Logarithmic forms and singular projective foliations (Q2184853)

Logarithmic \(1\)-forms are ubiquitous in the study of codimension one holomorphic foliations. They provide computable examples and their families give rise to irreducible components of moduli spaces of foliations [\textit{O. Calvo-Andrade}, Math. Ann. 299, No. 4, 751--767 (1994; Zbl 0811.58006)]. Much less is known in higher codimension and the article under review sheds some light into this area. It lays down several technical tools to study the space \(\mathcal{L}_q(\mathbf{d}, n)\) of codimension \(q\geq 2\) logarithmic foliations on the projective space \(\mathbb{P}^n\) with a given degree vector \(\mathbf{d} = (d_1, \dots , d_m)\). For \(q=2\) we have the following main result. Theorem. Fix natural numbers \(n, m \in \mathbb{N}_{>3}\), and a \(2\)-balanced vector of degrees \(\mathbf{d} = (d_1, \dots , d_m)\) with \(d = \sum_i d_i\). The variety \(\mathcal{L}_2(\mathbf{d}, n)\) is an irreducible component of the moduli space \(\mathcal{F}_2(d, \mathbb{P}^n)\). Furthermore, the scheme \(\mathcal{F}_2(d, \mathbb{P}^n)\) is generically reduced along this component, in particular at the points of \(\rho(\mathcal{U}_2(\mathbf{d}))\). The proof is based on the explicit computation of the Zariski tangent space via a natural parameterization \(\rho\). The main dificulty is to establish the surjectivity of the derivative of \(\rho\) and here it is used the \(2\)-balanced assumption which is defined as follows: Definition. We say that an \(m\)-tuple of degrees \(\mathbf{d} = (d_1, \dots , d_m)\) is \(k\)-balanced if for each \(I \subset \{1, \dots , m\}\) of size \(|I| = k\), the following inequality holds: \[ \sum_{i\in I} d_i < \sum_{l \notin I} d_l \] The article is well written and very readable despite its technical nature.