Status of the $S_8$ Tension: A 2026 Review of Probe Discrepancies

TL;DR

The paper evaluates the S8 tension between early-universe CMB constraints and late-time probes in light of the 2026 Combined CMB baseline (Planck+ACT+SPT-3G), which yields S8 = 0.836^{+0.012}_{-0.013}. It compiles measurements from 2019–2026 across weak lensing, galaxy clustering, clusters, and RSD, highlighting a bifurcated landscape: DES Y6 remains significantly low in S8 relative to the CMB, while KiDS Legacy aligns with the baseline. The analysis emphasizes survey-specific systematics (photo-z, intrinsic alignments, shear, and baryonic feedback) as major contributors to discrepancies, though some results hint at possible new physics beyond ΛCDM. The paper discusses interpretations including modified gravity, interacting dark sectors, and evolving dark energy, and outlines paths to resolution via joint analyses, future surveys (Euclid, CMB-S4), and refined modeling.

Abstract

The parameter $S_8 \equiv σ_8 (Ω_m/0.3)^{0.5}$ quantifies the amplitude of matter density fluctuations. A persistent discrepancy exists between early-universe CMB observations and late-universe probes. This review assesses the ``$S_8$ tension'' against a new 2026 baseline: a unified ``Combined CMB'' framework incorporating Planck, ACT DR6, and SPT-3G. This combined analysis yields $S_8 = 0.836^{+0.012}_{-0.013}$, providing a higher central value and reduced uncertainties compared to Planck alone. Compiling measurements from 2019--2026, we reveal a striking bifurcation: DES Year 6 results exhibit a statistically significant tension of $2.4σ$--$2.7σ$ \citep{DESY6}, whereas KiDS Legacy results demonstrate statistical consistency at $<1σ$ \citep{Wright2025}. We examine systematic origins of this dichotomy, including photometric redshift calibration, intrinsic alignment modeling, and shear measurement pipelines. We further contextualize these findings with cluster counts (where eROSITA favors high values while SPT favors low), galaxy-galaxy lensing, and redshift-space distortions. The heterogeneous landscape suggests survey-specific systematic effects contribute substantially to observed discrepancies, though new physics beyond $Λ$CDM cannot be excluded.

Figures (1)

• Figure 1: Compilation of $S_8$ measurements from major cosmological surveys (2019--2026). The horizontal gray band represents the $1\sigma$ confidence interval of the Combined CMB baseline (Planck + ACT DR6 + SPT-3G), centered at $S_8 = 0.836$ with asymmetric width $^{+0.012}_{-0.013}$. The dashed gray line indicates the central value. Note the systematic offset between DES results (red circles), which lie consistently below the CMB band, and KiDS Legacy results (blue circles), which have migrated upward into consistency with the CMB. The eROSITA cluster count measurement (orange triangle, upper region) uniquely favors a clustering amplitude exceeding the CMB baseline. Error bars represent $1\sigma$ uncertainties.