Singular sets of a class of locally conformally flat manifolds (Q816613)

Let \((M,g)\) be a Riemannian manifold of dimension \(n\geq 3\), and let \(\text{Ric}^g\) and \(R^g\) denote the Ricci and scalar curvatures respectively. The Schouten tensor is given by \[ A^g_{ij} = {1\over{n-2}}\left( \text{Ric}^g_{ij} -{1\over{2(n-1)}}R^g g_{ij} \right). \] As it can be easily seen for the usual decomposition of the Riemann tensor, the Schouten tensor does contain all conformal informations related to the curvature of the manifold. Let \(\lambda_1,\dots,\lambda_n\) denote the eigenvalues of \(A^g=g^{-1}A^g\). The \(\sigma_k\)-curvatures \(\sigma_k=\sigma_k(A^g)\) are the elementary symmetric functions of \(\lambda_1,\dots,\lambda_n\), i.e., \(\sigma_k=\sum_{1\leq i_1<\dots<i_k\leq n}\lambda_{i_1}\cdots\lambda_{i_k}\). The equation \(\sigma_k(A_g) = f\) is a fully nonlinear version of the prescribed scalar curvature problem. Indeed, for the case \(k = 1\) it reduces to the usual prescribed scalar curvature problem. In the paper under review the author studies the singular sets of complete, locally conformally flat metrics on domains \(\Omega \subset S^n\). The main results are the following: Let \(g\) be a complete metric on a domain \(\Omega\subset S^n\), conformal to the standard metric on \(S^n\). If \(\sigma_1(A^g)\geq c_0>0\) and \(\sigma_2(A^g),\dots,\sigma_k(A^g)\geq 0\) for some integer \(k\) with \(1\leq k<n/2\), then \(\dim_{\mathcal H}(\partial\Omega)\leq (n-2k)/2\). If in addition \(k>1\) and \(| R^g| +| \nabla R^g| \leq C_0\), then \[ \dim_{\mathcal H}(\partial\Omega) < (n-2k)/2. \] Here \(\dim_{\mathcal H}(\partial\Omega)\) denotes the Hausdorff dimension on \(\partial\Omega\). This bound on the singular set is closely related to the work of \textit{R. Schoen} and \textit{S. T. Yau} [Invent. Math. 92, 47--71 (1988; Zbl 0658.53038)]. The proof of this result is based on the work of \textit{S.-Y. A. Chang, F. Hang} and \textit{P. C. Yang} [Int. Math. Res. Not. 2004, No. 4, 185--209 (2004; Zbl 1137.53327)]; for the case \(k=2\) and an inductive formula showing that \(\sigma_k\) has an ``almost'' divergence structure, which allows suitable integral estimates to be derived.