Analysis for Mathematical Relativity

Analysis enters mathematical relativity through precise choices of space, topology, operator, and boundary data. This page maps those choices to the compactness, weak-existence, and regularity statements they permit; the diagnostic and worked preparation are in Analysis for Mathematical Relativity.

Helpful background. Calculus and Linear Algebra for Relativity supplies norms, duals, and integration by parts. PDEs and Green Functions for Gravitation supplies operator classification, data, and causal support.

Spaces, measures, and topology

A symbol such as $H^s$ is incomplete until its setting is named. State the domain or manifold, measure, scalar or tensor bundle, local or global character, homogeneous or inhomogeneous norm, and boundary trace. On a Riemannian manifold, also state the connection and volume form used in covariant derivatives and integration.

Object	What it controls	Data that cannot be omitted
$L^p(M,d\mu)$	size up to equality almost everywhere	measure, exponent, scalar or bundle norm
$W^{k,p}$ or $H^s$	weak derivatives or fractional regularity	domain, derivative definition, local/global choice, boundary trace
$\mathcal D'(\Omega)$	continuous linear action on smooth compactly supported tests	test space and its topology
finite measure	mass on measurable sets and weak-* limits	sigma algebra, sign or positivity, total-variation control

Distributions illustrate why the topology is part of the object: continuity is defined relative to convergence in $C_c^\infty(\Omega)$. Differentiation is always available by duality, but arbitrary nonlinear products are not. The definitions, support properties, and product limitations are developed in Hörmander 2003, Chs. 2–7, printed pp. 5–157. For the operational identities used with singular sources, see Distributions and Generalized Functions.

Weak convergence and compactness

For a Banach space $X$,

$$u_j\rightharpoonup u \quad\Longleftrightarrow\quad \ell(u_j)\longrightarrow\ell(u) \quad\text{for every }\ell\in X^*.$$

Strong convergence implies weak convergence, but the converse fails. A bounded sequence in a reflexive space has a weakly convergent subsequence; boundedness alone does not give a strongly convergent subsequence. These distinctions and the Hilbert-space example of an orthonormal sequence are recorded in Hunter, Def. 1.17 and Thm. 1.19, printed pp. 7–8.

Strong compactness requires an additional mechanism. For example, if $\Omega\subset\mathbb R^n$ is bounded and open, $1\le p\lt n$, and $1\le q\lt p^\ast=np/(n-p)$, bounded subsets of $W_0^{1,p}(\Omega)$ are precompact in $L^q$; see Hunter, §§ 3.7–3.10, printed pp. 59–75, especially Thm. 3.45. Translation on an unbounded domain and concentration at the critical exponent show why the domain and strict exponent range matter.

Weak limits pass through continuous linear functionals. They need not pass through products, norms, curvature terms, or other nonlinear maps. A limit argument must therefore supply strong compactness, a compensated structure, or an explicit defect measure.

Weak existence and elliptic regularity

Let $H$ be a Hilbert space and $a:H\times H\to\mathbb R$ a bilinear form. If

$$|a(u,v)|\le C\lVert u\rVert_H\lVert v\rVert_H, \qquad a(u,u)\ge c\lVert u\rVert_H^2 \quad(c\gt0),$$

then every $f\in H^*$ determines a unique $u\in H$ with $a(u,v)=f(v)$ for all $v\in H$. The continuity and coercivity hypotheses, not the differential expression alone, produce the conclusion; see Hunter, § 4.7, printed pp. 103–105, especially Thm. 4.20.

Boundary conditions are encoded in the trial and test spaces. If coercivity fails because the operator has a kernel, existence needs a compatibility condition and uniqueness requires a normalization or quotient. For example, the Laplacian on a closed connected manifold has constants in its kernel, so a source must have zero mean before inversion.

Regularity is a separate step. For $L=-\partial_i(a^{ij}\partial_j\,\cdot)$ with real symmetric uniformly elliptic $a^{ij}\in C^1(\Omega)$, a weak solution $u\in H^1(\Omega)$ with $f=Lu\in L^2(\Omega)$ lies in $H^2(\Omega')$ for every $\Omega'\Subset\Omega$. Global $H^{k+2}(\Omega)$ regularity for the homogeneous Dirichlet problem additionally requires $\Omega$ bounded with $C^{k+2}$ boundary, $a^{ij}\in C^{k+1}(\overline\Omega)$, $f\in H^k(\Omega)$, and $u\in H_0^1(\Omega)$. These interior and boundary statements are Hunter, §§ 4.11–4.12, printed pp. 110–116, especially Thms. 4.27 and 4.31. Corners, rough coefficients, incompatible traces, or a nontrivial kernel can invalidate a formally plausible gain.

Uses in mathematical relativity

Ringström develops the Sobolev, hyperbolic-energy, and geometric setup for the Einstein Cauchy problem in Ringström 2009, Chs. 3–9, printed pp. 21–91. The application still has to name its gauge, data class, differentiability, matter model, and spacetime domain.

Question	Detailed treatment	Check that must remain visible
Which regularity closes Einstein evolution?	Sobolev Thresholds for Einstein Evolution	gauge, norm, nonlinear products, and theorem scope
What is the weak equation?	Weak Formulations of Einstein’s Equations	trial/test spaces, integration by parts, constraints, and boundary terms
What survives a limiting sequence?	Compactness and Limits of Spacetimes	topology, gauge, noncollapse, and defect terms
When is rough curvature defined?	Distributional Curvature	admissible metric class and every nonlinear product

Best-known thresholds and open theorem frontiers belong in those dated treatments. This lookup supplies the analytic questions that must be answered before such a result can be used.

References

• Hörmander, Lars. The Analysis of Linear Partial Differential Operators I: Distribution Theory and Fourier Analysis. 2nd ed. Berlin: Springer, 2003. DOI.
• Hunter, John K. Notes on Partial Differential Equations. Revised June 18, 2014. University of California, Davis. Official PDF.
• Ringström, Hans. The Cauchy Problem in General Relativity. Zürich: European Mathematical Society, 2009. Publisher record.