[Paper Review] Boundary value problems for Dirac--type equations, with applications

Piotr T. Chruściel, Robert Bartnik|ArXiv.org|Jul 21, 2003

Spectral Theory in Mathematical Physics51 references19 citations

TL;DR

This paper establishes sharp regularity and Fredholm properties for first-order elliptic systems, including Dirac-type equations, with spectral and Lopatinski-Shapiro boundary conditions on compact and non-compact manifolds. It introduces weaker coefficient differentiability conditions than prior work, proves weak-to-strong regularity at the boundary via an H¹ identity and spectral theory, and derives solvability criteria using the adjoint problem, enabling applications to positive mass theorems in general relativity with asymptotically flat ends.

ABSTRACT

We prove regularity for a class of boundary value problems for first order elliptic systems, with boundary conditions determined by spectral decompositions, under coefficient differentiability conditions weaker than previously known. We establish Fredholm properties for Dirac-type equations with these boundary conditions. Our results include sharp solvability criteria, over both compact and non-compact manifolds; weighted Poincare and Schroedinger-Lichnerowicz inequalities provide asymptotic control in the non-compact case. One application yields existence of solutions for the Witten equation with a spectral boundary condition used by Herzlich in his proof of a geometric lower bound for the ADM mass of asymptotically flat 3-manifolds.

Motivation & Objective

• To establish existence and regularity for first-order elliptic systems with Dirac-type boundary conditions under weaker coefficient differentiability assumptions than previously known.
• To prove Fredholm properties and explicit solvability criteria for inhomogeneous boundary value problems on both compact and non-compact manifolds with compact boundary.
• To provide a framework for solving the Witten equation with spectral boundary conditions, relevant to geometric lower bounds on ADM mass in general relativity.
• To derive weighted Poincaré and Schr"odinger-Lichnerowicz inequalities to control behavior at infinity in the non-compact case.
• To unify pointwise (Lopatinski-Shapiro) and non-local (spectral projection) boundary conditions under a single analytical framework without requiring product metric structures on the boundary.

Proposed method

• Use of an H¹ identity (equation 5.15) as the central technical tool to establish boundary regularity for a broad class of first-order systems.
• Application of basic spectral theory to control the boundary operator, particularly through eigenfunctions of self-adjoint boundary operators.
• Reduction of the existence problem to a weak-strong regularity question via the Lax-Milgram lemma, showing that weak L² solutions of the adjoint problem are in fact strong H¹ solutions.
• Construction of a tubular neighborhood diffeomorphism using W^{k+1,p} partition of unity averaging to achieve coordinate normal forms with controlled metric decay near the boundary.
• Derivation of weighted Poincaré and Schr"odinger-Lichnerowicz inequalities in the non-compact case to ensure non-parabolicity at infinity and control asymptotic behavior.
• Use of the Schr"odinger-Lichnerowicz identity to derive H¹ estimates, which are essential for proving the required a priori inequalities in the non-compact setting.

Experimental results

Research questions

• RQ1 Under what minimal differentiability conditions on the coefficients and boundary metric can one establish regularity for Dirac-type equations with spectral boundary conditions?
• RQ2 What are the necessary and sufficient conditions for the solvability of inhomogeneous boundary value problems for first-order elliptic systems on compact and non-compact manifolds with compact boundary?
• RQ3 How can one prove weak-to-strong regularity for solutions at the boundary without assuming smooth coefficients or product-type metrics?
• RQ4 In what settings do weighted Poincaré and Schr"odinger-Lichnerowicz inequalities hold, and how do they ensure Fredholm properties in the non-compact case?
• RQ5 Can the spectral boundary condition used in Witten's proof of the positive mass theorem be rigorously justified within this framework?

Key findings

• The paper establishes boundary regularity for first-order elliptic systems with minimal coefficient differentiability, improving upon prior results that required smoother coefficients.
• It proves that the spectral boundary condition for the Dirac operator is elliptic in the sense of the authors' framework, validating its use in geometric analysis.
• The Fredholm property is established for both compact and non-compact manifolds with compact boundary, with solvability criteria involving solutions to the homogeneous adjoint problem.
• Weighted Poincaré inequalities are derived for manifolds with asymptotically flat or hyperbolic ends, ensuring control at infinity.
• The Schr"odinger-Lichnerowicz identity is used to derive H¹ estimates, which are essential for proving the weak-to-strong regularity result at the boundary.
• The framework successfully justifies the use of spectral boundary conditions in the Witten equation, leading to a proof of the positive mass theorem for asymptotically flat 3-manifolds.