(D3.8.1.3, Mar 2027) In the final year of Phase C, we will enable the DESC’s drive towards forward-modelling analysis approaches that naturally account for complex selection effects and have an accuracy ready for the extreme statistical
power of later LSST data releases. We will optimise the photo-z pipelines for speed, e.g., through AI techniques, to run on a large number of simulated datasets and interconnect the data processing through algorithmic links between photo-z, galaxy properties, and external data (e.g., infrared/UV bands). The DESC photo-z infrastructure will thus have considerable legacy value which we will apply to the Rubin Project’s photo-z ‘validation cooperative’.
From Phase C proposal