Minimal networks on S^2

Abstract: The minimal network problem is a classical topic in geometric measure theory and the calculus of variations, which aims to find networks of minimal length connecting given points. Most classical results are established in the Euclidean plane, while a complete theory for constant-curvature Riemannian manifolds remains to be developed. In this paper, we locally extend the theory of minimal networks and the calibration method from the Euclidean plane to the standard unit sphere (S^2). We redefine (\mathbb{R}^2)-valued co-vectors, differential forms, currents, and calibrations adapted to spherical geometry. Using exponential maps and local metric perturbation estimates, we prove that spherical minimal networks composed of great-circle arcs with (120^\circ) triple junctions are \textbf{locally length-minimizing only within sufficiently small geodesic balls} on (S^2), without obtaining global minimality results. Our work partially enriches the theory of minimal networks on constant-curvature spaces, and provides a theoretical reference and technical basis for future research on extending such results to higher-dimensional Riemannian manifolds and more general surfaces.

• The paper demonstrates that a Steiner network on S^2 with triple junctions at 120° is locally length-minimizing via adapted calibration methods.
• It extends classical tools like 1-rectifiable currents and calibration forms from the Euclidean setting to small convex spherical caps using local geometric analysis.
• The approach provides a rigorous framework for future studies on minimal networks in curved spaces, establishing explicit bounds for cap radii and curvature conditions.

Minimal Networks on Spheres: Local Theory and Calibration

Background and Theoretical Context

The study addresses the extension of minimal network theory—traditionally well-understood in the Euclidean setting—to spaces of constant positive curvature, specifically the standard unit sphere $S^2$. Minimal networks, particularly those conforming to triple-junction (Steiner) configurations with $120^\circ$ angles, are classical objects in the calculus of variations and geometric measure theory. In the plane, their structure and optimality are established via tools such as 1-rectifiable currents, geometric measure theory, and the calibration method. However, a complete analog on $S^2$ or general Riemannian manifolds remains incomplete, primarily due to the challenges of curvature, nontrivial topology, and lack of globally defined affine structures.

Extension of Minimal Network Theory to $S^2$

The paper formally generalizes the theoretical machinery from the Euclidean case to the sphere. It achieves this by first reconstructing the notions of co-vectors, differential forms, and currents, which are foundational to the functional-analytic approach to minimality, but now adapted to each spherical tangent space $T_p S^2$ for $p \in S^2$. All analysis is performed locally, exploiting the exponential map and normal coordinates to transfer geometric and variational structural results from $\mathbb{R}^2$ to appropriate regions of $S^2$.

A precise reformulation is given for:

• 1-rectifiable currents on $S^2$: Defined via integration of $\mathbb{R}^2$-valued 1-forms along rectifiable sets embedded in the sphere.
• Calibration forms: Now constructed as closed, comass-bounded 1-forms on $120^\circ$0, consistent on patches corresponding to each network edge, and globally patched compatibly at the triple junction using geometric identities satisfied uniquely at the $120^\circ$1 angles of the minimal network.
• Minimality criterion: If the network admits a calibration, it is locally minimizing for length among all admissible competitors within the same homology class and boundary data.

Main Results and Methodological Advances

The principal result is the rigorous demonstration that the Steiner network with three boundary points connected by geodesic arcs, intersecting at a single interior “Steiner point” (triple junction) at precise $120^\circ$2 angles, minimizes the geodesic length locally within sufficiently small spherical caps. The argument proceeds in several substantive steps:

1. Local Convexity and Chart Reduction: The relevant segment of $120^\circ$3 is taken to be a geodesic ball $120^\circ$4 with $120^\circ$5, ensuring convexity, simple connectedness, and local bi-Lipschitz equivalence to $120^\circ$6. This enables transfer of Euclidean geometric measure-theoretic arguments.
2. Calibration Construction: The traditional branch calibration approach is adapted. For each arc of the Steiner network, a local calibration is constructed using the parallel-transported unit tangent vectors, extended smoothly to a neighborhood. Compatibility at patching interfaces is ensured by the algebraic disappearance of cross terms at the $120^\circ$7 triple junction.
3. Verification of Calibration Axioms: The closedness of the constructed form is by construction, comass boundedness is ensured via the smallness of $120^\circ$8 (so metric distortion is within controlled bounds), and the calibration condition is satisfied exactly along the network.
4. Minimality Proof via Stokes' Theorem: For any admissible (rectifiable, boundary-matching) competitor inside the cap, integrating the calibration yields equality with the network length but an upper bound for the competitor, thus establishing the network’s local minimality.

Notable Claims

• Local (not global) minimality: The proof does not establish global minimality of Steiner networks on $120^\circ$9, but only within caps whose size is governed by the curvature and geometric configuration.
• Rigorous radius threshold: Explicit quantitative bounds are given for the admissible cap radius, ensuring strict convexity, triangle containment, and calibration patching.

Implications

Theoretical Significance

The result fills a recognized gap in minimal network theory for spaces of constant positive curvature, offering a robust and flexible framework readily extensible to more general Riemannian manifolds. Importantly, it demonstrates that calibration methods—so successful in the Euclidean context—can be adapted using local geometric analysis, at least in convex neighborhoods.

This local theory provides a critical foundation for subsequent advances:

• Extension to higher genus or higher-dimensional spaces,
• Analysis of global minimality and regularity under curvature constraints,
• Connections to Mumford-Shah-type free discontinuity problems and geometric optimization on manifolds.

Practical Effect

While direct practical applications may be limited due to the local nature of the result, the framework lays groundwork for algorithms and variational methods in computational geometry, manifold learning, and geometric optimization over curved spaces. Application domains include geospatial network design, geometric modeling, and the study of optimal connection problems constrained to surfaces (e.g., in physics, computer graphics, or biology).

Future Directions

The natural next steps include:

• Generalization to networks with more terminals or on manifolds with variable curvature.
• Extension of calibration-based arguments to achieve global minimality, likely requiring new tools in geometric measure theory incorporating global topology and curvature.
• Bridging to discrete geometry and computational methods for practical construction of minimal networks on curved surfaces.
• Rigorous analysis of regularity and possible singularity formation in the network structure beyond standard triple junction theory.

Conclusion

The paper establishes a rigorous local minimal network theory on the sphere $S^2$0 by extending and adapting calibration methods and geometric measure-theoretic techniques from the Euclidean plane. The result identifies precise curvature and convexity constraints under which spherical Steiner networks with $S^2$1 junctions are locally length-minimizing and provides a formal apparatus for calibration in the presence of constant curvature. This framework supports future theoretical advances and potential applications in geometric variational problems on manifolds, enriching both foundational geometric measure theory and applied geometric optimization (2604.04119).