Leaver’s Continued Fraction Method in Black Holes

• Leaver’s continued fraction method is a computational technique that determines quasinormal mode frequencies by matching Frobenius series expansions at singular points in black hole perturbation theory.
• It generalizes to discontinuous potentials by incorporating Israel–Lanczos–Sen junction conditions, leading to higher-order recurrence relations and accurate spectral shifts.
• The algorithm’s high-precision approach reveals universal asymptotic features in both low-lying and high-overtone modes, providing actionable insights for gravitational wave analysis.

Leaver’s continued fraction method is a computational technique for determining quasinormal mode (QNM) spectra in black hole perturbation theory, particularly effective for high overtone modes and systems described by linear differential equations with singularities or discontinuities. This method, originally formulated for smooth potentials, has been generalized to cases in which the effective potential exhibits discontinuities, significantly expanding its applicability in black hole physics and related fields (Li et al., 6 Feb 2026).

1. Fundamental Principles of Leaver’s Continued Fraction Method

Leaver’s paradigm addresses equations of the form

$$\frac{d^2 \Psi}{dr_*^2} + [\omega^2 - V_\mathrm{RW}(r)]\Psi = 0,$$

where $V_\mathrm{RW}(r)$ is the Regge–Wheeler potential for Schwarzschild black holes. Quasinormal modes are defined by boundary conditions of purely ingoing waves at the horizon and purely outgoing waves at infinity. The method employs a Frobenius series expansion centered at a regular singular point—typically the event horizon—after factorization of asymptotic behaviors at the physical boundaries. Introducing the variable $z = (r-1)/r \in (0,1)$ with $r=1$ as the horizon and $r \to \infty$ mapping to $z \to 1$, the wavefunction is decomposed as

$$\Psi(r) = (r-1)^{-i\omega}e^{i\omega r} \sum_{n=0}^\infty a_n z^n.$$

This ansatz leads to a three-term recurrence for the expansion coefficients $\{a_n\}$: $$\alpha_1 a_1 + \beta_1 a_0 = 0,\qquad \alpha_n a_n + \beta_n a_{n-1} + \gamma_n a_{n-2} = 0,\,\, n \geq 2,$$ with explicit forms for $\alpha_n$, $\beta_n$, $\gamma_n$ in terms of $\omega$, $\ell$, and $s$, the perturbation parameters.

2. Continued Fraction Condition and Solution Procedure

The unique determination of quasinormal frequencies relies on the requirement that the Frobenius series is regular at both the horizon and spatial infinity. Equating the direct ratio $a_1/a_0$ obtained from the lowest-order equation with its expression from the downward three-term continued fraction expansion,

$$\frac{a_1}{a_0} = \frac{-\gamma_2}{\beta_2 - \dfrac{\alpha_2 \gamma_3}{\beta_3 - \dfrac{\alpha_3\gamma_4}{\beta_4-\cdots}}},$$

results in a transcendental equation for $\omega$ in continued fraction form: $$\beta_1 - \frac{\alpha_1\gamma_2}{\beta_2 - \dfrac{\alpha_2\gamma_3}{\beta_3-\cdots}} = 0.$$ Truncation at large $n$ and application of root-finding techniques (including Nollert’s inversion trick for damped modes) yield the QNM spectrum efficiently for both low-lying and high-overtone modes.

3. Generalization to Discontinuous Potentials

When the effective potential $V_\mathrm{eff}(r)$ has a discontinuity (e.g., a step at $r=r_d$), a single Frobenius expansion is inadequate because the wave equation is piecewise-defined and continuity or junction conditions must be enforced at $r_d$. Integration of the master equation across the discontinuity leads to the Israel–Lanczos–Sen junction condition,

$$\lim_{\epsilon\to0^+}\left[\frac{\Psi'(r_*^d+\epsilon)}{\Psi(r_*^d+\epsilon)} - \frac{\Psi'(r_*^d-\epsilon)}{\Psi(r_*^d-\epsilon)}\right] = \kappa,$$

where $\kappa$ encodes the integrated jump in $V_\mathrm{eff}$, and for a finite step, $\kappa = 0$ corresponding to a vanishing Wronskian across the discontinuity. Wavefunctions are therefore expanded separately on either side: $$\Psi(r) = \begin{cases} e^{-i\omega r_*^-} \sum_{n=0}^\infty a_n^- \big((r-r_d)/r_d\big)^n, & r \leq r_d \ e^{+i\omega r_*^+} \sum_{n=0}^\infty a_n^+ \big((r-r_d)/r_d\big)^n, & r > r_d \end{cases}$$ Each expansion leads to higher-order (five- or six-term) recurrences, directly reflecting the non-analyticity introduced by the discontinuity.

4. Implementation of the Modified Continued Fraction Algorithm

The modified procedure involves the calculation and matching of two separate continued fraction expansions for $a_1^-/a_0^-$ and $a_1^+/a_0^+$. These are substituted into the linear relation derived from the junction (Wronskian) condition. The transcendental equation for $\omega$ is thus encoded in a continued fraction whose coefficients are “tilded” to represent the structure from both regions and the matching constraint: $$\widetilde\beta_0 - \frac{\widetilde\alpha_0\widetilde\gamma_1}{\widetilde\beta_1 - \dfrac{\widetilde\alpha_1\widetilde\gamma_2}{\widetilde\beta_2 -\cdots}} = 0.$$ The solution algorithm consists of:

• Selecting a trial $\omega$ and evaluating both continued fractions to truncation depth $N$.
• Forming the Wronskian combination and varying $\omega$ (e.g., via Müller's method) until it vanishes.
• Employing Nollert’s inversion for high overtone ($n \gg 1$) convergence, where the $n$-th inversion yields the $n$-th root. High-precision arithmetic is required as $\operatorname{Im}\omega$ increases, with working precision $\gtrsim 2n$ digits, and depth $N \sim O(n)$.

5. Numerical Behavior and Spectral Features

For discontinuous potentials, such as a two-sided Regge–Wheeler step with differing horizon radii ($r_h^\pm$) across $r_c$, Leaver’s generalized method computes QNM spectra with up to $n \sim 2000$ modes at high precision (Li et al., 6 Feb 2026). Observed features include:

• Low-lying modes coincide (within $>6$ digits) with established matrix and Prony methods.
• High overtone modes are substantially shifted: their distribution lies along straight lines in the complex $\omega$ plane. The slope $\Delta\omega_I/\Delta\omega_R$ and spacing $\Delta\omega_R \simeq \pi/r_c$ match analytic expectations for echo-like, discontinuous potentials.
• The deformation of the asymptotic QNM spectrum by the discontinuity is largely independent of the detailed profile of the potential, suggesting universality in the resulting spectral instability. A plausible implication is that the cumulative contribution of high overtone QNMs, deformed by potential discontinuities, may have observable signatures in gravitational wave signals.

6. Significance and Future Directions

Leaver’s continued fraction method, extended via the incorporation of Israel–Lanczos–Sen conditions and local expansions about discontinuities, enables the systematic exploration of quasinormal spectra in non-smooth spacetime backgrounds. The algorithm’s high-precision and asymptotic coverage allow investigation of spectral instability phenomena in gravitational ringdown physics and black hole perturbation scenarios with potential “echoes” or sharp features. The observed agreement of QNM asymptotics across disparate discontinuous potentials, such as modified Regge–Wheeler and Pöschl–Teller forms, points to robust, generic structures in black hole spectral response (Li et al., 6 Feb 2026). Future work may further clarify observational consequences and potential universality classes of QNM spectra in modified gravity and astrophysical settings.