Optimize fine-grained signal-to-noise cuts for higher-order statistics

Optimize fine-grained signal-to-noise-ratio cuts for the starlet peak-count and starlet \ell_1-norm statistics to balance the removal of baryon-contaminated bins against the retention of uncontaminated cosmological information.

Background

The analysis finds that baryonic feedback is concentrated in particular signal-to-noise regions, especially the positive tail, but adopts a conservative strategy of removing complete wavelet scales rather than individual bins. This whole-scale removal is robust but may discard clean information. The authors explicitly defer optimization in signal-to-noise space, leaving unresolved how to select contaminated bins while preserving the maximum cosmological constraining power.

References

We leave the complex optimization of fine-grained SNR-space cuts for future investigation.

— Mitigating baryonic effects in weak lensing with higher-order statistics  (2609.09131 - Tersenov et al., 8 Sep 2026) in Section 3.2, subsection “Higher-order statistics”

We leave the exploration of non-dyadic filter banks and of cuts in SNR to future work.

— Mitigating baryonic effects in weak lensing with higher-order statistics  (2609.09131 - Tersenov et al., 8 Sep 2026) in Section 5.2, subsection “Determining robust scale cuts”