Compactness in spaces of inner regular measures and a general portmanteau lemma (Q2518346): Difference between revisions

Let \({\mathcal S}\) be a lattice of subsets of a non-empty set \(\Omega\) such that \(\Omega\) and \(\emptyset\) belong to \({\mathcal S}\) and \({\mathcal S}\) is closed (also called stable) under countable intersections. Let \(\sigma({\mathcal S})\) be the \(\sigma\)-algebra generated by \({\mathcal S}\) and let \({\mathcal M}_f(\Omega,{\mathcal S})\) be the set of all non-negative real-valued measures defined on \(\sigma({\mathcal S})\) which are inner regular with respect to \({\mathcal S}\), that is, \(\mu(A)= \sup\{\Phi(S): S \in{\mathcal S}\), \(S\subseteq A \}\) holds for all \(A \in \sigma({\mathcal S})\) and for all \(\mu \in{\mathcal M}_f(\Omega,{\mathcal S})\). \({\mathcal M}_f(\Omega,{\mathcal S})\) is equipped with the weakest topology \(\tau_w\) such that the mapping \(\Psi_A :{\mathcal M}_f(\Omega,{\mathcal S}) \rightarrow \mathbb{R}\) given by \(\Psi_A = \mu(A)\) \(\forall\) \(\mu \in{\mathcal M}_f(\Omega,{\mathcal S})\) is upper semicontinuous for each \(A \in{\mathcal S}\), and \(\Psi_\Omega\) is continuous. In order to study the relatively compact subsets of \({\mathcal M}_f(\Omega,{\mathcal S})\), the author first considers a sublattice \(\overline{{\mathcal S}}\) of \({\mathcal S}\) having similar properties as \({\mathcal S}\) and then imposes certain conditions on \({\mathcal S}\) relating to \(\overline{{\mathcal S}}\). Under these conditions and with the help of a generalization of the classical Portmanteau lemma (this generalization is also a significant part of this paper), the author gives several characterizations of relatively compact subsets in the space \(({\mathcal M}_f(\Omega,{\mathcal S}), \tau_w)\). As pointed out by the author himself, this paper may well be considered as a continuation of \textit{F. Topsoe}'s seminal paper entitled ``Compactness in spaces of measures'' [Stud. Math. 36, 195--212 (1970)]. Also the investigations carried out in this paper under review rely essentially on the inner extension theory for measures which has been studied extensively and vigorously by H. König. Finally, the reviewer recommends this paper highly to the researchers interested in studying the topological measure theory in an abstract and general setting.