What precision threshold determines which scattering regime requires the detailed model versus fast approximation?

Your question asks about a quantitative threshold—a precision level that determines when to switch from a fast approximation to a detailed model in scattering calculations. This is a practical computational question, and the preprints touch on related issues, but they don't provide a single universal threshold.

Here's what the archive does address:

Hybrid switching strategies: The most directly relevant work describes a tunable parameter that controls the switch. In the hybrid Monte Carlo/deterministic transport method [AG-2025.11-956], a parameter *N_s* governs how many scattering events are handled in the Monte Carlo leg before handing off to a deterministic solver; the choice of *N_s* reflects a trade-off between accuracy and computational cost, though no universal precision threshold is specified.

Multi-regime frameworks: Several papers handle scattering across multiple regimes—near-field, mid-field (Fresnel), and far-field—rather than switching at a single threshold. The PSTD method [AG-2024.03-1516] computes the full wave function including all regions and validates accuracy against the partial wave method, but again, no specific precision cutoff is given.

Regime transitions in physics: In black hole scattering [AG-2024.06-217], the scattering angle diverges near a critical impact parameter, and self-force corrections become singular; resummation techniques improve agreement with numerical results in the strong-field regime. For photon transport [AG-2024.05-347], the PDF transitions from free-streaming to diffusion, but the switching criterion is physical (Lorentz factor of the flow) rather than a precision threshold.

Bottom line: The archive doesn't provide a canonical precision threshold (e.g., "switch when error exceeds 1%") that applies across scattering problems—the choice depends on the physics, geometry, and computational budget of each problem.