PaperlayerView exact revision on arXiv

Source-linked AI summary

Picard Iteration for the Characteristic Initial Value Problem in Einstein Equations

Shengrong Wu

arXiv:2609.08421v1gr-qccs.NAmath.NAphysics.comp-ph

TL;DR

Direct characteristic evolutions with strongly nonspherical, freely prescribed initial geometry remain relatively scarce. The paper develops a geometry-ordered Picard iteration with independent residual and consistency checks; its experiments include exact benchmarks and numerical evidence for a trapped region and apparent horizon.

  • Problem

    Fully angular double-null evolutions with strongly nonspherical, freely prescribed initial geometry are relatively scarce among characteristic numerical relativity studies.

    Source: §1 · p. 2

  • Method

    A first-order Picard iteration for vacuum and Einstein–massless-scalar equations follows their geometric dependencies and combines characteristic constraint solves with independent residual and metric–connection checks.

    Source: §1 · pp. 2–3, §6 · p. 26

  • Results

    The experiments cover exact geometries and nonspherical characteristic data, and the final experiment reconstructs an apparent-horizon tube whose v = 0.04 graph changes by 1.50 × 10−7 under coordinate refinement.

    Source: §6 · p. 26

  • Takeaways & Limitations

    The numerical framework handles exact benchmarks and characteristic data with no explicit interior solution, including nonspherical low-regularity data and a numerical trapped-region reconstruction.

    Source: §6 · p. 26

    Residual reporting excludes endpoint values at singular coordinate Jacobians and treats unprotected near-endpoint and interface values as differentiation diagnostics, not interior accuracy statistics.

    Source: §3.2 · p. 12, §4.5 · p. 20

Abstract

from arXiv · show

We present an iteration algorithm for vacuum and Einstein scalar-field equations in double-null gauge, which transform the non-linear PDE into systems of ODE. The numerical realization combines characteristic constraint solves, LGL spectral elements, pole-free spherical operators, Galerkin projection, and independent first-order residual and consistency checks.

1 Introduction

The paper addresses the scarcity of fully angular double-null evolutions with strongly nonspherical, freely prescribed initial geometry. It develops a geometry-ordered Picard iteration and tests it on vacuum and Einstein–scalar problems, including a numerical apparent-horizon reconstruction.

1 Introduction

  • Fully angular double-null evolutions with strongly nonspherical, freely prescribed initial geometry remain relatively scarce, despite prior nonspherical numerical work.The characteristic formulation is natural when radiation, null focusing, and trapped-surface formation are central.Source: §1 · p. 2
  • The fourth experiment prescribes a substantial nonspherical initial-metric perturbation, with outgoing initial-hypersurface norms of 7.50960 in L2(S) and 2.53468 in L∞(S).The data are freely prescribed subject to characteristic constraints and corner compatibility.Source: §1 · p. 2
  • The first-order Picard iteration covers Einstein vacuum and Einstein–massless-scalar equations, with updates ordered by the characteristic equations’ geometric dependencies.It adapts the An and Wu construction [2] into an executable numerical map, using weighted variables, discrete characteristic traces, projection, relaxation, and residual audits.Source: §1 · pp. 2–3
  • The implementation combines characteristic constraint solves, null-coordinate spectral elements, pole-free spherical operators, angular projection, and independent residual and consistency checks.Formula checks, exact-solution tests, fresh derivatives of stored weighted fields, and separate metric–connection checks address executable-formula and iteration-design difficulties.Source: §1 · p. 3
  • The final experiment finds both future null expansions strictly negative on a section and reconstructs an 18-section apparent-horizon tube over 0.00441 ≤v ≤0.05.The reconstructed tube is tied to the chosen incoming-null foliation and does not prove either motivating theorem.Source: §1 · pp. 3–4

2 Equations and Theoretical Analysis

  • The equations section fixes double-null conventions and introduces the metric variables, connection components, and transport equations used in the numerical construction.The analytic background includes local-existence results for the characteristic problem.Source: §2 · p. 4, §2.1 · p. 4

2.2 Local existence

  • Local existence extends along portions of the initial null hypersurfaces where the required norms remain controlled; larger rectangles require corresponding a priori bounds throughout.The stated vacuum result follows the characteristic existence framework described in the section.Source: §2.2 · p. 6

2.3 Residual control

  • For asymmetric Einstein equations without explicit solutions, the paper measures approximation error using Ricci residues, with metric differences controlled by curvature-residue differences.The stated residual is ∥G −T∥L2(Su,v).Source: §2.3 · p. 6

We assume moreover

The proof uses bootstrap and energy estimates to control the iterates, with a scalar estimate illustrating the resulting small-data bound.

We assume moreover

  • For sufficiently small ε, a solution (g, b, Ω) to EVE exists on (0, 1) × (0, 1) × S² with the stated initial data.The smallness threshold is written as ε = ε(C0).Source: We assume moreover · p. 7
  • With B and D fixed independently of ε, the scalar bootstrap estimate gives O(ε) = o(ε^1/2) on a fixed finite region.Choosing B ≥ 4C0 and D ≥ 4C0 improves the bound to (B/2) exp(D(u + v))ε.Source: We assume moreover · p. 7
  • The energy estimate bounds the square of the full bootstrap norm by C(1 + B²/D) exp(2D(u + v))ε², then improves the bootstrap bound to (B/2) exp(D(u + v))ε.The proof sketch chooses B² ≥ 8C and then D ≥ 8C.Source: We assume moreover · p. 8

3 Picard Iteration Design and Numerical Strategies

The numerical Picard scheme preserves the geometric dependency order of the characteristic equations while adapting their transport equations for numerical integration. Its Einstein scalar-field setup prescribes data on intersecting null hypersurfaces and imposes trace requirements during iteration.

3 Picard Iteration Design and Numerical Strategies

  • The numerical scheme freezes nonlinear coefficients at the preceding iterate and solves transport equations in geometrically distinguished null directions, modifying equivalent forms for integration.It is an executable numerical adaptation of the approximation scheme in [2], not a literal transcription.Source: §3 · p. 8

3.1 Picard iteration for Einstein equations

  • For Einstein scalar-field initial data, g, b, and Ω are prescribed along H0; along H−1, Ωe4ϕ(−1,v) and Ωω(−1,v) compute ϕ and Ω, then g and Ωχ are prescribed by solving.Source: §3.1 · p. 9
  • At every iteration, the listed geometric and scalar quantities are required to retain their prescribed traces at v = 0.The trace list includes χ, tr χ, ω, and scalar-field derivatives.Source: §3.1 · p. 10

3. Construction of Ω(i+1), ∇A log Ω(i+1), and (Ωω)(i+1)

The passage notes an equivalent expression for constructing (Ωω)^(i+1).

3. Construction of Ω(i+1), ∇A log Ω(i+1), and (Ωω)(i+1)

  • The construction of (Ωω)^(i+1) is stated to be equivalent to another expression.The passage gives only a fragment of the displayed relation.Source: §3 · p. 11

4. Construction of ζ(i+1), b(i+1), and (Ωω)(i+1)

The update rules specify the field type for ζ, preserve a characteristic trace, use the updated shift for the incoming lapse, and restore traces after projection and relaxation.

4. Construction of ζ(i+1), b(i+1), and (Ωω)(i+1)

  • ζ^(i+1) is treated as a vector field rather than a 1-form in the update.Source: §4 · p. 11
  • The update keeps b^(i+1)A(u, 0) equal to the prescribed bA(u, 0) trace.Source: §4 · p. 11
  • The updated shift is used when constructing the incoming lapse coefficient.Source: §4 · p. 11
  • After angular projection and relaxation, the numerical implementation restores the prescribed traces.Source: §4 · p. 11

5. Construction of g(i+1) and (Ωχ)(i+1)

This section introduces the system used to define the next metric iterate.

5. Construction of g(i+1) and (Ωχ)(i+1)

  • The next metric iterate is introduced through a system involving g, tr χ, and bχ.The supplied passage names the system but does not specify its equations or update procedure.Source: §5 · p. 11

6. Construction of (Ωχ)(i+1) and (Ωe3ϕ)(i+1)

The section describes constructions for the next iterate, curvature-residue checks, and numerical discretizations in the null and angular coordinates.

6. Construction of (Ωχ)(i+1) and (Ωe3ϕ)(i+1)

  • The vacuum construction follows by removing the ϕ-terms from the Einstein scalar-field equation construction.The passage says the vacuum construction is therefore omitted.Source: §6 · p. 12

3.2 Curvature residues

  • Curvature residues are computed from intrinsic Lorentzian-manifold equations to avoid precision loss from second derivatives.The checks use first null derivatives of stored connection variables, and exclude endpoint values at singular coordinate Jacobians.Source: §3.2 · p. 12

3.3 Legendre–Gauss–Lobatto spectral elements

  • Legendre–Gauss–Lobatto elements represent fields with element polynomials whose shared endpoints and interface derivatives support local refinement.A coordinate change makes profiles proportional to v^δ linear in the computational coordinate s without regularizing the physical solution.Source: §3.3 · p. 13

3.4 Tensor spherical-harmonic Galerkin discretization

  • Pole-free angular discretization uses typed spherical harmonics and projects nonlinear right-hand sides from a larger work space into the retained space.This retained/work separation reduces angular aliasing while keeping the evolved state in a declared finite-dimensional geometric space.Source: §3.4 · pp. 13–14

4 Results for Einstein Vacuum Equations

The vacuum tests show accurate metric reproduction in regular Schwarzschild and Kerr domains and across the Schwarzschild horizon, but accuracy degrades near the Schwarzschild singularity. Nonspherical tests use independent first-order curvature residuals and consistency checks, with crossed shears producing singular limiting initial curvature.

4.1 Regular Schwarzschild and Kerr benchmarks

  • For rotating Kerr cases, mean section-metric error remains of order 10−6; Schwarzschild and zero-spin Kerr agree with the explicit metric essentially to floating-point accuracy.The larger rotating-case error is associated with numerical angular reconstruction of the Kerr optical map.Source: §4.1 · p. 15, Table 1 · p. 15

4.2 Schwarzschild horizon in static and Kruskal coordinates

  • Across the horizon, the Kruskal calculation has mean section-metric error 1.12 × 10−12, while static-coordinate metric reproduction remains at roundoff.Derivative-based curvature diagnostics are more sensitive to static lapse degeneration, a coordinate-conditioning effect rather than a curvature singularity at r = 2M.Source: §4.2 · p. 16, Table 2 · p. 17

4.3 Schwarzschild interior and approach to r = 0

  • Near r = 0, the iteration fails to settle within 20 sweeps from ϵ = 2−3, and the closest case has maximum relative radius error 16.43 and section-metric error 3.03 × 10².The stabilized finite output is not evidence of convergence at the Schwarzschild singularity.Source: §4.3 · p. 18

4.4 Strong outgoing short pulse and zero control

  • The strong outgoing pulse is a substantial metric perturbation, while the exact Minkowski control has an L∞ residual floor of 1.973 × 10−9.The contrast with the strong-pulse residual isolates error associated with evolving the nonspherical datum; independent metric–connection defects are 1.75 × 10−9 incoming and 2.77 × 10−7 outgoing.Source: §4.4 · pp. 18–19

4.5 Crossed characteristic shears

  • Crossed characteristic shears vanish continuously at the corner but produce unbounded limiting initial curvature in both null directions.The large unprotected-grid residuals reflect differentiation errors near fractional-power endpoints and element interfaces, so the protected-mask values are the interior statistic.Source: §4.5 · pp. 19–20

5 Results for Einstein Scalar-field Equations

The Einstein–scalar experiments span exact Fisher–JNW benchmarks, nonspherical characteristic data, and a scalar-pulse evolution that numerically identifies a trapped region and reconstructs an apparent-horizon tube.

5 Results for Einstein Scalar-field Equations

  • The experiments test exact Fisher–JNW metric errors, nonspherical data using independent curvature residuals, and a stronger scalar pulse using an overlapping inner atlas to locate trapping and reconstruct its horizon.Experiment 7 has no explicit reference spacetime, so accuracy is assessed through independently differentiated first-order curvature residuals.Source: §5 · pp. 20–21, §5.2 · p. 23

5.1 Fisher–Janis–Newman–Winicour benchmarks

  • The ν = 0.2 Fisher–JNW case retains a maximum absolute metric error below 5.7 × 10−10, although error increases as ν moves away from the Schwarzschild limit.The domain remains outside the curvature singularity, and the finest grid has 49 × 49 coordinate nodes.Source: §5.1 · p. 22

5.2 Nonspherical scalar characteristic data

  • Across three coordinate runs, the protected maximum residual decreases from 6.25 × 10−2 to 1.07 × 10−2; angular refinement lowers the protected Einstein and scalar-wave residual maxima sum from 3.99×10−2 to 9.79 × 10−3.The reported statistics use the protected region, while Figure 2 displays the finest-coordinate audit.Source: §5.2 · p. 23, Table 7 · p. 23, Figure 2 · p. 24

5.3 Scalar-pulse formation of a trapped region and apparent horizon

  • Both future null expansions are strictly negative on the entire reference sphere, and an independent angular control gives −0.0802016 and −4.27999, respectively.The initial faces contain no trapped section; trapped sections are sampled across several atlas patches.Source: §5.3 · pp. 24–25
  • Eighteen verified MOTSs form an apparent-horizon tube over 0.00441 ≤v ≤0.05, with coordinate refinement changing the v = 0.04 graph by 1.50 × 10−7 in maximum norm.The independent higher-angular-band evolution changes its mean location by 7.80×10−10; degree four is needed to resolve the nonspherical correction.Source: §5.3 · pp. 25–26

6 Conclusion and Outlook

The paper develops a first-order Picard iteration for vacuum and Einstein–massless-scalar equations in double-null gauge, then tests it on exact and nonspherical data. The experiments provide numerical evidence for a trapped region and reconstruct an apparent-horizon tube, while identifying reduced accuracy near curvature singularities as an area for further work.

Method

  • The method combines characteristic constraint solves, spectral elements in both null coordinates, pole-free angular operators, and independent first-order residual audits with metric–connection consistency checks.It provides a unified framework for exact benchmarks and characteristic data without an explicit interior solution.Source: §6 · p. 26

Results

  • Exact tests show small metric errors in regular domains and across a Schwarzschild horizon, while nonspherical scalar data show coordinate convergence on the protected region.The experiments also include vacuum perturbations and crossed characteristic data with singular limiting initial curvature.Source: §6 · p. 26
  • The final experiment provides numerical evidence for a trapped region, with both future null expansions strictly negative on the entire sphere at the identified section.The reconstructed horizon consists of 18 MOTSs, with independent coordinate and angular controls reported.Source: §6 · p. 26
  • Eighteen independently reconstructed MOTSs form a sampled apparent-horizon tube over 0.00441 ≤v ≤0.05, with coordinate refinement changing the v = 0.04 graph by 1.50 × 10−7 in maximum norm.An independent higher-angular-band evolution changes the mean location by 7.80×10−10.Source: §6 · p. 26

Limitations and outlook

  • Improved coordinates, adaptive refinement, and more robust iteration strategies are needed to improve approximation near curvature singularities.The outlook also calls for adaptive refinement toward the singular endpoint and studies of trapping thresholds under varied initial-data amplitudes.Source: §6 · p. 26
Loading 2609.08421v1…