Poster
2025 workshop program
Evaluation and Modeling of Spatial Selection Effects in Photometric Redshifts
The work
Abstract
Photometric redshift (photo‑z) uncertainties are a key systematic in Stage IV weak lensing surveys. While traditional methods rely on spectroscopic calibration, spatial selection effects—caused by inhomogeneous survey depth and observational conditions—can induce coherent biases in redshift estimates across the sky. We present a Bayesian forward modeling approach to account for these spatial selection effects, using a Log-Gaussian Cox Process to model spatial structure and a Dirichlet-multinomial likelihood for redshift inference. Our method is validated on MICE Grand Challenge simulations and assessed for its impact on cosmological constraints from 3×2-point statistics.