Speaker
Description
Current and next-generation wide-field surveys, such as the Dark Energy Survey (DES) and the upcoming Legacy Survey of Space and Time (LSST), will rely heavily on joint 3x2pt analyses, combining cosmic shear, galaxy-galaxy lensing, and galaxy clustering, to place stringent constraints on cosmological parameters. In this context, the selection and characterization of optimal galaxy samples is a major area of research and one promising choice is to use the same galaxy sample for lenses and sources. This reduces the systematics parameter space that describes the uncertainties related to galaxy samples. Such a “lens-equal-source” analysis significantly improves the self-calibration of photo-z systematics, leading to improved cosmological constraints.
In this talk, we present a novel lens-equal-source 3x2pt pipeline applied to synthetic DES Year 6 data. To break parameter degeneracies and mitigate systematic effects, we introduce an informed prior on large-scale galaxy bias, including the effects of assembly bias. This physics-informed prior is calculated with the Halo Occupation Distributions (HOD) constructed on physically motivated UniverseMachine catalogs. Deep field observations with synthetic source injection are used to mimic realistic galaxy populations. By sampling across varying cosmologies, our prior naturally captures the complex, multidimensional covariance between galaxy bias, photo-z errors, and underlying cosmology.
We will discuss the integration of this prior directly into the likelihood inference using the Cocoa framework. Finally, we will demonstrate how leveraging high-fidelity, physics-informed priors in a lens-equal-source framework successfully stabilizes MCMC chains and tightens parameter constraints.
| Other topic / keywords: | weak gravitational lensing; galaxy survey science |
|---|