BCL Parametrization in Heavy Meson Decays

• BCL parametrization is a rigorous, model-independent framework that maps the q² range via a conformal transformation to ensure rapid series convergence.
• It integrates pole factorization, kinematic constraints, and unitarity bounds to seamlessly combine lattice QCD, LCSR, and experimental data.
• The approach underpins global fits to extract CKM matrix elements and predict observables in semileptonic and rare heavy meson decay processes.

The Bourrely-Caprini-Lellouch (BCL) parametrization is a rigorous, model-independent framework for representing the $q^2$-dependence of hadronic form factors in heavy meson decays. Built upon analyticity, crossing symmetry, and QCD unitarity, the BCL formalism is now standard in flavor physics, underpinning global fits for semileptonic and rare $B\to\pi$, $B\to K^{(*)}$, $B\to D^{(*)}$, and $B\to A$ transitions as well as for pion electromagnetic and other form factors. The approach employs a conformal map $z(q^2)$ to compress the entire physical $q^2$ region into a small interval, ensuring rapid convergence of the truncated $z$-series. It enforces pole structure when resonant states are known and includes threshold and kinematic constraints to guarantee physical behavior. This methodology enables seamless interpolation across lattice QCD, light-cone sum rule (LCSR), and experimental data, with full uncertainty propagation and compatibility with unitarity bounds.

1. Mathematical Structure and Conformal Variable

The BCL parameterization maps the $q^2$-plane via a conformal transformation

$$z(q^2,t_0) = \frac{\sqrt{t_+ - q^2}\;-\;\sqrt{t_+ - t_0}}{\sqrt{t_+ - q^2}\;+\;\sqrt{t_+ - t_0}},$$

where $B\to\pi$0 (for two-meson transitions) and $B\to\pi$1 is a free parameter typically chosen to minimize $B\to\pi$2 over the semileptonic physical domain, e.g. $B\to\pi$3, with $B\to\pi$4. This mapping compresses the entire physical $B\to\pi$5 range into a domain with $B\to\pi$6 for $B\to\pi$7 and similar transitions, enhancing series convergence and numerical stability (Du et al., 2013, Gustafson et al., 2018, Li et al., 12 Dec 2025, Cui et al., 2022, Gao et al., 2024).

2. Series Expansion, Pole Structure, and Kinematic Constraints

Form factors are expressed in terms of $B\to\pi$8 as follows:

• Vector and tensor form factors ($B\to\pi$9, $B\to K^{(*)}$0):

$B\to K^{(*)}$1

where the explicit pole at $B\to K^{(*)}$2 (usually the lowest-lying vector resonance) is factored out to enforce correct analytic behavior.

• Scalar form factor ($B\to K^{(*)}$3):

$B\to K^{(*)}$4

since no nearby pole appears for $B\to K^{(*)}$5.

• Generalization to other transitions: For $B\to K^{(*)}$6, $B\to K^{(*)}$7, and $B\to K^{(*)}$8, analogous BCL forms include appropriate pole factors for each form factor species, and additional endpoint or kinematic constraints as dictated by QCD, e.g., $B\to K^{(*)}$9, $B\to D^{(*)}$0, or $B\to D^{(*)}$1 (Gao et al., 2024, Di et al., 18 Apr 2025, Cui et al., 2022).

The endpoint subtraction term $B\to D^{(*)}$2 ensures vanishing slopes or endpoint values when required by QCD or phase-space considerations (e.g., $B\to D^{(*)}$3 near $B\to D^{(*)}$4 for $B\to D^{(*)}$5), and the kinematic constraint $B\to D^{(*)}$6 is imposed by construction or as a parameter-reducing relation.

3. Unitarity and Analyticity Bounds

Analyticity, via the dispersion relation for the correlator of two weak currents, combined with QCD unitarity, leads to bounds on the BCL series coefficients: $B\to D^{(*)}$7 where $B\to D^{(*)}$8 is fixed by the QCD operator product expansion of the relevant two-point function. For some processes (e.g., pion electromagnetic form factor), the nonperturbative value of the susceptibility can be determined directly from lattice data. In practical fits, these bounds are often implemented as soft Gaussian priors or as hard constraints (Du et al., 2013, Simula et al., 2023, Cui et al., 2022, Gustafson et al., 2018). For scalar form factors without a nearby pole, the bound is weaker due to less constraining analytic input.

4. Global Fitting Methodology and Statistical Treatment

BCL coefficients are determined through global fits simultaneously incorporating:

• High-$B\to D^{(*)}$9 lattice QCD points (with full covariance propagation)
• Low-$B\to A$0 LCSR points or synthetic data (with correlated systematic errors)
• Experimental measurements (e.g., binned partial branching fractions)

For continuous curves (e.g., after chiral-continuum matching in lattice QCD), a functional $B\to A$1 is minimized: $B\to A$2 where $B\to A$3 is the covariance kernel. For discrete data, a $B\to A$4 sum over data points with the full covariance matrix is minimized, including theory–data cross-correlations and any model-dependent parameters (e.g., $B\to A$5-meson LCDA inverse moment $B\to A$6) (Li et al., 12 Dec 2025, Kang et al., 21 Dec 2025, Cui et al., 2022, Gao et al., 2024, Di et al., 18 Apr 2025).

The fits typically employ truncation orders $B\to A$7 or $B\to A$8, confirmed through convergence checks and stability tests as higher terms are included. Overparameterization is avoided, as seen in explicit studies where introducing superfluous parameters leads to non-Gaussian tails and poor predictive power in extrapolation regions (Gustafson et al., 2018, Simons et al., 2022).

5. Comparison to Alternative Parametrizations and Predictive Stability

BCL is systematically compared to the older Caprini-Lellouch-Neubert (CLN) and Boyd-Grinstein-Lebed (BGL) approaches:

• CLN: Small number of parameters, strong HQET/kinematic assumptions, prone to under-fitting and underestimated uncertainties outside the fit region.
• BGL: Outer function and multiple Blaschke factors, admits maximal generality but can cause substantial overfitting in constrained kinematic windows unless the dispersive bound is strictly imposed.
• BCL: Implements the minimal number of subtractions to enforce correct threshold or endpoint behavior, straightforward pole factor(s), and no outer function, yielding superior or comparable fits and extrapolation properties for $B\to A$9, $z(q^2)$0, and $z(q^2)$1 decays (Gustafson et al., 2018, Simons et al., 2022, Simula et al., 2023). For example, in direct predictive comparison, the two-parameter BCL yields consistently lower out-of-region $z(q^2)$2 metrics than the BGL in extrapolations from lattice to experiment.

6. Applications, Fit Results, and Phenomenological Impact

The BCL parametrization underpins the combination of lattice, LCSR, and experimental data in extracting CKM matrix elements, $z(q^2)$3-hadron light-cone distribution amplitude parameters, and predicting observables for SM and BSM phenomenology:

• $z(q^2)$4 extraction: Latest global fits employing the BCL expansion yield $z(q^2)$5 (Li et al., 12 Dec 2025) and $z(q^2)$6 (Kang et al., 21 Dec 2025), with full error propagation and systematics dominated by theory input at $z(q^2)$7.
• Form factor coefficients: Stable, correlated determinations of $z(q^2)$8 (or $z(q^2)$9) for each form factor; for $q^2$0 vector, values at $q^2$1 are $q^2$2, $q^2$3, $q^2$4 (Li et al., 12 Dec 2025).
• Rare process predictions: The BCL coefficients fully propagate to observables including differential and integrated branching ratios, lepton flavor universality ratios (e.g., $q^2$5), polarization fractions, and angular observables in $q^2$6, $q^2$7, $q^2$8 decays (Li et al., 12 Dec 2025, Li et al., 2014, Gao et al., 2024, Di et al., 18 Apr 2025).
• Pion charge radius: BCL fits to $q^2$9 data yield $z$0 fm, in excellent agreement with alternative dispersive approaches (Simula et al., 2023).

Channel	$z$1	$z$2	$z$3	$z$4	Reference
$z$5 vector	3	0.408	–0.509	–0.12	(Li et al., 12 Dec 2025)
$z$6 scalar	2	0.495	–1.411	–	(Li et al., 12 Dec 2025)
$z$7 $z$8	3	$z$9	$q^2$0	$q^2$1	(Gao et al., 2024)
$q^2$2	4+	1	$q^2$3	$q^2$4 …	(Simula et al., 2023)

The table lists representative BCL coefficients; precise pole masses and coefficient covariance matrices are given in the supplementary material of each cited work.

7. Current Best Practices and Theoretical Uncertainties

Best practices in BCL applications include:

• Truncation order $q^2$5 for most current lattice+LCSR+experiment fits, with crosschecks for stability under $q^2$6.
• Minimization of $q^2$7 by optimal choice of $q^2$8.
• Imposing unitarity/dispersive relations as loose priors unless strongly constraining.
• Simultaneous fitting to all available data (lattice, LCSR, experiment), with full covariance propagation, and explicit incorporation of key theory inputs (e.g., $q^2$9).
• Systematic uncertainty assessment via variation in input parameters, model parameters (e.g., LCSR logarithmic moments), and truncation order. Dominant uncertainties in $$z(q^2,t_0) = \frac{\sqrt{t_+ - q^2}\;-\;\sqrt{t_+ - t_0}}{\sqrt{t_+ - q^2}\;+\;\sqrt{t_+ - t_0}},$$0 typically arise from LCDA parameters and higher-twist LCSR corrections, not the BCL truncation itself (Li et al., 12 Dec 2025, Kang et al., 21 Dec 2025).

A plausible implication is that the BCL formalism, due to its rapid convergence, mild parameterization-induced uncertainties, and rigorous theoretical underpinnings, offers a robust and future-compatible standard for global fits in heavy flavor phenomenology across both semileptonic and rare decay channels.