Intersection of positive closed currents of higher bidegree (Q516609): Difference between revisions

Let \((X,\omega)\) be a compact Kähler manifold of dimension \(n\) with a Kähler form \(\omega\). Let \(T\) and \(S\) be two positive closed currents on \(X\) of bidegrees \((p,p)\) and \((q,q)\), \(p+q\leq n\), respectively. Demailly posed the problem to define the intersection \(T\wedge D\) for currents of higher degree. \textit{T.-C. Dinh} and \textit{N. Sibony} [J. Algebr. Geom. 19, No. 3, 473--529 (2010; Zbl 1202.32033)] solved the problem using the theory of superpotentials. Let \(\mathcal D_p\) be the real vector space of closed positive \((p,p)\)-currents on \(X\), and let \(\mathcal D_p^0\) be the subspace of \(\mathcal D_p\) of currents belonging to the class \(\{0\}\) in \(H^{p,p}(X,\mathbb R)\). Let \(\{\alpha\}=(\{\alpha_1\},\dots,\{\alpha_k\})\) be a basis of \(H^{p,p}(X,\mathbb R)\). For \(R\in \mathcal D_{n-p+1}^0\), let the \((n-p,n-p)\)-current \(U_R\) be its \(\alpha\)-normalized potential, i.e., \(dd^cU_R=R\) and \(\left<U_R,\alpha_j\right>=0\), for \(j=1,\dots,k\). Let \(T\in \mathcal D_p\), then its \(\alpha\)-normalized superpotential \(U_T\) is the function defined on smooth \(R\in \mathcal D_{n-p+1}^0\) by \[ U_T(R)=\left <T,U_R\right>, \] where \(U_R\) is an \(\alpha\)-normalized smooth potential of \(R\). Moreover we say that \(T\) has a continuous superpotential if \(U_T\) can be extended to a function on \(\mathcal D_{n-p+1}^0\), also denoted by \(U_T\), which is continuous with respect to the norm \(||\cdot||_*\) defined by \[ ||R||_*=|\left<R,\omega^{p-1}\right>|. \] For \(S\in \mathcal D_q\) and for \(T\in \mathcal D_p\) with continuous superpotential and such that \(\{T\}=\sum_{j=1}^ka_j\{\alpha_j\}\), one can define the intersection \[ \left<T\wedge S,\Phi\right>:=U_T(dd^c\Phi\wedge S)+\sum_{j=1}^ka_j\left<\alpha_j,\Phi\wedge S\right>, \tag{1} \] where \(\Phi\) is a real smooth \((n-p-q,n-p-q)\)-form on \(X\). The aim of this paper is to prove that if \(T\) has continuous superpotential, then the intersection \(T\wedge S\) is a positive closed current of bidegree \((p+q,p+q)\). To do so the author provides an alternative definition for the intersection of currents. Let \(\pi_j\), \(j=1,2\), be the projections from \(X\times X\) to the first and second variables, and let \([\Delta]\) be the current of integration over the diagonal \(\Delta\subset X\times X\). Let \(\widehat {X\times X}\) be the blow-up of \(X\times X\) along \(\Delta\) with the natural projection \(\Pi:\widehat {X\times X}\to X\times X\). Then we can define \[ T\wedge S=(\Pi_j)_*\left(\widehat{T\otimes S}\wedge \widehat {\omega}^{n-1}\wedge [\widehat {\Delta}]\right), \tag{2} \] where \(\Pi_j=\pi_j\circ \Pi\) (\(j=1,2\)), \(\widehat{T\otimes S}=\Pi^*\left(T\otimes S\right)=\Pi^*\left(\pi_1^*(T)\wedge\pi_2^*(S)\right)\), \(\widehat{\Delta}=\Pi^{-1}(\Delta)\) and \(\widehat {\omega}\) is a Kähler form on \(\widehat {X\times X}\) such that \(\Pi_*(\widehat {\omega}^{n-1}\wedge[\widehat{\Delta}])=[\Delta]\). It is proved in the paper that Definition (2) of the intersection of two currents (under the assumption that \(T\) has a continuous superpotential) is equivalent to Definition (1) and gives a well-defined positive closed current.