Necessary conditions for interpolation by multivariate polynomials (Q2129492)

Given a measure \(\mu\) in \(\mathbb R^n\) consider the space \(\mathcal P_k\) of polynomials of total degree at most \(k\) in \(n\)-variables with the natural scalar product in \(L^2(\mu)\). The space \(\mathcal P_k\) is a reproducing kernel Hilbert space with kernel \(K_k(\mu,x,\cdot)\in \mathcal P_k\). For \(x\in \mathbb R^n\) let \(\kappa_{k,x}\) denote the normalized reproducing kernel, i.e., \(\kappa_{k,x}(\mu,y) = K_k(\mu,x,y)/\sqrt{K_k(\mu,x,x)}\). A sequence \(\Lambda=\{\Lambda_k\}\) of finite sets of points in \(\mathbb R^n\) is said to be interpolating for the sequence of spaces \((\mathcal P_k, L^2(\mu))\) if there exists a constant \(C>0\), independent of \(k\), such that \[ \frac1{C} \sum_{\lambda\in \Lambda_k} |c_\lambda|^2 \le \|\sum_{\lambda\in\Lambda_k} c_\lambda \kappa_{k,\lambda}\|^2 \le C\sum_{\lambda\in\Lambda_k} |c_\lambda|^2. \] \textit{R. J. Berman} and \textit{J. Ortega-Cerdà} [Am. J. Math. 140, No. 3, 789--820 (2018; Zbl 1396.32014)] obtained necessary geometric conditions for sampling families in bounded smooth convex sets with weights satisfying two technical conditions. In the present paper the authors obtain necessary geometric conditions for interpolating families in the following two cases: 1) \(\Omega\) is a smooth, bounded, convex domain and \(d\mu(x) = \chi_{\Omega}(x) dV(x)\), and 2) \(\Omega =\mathbb B=\{x\in \mathbb R^n: |x|\le 1\}\), and measures of the form \(d\mu(x) = (1-|x|^2)^{a-\frac12} dV(x)\) where \(a\ge 0\). The geometric conditions that appear in this type of problem are of three types: A separation condition, a Carleson-type condition, and a density condition.