The many flavours of photometric redshifts

For more that seventy years, the measurements of fluxes of galaxies at different wavelengths and derived colours have been used to estimate their corresponding cosmological distances. From the fields of galaxy and AGN evolution to precision cosmology, the number of scientific projects relying on such distance measurements, called photometric redshifts, have exploded. The benefits of photometric redshifts is that all sources detected in photometric images can have some distance estimates relatively cheaply. The major drawback is that these cheap estimates have a low precision when compared with the resource-expensive spectroscopy. The methodology to estimate redshifts has been through several major revolutions throughout the last decades, triggered by increasingly more stringent requirements on the photometric redshift accuracy. Here, we review the various techniques to obtain photometric redshifts, from template-fitting to machine learning and hybrid systems. We also describe the state-of-the-art results on current extra-galactic samples and explain how survey strategy choices impact redshift accuracy. We close the review with a description of the photometric redshifts efforts planned for upcoming wide field surveys, which will collect data on billions of galaxies, aiming to solve the most exciting cosmological and astrophysical questions of today.