Les sources de la reionisation cosmique vues par l'instrument VLT/MUSE et par effet de lentille gravitationnelle
Doctorant à l'IRAP
Résumé de la thèse :
Reionization is the last change of state of the Universe which made its entire hydrogen content transition from a neutral to a completely ionized state. This rapid transition and heating of the gas content had major consequences on the formation and evolution of structures which makes of reionization a key element to understand the present day Universe.
In our current understanding, reionization was mostly done by z ∼ 6 and the sources responsible for this transition are likely faint, low mass and star-forming galaxies (SFGs). One way to study this population is to determine the Luminosity Function (LF) of galaxies selected from their Lyman-alpha emission and assess their ionizing flux density. However, most recent studies and their conclusions are in general limited by the lowest flux that can be reached with the current observational facilities.
One of the major goals of the work presented in this manuscript is the study of the Lyman-alpha emitters (LAE) LF using deep observations of strong lensing clusters with the VLT/MUSE instrument. MUSE is a large field of view Integral Field Unit (IFU) ideal to work on the galaxy LF since it allows a complete and blind selection of LAEs without any photometric preselection. In addition, MUSE provides a continuous redshift range of 2.9 < z < 6.7 for LAEs that just overlaps with the end of the reionization era, making MUSE well suited to provide constraints on the contribution of LAEs to reionization.
The combined usage of large IFU data cubes and lensing fields makes this analysis computationally challenging. To get around this difficulty, we have developed new methods to account for the contribution of each individual LAE, including the effective-volume and completeness determi- nations. The volume computation is based on the simulation of the detection process of individual LAEs in the source plane reconstruction of MUSE cubes to account for both the lensing effects and the individual spatial and spectral profiles of LAEs. Throughout this work special care was given to faint and highly magnified LAEs since they are the key to access the very faint end of the LAE LF. To the cost of a significant increase in complexity and a lower volume of Universe explored, both due to the lensing effect, we build the LAE LF using a Monte-Carlo process to account for all possible sources of uncertainties, for a population of 152 LAEs with 39 < log LLyα [erg s−1] < 43 selected across four MUSE fields of view.
The LFs resulting form this analysis set an unprecedented level of constraint on the shape of the faint end and reached down to log L = 40.5 . We conclude, making no assumption on the escape fraction of Lyα emission, that the LAE population has a similar level of contribution to the total ionising emissivity as the UV-selected galaxies (LBGs) at z ∼ 6. Depending on the relative intersection of these two populations, the SFGs could produce enough ionising flux to fully ionise the Universe by z ∼ 6.
In the continuity of this work on the LAE LF, we investigate the effect of the selection method on the conclusions mentioned above. A better characterisation of the bias intrinsic to the LAE and LBG selection processes is needed to reach a more robust assessment of the complete population of high redshift SFGs. To this end, we implemented a blind and systematic search of both LAEs and LBGs behind lensing field A2744 using the deep MUSE observations in combination with public data of the Hubble Frontier Fields (HFF) program. The results have shown that the observed proportion of LAEs increases significantly among UV-faint galaxies and at increasing redshift.
Composition du jury de thèse :
- Roser Pelló Directrice de thèse IRAP
- David Elbaz Rapporteur CEA Saclay
- Olivier Le Fèvre Rapporteur LAM
- Johan Richard Examinateur CRAL
- Laurence Tresse Examinatrice CRAL
- Jean-Paul Kneib Examinateur EPFL
- Geneviève soucail Examinatrice IRAP