Joint neutrino oscillation analysis from the T2K and NOvA experiments

This paper presents a complete and well-supported joint analysis of neutrino oscillation data from the T2K and NOvA experiments. The argument is fully developed, with detailed explanations of the methodology, including flux modeling, detector correlations, and statistical inference using MCMC frameworks. All key elements, such as variables, assumptions, and limitations, are explicitly stated, and the analysis addresses edge cases like mass ordering ambiguities and systematic uncertainties through rigorous tests. The work achieves its goals of deriving precise constraints on parameters like Δm²_32 and δ_CP within the three-flavor paradigm, demonstrating strong internal logic and mathematical rigor without skipping steps.

Mathematically, the manuscript is largely consistent in its use of standard neutrino-oscillation parameter definitions and in the limited explicit formulae it presents (oscillation phase and J_CP). The Bayesian logic of combining two likelihoods with priors, exploring via MCMC, and summarizing via HPD credible intervals and Bayes factors is coherent and internally consistent. The main rigor limitation is that many essential statistical/mathematical objects are only described qualitatively rather than specified (explicit likelihood, prior over discrete ordering, precise definition of alternative priors like “flat in sinδ_CP,” and the posterior-predictive p-value discrepancy). As a result, while no clear mathematical errors are visible in the excerpted equations/definitions, several headline quantitative outputs are not verifiable from the manuscript text alone, and some prior/model-index subtleties could materially affect Bayes factors unless explicitly defined.

This is a mathematically rigorous analysis that successfully combines data from two major neutrino experiments within a unified Bayesian framework. The statistical methodology is sound, with proper treatment of parameter estimation, hypothesis testing, and uncertainty quantification. The internal logic is consistent throughout, from the theoretical foundations through the statistical analysis to the final results. The authors demonstrate careful attention to potential systematic biases and correlations between experiments, with extensive validation studies supporting their methodological choices. While some mathematical details are omitted (likely due to space constraints and reliance on established formalism), the presented mathematics is valid and the logical flow from assumptions to conclusions is coherent.

Within its stated framework, this is a logically coherent Bayesian joint-analysis paper. The central inferential claims—improved precision on |Δm^2_32|, weak ordering preference, octant preference, and conditional evidence for CP violation under inverted ordering—follow consistently from the reported posterior summaries and Bayes factors. I do not find internal contradictions in how the paper moves between marginalized and ordering-conditional statements, nor in how it interprets the combined T2K and NOvA information. Mathematically, the work is credible but not fully transparent. The submission relies heavily on external analyses and software implementations for the detailed likelihood construction, detector/systematic parameterizations, and numerical inference. As a result, the paper supports the conclusions at a high level but does not provide enough explicit equations or derivational detail for a full independent mathematical audit. The main vulnerability is not an evident error, but under-specification of the formal statistical machinery underlying the final numbers.

This paper presents a methodologically sound and logically robust joint analysis of data from the T2K and NOvA experiments. From a mathematical and logical perspective, the work is of the highest quality. The statistical framework is standard for the field and is implemented with an exceptional degree of rigor, most notably through the cross-verification of results using two independent MCMC fitting packages. The authors directly confront the most significant logical challenge in combining these datasets: the treatment of correlated systematic uncertainties. Their approach to this challenge is exemplary. They provide clear physical reasoning for the independence of most uncertainties and perform a series of sophisticated tests to justify treating the complex cross-section uncertainties as uncorrelated for this analysis. These tests, including the 'nightmare' parameter study, convincingly argue that any neglected correlations are subdominant to the current statistical uncertainties and do not bias the main conclusions. The resulting parameter constraints are therefore derived from a logically consistent and mathematically valid framework.

This is an exceptionally complete scientific analysis that represents a landmark achievement in neutrino oscillation physics. The joint analysis of T2K and NOvA data addresses one of the most technically challenging aspects of combining results from different experiments - the proper treatment of systematic uncertainties and potential correlations. The authors provide exhaustive studies of flux correlations, detector modeling, cross-section uncertainties, and develop 'nightmare' scenarios to test robustness. The statistical framework is rigorous, employing proper Bayesian methods with comprehensive goodness-of-fit validation. All experimental details, parameter definitions, prior choices, and analysis limitations are clearly presented. The work fully delivers on its stated objectives of providing improved precision on oscillation parameters and testing for CP violation, while honestly acknowledging areas where definitive conclusions cannot yet be drawn (such as mass ordering determination). The level of methodological rigor and transparency sets a high standard for collaborative analyses in particle physics.

This is a very complete experimental paper. It clearly states its objective, describes the combined Bayesian analysis framework, defines the oscillation parameters under study, and gives enough methodological structure to understand how the T2K and NOvA likelihoods are merged and interpreted. Importantly, the paper does not stop at presenting best-fit values and intervals; it systematically addresses the key methodological risk in such a combination—shared or mismodeled systematics—through dedicated correlation studies, 'nightmare' scenarios, out-of-model tests, and posterior-predictive goodness-of-fit checks. The conclusions are correspondingly well supported within the paper’s assumptions. Claims are carefully phrased, especially around CP violation and mass ordering, and the manuscript distinguishes robust findings (improved precision on |Δm^2_32|, no strong mass-ordering preference) from conditional ones (evidence for leptonic CP violation if inverted ordering is assumed true). The main weakness is not logical incompleteness but limited external reproducibility due to non-public code and reliance on prior collaboration analyses for some implementation specifics. Overall, however, the submission is exceptionally complete by the standards of a large-collaboration experimental paper.

This paper represents a highly complete and well-supported experimental analysis. Its primary strength lies in the meticulous and transparent handling of systematic uncertainties, a critical challenge for any joint analysis between distinct experiments. The authors justify their approach with dedicated, non-trivial studies, clearly state their assumptions and priors, and rigorously cross-check their results using two independent fitting frameworks. The work successfully achieves its stated goal of combining data from the T2K and NOvA experiments to produce world-leading constraints on neutrino oscillation parameters. The reporting is comprehensive, with a detailed methods section and extensive supplementary material that address all key aspects of the analysis, including goodness-of-fit and robustness checks. From a completeness perspective, the paper is exemplary.

This landmark paper presents the first joint analysis of data from the T2K and NOvA neutrino oscillation experiments, demonstrating how complementary experimental designs can be leveraged to achieve unprecedented precision in fundamental physics measurements. The work achieves the world's best precision on the neutrino mass-squared splitting |∆m²₃₂| and provides tantalizing evidence for CP violation in the lepton sector, contingent on the inverted mass ordering being correct. Beyond the physics results, the paper introduces valuable methodological innovations for handling systematic uncertainties across different experimental frameworks. The containerization approach for integrating disparate analysis software and the comprehensive correlation studies establish new standards for collaborative analyses in particle physics. While some modeling assumptions merit continued scrutiny, the rigorous validation studies and transparent documentation of limitations make this an exemplary contribution to experimental neutrino physics.

This is a strong experimental-analysis submission whose main scientific value lies in combining two complementary long-baseline neutrino datasets in a genuinely joint inference framework. The work is highly testable because it produces explicit numerical constraints on oscillation parameters and a conditional claim about CP violation that future data can directly confirm or refute. The paper is also commendably careful in validating its own statistical machinery through cross-framework checks, goodness-of-fit tests, pseudo-data studies, and robustness tests against discrete model variations. Its originality is substantial at the level of experimental methodology and collaboration-scale synthesis, even though it does not propose a new oscillation theory. Communication is mostly strong, especially in explaining why the two experiments complement each other, though some of the systematic-correlation logic and the interpretation of conditional Bayesian claims could be made even more explicit. Overall, this is a scientifically solid and well-communicated joint analysis with clear empirical content and meaningful novelty in execution.

The atmospheric mass-squared splitting is |\Delta m^2_{32}| = 2.43^{+0.04}_{-0.03}\times10^{-3} eV^2 (normal ordering) and |\Delta m^2_{32}| = 2.48^{+0.03}_{-0.04}\times10^{-3} eV^2 (inverted ordering).

Falsifiable if: A future independent high-precision measurement (e.g., JUNO, DUNE, Hyper-K, or upgraded reactor/long-baseline experiments) returns a value of |\Delta m^2_{32}| outside these 1σ (or specified credible) intervals with significance >3σ.

The CP phase δ_CP is confined to a 1σ credible interval of [-0.81π, -0.26π] (marginalized over ordering) and to 3σ intervals of [-1.38π, 0.30π] (normal ordering) and [-0.92π, -0.04π] (inverted ordering).

Falsifiable if: Future measurements of δ_CP exclude these credible intervals (i.e., observed δ_CP values fall outside the stated 3σ ranges) or provide a credible interval that does not overlap the currently reported ranges at >3σ.

If the inverted mass ordering is the true ordering, the joint data exclude CP-conserving values (δ_CP = 0, π) at the 3σ level, providing evidence of leptonic CP violation under that ordering assumption.

Falsifiable if: Either (a) future experiments determine the mass ordering to be inverted but measure δ_CP consistent with 0 or π within the 3σ credible region, or (b) future data establish the normal ordering with high significance, removing the inverted-ordering interpretation and thus the stated evidence.

There is no strong global preference for normal versus inverted mass ordering in the combined data (Bayes factor ~1.3 with the reactor θ13 constraint).

Falsifiable if: Subsequent data produce a Bayes factor or other ordering-discriminating test statistic that significantly favors one ordering (e.g., Bayes factor > 10 or a frequentist significance >3σ), thereby resolving the ordering.

The fit weakly prefers the upper octant for θ23 (sin^2θ23 > 0.5) with a Bayes factor ≈ 3.5 when the reactor constraint is applied.

Falsifiable if: Future precise measurements constrain sin^2θ23 to the lower octant (sin^2θ23 < 0.5) with significance exceeding the present Bayes-factor indication (e.g., >3σ), or provide Bayes factors strongly favoring the lower octant.