Asymptotics and Matched Expansions

An asymptotic formula is a statement about a limit, a domain, and a controlled remainder—not merely a formal series. This record identifies which scale-separation method fits a gravity calculation and what must be checked before its approximation is used outside the derivation.

Helpful background. Differential Equations for Relativity supplies endpoint and perturbation methods; PDEs and Green Functions for Gravitation supplies domains, boundary data, and Green operators.

Asymptotic scales and remainders

Let $\varepsilon\to0^+$ in a stated domain, with every other variable declared fixed. An asymptotic scale satisfies

$$\phi_{n+1}(\varepsilon)=o\!\left(\phi_n(\varepsilon)\right).$$

A useful finite-order statement is

$$f(\varepsilon) =\sum_{n=0}^{N-1}a_n\phi_n(\varepsilon)+R_N(\varepsilon), \qquad R_N=O\!\left(\phi_N\right).$$

The limit, real interval or complex sector, parameter range, norm, and uniformity are part of this statement. The infinite formal series need not converge, and its coefficients do not determine exponentially small terms without additional information. Poincaré, uniform, and generalized expansions are distinguished in NIST DLMF, §§ 2.1(iii)–(v).

Structure in the problem	Appropriate method	Required check
One regular scale	Regular perturbation	Substitute the truncation, bound the residual and remainder, and test uniformity on the full domain
A reduced equation loses data or changes balance	Inner and outer expansions	Find a nonempty overlap, match the same quantity, and build a uniform composite approximation
Slow modulation creates secular terms	Multiple scales or averaging	Derive the solvability condition and reconstruct the original variables without secular growth
A rapidly varying phase solves a differential equation	WKB or geometric optics	State branch and flux conventions; replace the approximation near turning points
A large parameter appears in an integral	Laplace, stationary phase, or steepest descent	Include endpoints and all relevant critical points; state contour, sheet, and Stokes sectors
A divergent series is to be continued	Named resummation prescription	State the prescription and ambiguities, then compare with independent analytic or numerical information

The integral, boundary-layer, WKB, and multiple-scale constructions are organized together in Bender and Orszag 1999, Chs. 6 and 9–11, pp. 247–316 and 419–576. For contour asymptotics and numerical use, NIST DLMF treats Laplace and stationary-phase methods in §§ 2.3(iii)–(iv), saddles and coalescing critical points in §§ 2.4(iv)–(vi), and remainder failure and Stokes changes in §§ 2.11(i), (iv).

Inner, outer, and common parts

Suppose an outer approximation is valid at fixed $x$ away from $x_0$, while an inner approximation holds at fixed

$$X=\frac{x-x_0}{\delta(\varepsilon)}, \qquad \delta(\varepsilon)\to0^+.$$

Matching requires a nonempty region such as

$$\delta(\varepsilon)\ll\lvert x-x_0\rvert\ll L,$$

where both approximations can be re-expanded in the same variables, normalization, and gauge. If their common expansion is $f_{\mathrm{common}}$, a standard composite form is

$$f_{\mathrm{comp}} =f_{\mathrm{outer}}+f_{\mathrm{inner}}-f_{\mathrm{common}}.$$

The subtraction prevents double counting. An arbitrary matching point may simplify algebra, but leading dependence on it signals inconsistent normalization, a missing term, or no true overlap. The residual and error should be tested in the outer region, inner region, and overlap separately, then bounded uniformly where the composite approximation is claimed.

In the small-body limit, body mass and radius shrink relative to an external curvature scale while compactness is held fixed; matching must preserve the buffer region, multipole normalization, gauge, and order-by-order remainder. That application is set out in Pound and Wardell 2024, § 5.1, pp. 40–42 and developed on Matched Expansions and the Small-Body Limit.

Uniformity and failure tests

Exchange consequential limits: small parameter, near-zone, late-time, frequency, extremality, and observation point. Different answers expose nonuniformity rather than a paradox. A vanishing leading coefficient, coincident powers, or a turning point can promote subleading terms or require logarithmic and uniform comparison functions. Stokes crossings can switch exponentially small contributions without changing the formal power series. Secular terms require a new slow variable; resummation supplies an additional definition, not a proof of the original series or its error.

Use Matched Asymptotic Expansions Across the Throat for a black-hole connection problem with noncommuting limits. Black-Hole Perturbation Theory supplies the physical normalizations. Asymptotic Analysis and Radiation Theorems supplies the geometric hypotheses and distinguishes a formal asymptotic hierarchy from a theorem controlling its remainder.

References

• Bender, Carl M., and Steven A. Orszag. Advanced Mathematical Methods for Scientists and Engineers I: Asymptotic Methods and Perturbation Theory. New York: Springer, 1999. DOI 10.1007/978-1-4757-3069-2.
• National Institute of Standards and Technology. 2026. Digital Library of Mathematical Functions, Chapter 2, “Asymptotic Approximations.” Release 1.2.7 of June 15, 2026. § 2.1, § 2.3, § 2.4, and § 2.11.
• Pound, Adam, and Barry Wardell. “Black Hole Perturbation Theory and Gravitational Self-Force.” In Handbook of Gravitational Wave Astronomy. Singapore: Springer, 2024. DOI 10.1007/978-981-15-4702-7_38-1.