and of 1. This original study based on our batch of supernovæ, us finally made it possible to make two estimates independent of $\Omega _ { \rm M}$. This analysis is presented in chapter 10 .