We introduce the notion of the Bruce-Roberts number for holomorphic 1-forms relative to complex analytic varieties. Our main result shows that the Bruce-Roberts number of a 1-form \omega with respect to a complex analytic hypersurface X with an isolated singularity can be expressed in terms of the Ebeling–Gusein-Zade index of \omega along X, the Milnor number of \omega and the Tjurina number of X. This result allows us to recover known formulas for the Bruce-Roberts number of a holomorphic function along X and to establish connections between this number, the radial index, and the local Euler obstruction of \omega along X. Moreover, we present applications to both global and local holomorphic foliations in complex dimension two.
Let \omega be the germ of a holomorphic 1-form with isolated singularity at 0\in\mathbb{C}^n, n\geq 2. Let X be the germ of a complex analytic hypersurface with an isolated singularity at 0\in\mathbb{C}^n. Assume that X is not invariant by \omega. Then
where \tau_0(X) is the Tjurina number of (X,0).
Let X be an isolated hypersurface singularity defined by \phi :(\mathbb{C}^n,0)\to (\mathbb{C}, 0) and let f\in\mathcal{O}_n has an isolated singularity over (X,0). Then
where \mu_0(X) is the Milnor number of X and \mu_0(\phi,f) is the Milnor number of the isolated complete intersection singularity defined by (\phi,f), in the sense of Hamm [Hamm].
Let \omega be a germ of a holomorphic 1-form with isolated singularity at 0\in\mathbb{C}^n, n\geq 2. Let X be a germ of a complex analytic hypersurface with an isolated singularity at 0\in\mathbb{C}^n. Assume that X is not invariant by \omega. Then
where M_{\ell} is the Milnor fiber of the generic linear function \ell:\mathbb{C}^n\to\mathbb{C} on X and \bar{\chi}(M_{\ell})=\chi(M_{\ell})-1.
Let \omega be a germ of a holomorphic 1-form with an isolated singularity at 0\in\mathbb{C}^n, n\geq 2. Let X be a germ of a complex analytic hypersurface with an isolated singularity at 0\in\mathbb{C}^n. Assume that X is not invariant by \omega. Then
The index \operatorname{Ind}_{\operatorname{GSV} }(\omega;V,0) of the 1-form \omega on the ICIS V at the origin is the degree of the map
The germ \omega of a 1-form at the point p is called radial if, for all \epsilon>0 small enough, the 1-form is positive on the outward normals to the boundary of the \epsilon-neighborhood of the point p.
The radial index \operatorname{Ind}_{\operatorname{rad}}(\omega;V,p) of the 1-form \omega at the point p is
where \operatorname{ind}(\tilde{\omega}|_{V_{(q)}};V_{(q)},q) is the usual index of the restriction of the 1-form \tilde{\omega} to the stratum V_{(q)}, see [Ebeling2023, Definition 5.2.3].
For a 1-form \omega on an ICIS (V,0) we have
where \mu_0(V) denotes the Milnor number of (V,0), in the sense of Hamm [Hamm].
The local Euler obstruction \operatorname{Eu}(\omega; V,0) of the 1-form \omega on V at 0\in\mathbb{C}^N is the obstruction to extend the non-zero section \widehat{\omega} from the pre-image of a neighborhood of the sphere S_{\epsilon}=\partial B_{\epsilon} to the pre-image of its interior. More precisely, its value (as an element of the cohomology group H^{2n}(\nu^{-1}(V\cap B_{\epsilon}),\nu^{-1}(V\cap S_{\epsilon}),\mathbb{Z})) on the fundamental class of the pair [(\nu^{-1}(V\cap B_{\epsilon}),\nu^{-1}(V\cap S_{\epsilon}))].
Let (V,0)\subset(\mathbb{C}^N,0) have an isolated singularity at 0\in\mathbb{C}^N and let \ell:\mathbb{C}^N\to \mathbb{C} be a generic linear function. Then
where M_{\ell} is the Milnor fiber of the linear function \ell on V. In particular, if f is a germ of holomorphic function with an isolated critical point on (V,0), then
where M_f is the Milnor fiber of f.
Let f_1, \ldots, f_m,g,p_1,\ldots, p_n \in \mathcal{O}_n, where f_i and g are relatively prime, for any i \in \{ 1,\ldots,m \}. Then
Observe that
with \mathcal{O}=\displaystyle\frac{\mathcal{O}_n}{\langle f_1,\ldots,f_m\rangle} and r_i'=r_i+\langle f_1,\ldots,f_m\rangle, for any i \in \{ 1,\ldots,n \} and any r_i \in \mathcal{O}_n. The proof is completed using the following exact sequence:
where \sigma(z'+\langle p_1',\ldots,p_n'\rangle)=g'z'+\langle g'p_1',\ldots,g'p_n'\rangle and \delta(z'+\langle g'p_1',\ldots,g'p_n'\rangle)=z'+\langle g'\rangle for any z' \in \mathcal{O}.
We consider \Theta_X^T the \mathcal{O}_n-submodule of \Theta_X generated by the trivial vector fields, that is, \Theta_X^T is generated by
where \phi:(\mathbb{C}^n,0)\to (\mathbb{C},0) is the defining equation of (X,0).
The following sequence of \mathbb{C}-vector spaces is exact
where \beta is induced by the inclusion \omega(\Theta_X^T)\subset\omega(\Theta_X), and \cdot\omega is the evaluation map.
It is clear that the sequence is well-defined. Let us now prove that it is an exact sequence. Suppose f\in\operatorname{Im}(\omega). Then, there exists \eta\in\Theta_X such that f =\omega(\eta+\Theta_X^T)=\omega(\eta) +\omega(\Theta^T_X). Hence,
Since \eta\in\Theta_X, we have \omega(\eta)\in\omega(\Theta_X), and thus, \beta(\omega(\eta))=\omega(\eta)\in\omega(\Theta_X). Consequently, it follows that f\in \operatorname{Ker}(\beta). On the other hand, consider g\in\operatorname{Ker}(\beta). Then, g = f +\omega(\Theta_X^T), where f\in \mathcal{O}_n and \beta(g)\in\omega(\Theta_X). This implies
therefore, f\in\omega(\Theta_X), i.e., f = \omega(\xi), where \xi\in\Theta_X. Thus, we have f + \omega(\Theta_X) = \omega(\xi +\Theta_X^T), and therefore g\in\operatorname{Im}(\omega). Consequently, \operatorname{Ker}(\beta)=\operatorname{Im}(\omega). This establishes the exactness of the sequence and verifies the validity of the lemma.
It is clear that the sequence is well-defined. Let us now prove that it is an exact sequence. Suppose f\in\operatorname{Im}(\omega). Then, there exists \eta\in\Theta_X such that f =\omega(\eta+\Theta_X^T)=\omega(\eta) +\omega(\Theta^T_X). Hence,
Since \eta\in\Theta_X, we have \omega(\eta)\in\omega(\Theta_X), and thus, \beta(\omega(\eta))=\omega(\eta)\in\omega(\Theta_X). Consequently, it follows that f\in \operatorname{Ker}(\beta). On the other hand, consider g\in\operatorname{Ker}(\beta). Then, g = f +\omega(\Theta_X^T), where f\in \mathcal{O}_n and \beta(g)\in\omega(\Theta_X). This implies
therefore, f\in\omega(\Theta_X), i.e., f = \omega(\xi), where \xi\in\Theta_X. Thus, we have f + \omega(\Theta_X) = \omega(\xi +\Theta_X^T), and therefore g\in\operatorname{Im}(\omega). Consequently, \operatorname{Ker}(\beta)=\operatorname{Im}(\omega). This establishes the exactness of the sequence and verifies the validity of the lemma.
Let X be an isolated hypersurface singularity defined by \phi :(\mathbb{C}^n,0)\to (\mathbb{C}, 0) and let f\in\mathcal{O}_n with an isolated singularity over (X,0). Then,
Indeed, consider the 1-form \omega given by \omega= df. By Theorem 1, we have:
Since \mu_0(f) =\mu_0(df), the proof concludes by applying formula (?)
and substituing this into equation (Proof).
Let \omega be a germ of a holomorphic 1-form with isolated singularity at 0\in\mathbb{C}^n, n\geq 2. Let X be a germ of a complex analytic hypersurface with an isolated singularity at 0\in\mathbb{C}^n. Assume that X is not invariant by \omega. Then
where M_{\ell} is the Milnor fiber of the generic linear function \ell:\mathbb{C}^n\to\mathbb{C} on X and \bar{\chi}(M_{\ell})=\chi(M_{\ell})-1.
We have \mu_{BR}(\omega,X)=\operatorname{Ind}_{\operatorname{GSV} }(\omega;X,0)+\mu_0(\omega)-\tau_0(X) by Theorem 1. On the other hand, \operatorname{Ind}_{\operatorname{GSV} }(\omega;X,0)=\operatorname{Ind}_{\operatorname{rad} }(\omega;X,0)+\mu_0(X) by Proposition 3.1. Therefore,
The proof concludes by applying Proposition 3.2 to (X,0) and substituting into equation (Proof).
Let X=\{\phi=0\}, where \phi:(\mathbb{C}^3,0) \rightarrow (\mathbb{C},0) is given by \phi(x,y,z)=x^3+yz^2+y^3+xy^4. Using [Tajima2021, Example 9], we have
Let \omega=zdx+xdy+ydz a germ of a holomorphic 1-form with isolated singularity at 0 \in \mathbb{C}^3. Since
We deduce that X is not invariant by \omega. Using Singular ([DGPS]) to compute the indices, we verify that Theorem 1 holds, since \mu_{BR}(\omega,X)=14, \operatorname{Ind}_{\operatorname{GSV} }(\omega;X,0)=21, \ \mu_0(\omega)=1 and \tau_0(X)=8.
Let \omega=\displaystyle\sum^n_{i=1}A_idx_i\in\Omega^1(\mathbb{C}^n,0) and let g \in \mathcal{O}_n be such that \dim V((\omega,dg)) = 1 and V(\omega) = \{0\} (here (\omega):=\langle A_1,\ldots,A_n\rangle and (\omega,dg) is the ideal in \mathcal{O}_n generated by the maximal minors of the matrix of \left(A_i,\frac{\partial g}{\partial x_i}\right)). Consider the following matrices
where \lambda, \mu \in \mathcal{O}_n. Let M, M' be the submodules of \mathcal{O}^2_n generated by the columns of A, A' respectively. If I_2(A) = I_2(A') then M = M'.
Let \omega=\displaystyle\sum^n_{i=1}A_i(x)dx_i be a germ of a holomorphic 1-form with isolated singularity at 0\in\mathbb{C}^n, n\geq 2, where A_i\in\mathcal{O}_n. Let X=\{\phi=0\} be a germ of a complex analytic hypersurface with an isolated singularity at 0\in\mathbb{C}^n. Assume that X is not invariant by \omega. Then
(i) \dfrac{\Theta_X}{\Theta_X^T}\cong\dfrac{\omega(\Theta_X)+I_X}{\omega(\Theta_X^T)+I_X};
(ii) \omega(\Theta_X)\cap I_X=(\omega)\cdot I_X;
(iii) \dfrac{\mathcal{O}_n}{(\omega)}\cong\dfrac{\omega(\Theta_X)+I_X}{\omega(\Theta_X)},
where (\omega):=\langle A_1,\ldots,A_n\rangle and I_X=\langle\phi\rangle.
\;
(i) The homomorphism \Psi : \Theta_X \to \omega(\Theta_X) + I_X defined by \Psi(\xi) = \omega(\xi) induces the isomorphism
In fact, it is enough to show that \operatorname{Ker}(\overline{\Psi})=\Theta^T_X. Let \xi+\Theta_X^T\in \operatorname{Ker}(\overline{\Psi}) then \omega(\xi)\in\omega(\Theta_X^T)+I_X, that is, there exist \eta \in \Theta^T_ X and \mu, \lambda \in \mathcal{O}_n, such that
then
and
Therefore
and since \phi is regular in \dfrac{\mathcal{O}_n}{(\omega, d\phi)} then
By Lemma 5.1, \lambda \in J\phi and using [Nuno2020, Lemma 3.1], \xi \in \Theta^T_X. Hence, \xi+\Theta_X^T\in\Theta_X^T.
(ii) Let \xi\in \omega(\Theta_X)\cap I_X, then there exist \eta\in \Theta_X and \mu,\lambda\in \mathcal{O}_n such that
Using the same argument as the proof of item (i), we obtain \mu\in (\omega). Hence, \xi\in(\omega)\cap I_X. Conversely, let \xi\in (\omega)\cap I_X then there exist \alpha_1,\ldots,\alpha_n,\mu\in\mathcal{O}_n such that
Taking
we obtain \xi\in \omega(\Theta_X)\cap I_X.
(iii) Consider the following module isomorphism:
Then, by the previous isomorphism and item (ii), we obtain
If X=\{\phi=0\} is a germ of a complex analytic hypersurface with an isolated singularity at 0\in\mathbb{C}^n, by [Tajima, Theorem 2.3] or [Nuno2020, Corollary 3.4], then
Therefore, by Theorem 5.2 item (i), we have
If we take \omega=df in the Corollary 2, we recover the formula
which is found in Section 3 of [Lima2021].
Let \omega be a germ of a holomorphic 1-form with isolated singularity at 0\in\mathbb{C}^n, n\geq 2. Let X be a germ of a complex analytic hypersurface with an isolated singularity at 0\in\mathbb{C}^n. Assume that X is not invariant by \omega. Then
Let X be a germ of a complex analytic hypersurface with an isolated singularity at 0\in\mathbb{C}^n defined by \phi:(\mathbb{C}^n,0)\to (\mathbb{C}, 0) and f\in\mathcal{O}_n be a function germ such that \mu_{BR}(f, X) <\infty. Then (\phi, f) defines an ICIS and
Consider the 1-form \omega given by \omega= df. By Theorem 5.3, we have
The proof follows by applying the Lê-Greuel formula (see equation (?)):
Let \mathcal{F} be a germ of a singular foliation at (\mathbb{C}^2,0), and let X be a germ of a reduced curve at (\mathbb{C}^2,0). Assume that X is not invariant by \mathcal{F}. Then
Consider the foliation \mathcal{F} defined by the 1-form \omega=A\,dx+B\,dy, where A, B\in\mathcal{O}_2. Let X be the complex analytic curve defined by the function \phi:(\mathbb{C}^2,0)\to (\mathbb{C},0). From equation (?) it follows that
The vector field v generating \mathcal{F} is given by v=-B\frac{\partial}{\partial x}+A\frac{\partial}{\partial y}. Therefore, by definition,
Hence,
by Theorem 1.
Consider the curve X given by X=\{ \phi=y^p-x^q=0 \}, and let \mathcal{F} be a foliation defined by the 1-form \omega=\lambda xdy+ydx, with \lambda \neq -\displaystyle\frac{p}{q}. Then, by [Saito, Example 1], we have:
Note that X is not invariant by \mathcal{F}, since
and \lambda \neq -\displaystyle\frac{p}{q}. Thus, we have \mu_0(\mathcal{F})=1, \operatorname{tang}(\mathcal{F},X,0)=pq, and \tau_0(X)=(p-1)(q-1). Computing the Bruce-Roberts number in this case, we find
which indeed satisfies equation (6.1).
Consider the foliation \mathcal{F}_\omega (Suzuki’s foliation) defined by the 1-form
and the foliation \mathcal{F}_\eta defined by the 1-form
The foliations \mathcal{F}_\omega and \mathcal{F}_\eta are topologically conjugate ([CervMattei, Chapter 2, Part 3]). Using the curve X from Example 6.1, with p=7 and q=3, i.e., X=\{\phi=0\}, with \phi=y^7-x^3, we have \mu_{BR}(\mathcal{F}_\omega,X)=\mu_{BR}(\mathcal{F}_\eta,X)=17, \mu_0(\mathcal{F}_\omega)=\mu_0(\mathcal{F}_\eta)=5, \operatorname{tang}(\mathcal{F}_\omega,X,0)=\operatorname{tang}(\mathcal{F}_\eta,X,0)=24, and \tau_0(X)=12. Thus, for both foliations, (6.1) is satisfied.
Let \mathcal{F} be a germ of a singular foliation at (\mathbb{C}^2,0), and let X be a germ of a reduced curve at (\mathbb{C}^2,0) defined by \phi: (\mathbb{C}^2, 0) \rightarrow (\mathbb{C},0). Suppose that X is irreducible and not invariant by \mathcal{F}. Then
where \varphi is a Puiseux parametrization of X.
Suppose that \mathcal{F} is defined by the vector field v = -B(x,y)\frac{\partial}{\partial{x}} + A(x,y)\frac{\partial}{\partial{y}}. Then, it follows from the definition the tangency order that
where \varphi(t)=(x(t),y(t)) is a Puiseux parametrization of X. Without lost of generality, we can suppose that y(t) \neq 0. Since \phi(x(t),y(t))=0, we obtain
Hence,
Thus, we get
Now, from [EveliaPloski, Proposition 2.1, item 3], we obtain
where \mathcal{P}_l(\phi)=\frac{\partial \phi}{\partial x} \frac{\partial l}{\partial y}-\frac{\partial \phi}{\partial y} \frac{\partial l}{\partial x} is the polar of X with respect to l \in \mathbb{C}[x,y], assuming that l does not divide f. Let l=ay-bx. Computing the items on the right side of (\ref{teissier}), we have
and
Finally, by rewriting (Proof) using that and (Proof), we get
Let X=\{\phi=0\} be a germ of an irreducible reduced curve and f\in\mathcal{O}_2 is a germ with an isolated singularity over (X,0). Note that applying Proposition 6.2 to foliation \mathcal{F}:\omega=df, we obtain an expression to compute the Milnor number of the isolated complete intersection singularity defined by (\phi,f) in the sense of Hamm [Hamm]. Indeed, it follows from (?) that
Hence, \mu_0(\phi,f)= \text{ord}_{t=0}\ \varphi^* (df), where \varphi is a Puiseux parametrization of X.
Let \mathcal{F} be a germ of a singular holomorphic foliation at (\mathbb{C}^2,0), let X be a germ of an irreducible reduced curve and let \pi:\tilde{\mathbb{C}}^2 \rightarrow (\mathbb{C}^2,0) be the blow-up at (\mathbb{C}^2,0). Assume that X is not invariant by \mathcal{F}, \tilde{\mathcal{F}}:=\pi^* \mathcal{F}, and q \in \pi^{-1}(0) \cap \tilde{X}, where \tilde{\mathcal{F}} and \tilde{X} are the strict transforms of \mathcal{F} and X respectively. Denote by m the multiplicity of X in its Puiseux parametrization, and denote by \nu the algebraic multiplicity of the foliation \mathcal{F} at 0\in\mathbb{C}^2. Then, we have the following statements:
(a) If \pi is non-dicritical, then
(b) If \pi is dicritical, then
We write \mathcal{D}=\dim_{\mathbb{C}} \displaystyle\frac{\tilde{\sigma}^* \Omega_{\tilde{X}}}{\sigma^* \Omega_{X}}, where \tilde{\sigma}:(\overline{X},0) \rightarrow (\tilde{X},0) is the normalization, and \sigma=\pi\circ\tilde{\sigma}.
Let \omega be the 1-form defining \mathcal{F} and let \tilde{\omega} be the 1-form defining \tilde{\mathcal{F}}. Applying Proposition 6.2 to the Corollary 6.1, we have
and similarly,
where \varphi and \tilde{\varphi} are the Puiseux parametrization of X and \tilde{X}, respectively.
The proof is obtained using results that show how the indexes on the right side of equation (Proof) change through a blow-up. From [CanoCerv, Proposition 4.13], we have
and, from [Wang, p. 4], we have
Now, from [FPGBSM, Section 2], evaluating the 1-form that defines \tilde{\mathcal{F}} in \tilde{\varphi} and taking orders, we get
Substituting the above formulas into (Proof), we conclude the proof of the Proposition.
Let \mathcal{F} be a foliation defined by the 1-form \omega=2xdy-3ydx, and let X be the curve given by X=\{ \phi=y^2-x^5=0 \}. Observe that X is not invariant by \mathcal{F}. If \pi is a blow-up at 0\in\mathbb{C}^2, in local coordinates, we have
where \tilde{\mathcal{F}}_1 is obtained using the local chart \pi(x,t)=(x,tx), and \tilde{\mathcal{F}}_2 is obtained using the local chart \pi(u,y)=(uy,y). Now, we consider \tilde{X}=\{ t^2-x^3=0 \}, since
Since \tilde{X} is not invariant by \tilde{\mathcal{F}}_1, set \tilde{\mathcal{F}}=\tilde{\mathcal{F}}_1. Thus, \tilde{\mathcal{F}} is defined by the 1-form \tilde{\omega}, given by
Using Singular ([DGPS]), we get \mu_{BR} (\mathcal{F},X,0)=7, \mu_{BR} (\tilde{\mathcal{F}},\tilde{X},q)=5, \nu=1, m=2, \displaystyle\sum_{\substack{p \in \pi^{-1}(0) \\ p \neq q}} \mu_p(\tilde{\mathcal{F}})=\mu_0(\tilde{\mathcal{F}}_2)=1 and \mathcal{D}=1. Note that
Since \pi is non-dicritical, item (a) of Proposition 6.3 is satisfied.
Consider the foliation \mathcal{F} defined by the 1-form
and let X be the curve not invariant by \mathcal{F} given by X=\{ \phi=y^3-x^7=0 \}. Again, by the local of the coordinates of the blow-up \pi at 0\in\mathbb{C}^2, we have
where \tilde{\mathcal{F}}_1 and \tilde{\mathcal{F}}_2 are obtained using the local charts \pi(x,t)=(x,tx) and \pi(u,y)=(uy,y), respectively. Set \tilde{X}=\{ t^3-x^4=0 \}, since
Since \tilde{X} is not invariant by \tilde{\mathcal{F}}_1, set \tilde{\mathcal{F}}=\tilde{\mathcal{F}}_1. In that way, \tilde{\mathcal{F}} is defined by the 1-form
Again, by Singular [DGPS], we obtain \mu_{BR} (\mathcal{F},X,0)=56, \mu_{BR} (\tilde{\mathcal{F}},\tilde{X},q)=9, \nu=5, m=3, \displaystyle\sum_{\substack{p \in \pi^{-1}(0) \\ p \neq q}} \mu_p(\tilde{\mathcal{F}})=\mu_0(\tilde{\mathcal{F}}_2)=0 and \mathcal{D}=3. Then,
Since \pi is dicritical, item (b) of Proposition 6.3 is satisfied.
Let \mathcal{F} be a germ of a singular holomorphic foliation at 0\in\mathbb{C}^2. Let X be a germ of a reduced curve at 0\in\mathbb{C}^2. Assume that X is not invariant by \mathcal{F}. Then
We can use Singular [DGPS] to compute the relative Bruce-Roberts number for the previous examples:
In Example 6.1, we have \mathcal{F}:\omega=\lambda xdy+ydx=0, \ \lambda \neq -p/q, and X=\{ \phi=y^p-x^q=0 \}. As we needed numerical examples, for \lambda=1, we have
In both cases, \mu_{BR}^-(\mathcal{F},X)=p+q-1=\mu_{BR}(\mathcal{F},X)-\mu_0(\mathcal{F}).
In Example 6.2, we have \mathcal{F}_{\omega}:\omega=(y^3+y^2-xy)dx-(2xy^2+xy-x^2)dy=0, \mathcal{F}_{\eta}:\eta=(2y^2+x^3)dx-2xydy=0 and X=\{ \phi=y^7-x^3=0 \}, with \mu_{BR}(\mathcal{F}_{\omega},X)=\mu_{BR}(\mathcal{F}_{\eta},X)=17 and \mu_0(\mathcal{F}_{\omega})=\mu_0(\mathcal{F}_{\eta})=5. Then
In each case, we have \mu_{BR}^-(\mathcal{F},X)=\mu_{BR}(\mathcal{F},X)-\mu_0(\mathcal{F}), satisfying Corollary 6.4.
Let \mathcal{F} be a germ of a singular holomorphic foliation at 0\in\mathbb{C}^2. Let X be a germ of a reduced curve at 0\in\mathbb{C}^2. Assume that X is not invariant by \mathcal{F}. Then
where \nu denotes the algebraic multiplicity of \mathcal{F}.
Let \mathcal{F} be a germ of a singular holomorphic foliation at (\mathbb{C}^2,0), let X be a germ of an irreducible reduced curve and let \pi:\tilde{\mathbb{C}}^2 \rightarrow (\mathbb{C}^2,0) be the blow-up at (\mathbb{C}^2,0). Assume that X is not invariant by \mathcal{F}, \tilde{\mathcal{F}}:=\pi^* \mathcal{F}, and q \in \pi^{-1}(0) \cap \tilde{X}, where \tilde{\mathcal{F}} and \tilde{X} are the strict transforms of \mathcal{F} and X respectively. Denote by m the multiplicity of X in its Puiseux parametrization, and denote by \nu the algebraic multiplicity of the foliation \mathcal{F} at 0\in\mathbb{C}^2. Then, we have the following statements:
(a) If \pi is non-dicritical, then
(b) If \pi is dicritical, then
We write \mu_{BR}^-(\mathcal{F},X,p) to denote the relative Bruce-Roberts number around the point p. Moreover, we have \mathcal{D}=\dim \displaystyle\frac{\tilde{\sigma}^* \Omega_{\tilde{X}}}{\sigma^* \Omega_{X}}, where \tilde{\sigma}:(\overline{X},0) \rightarrow (\tilde{X},0) is the normalization, and \sigma=\pi\circ\tilde{\sigma}.
[CLS, Theorem 4] Given a germ of a non-dicritical holomorphic foliation \mathcal{F} at 0\in\mathbb{C}^2 one has \mu_0(\mathcal{F})\geq\mu_0(C), and the equality holds if and only if \mathcal{F} is a generalized curve foliation.
Let \mathcal{F} be a germ of a non-dicritical holomorphic foliation at 0\in\mathbb{C}^2. Let X be a germ of a reduced curve at 0\in\mathbb{C}^2, and C=Sep_0(\mathcal{F})=\{f=0\} be a reduced equation of Sep_0(\mathcal{F}). Assume that X is not invariant by \mathcal{F}. Then
and the equality holds if and only if \mathcal{F} is a generalized curve foliation.
Let \mathcal{F} be a holomorphic foliation on a compact complex surface S, and let X\subset S be a compact curve, none of whose components are invariant by \mathcal{F}. Then
where \chi(X)=-K_{S}\cdot X-X\cdot X is the virtual Euler characteristic of X, and \displaystyle\tau(X)=\sum_{p\in X}\tau_p(X) is the global Tjurina number of X.
For each p\in \textsf{Sing}(\mathcal{F})\cap X, we have
by Corollary 6.1. On the other hand, according to [Brunella-book, Propositions 2.1 and 2.2] we get
and
We complete the proof by summig over p\in\textsf{Sing}(\mathcal{F})\cap X and substituting the above formulas into equation (Proof).
Let \mathcal{F} be a holomorphic foliation on a compact complex surface S, and let X\subset S be a compact curve, none of whose components are invariant by \mathcal{F}. Then
Since \mu_{BR}(\mathcal{F},X,p)\geq 0 for all p\in\textsf{Sing}(\mathcal{F})\cap X, we obtain an upper bound for the global Tjurina number of X:
by Theorem 7.1.
Let \mathcal{F} be a holomorphic foliation on a compact complex surface S, and let X\subset S be a compact curve, none of whose components are invariant by \mathcal{F}. Then
The text on this page is from arXiv:2409.01237v1 by Pedro Barbosa, Arturo Fernández-Pérez and Víctor León, under the licence CC BY 4.0. The corpus made these changes: it read the items out of the LaTeX source, it expanded the macros of the paper, it put each proof after its statement, and it converted the text to HTML. The formulas are the TeX of the paper, and KaTeX renders them. A formula or a reference can look different from the PDF. The page is published under CC BY 4.0.
The record of each item is in analysis/objects/2409/2409.01237.jsonl.