Multi-Probe Cosmology Tension Console V3.0
Every independent observational constraint on the fundamental cosmological parameters, aggregated onto unified forest plots. CMB, BAO, supernovae, weak lensing, cluster counts, distance ladders, megamasers, lensed quasars, and gravitational-wave standard sirens, with live pairwise tension calculation. Surfaces the Hubble tension (~5 sigma) and the S_8 tension (~2.7 sigma) directly from current published values. This module is the only open source instrument that may be embedded off-site with the condition that the copyright stays in place. Email info@astrophyzix.org for a licence to be issued a working version for your platform. The research notes for this module can be found on Academia.edu — (Read Notes)
Every measurement in this console has been verified against its source paper. Each entry carries an arXiv identifier (visible in the Probes tab) so the reader can confirm the central value and uncertainty independently. Entries that could not be matched to an exact published value have been removed rather than approximated.
Last verification sweep: 24 May 2026. New papers appear continuously; the curated set reflects the literature as of that date. Categories represented: CMB primary anisotropies and lensing, BAO geometry, distance-ladder standard candles (Cepheid, TRGB, JAGB), megamaser geometry, time-delay cosmography of lensed quasars, gravitational-wave standard sirens, cosmic shear, and X-ray cluster abundances.
Consilience (Whewell 1840) is the convergence of independent lines of evidence on a single answer. In modern cosmology, the same parameter, for example the Hubble constant H_0, is measured by physically and methodologically independent probes: the cosmic microwave background, baryon acoustic oscillations, Type Ia supernovae, gravitationally-lensed quasars, megamaser kinematics, gravitational-wave standard sirens, surface brightness fluctuations, and the Cepheid + TRGB + JAGB distance ladder.
If the underlying cosmological model is correct, all probes should converge on the same value within their stated uncertainties. Where they disagree systematically, the discrepancy is a tension, which either signals unaccounted systematics in one or more probes or, more excitingly, points to physics beyond the standard Lambda-CDM model.
- Early Universe (CMB, BAO, BBN) :: anchored at recombination (z ~ 1100) or earlier. Sensitive to the sound horizon at the drag epoch, which depends on the energy contents of the early universe. Model-dependent: assumes Lambda-CDM (or whatever the underlying expansion history is). Planck, ACT, SPT, DESI, BBN abundances.
- Late Universe (distance ladders) :: anchored at z = 0 via geometric parallaxes, then extended outward by standard candles. Cepheids, TRGB, JAGB, Mira variables, Type Ia supernovae. Largely model-independent for H_0; assumes only that the local universe is described by a Friedmann metric. SH0ES, CCHP, Carnegie-Chicago.
- Standard Sirens (GW) :: gravitational-wave inspirals give the luminosity distance directly from the waveform amplitude, without a distance ladder. Combined with an electromagnetic redshift (when available) or statistical galaxy catalog matching, they yield H_0. GW170817 was the first. Independent of both early- and late-universe systematics.
- Lens Time Delays :: a lensed quasar produces multiple images. Time delays between image variability give an absolute distance scale combined with a lens mass model. H0LiCOW, TDCOSMO. Sensitive to lens model assumptions.
- Megamaser Cosmology :: water masers orbiting supermassive black holes in active galactic nuclei give geometric distances out to ~150 Mpc. Megamaser Cosmology Project (MCP). Pure geometric, no distance ladder.
- Weak Lensing + Cluster Counts (S_8) :: late-universe matter clumpiness probed by cosmic shear of background galaxies (DES, KiDS, HSC) or X-ray cluster abundance (eROSITA). Constrains the combination S_8 = sigma_8 * sqrt(Omega_m / 0.3).
For two independent measurements with central values x_1 and x_2 and Gaussian uncertainties sigma_1 and sigma_2, the tension is:
N_sigma = |x_1 - x_2| / sqrt(sigma_1^2 + sigma_2^2)
This assumes Gaussian, uncorrelated errors. Real cosmological measurements have correlated systematics in places (calibration cross-overs, shared distance anchors), but for headline tensions between cleanly independent probes the formula is the standard reporting convention.
The Early vs Late headline tension on each parameter uses the most-cited representative measurement from each camp: Planck 2018 for early-universe H_0, SH0ES R22 for late-universe H_0. The figure of 5 sigma quoted in news coverage refers exactly to this pair.
- Hubble tension (H_0, ~5 sigma) :: between Planck-CMB and the SH0ES Cepheid + SNIa distance ladder. The most statistically significant disagreement in cosmology. Proposals include early dark energy, modified neutrino physics, varying gravity, or unresolved systematic floors in one or both methods. JWST has confirmed the Cepheid distances, narrowing the systematic-error window.
- S_8 tension (~2.7 sigma) :: between Planck-CMB and ground-based weak lensing surveys (KiDS, DES, HSC). The late universe appears slightly less clumpy than the early universe predicts. Could indicate non-standard dark matter, evolving dark energy, or unmodeled baryonic feedback.
- w_0 - w_a evidence from DESI (~3 sigma) :: the first dataset where BAO + SNe + CMB shows mild preference for an evolving dark energy equation of state (w_0 > -1, w_a < 0) over a pure cosmological constant. Independent confirmation pending.
- The shaded vertical band in the centre is the early-universe combined range (1-sigma).
- The second band further right or left is the late-universe combined range. The late band appears only when at least one late-universe probe is in view. With the current dataset it is visible for H_0, S_8, and w_0; for Omega_m the entries are all early-universe or cluster, so only the early band shows.
- If the two bands overlap there is no tension. If they are disjoint and far apart, the tension is real and large.
- Dots inside their colour-class band agree with their camp; dots outside agree with the opposite camp and are the interesting outliers (for example JAGB sits with early universe rather than with the rest of the distance-ladder team).
- Parameter :: switches the entire dataset between H_0, S_8, Omega_m, w_0.
- Sort :: re-orders the forest plot rows. By value highlights the early-late split; by year shows how measurements have moved over time; by precision puts the smallest error bars at the top.
- Category chips :: tap (mobile) or click (desktop) to hide or show each probe class. The headline early-vs-late tension number above always uses the fixed representative pair regardless of filters.
- Combined tension (top-right corner of the plot) :: inverse-variance weighted mean of every currently-visible early-universe probe versus every currently-visible late-universe probe. Updates live as you tap the chips. The headline number above the plot is the fixed Planck-vs-SH0ES style pair; this combined number includes every probe in the active filter and so reflects the full visible dataset rather than two chosen anchors.
- Probes tab :: full details for every measurement, including method description, year, citation.
- The tension formula assumes Gaussian, uncorrelated uncertainties. Some pairs share calibration anchors (Gaia parallaxes, NGC 4258 distance) and are not strictly independent. Tensions among such pairs are slight over-estimates.
- Numbers are published central values from the cited papers. The literature evolves; this console snapshot reflects the state of constraints as of the build year stamped in the header.
- The headline early-vs-late tension uses the most-cited representative measurement from each camp. Alternative choices (for example DESI BAO + BBN versus CCHP TRGB) give different and generally lower tension values.
- Beyond-Lambda-CDM extensions can resolve some tensions while opening others. This console reports the tensions as measured within the standard model. Model space exploration belongs in a separate falsifiability module.