Map-level implementation of remaining TATT intrinsic-alignment terms

Develop a map-level implementation of the remaining terms of the tidal alignment and tidal torquing (TATT) intrinsic alignment model for weak-lensing analyses, enabling forward modeling of intrinsic-alignment contributions directly on convergence/shear maps rather than relying solely on the non-linear alignment (NLA) approximation or a partial TATT coupling term.

Background

Intrinsic galaxy alignments are a key astrophysical systematic in weak gravitational lensing that can bias cosmological constraints. The TATT (tidal alignment and tidal torquing) model extends simpler prescriptions like the non-linear alignment (NLA) model by including additional terms capturing quadratic alignment and cross-correlation components.

In this analysis the authors implement the NLA model with an effective redshift-dependent amplitude and include a single TATT coupling term (b_TA). However, a complete map-level implementation of the remaining TATT terms is currently absent, limiting the realism of forward-modeled intrinsic-alignment signals in their simulation-based inference pipeline.

A full map-level TATT implementation would allow intrinsic-alignment effects to be projected consistently at the field (map) level across tomographic bins, improving robustness for non-Gaussian, map-based cosmological summaries and probe combinations.

References

Currently, a map-level implementation of the remaining terms of the TATT model is an open research question and therefore not included in this work.

Dark Energy Survey Year 3 results: Simulation-based $w$CDM inference from weak lensing and galaxy clustering maps with deep learning. I. Analysis design  (2511.04681 - Thomsen et al., 6 Nov 2025) in Subsubsection “Intrinsic Alignment” (Section sec:intrinsic_alignment)