Reconstruction of the photometric redshift

• Grant number: 21/08345-3
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: December 01, 2021
• End date: November 30, 2022
• Field of knowledge: Physical Sciences and Mathematics - Physics - Elementary Particle Physics and Fields
• Principal Investigator: Elcio Abdalla
• Grantee: Larissa Freire de Jesus Costa
• Host Institution: Instituto de Física (IF). Universidade de São Paulo (USP). São Paulo , SP, Brazil
• Associated research grant: 14/07885-0 - The BINGO telescope: a new 21 cm window for exploring the dark universe and other astrophysics, AP.ESP

Abstract

The study of galaxies provides knowledge about various aspects of the Universe that go beyond individual analysis for these objects. An important piece of information resides in its emitted radiation, which suffers change in the spectrum due mainly to the expansion of the Universe. Measuring this redshift of light is costly when done by spectroscopic approaches. However, the current project aims to apply machine-learning techniques for manipulating artificial neural networks so that they can predict a function and estimate the redshifts of a photometric catalog based on the colors of the galaxies in this survey, that, from a set of spectroscopic data, in the same region, whose redshifts are already known. The function will be obtained through redshift estimation algorithms for photometric data that also make use of Gaussian Processes. With considerable calibration of errors, the result of this research will allow the reconstruction of numerical densities from galaxies by redshift to other wider areas of the sky. (AU)