Transitioning from Stage-III to Stage-IV: Cosmology from galaxy$\times$CMB lensing and shear$\times$CMB lensing

Abstract: We examine the cosmological constraining power from two cross-correlation probes between galaxy and CMB surveys: the cross-correlation of lens galaxy density with CMB lensing convergence $\langle\delta\kappa\rangle$, and source galaxy weak lensing shear with CMB lensing convergence $\langle\gamma\kappa\rangle$. These two cross-correlation probes provide an independent cross-check of other large-scale structure constraints and are insensitive to galaxy-only or CMB-only systematic effects. In addition, when combined with other large-scale structure probes, the cross-correlations can break degeneracies in cosmological and nuisance parameters, improving both the precision and robustness of the analysis. In this work, we study how the constraining power of $\langle\delta\kappa\rangle+\langle\gamma\kappa\rangle$ changes from Stage-III (ongoing) to Stage-IV (future) surveys. Given the flexibility in selecting the lens galaxy sample, we also explore systematically the impact on cosmological constraints when we vary the redshift range and magnitude limit of the lens galaxies using mock galaxy catalogs. We find that in our setup, the contribution to cosmological constraints from $\langle\delta\kappa\rangle$ and $\langle\gamma\kappa\rangle$ are comparable in the Stage-III datasets; but in Stage-IV surveys, the noise in $\langle\delta\kappa\rangle$ becomes subdominant to cosmic variance, preventing $\langle\delta\kappa\rangle$ to further improve the constraints. This implies that to maximize the cosmological constraints from future $\langle\delta\kappa\rangle+\langle\gamma\kappa\rangle$ analyses, we should focus more on the requirements on $\langle\gamma\kappa\rangle$ instead of $\langle\delta\kappa\rangle$. Furthermore, the selection of the lens sample should be optimized in terms of our ability to characterize its redshift or galaxy bias instead of its number density.