On the syzygies and Hodge theory of nodal hypersurfaces (Q2364520)

Let \(S={\mathbb{C}}[x_0,\dots, x_n]\) be the polynomial ring and \(S_r\) the vector space of homogeneous polynomials in \(S\) of degree \(r\). For any polynomial \(f\in S_r\), let \(J_f\subset S\) be the Jacobian ideal spanned by the partial derivatives \(f_0, f_1,\dots, f_n\) of \(f\) with respect to \(x_0,\dots, x_n\) respectively. The corresponding graded Milnor (or Jacobian) algebra is defined by \[ M(f)=S/J_f. \] Let \(K^*(f)\) be the Koszul complex of \(f_0,\dots, f_n\) with the natural grading \(|x_j|=|dx_j|=1\) defined by \[ K^*(f): 0\rightarrow \Omega^0\rightarrow \Omega^1\rightarrow \dots\rightarrow \Omega^{n+1}\rightarrow 0 \] with all the arrows given by the wedge product: \(df=f_0dx_0+f_1dx_1+\dots+f_ndx_n\). Then the Milnor algebra \(M(f)\) can be seen as the top cohomology \(H^{n+1}(K^*(f))\). In this paper, the author first gives a new bound in the case n odd, which is optimal at least for nodal hypersurfaces of even dimension up to 10. Theorem 1. Let \(D: f=0\) be a nodal surface of degree \(d>2\) in \({\mathbb{P}}^n\). If \(n=2n_1+1\) is odd, then \(H^n(K^*(f))_m=0\) for any \(m\leq (n_1+1)d-[\frac{d}{2}]-1\). As an application of the above theorem, the author proves Corollary. Let \(D: f=0\) be a nodal surface of degree \(d>2\) in \({\mathbb{P}}^3\). Then the set of nodes imposes independent conditions on hypersurfaces of degree \(d+[\frac{d}{2}]-3\). Let \(d\) be the degree of surfaces \(D\subset {\mathbb{P}}^3 \) and \[ N=\frac{(d+3)!}{3!d!}-1. \] There is a locally closed subscheme \(V_{d, n}\subset {\mathbb{P}}^N \) parametrizing the surfaces \(D\). The author proves the following new results. Theorem 2. Fix a degree \(d\in \{5, 6, 7\}\) and a positive integer \(n\). (1) The variety \(V_{d, n}\) is a smooth locally closed subscheme of \( {\mathbb{P}}^N\). If \(V_{d, n}\neq \emptyset\), then \(V_{d, n}\) has pure dimension \(N-n\). (This part was obtained by \textit{R. Kloosterman} [``Nodal surfaces with singular or too large deformation spaces'', \url{arXiv:1612.0672}].) (2) If \(V_{d, n}\neq \emptyset\), then \(V_{d, n'}\neq \emptyset\), for any positive integer \(n'\leq n\). (3) The surface \(X\) is unobstructed, i.e., the complete local ring \(R(X)\) is regular for any nodal surface \(D\in V_{d, n} \), where \(X\) is the minimal resolution of a surface \(D\) corresponding to a point in \(V_{d, n}\) and \(R(X)\) is the formal moduli space of \(X\). To prove Theorem 1, the author uses the same idea in the joint work with \textit{G. Sticlaru} [``Computing the monodromy and pole order filtration on Milnor fiber cohomology of plane curves'', \url{arXiv:1609.06818}]: the interplay between Hodge filtrations, pole order filtrations and some spectral sequences. In the end of the paper, the author discusses the application of the above results and proves that an algebraic description of some of the graded pieces \(Gr^p_FH^n(U, {\mathbb{C}})\) of the top cohomology group \(H^n(U, {\mathbb{C}})\) with respect to the Hodge filtration \(F^*\) in the joint work with \textit{M. Saito} and \textit{L. Wotzlaw} [Mich. Math. J. 58, No. 3, 603--625 (2009; Zbl 1192.14009)] holds in many cases.