CosmoVerse White Paper

• The paper presents CosmoVerse, a research program designed to resolve key cosmological tensions like the H₀ and S₈ discrepancies through coordinated observational and theoretical efforts.
• It outlines strategic scientific objectives and comprehensive mapping techniques that leverage next-generation surveys alongside advanced statistical-computational methodologies.
• The work introduces novel approaches such as hierarchical Bayesian modeling, likelihood-free inference, and Artificial Cosmogenesis to robustly test extensions to the ΛCDM framework.

CosmoVerse is a research program and conceptual framework focused on resolving leading observational and theoretical tensions in cosmology, notably the discrepancies between key measurements of cosmological parameters. The project unites next-generation surveys, advanced theory, and novel statistical-computational approaches to interrogate both the standard cosmological model and emerging extensions. It provides both strategic scientific objectives and methodological recommendations for the 2020s and beyond, addressing precision and robustness demands posed by the increasingly precise cosmological data landscape (Valentino et al., 2 Apr 2025).

1. Scientific Objectives and Core Projects

CosmoVerse identifies several principal goals for the coming decade:

• H₀ Tension: A primary objective is to resolve the 5–6σ discrepancy between local direct measurements of the Hubble constant ($H_0 \approx 73$ km/s/Mpc; e.g., Cepheid and SN Ia distance-ladder) and the value inferred from early-Universe CMB data ($H_0 \approx 67$ km/s/Mpc; e.g., Planck). Distinguishing systematic effects from new physics is central to this endeavor.
• S₈ Tension: CosmoVerse aims to clarify the 2–3σ lower amplitude of matter fluctuations ($S_8 \equiv \sigma_8 \sqrt{\Omega_m/0.3}$) inferred from weak lensing and galaxy clusters versus the CMB.
• Comprehensive Cosmological Mapping: Mapping the full expansion and structure growth history out to $z \approx 4$–6 to search for and characterize possible deviations from $\Lambda$CDM.
• Consistency and Anomalies: Examining the consistency of the standard model via new measurements of spatial curvature, the integrated Sachs–Wolfe (ISW) effect, and CMB anomalies.
• Data Science Innovation: Systematically developing and deploying advanced statistical, computational, and simulation-based techniques to maximize extraction of cosmological information and robust error estimation.

Supporting these aims, CosmoVerse coordinates several flagship projects and observational campaigns, including:

• CMB-S4, Simons Observatory, LiteBIRD (for temperature, polarization, lensing, $B$-mode science)
• DESI, Euclid, LSST, Roman (for BAO, RSD, lensing, and cluster abundances)
• Joint gravitational-wave standard-siren efforts (e.g., with LIGO/Virgo/KAGRA, LISA)
• JWST programs for local distance ladder systematics
• Refinement and extension of theoretical tools (Boltzmann codes, hydrodynamic simulations, analytical peak statistics) (Valentino et al., 2 Apr 2025).

2. Principal Cosmological Probes and Tensions

Type Ia Supernovae and Distance Ladder

Direct $H_0$ determinations using Cepheid-calibrated SN Ia (e.g., SH0ES: $H_0 = 73.17 \pm 0.86$ km/s/Mpc) persistently exceed the CMB-inferred value by $5$–$6\sigma$. Alternative calibrators (TRGB, Mira, masers, standard sirens) consistently support $H_0 \gtrsim 70$ km/s/Mpc. Systematic uncertainties in local SN are constrained to the sub-percent level.

Baryon Acoustic Oscillations

BAO measurements constrain cosmological distances relative to the sound horizon $r_s$. Combined with BBN, BAO yields $H_0 \approx 68.5 \pm 0.8$ km/s/Mpc. Tension with local measurements persists at $\sim2$–3$\sigma$ and is sensitive to assumed priors.

Cosmic Microwave Background Anisotropies

Planck 2018 provides $H_0 = 67.27 \pm 0.60$ km/s/Mpc, $S_8 = 0.834 \pm 0.016$. SPT-3G and ACT results closely agree. Multipole-level anomalies (Alens, large-angle correlations) are observed at $2$–3$\sigma$.

Weak Lensing and Large-Scale Structure

Tomographic shear surveys (KIDS-1000, DES Y3, HSC Y3) yield $S_8$ values of $0.76$–$0.79$, $2$–$3\sigma$ below Planck. Three-point and full-shape analysis reinforce this tension. RSD measurements of $f\sigma_8(z)$ are broadly compatible with Planck, with mild outliers.

Other Probes

Cluster counts, BAO+RSD, and gravitational-wave standard sirens (GW170817: $H_0 \approx 70 \pm 6$) consistently indicate persisting tensions. Novel methods (e.g., strong-lens time delays, new standard candles) and multi-wavelength cross-checks are now routine in the overall assessment (Valentino et al., 2 Apr 2025).

3. Systematic Uncertainties in Current Probes

Systematic risks are critically evaluated for every primary probe:

Probe	Dominant Systematics	Residual Uncertainty
SN Ia	Calibration, selection bias	$\lesssim 1$\% in $H_0$
BAO	Nonlinear, galaxy bias	$\lesssim 0.2$\% in $D_V/r_s$
CMB	Beam, foreground subtraction	$\lesssim 0.1$\% in spectra
Weak lensing	Shape measurement, photo-z	$\lesssim 0.5$\% in $S_8$
RSD/Clustering	Galaxy bias, velocity dispersion	$\lesssim 5$\% in $f\sigma_8$
Cluster counts	Mass calibration	$\lesssim 10$\% in cluster mass
Standard sirens	Calibration, inclination-redshift	$\lesssim 3$\% per event

Residual systematics are generally subdominant to the magnitude of observed tensions, particularly for SN Ia and weak lensing. This suggests that while systematics require ongoing mitigation and calibration, new physical mechanisms may be required to fully resolve statistical discrepancies (Valentino et al., 2 Apr 2025).

4. Theoretical Extensions Beyond ΛCDM

CosmoVerse explores a range of models for new physics motivated by persistent tensions:

• Extra Relativistic Species: $\Delta N_{\rm eff}$ alters pre-recombination expansion, potentially reconciling $H_0$ values. Current bounds ($N_{\rm eff}=2.99\pm0.17$) limit the effect.
• Early Dark Energy (EDE): Transient energy injection near matter-radiation equality (e.g., axion-like fields) increases $H_0$ by reducing $r_s$, but tends to exacerbate $S_8$ tension.
• Dynamical and Interacting Dark Energy: Models with time-varying $w(z)$, e.g., CPL parametrization, can cross the phantom divide and modify expansion at late times.
• Interacting Dark Sector: Dark matter–dark energy couplings (e.g., $Q = \Gamma\rho_{\rm DM}$) can relieve both tensions in certain regimes.
• Modified Gravity: $f(R)$, Horndeski, massive gravity, and teleparallel models test the consistency of general relativity on cosmological scales.
• Exotic Dark Matter: Decaying DM, ultralight axions, and models with varying constants are considered; only decays with $\tau \sim 10^3$ Gyr or ultra-light mass produce competing effects.
• Primordial Magnetic Fields, Power Spectrum Features, Local Inhomogeneities: Each is quantified for its efficacy in shifting cosmological inferences, typically insufficient to resolve the full anomalies (Valentino et al., 2 Apr 2025).

Key tests for these scenarios are specified for upcoming surveys, with forecasted errors cited (e.g., CMB-S4 $\sigma(N_{\rm eff}) \lesssim 0.02$, Euclid S8/H0 percent-level precision).

5. Advanced Data Analysis and Inference Approaches

CosmoVerse stresses that traditional likelihood and $\chi^2$ analyses are insufficient for the high-dimensional, correlated, and non-Gaussian structure of forthcoming datasets. Recommended methodological developments include:

• Hierarchical Bayesian modeling using hyper-parameters to combine diverse datasets and treat tension as a statistical quantity.
• Likelihood-free inference (e.g., Approximate Bayesian Computation, simulation-based inference) for complex models and forward modeling pipelines.
• Machine-learning surrogates (emulators) for rapid nonlinear predictions of $P(k)$, the bispectrum, and higher-order statistics.
• Gaussian processes for non-parametric expansion of $H(z)$ or $w(z)$ directly from data.
• Compression techniques (KL, MOPED) and multi-fidelity emulation to maximize computational efficiency and information yield.
• Active learning strategies to optimally sample uncertain parameter regimes.
• Deployment of deep learning for CMB foreground removal, auto-encoders for anomaly detection, and graph neural networks for large-scale structure likelihoods (Valentino et al., 2 Apr 2025).

6. Integration with Next-Generation Surveys and Future Prospects

Implementation of the CosmoVerse agenda is coordinated with global high-precision surveys, including:

• CMB-S4, LiteBIRD, PICO, and proposed CMB-HD for temperature, polarization, lensing, and spectral distortion science at the microkelvin level.
• LSST, Euclid, DESI, Roman for optical and NIR mapping, enabling dense tomographic weak lensing and comprehensive BAO/RSD studies.
• Gravitational wave observatories (Advanced LIGO/Virgo, LISA, Cosmic Explorer, Einstein Telescope) for standard-siren cosmology.
• JWST for local calibrator studies and host-galaxy systematics.

Planned developments include refinement of analysis pipelines, inter-survey calibration standards, advanced data-sharing infrastructure, and open simulation/modeling tools. The platform emphasizes collaborative cross-survey analysis, joint working groups, and robust systematics control.

A plausible implication is that by systematically combining heterogeneous, overlapping datasets, CosmoVerse will improve constraints not only on standard cosmological parameters, but also on the parameter space of new physics models subject to realistic survey systematics and statistical degeneracies (Valentino et al., 2 Apr 2025).

7. Theoretical Foundations: Artificial Cosmogenesis and Parameter-Space Exploration

A foundational component of CosmoVerse is the methodology of Artificial Cosmogenesis, in which systematic computer simulations of the space of possible universes are used to address two central cosmological challenges: the robustness of emergent complexity and the quantification of fine-tuning (Vidal, 2012).

• The space of possible universes ($M$) is defined by parameter vectors (e.g., fundamental constants, cosmological densities), dynamical laws, and initial conditions.
• The Cosmic Evolution Equation (CEE) generalizes the Drake Equation to cosmic outcomes (atoms, stars, life, intelligence) across $M$, operationalized with outcome detectors and simulation ensembles.
• Evaluation metrics include coverage of $M$, robustness and fine-tuning indices, computational efficiency, and open-endedness of complexity evolution.
• The architecture includes a universe generator, modular physics engines, statistical and agent-based modules for chemistry/biology, and advanced analysis frameworks for mapping outcome likelihoods and fitness landscapes.

This simulation-driven paradigm underscores CosmoVerse’s capacity not only to explore phenomenological tensions but to systematically interrogate the foundations of cosmological structure and its sensitivity to fundamental physical parameters (Vidal, 2012).