The total geodesic curvature and the $(2+1)$-dimensional hyperbolic mass

We consider a Jordan domain diffeomorphic to a closed two-dimensional disk with a smooth boundary. Assuming the Gauss curvature of the domain has a negative lower bound, the Gauss-Bonnet formula provides an upper bound for the total geodesic curvature of the boundary curve. This bound, however, inherently depends on the interior geometry of the region. In this paper, we derive an upper bound for the total geodesic curvature expressed solely in terms of the boundary data. Notably, the proof is connected to the positivity of the hyperbolic Hamiltonian mass in the (2+1)-dimensional gravity theory.

💡 Research Summary

The paper investigates a Jordan domain Ω diffeomorphic to a closed disk, equipped with a smooth Riemannian metric g whose Gaussian curvature satisfies the lower bound K_g ≥ −1. Classical Gauss–Bonnet yields the inequality
∫_Σ k ds = 2π − ∫_Ω K_g dV ≤ 2π + |Ω|_g,
where Σ = ∂Ω and k is the geodesic curvature with respect to the outward normal. This bound depends on the interior area |Ω|_g, which the authors aim to eliminate.

The key idea is to embed the boundary curve Σ isometrically as a round circle F(Σ) of radius r₀ in the standard hyperbolic plane H² with metric g_{−1}=dr²/(1+r²)+r² dϕ². The geodesic curvature of the reference circle is explicitly ˆk = 1/√(1+r₀²). The authors then compare the physical metric g with the hyperbolic reference via a conformal factor u², i.e. g = u² g_{−1}. Imposing scalar curvature R(g_u)=−2 leads to a nonlinear parabolic‑type PDE (2.4) for u(r,ϕ). Using the maximum principle and suitable barrier functions f₁(r), f₂(r), they prove that u stays between these barriers, and as r→∞ the quantity v_∞(ϕ)=lim_{r→∞} r²(u−1) exists as a smooth function on S¹.

With these preparations they define the Brown–York–Shi–Tam (BYST) quasi‑local mass for the boundary Σ: m_{BYST}(Σ)= (1/π)∫_0^{2π} (ˆk − k) √(1+r₀²) dϕ = (1/π)∫0^{2π} v∞(ϕ) dϕ. Thus the excess of the reference curvature over the physical curvature is interpreted as a mass.

To prove non‑negativity of this mass, the authors glue Ω to the exterior of the reference circle in H², forming a Lipschitz manifold ˜M = Ω ∪ (H² \ B_{r₀}) that retains scalar curvature −2 and is asymptotically hyperbolic. They introduce a Killing connection and Dirac operator, and construct a Killing‑Dirac harmonic spinor φ (̂D φ = 0) asymptotic to a Killing spinor at infinity. Applying a Witten‑type spinor identity yields the Hamiltonian mass H₀ = (1/2π)∫_{∂˜M} μ dϕ ≥ 0, where μ is the trace of the mass‑aspect tensor. A monotonicity formula shows that the quantity m(r)= (1/π)∫0^{2π} r(k₁−k_u)/√(1+r²) dϕ is non‑increasing in r and converges to H₀ as r→∞. Consequently m{BYST}(Σ) ≥ 0, and equality forces H₀=0, which implies that ˜M is the exact hyperbolic plane and Ω is a geodesic disk in H². This establishes the equivalence between the geometric inequality ∫_Σ k ds ≤ ∫_Σ ˆk ds = 2π √(1+r₀²) and the positive mass theorem in (2+1)‑dimensional hyperbolic gravity.

The final section studies the BYST mass for large ellipses Σ_R in the time‑symmetric slices of the BTZ black hole family g_m = dr²/(r²−m) + r² dϕ². For m>0 the Hamiltonian mass equals m+1; for m=−1 the metric reduces to the standard hyperbolic plane. Ellipses are parametrized by a deformation parameter ε, and the authors compute the asymptotic expansion of the physical curvature k and the reference curvature ˆk. They find that the limit m_∞(m,ε)=lim_{R→∞} m_{BYST}(Σ_R) depends on ε. When ε=0 (i.e., circles), m_∞ = m+1, reproducing the Hamiltonian mass. For ε≠0, extra angular terms appear, and the limit does not coincide with H₀, showing that the BYST mass only captures the global mass for round boundaries.

In summary, the paper provides a purely boundary‑based upper bound for the total geodesic curvature of negatively curved surfaces, interprets this bound as a quasi‑local mass, proves its non‑negativity via a spinor‑based positive mass theorem in (2+1)‑dimensional hyperbolic gravity, and explores the limitations of this quasi‑local mass on non‑circular boundaries.