Primordial Tensor Backgrounds

Primordial tensor backgrounds can arise from vacuum inflation, sourced inflation, phase transitions, defects, scalar perturbations, or nonstandard post-inflationary eras. Their spectra need not resemble one power law across the CMB, pulsar-timing, space-interferometer, and ground-interferometer bands. Through the evidence cutoff of 2026-08-20, CMB B modes and ground-based cross correlations provide model-specific upper limits, while pulsar timing provides strong evidence for a nanohertz background of unresolved origin. No signal has been identified as primordial. The observational null or detection in one band cannot be extrapolated across decades of frequency without a source and transfer model.

Helpful background. Inflationary Tensor Backgrounds derives primordial tensor production, transfer functions, and observable spectra. This record compares the dated status across bands and keeps source attribution separate from background detection.

Detection, attribution, and cross-band consistency

The finite proposition asks whether any statistically characterized stochastic tensor signal has been shown, after foreground, noise, and source comparison, to originate in the early Universe, and which primordial spectra are constrained in each observed band. A CMB limit on the tensor-to-scalar ratio $r$ at a pivot scale constrains a specified nearly scale-invariant family; it is not a bound on an arbitrarily blue or featured ground-band spectrum.

The answer depends on source mechanism, spectral tilt and features, reheating and thermal transfer, polarization and parity, anisotropy, foregrounds, direct-band likelihood versus integrated radiation bounds, and whether the claim concerns detection, compatibility, or attribution.

Assessment at 2026-08-20	Meaning
Question status	Open: PTA evidence for a background is strong, while primordial identity and a standard inflationary tensor signal remain undetected
Evidence status	Insufficient evidence for primordial attribution; supported for the stated model-specific upper limits
Confidence	Moderate at record level; high that no primordial identification is established, with conditional cross-band model translations
Domain	BK18 CMB likelihood, NANOGrav 15-year source comparisons, LVK O1–O4a isotropic and early-Universe searches, and explicit transfer models
What could change the conclusion	A new CMB polarization likelihood, PTA source-separation release, LVK stochastic update, or a cross-band model predicting independently observed spectra

A starting project: reconstruct one published tensor constraint

Choose this route if you can explain a CMB angular power spectrum, foreground marginalization, and a one-sided credible interval. Start with the linked inflationary-tensor background and then the NASA LAMBDA 2021 BK18 release, BK18_r_likelihood and BK18_Components. The former reproduces the black marginalized curve in the upper-left panel of BK-XIII Fig. 4; the latter provides the component decomposition in Fig. 16. Extract the parameter measure, foreground treatment, and plotted normalization before comparing any numbers.

Replot the tabulated curve, normalize it with the documented prior, and reconstruct its reported upper interval. Check sensitivity to interpolation and integration resolution. Compare the component products to understand which angular scales and frequency channels distinguish tensor power from dust and synchrotron. This is a reproduction of released inference products; it does not rerun map calibration or the multifrequency likelihood. The small tabulated files permit a laptop project. Changing foreground assumptions requires the full likelihood and advice from someone experienced with CMB polarization, not a reweighting of the one-dimensional curve.

Deliver the plotting and integration code, input checksums, and a precise account of what was and was not reproduced. A candidate extension is a forecast or likelihood study for one physically motivated spectral feature with a separately specified transfer history. Establish novelty and product suitability before committing: a nonstandard shape may invalidate the compressed likelihood.

Stop any direct cross-band extrapolation when the proposed source leaves the assumptions of the CMB template. Move to the appropriate native likelihood and forward model. The useful scientific question is which additional measurement could distinguish that fixed source from foregrounds; combining unrelated upper-limit curves cannot answer it.

What each observational window says

In the CMB window, the BICEP/Keck BK18 multifrequency analysis finds no primordial B-mode detection and supplies the latest published collaboration likelihood for standard $r$. Data, component separation, the tensor result, and checks are given in §§II–V and Fig. 4 BICEP/Keck Collaboration 2021, §§II–V. Dust, synchrotron, lensing, sky selection, and the assumed tensor spectrum remain part of the statement.

Later BICEP Array observations and delensing improve expected sensitivity, but the collaboration’s 2024 overview presents them as an observing program and forecast rather than a replacement detection likelihood BICEP/Keck Collaboration 2024, §§2–6. At the cutoff, BK18 remains the appropriate published product; projected performance through 2027 is not data.

In the nanohertz window, NANOGrav’s 15-year correlation analysis provides direct evidence for a Hellings–Downs-correlated gravitational-wave background Agazie et al. 2023, §§III–VI and Fig. 3. The separate new-physics paper compares supermassive-black-hole binaries and several primordial templates and explicitly stops short of primordial attribution Afzal et al. 2023, §§3–7. Source priors, pulsar red noise, ephemeris, and correlation modeling remain consequential.

In the ground-interferometer window, the current O1–O4a isotropic cross-correlation search finds no statistically significant background LVK 2025, §§II–VI. A companion early-Universe analysis translates the same cross spectra into bounds on selected inflationary, stiff-era, scalar-induced, defect, and parity-violating models LVK 2025, model sections and conclusion. Those translations share data and are correlated even when their theory templates differ.

Results and open questions at 2026-08-20

Claim or evidence	What it supports	What it does not establish
BK18 multifrequency B-mode likelihood	A strong upper limit on standard nearly scale-invariant primordial tensors	Arbitrary blue, featured, or small-scale spectra
Later BICEP/Keck observing overview	Improved future sensitivity and delensing program	A post-BK18 detection or released replacement likelihood
NANOGrav 15-year Hellings–Downs correlations	Direct evidence for a nanohertz gravitational-wave background	Primordial origin or one preferred source class
NANOGrav 15-year source-model fits	Compatibility of several astrophysical and primordial templates	Independent detection evidence or a preferred primordial mechanism
LVK O1–O4a isotropic null	Ground-band upper limits under stated spectra and isotropy	Absence of narrow or out-of-band primordial sources
LVK early-Universe translations	Constraints on selected physical spectra and thermal histories	A model-independent cross-band envelope
Broad production review	A common language for mechanisms and transfer functions	Current numerical likelihood values

The CMB and LVK nulls are genuine measurements of their bands; the PTA result is a direct evidence row for a Hellings–Downs-correlated nanohertz background. Attribution requires foreground discrimination, not majority voting across the rows. The separation in frequency means that physical consistency depends on a model’s full spectrum and expansion history.

Competing sources and cross-band models

The inflationary interpretation uses CMB-scale tensor production and a transfer function to predict present spectra. Standard slow-roll-like shapes are tightly constrained by B modes, but sourced, blue, broken, or reheating-dependent spectra can shift power into other bands. Phase transitions, defects, and scalar-induced waves have different shapes and polarizations. Caprini and Figueroa develop these mechanisms and the propagation framework in §§2–9 2018, §§2–9.

The astrophysical interpretation is strongest in the PTA band, where a population of supermassive-black-hole binaries naturally supplies a background but has uncertain demographics and spectral structure. Compact binaries form the leading ground-band foreground. CMB foregrounds are Galactic dust, synchrotron, and lensing. A convincing primordial claim must beat the relevant competitor in its own likelihood and predict a second observable or band before seeing it.

Counterexamples, null results, and failed routes

The PTA background is a counterexample to saying that all stochastic searches are null, yet it is also a counterexample to equating a detected background with primordial tensors. The LVK null rules out selected spectra that enter its band, not a narrow nanohertz source. BK18 rules out part of the standard-$r$ family, not an arbitrary spectrum enhanced far from the CMB pivot.

The failed route is to draw one power law through limits and detections across many decades without deriving its reheating, changes in relativistic degrees of freedom, source duration, and transfer history. Integrated BBN or CMB radiation bounds add further model dependence and cannot replace a direct-band likelihood.

What would change the assessment

A primordial attribution would become supported if foreground-controlled data preferred one early-Universe template over flexible astrophysical competitors and the same parameter posterior predicted an independently measured feature—another band, polarization, anisotropy, or non-Gaussian statistic. A new BICEP/Keck likelihood could update the CMB row; a PTA release with sharper spatial correlations and source separation could update attribution; an LVK or future space-interferometer detection could open a new direct band. Revised thermal transfer or foreground modeling can also materially change theory translations without changing the underlying data.

Research checklist

• Define the source mechanism, tensor normalization, spectral shape, polarization, and anisotropy.
• Specify the expansion and reheating history, relativistic degrees of freedom, and every transfer-function convention.
• Use the likelihood native to each band rather than a plotted envelope alone.
• Pin CMB maps, beams, bandpasses, foreground components, sky cuts, and delensing treatment.
• Pin PTA timing files, clocks, ephemeris, pulsar noise, spatial correlations, and astrophysical population priors.
• Pin interferometer calibration, cross spectra, stationarity cuts, correlated-noise treatment, and spectral template.
• Track shared data when comparing theory translations.
• Require an independently predicted discriminator before claiming primordial origin.

Source and data map

BK18 and the 2024 collaboration overview set the current CMB status. Agazie et al. provide the direct PTA background-evidence analysis, while Afzal et al. provide the source comparison. LIGO-P2500349-v9 supplies the O1–O4a isotropic result, LIGO-G2502630-v1 supplies its official cross-spectral data product, and LIGO-P2500150-v13 supplies correlated early-Universe translations. Caprini and Figueroa provide the cross-mechanism framework.

NASA LAMBDA BK18 products include bandpowers, bandpasses, component spectra, a tabulated marginalized $r$ likelihood, a CosmoMC package, and chains; the tabulated curve is not the full multifrequency likelihood. NANOGrav timing v1.0.1 is DOI 10.5281/zenodo.8104459. The official LIGO-G2502630-v1 data release provides O1, O2, O3, and O4a cross spectra, posterior samples, upper-limit arrays, power-law-integrated curves, and scripts for the isotropic result. No end-to-end rerun was performed. No separate durable model-likelihood environment for LIGO-P2500150-v13 was verified, and no later published BICEP/Keck tensor likelihood than BK18 was found.

References

• Afzal, A., et al. 2023, “The NANOGrav 15-year Data Set: Search for Signals from New Physics,” Astrophysical Journal Letters 951, L11, DOI 10.3847/2041-8213/acdc91.
• Agazie, G., et al. 2023, “The NANOGrav 15-year Data Set: Evidence for a Gravitational-Wave Background,” Astrophysical Journal Letters 951, L8, DOI 10.3847/2041-8213/acdac6.
• BICEP/Keck Collaboration. 2021, “Improved Constraints on Primordial Gravitational Waves Using Planck, WMAP, and BICEP/Keck Observations through the 2018 Observing Season,” Physical Review Letters 127, 151301, DOI 10.1103/PhysRevLett.127.151301.
• BICEP/Keck Collaboration. 2024, “Constraining Inflation with the BICEP/Keck CMB Polarization Experiments,” arXiv:2405.19469v2, arXiv.
• Caprini, C., and Figueroa, D. G. 2018, “Cosmological Backgrounds of Gravitational Waves,” Classical and Quantum Gravity 35, 163001, DOI 10.1088/1361-6382/aac608.
• LIGO Scientific Collaboration, Virgo Collaboration, and KAGRA Collaboration. 2025, “Upper Limits on the Isotropic Gravitational-Wave Background from the First Part of LIGO, Virgo, and KAGRA’s Fourth Observing Run,” LIGO-P2500349-v9, arXiv:2508.20721v1, official record.
• LIGO Scientific Collaboration, Virgo Collaboration, and KAGRA Collaboration. 2025, “Data for Upper Limits on the Isotropic Gravitational-Wave Background from the First Part of LIGO, Virgo, and KAGRA’s Fourth Observing Run,” LIGO-G2502630-v1, official data release.
• LIGO Scientific Collaboration, Virgo Collaboration, and KAGRA Collaboration. 2025, “Cosmological and High Energy Physics Implications from Gravitational-Wave Background Searches in LIGO–Virgo–KAGRA’s O1–O4a Runs,” LIGO-P2500150-v13, arXiv:2510.26848v2, official record.
• NASA LAMBDA. 2021, “BICEP/Keck, October 2021 Release Data Products (BK18).” Bandpowers, marginalized tensor likelihood, component spectra, likelihood code, and chains. DOI 10.71662/5etq-eh39; release and product descriptions.