A note on the approximation of plurisubharmonic functions (Q2499665): Difference between revisions

Let \(\varOmega\Subset\mathbb C^n\) be a hyperconvex domain. Let \(\mathcal F^a(\varOmega)\) be the set of all negative \(u\in\mathcal{PSH}(\varOmega)\) for which there exists a sequence \((u_j)_{j=1}^\infty\) of negative bounded plurisubharmonic functions in \(\varOmega\) such that: (i) \(u_j\searrow u\), (ii) \(\sup_{j\in\mathbb N}\int_\varOmega(dd^cu_j)^n <+\infty\), (iii) \(u_j\) tends to \(0\) at \(\partial\varOmega\), and \((dd^cu)^n\) vanishes on all pluripolar subsets of \(\varOmega\). The main result of the paper is the following theorem. Let \(\varOmega\Subset\mathbb C^n\) be a strongly hyperconvex domain and let \((\varOmega_j)_{j=1}^\infty\) be a sequence of hyperconvex domains such that \(\varOmega\subset\varOmega_{j+1}\subset\varOmega_j\), \(\varOmega=\text{int}\bigcap_{j=1}^\infty\varOmega_j\), and for each open set \(D\supset\overline\varOmega\) there exists a \(j\) with \(\varOmega_j\subset D\). Then for every \(\varphi\in\mathcal F^a(\varOmega)\) there exists a sequence \(\varphi_j\in\mathcal F^a(\varOmega_j)\), \(j=1,2,\dots\), such that \(\varphi_j\nearrow\varphi\) pointwise in \(\varOmega\).