Galerkin approximation for elliptic PDEs on spheres (Q1764874): Difference between revisions

The Galerkin method is used, with the approximate solution being spanned by spherical functions (SBF), for the elliptic partial differential equation (PDE) \(-\Delta u + \omega^{2} u = f\) on \(S^{n}\subset\mathbb{R}^{n+1}\), where \(\omega\) is a non-zero real constant, \(\Delta\) is the Laplace-Beltrami operator on \(S^{n}\) and \(f \in C^{2k}(S^{n})\) for some \(k \geq 1\). The approximate solution is constructed from a linear combination of shifts of a radial basis function on a scattered set of points. The main result is a Sobolev type error estimate for \(u-u_h\), where \(u_h\) is the finite element approximation of \(u\), constructed using SBFs satisfying certain regularity conditions. Related results can be found in the paper of \textit{F. J. Narcowich, J.D. Ward} [SIAM J. Math. Anal. 33(6), 1393--1410 (2002; Zbl 1055.41007)] and of \textit{H. Wendland} [J. Approx. Theory 93, 258--272 (1998; Zbl 1020.65084)].