Introduction
Five-dimensional N = 1 \mathcal{N}=1 N = 1 superconformal field theories (SCFTs) provide a particularly rich meeting point between strongly-coupled quantum field theory, string theory, and algebraic geometry. The geometric realization in M-theory on singular Calabi–Yau threefolds [1 , 2 ] is particularly insightful as it provides an alternative way, using geometry, to potentially classify such theories. Several complementary geometric constructions have been used to explore and map out the vast landscape of such theories, including toric Calabi–Yau threefolds [3 ] , using constraints from the local surface geometry [4 , 5 , 6 ] , using M-/F-theory duality which realizes canonical singularities starting with elliptically fibered Calabi–Yau threefolds with non-minimal singular fibers [7 , 8 , 9 ] , isolated hypersurface singularities [3 , 10 , 11 , 12 ] , or non-toric orbifolds [13 ] . Complementary to these geometric approaches, webs of ( p , q ) (p,q) ( p , q ) 5-branes in Type IIB string theory, possibly ending on 7-branes or in the presence of orientifold planes, have been used extensively to construct 5d SCFTs and to map out their gauge-theory deformations, UV dualities, Higgs branches and RG flows [14 , 15 , 16 , 17 , 18 , 19 , 20 , 21 , 22 , 23 , 24 ] , with the predicted enhanced global symmetries cross-checked against the superconformal index [14 , 25 ] .
M-theory sets up a dictionary between the SCFT and geometric data as follows: M-theory on the canonical singularity corresponds to the SCFT. The crepant resolution of the singularity provides a geometric description of the extended Coulomb branch, with Kähler parameters identified with Coulomb-branch parameters and mass deformations. The low-energy effective theory on the extended Coulomb branch is encoded by a cubic prepotential, whose coefficients are given geometrically by triple intersection numbers of divisors in the resolved Calabi–Yau threefold.
For toric Calabi–Yau threefolds this relation is particularly concrete. The geometry is encoded by a two-dimensional toric polygon, and, after choosing a triangulation, divisors, curves, and their intersection numbers can be extracted directly from the combinatorics of the polygon. This provides an efficient route from simple discrete data to the five-dimensional cubic prepotential, see for example [3 , 26 , 27 ] .
A large class of five-dimensional SCFTs also admits a complementary, dual description in terms of webs of ( p , q ) (p,q) ( p , q ) 5-branes in Type IIB string theory [28 , 29 , 30 ] . For theories engineered by toric Calabi–Yau threefolds, the two descriptions are directly related: the brane web is dual to the toric polygon, and the M-theory and Type IIB constructions engineer the same five-dimensional theory [29 , 31 ] . More general brane webs naturally involve [ p , q ] [p, q] [ p , q ] 7-branes and are no longer described by ordinary toric polygons. Generalized toric polygons (GTPs) provide a combinatorial framework for these more general configurations [22 , 24 , 32 , 33 ] . The corresponding Calabi–Yau geometries, referred to as generalized toric Calabi–Yau threefolds , are generically non-toric, so that the familiar toric dictionary and the standard combinatorial rules for computing intersection numbers no longer apply directly.
Brane webs and GTPs provide powerful combinatorial descriptions of these five-dimensional theories, which allow one to probe indirectly certain aspects of generalized toric Calabi–Yau threefolds, see for instance [34 , 35 ] for computations of BPS invariants using this combinatorics via a generalization of the topological vertex, and [36 ] for associated integrable systems. Explicit algebraic descriptions of the generalized toric Calabi–Yau threefolds have been developed for particular classes of GTPs through complex-structure deformations of toric geometries [37 , 38 ] . Other approaches explore the relation between Hanany–Witten transitions and polytope mutations [39 ] , the role of primitive T T T -cones as building blocks of GTP tessellations [40 ] , and the computation of Hilbert and Ehrhart series from GTP data [41 , 42 ] .
For consistent 5-brane webs with 7-branes, a general algebro-geometric construction of generalized toric Calabi–Yau threefolds associated to a GTP was obtained in [33 ] , building on [43 , 44 , 45 ] . This approach is expanded upon in [46 ] to provide an explicit patch by patch description. Once the geometry is explicitly defined, it becomes a mathematically well-posed problem to compute the triple intersection numbers of its divisors and, in particular, to determine whether the cubic prepotential can be recovered directly from the combinatorial data of the GTP.
The purpose of this paper is to answer this question. We determine the relevant intersection theory of the generalized toric Calabi–Yau threefolds constructed in [33 ] and show that their triple intersection numbers can be reconstructed directly from the combinatorial data of the GTP. The result is a finite algorithm that extends the computational power of toric methods to this class of non-toric Calabi–Yau threefolds.
Let us briefly describe the geometric input. Starting from a GTP with a fixed polyhedral subdivision, the construction of [33 ] produces an associated Symington polygon by a sequence of cuts and integral-affine identifications. The Symington polygon is an integral-affine manifold with singularities, together with an induced unimodular triangulation. The vertices and edges of this triangulation encode a collection of log Calabi–Yau surfaces and the curves along which they are glued. Their union forms a normal crossings surface that admits a one-parameter smoothing. Then, the generalized toric Calabi–Yau threefold associated with the GTP is obtained as the total space of this one-parameter smoothing. In the toric case, the Symington polygon is an ordinary toric polygon and the construction reduces to the familiar toric geometry.
In addition to the components of the central fiber, the geometry contains further non-compact divisors, which we call exuberant divisors . These are encoded combinatorially by so-called E E E -lines associated with the GTP. An E E E -line determines an exuberant divisor together with its intersection curve with the central fiber. The divisor and curve data relevant to the intersection problem are therefore encoded jointly by the subdivided Symington polygon and the E E E -lines inherited from the GTP.
Our main result is that this combinatorial information suffices to reconstruct the intersection data relevant for five-dimensional physics. Intersections among the components of the central fiber are determined from the vertex–edge–triangle incidence relations of the unimodular subdivision, together with the integral-affine lattice directions of adjacent edges and the S L ( 2 , Z ) SL(2,\mathbb{Z}) S L ( 2 , Z ) monodromies across cuts. Intersections involving exuberant divisors are controlled by the primitive directions and incidence data of the E E E -lines. The coefficients not fixed by these rules are determined by compatibility conditions following from associativity of the intersection product. We show that the resulting compatibility system has a unique solution and determines all intrinsic triple intersection numbers. These intersection numbers determine the five-dimensional cubic prepotential. This gives a combinatorial algorithm for computing the prepotential directly from the GTP, extending the toric computation to the generalized toric Calabi–Yau threefolds constructed in [33 ] .
We provide a public implementation of our algorithm in Mathematica, available on GitHub,1 and briefly described in Appendix A. It takes the GTP as input and returns all triple intersection numbers.
The organization of the paper is as follows. § 2 reviews five-dimensional SCFTs, their geometric engineering by Calabi–Yau threefolds, and toric intersection theory. § 3 reviews ( p , q ) (p,q) ( p , q ) 5-brane webs, 7-branes, and GTPs. § 4 reviews the construction of generalized toric Calabi–Yau threefolds following [33 ] and the associated divisor and curve data. § 5 develops the intersection theory, and § 6 gives the combinatorial algorithm for computing the triple intersection numbers. We conclude with a series of examples in § 7 .
Canonical Singularities and 5d SCFTs
In this section we review some salient properties of five-dimensional N = 1 \mathcal{N}=1 N = 1 SCFTs and their geometric engineering that will be needed throughout the paper. We first review the Coulomb and extended Coulomb parameter spaces. For each chamber of the extended Coulomb branch, we then discuss the cubic prepotential and its relation to the effective gauge coupling and Chern–Simons terms. We then review the geometric engineering of such theories in M-theory on resolved Calabi–Yau threefold singularities, where Coulomb and mass parameters are identified with Kähler parameters and the cubic prepotential is determined by triple intersection numbers of divisors in the Calabi–Yau threefold. Finally, we review the toric case, in which these intersection numbers can be computed directly from the combinatorics of a triangulated toric polygon.
Coulomb Branches and Prepotentials
Consider a five-dimensional N = 1 \mathcal{N}=1 N = 1 superconformal field theory (SCFT). Its Coulomb branch is the locus in the moduli space of vacua along which the S U ( 2 ) R SU(2)_{\mathrm{R}} S U ( 2 ) R R-symmetry remains unbroken. The rank r r r of the SCFT is defined as the real dimension of its Coulomb branch. At a generic point on the Coulomb branch, the low-energy effective theory is an abelian U ( 1 ) r U(1)^r U ( 1 ) r gauge theory comprising r r r vector multiplets. Locally, the Coulomb branch is parametrized by the vacuum expectation values ϕ i \phi^i ϕ i , i = 1 , … , r i=1,\ldots,r i = 1 , … , r , of the real scalar fields in these vector multiplets.
In addition to moving on the Coulomb branch, one may turn on real mass deformations. Let G F G_{\mathrm{F}} G F denote the flavor symmetry of the SCFT and let m α m^\alpha m α , α = 1 , … , f \alpha=1,\ldots,f α = 1 , … , f , be real mass parameters associated with its Cartan subalgebra. We refer to the space locally parametrized by
t A = ( ϕ i , m α ) (1) t^A=(\phi^i,m^\alpha) \tag*{(1)} t A = ( ϕ i , m α ) (1)
as the extended Coulomb parameter space. Fixing the masses gives the Coulomb branch of the corresponding mass-deformed theory, while setting them to zero gives the Coulomb branch of the undeformed SCFT.
The extended Coulomb parameter space is divided into finitely many chambers. Inside each chamber, the low-energy description by a U ( 1 ) r U(1)^r U ( 1 ) r gauge theory is smooth and the two-derivative effective action is encoded by a cubic prepotential:
F ( t ) = 1 6 c A B C t A t B t C (2) \mathcal{F}(t)=\frac{1}{6}c_{ABC}t^At^Bt^C \tag*{(2)} F ( t ) = 6 1 c A B C t A t B t C (2)
with coefficients c A B C c_{ABC} c A B C constant within the chamber. For fixed mass parameters, the two-derivative effective Lagrangian is of the form
L e f f = − 1 2 G i j ( t ) d ϕ i ∧ ⋆ d ϕ j − 1 2 G i j ( t ) F i ∧ ⋆ F j + 1 24 π 2 c i j k A i ∧ F j ∧ F k + ⋯ (3) \mathcal{L}_{\mathrm{eff}}=-\frac{1}{2}G_{ij}(t)\,\mathrm{d}\phi^i\wedge\star\mathrm{d}\phi^j-\frac{1}{2}G_{ij}(t)\,F^i\wedge\star F^j+\frac{1}{24\pi^2}c_{ijk}A^i\wedge F^j\wedge F^k+\cdots \tag*{(3)} L eff = − 2 1 G ij ( t ) d ϕ i ∧ ⋆ d ϕ j − 2 1 G ij ( t ) F i ∧ ⋆ F j + 24 π 2 1 c ij k A i ∧ F j ∧ F k + ⋯ (3)
where ϕ i \phi^i ϕ i are the real scalar fields, A i A^i A i are the U ( 1 ) U(1) U ( 1 ) gauge fields of the vector multiplets, F i = d A i F^i=\mathrm{d}A^i F i = d A i , with the effective inverse squared gauge-coupling matrix
G i j ( t ) = ∂ 2 F ∂ ϕ i ∂ ϕ j = c i j k ϕ k + c i j α m α (4) G_{ij}(t)=\frac{\partial^2\mathcal{F}}{\partial\phi^i\partial\phi^j}=c_{ijk}\phi^k+c_{ij\alpha}m^\alpha \tag*{(4)} G ij ( t ) = ∂ ϕ i ∂ ϕ j ∂ 2 F = c ij k ϕ k + c ij α m α (4)
and Chern–Simons coupling constants
c i j k = ∂ 3 F ∂ ϕ i ∂ ϕ j ∂ ϕ k . (5) c_{ijk}=\frac{\partial^3\mathcal{F}}{\partial\phi^i\partial\phi^j\partial\phi^k}. \tag*{(5)} c ij k = ∂ ϕ i ∂ ϕ j ∂ ϕ k ∂ 3 F . (5)
For instance, consider a chamber of the extended Coulomb branch in which the 5d theory provides a UV completion of a non-abelian N = 1 \mathcal{N}=1 N = 1 gauge theory with gauge group G g a u g e G_{\mathrm{gauge}} G gauge of rank r r r and with matter hypermultiplets in irreducible representations R f \mathbf{R}_f R f of G g a u g e G_{\mathrm{gauge}} G gauge . At a generic point of the chamber, the gauge group is broken to its Cartan subgroup U ( 1 ) r U(1)^r U ( 1 ) r . The real mass parameters m α m^\alpha m α of the 5d SCFT are the inverse square gauge couplings 1 2 g i j 2 \frac{1}{2g_{ij}^2} 2 g ij 2 1 and the matter masses m f m_f m f from the viewpoint of the non-abelian gauge theory. Then, the prepotential is [2 , 26 ]
F = 1 2 g i j 2 ϕ i ϕ j + k i j k 6 ϕ i ϕ j ϕ k + 1 12 ( ∑ α ∈ Δ ∣ α ⋅ ϕ ∣ 3 − ∑ f ∑ λ ∈ W f ∣ λ ⋅ ϕ + m f ∣ 3 ) (6) \mathcal{F}=\frac{1}{2g_{ij}^{2}}\phi^{i}\phi^{j}+\frac{k_{ijk}}{6}\phi^{i}\phi^{j}\phi^{k}
+\frac{1}{12}\left(\sum_{\alpha\in\Delta}|\alpha\cdot\phi|^{3}-\sum_{f}\sum_{\lambda\in W_f}|\lambda\cdot\phi+m_f|^{3}\right) \tag*{(6)} F = 2 g ij 2 1 ϕ i ϕ j + 6 k ij k ϕ i ϕ j ϕ k + 12 1 α ∈ Δ ∑ ∣ α ⋅ ϕ ∣ 3 − f ∑ λ ∈ W f ∑ ∣ λ ⋅ ϕ + m f ∣ 3 (6)
where k i j k k_{ijk} k ij k denote classical Chern–Simons coupling constants for the non-abelian gauge field, Δ \Delta Δ is the set of non-zero roots of G g a u g e G_{\mathrm{gauge}} G gauge , and W f W_f W f denotes the set of weights of the representation R f \mathbf{R}_f R f . From the non-abelian gauge-theory perspective, (6) is the exact perturbative prepotential: the non-classical terms are the one-loop contributions obtained by integrating out the massive W-bosons and matter fields. Note that a complementary approach is provided by the complete prepotential proposed in [47 ] , which reduces to the perturbative prepotential in the appropriate gauge-theory chambers and makes the enhanced global symmetry manifest in suitable Coulomb-branch variables.
Geometric Engineering Dictionary
Five-dimensional SCFTs can be engineered by considering M-theory on a non-compact Calabi–Yau threefold with a canonical singularity X \mathcal{X} X . An in-depth discussion of the gauge theory phases from geometry can be found in [2 , 7 , 8 , 18 , 26 , 48 , 49 ] . Here we focus on some of the salient features relating to the intersection theory.
Chambers of the extended Coulomb parameter space correspond to Kähler cones of the crepant resolutions
π : X ~ ⟶ X (7) \pi:\widetilde{\mathcal{X}}\longrightarrow\mathcal{X} \tag*{(7)} π : X ⟶ X (7)
of X \mathcal{X} X . Let D A D_A D A denote a basis of divisor classes in X ~ \widetilde{\mathcal{X}} X and let ω A \omega_A ω A be their dual two-forms. Reduction of the eleven-dimensional three-form gauge field C 3 C_3 C 3 with curvature G 4 = d C 3 G_4=dC_3 G 4 = d C 3 gives
C 3 ∼ A A ∧ ω A + ⋯ , G 4 ∼ F A ∧ ω A + ⋯ , F A = d A A . (8) C_3\sim A^A\wedge\omega_A+\cdots,\qquad G_4\sim F^A\wedge\omega_A+\cdots,\qquad F^A=dA^A. \tag*{(8)} C 3 ∼ A A ∧ ω A + ⋯ , G 4 ∼ F A ∧ ω A + ⋯ , F A = d A A . (8)
For compact divisor classes, the corresponding A A A^A A A are dynamical gauge fields. In the local non-compact limit, the modes associated with non-compact divisors are interpreted as background flavor fields.
The field-theory parameters t A t^A t A are identified with Kähler parameters. In the same divisor basis, introduce the divisor-valued Kähler parameter
S = t A D A . (9) S=t^A D_A. \tag*{(9)} S = t A D A . (9)
The coefficients associated with compact divisor classes are the Coulomb parameters,
t i = ϕ i , (10) t^i=\phi^i, \tag*{(10)} t i = ϕ i , (10)
while the independent coefficients associated with non-compact divisor classes are real mass parameters.
An M2-brane wrapped on a compact curve C ⊂ X ~ C \subset\widetilde{\mathcal X} C ⊂ X gives a BPS particle. Its charges and mass are
q A = D A ⋅ C , M C = ∫ C J , (11) q_A = D_A \cdot C,\qquad M_C = \int_C J, \tag*{(11)} q A = D A ⋅ C , M C = ∫ C J , (11)
where J J J is the Kähler form, with [ J ] = − S [J] = -S [ J ] = − S . Here and in the following, we will always assume that the intersections ⋅ \cdot ⋅ are in the Calabi–Yau threefold, unless otherwise stated. A wall of the extended Coulomb parameter space is reached when
Vol ( C ) = ∫ C J = 0. (12) \operatorname{Vol}(C) = \int_C J = 0. \tag*{(12)} Vol ( C ) = ∫ C J = 0. (12)
When the contraction is floppable, crossing the wall replaces C C C by the flopped curve and moves to an adjacent Kähler chamber [50 ] . See [51 ] for an extension of these gauge-theoretic methods for higher-length flops.
The five-dimensional Chern–Simons couplings ∫ A ∧ F ∧ F \int A \wedge F \wedge F ∫ A ∧ F ∧ F in (3) arise by dimensional reduction from the eleven-dimensional Chern–Simons coupling in the eleven-dimensional supergravity describing the low-energy limit of M-theory:
∫ C 3 ∧ G 4 ∧ G 4 . (13) \int C_3 \wedge G_4 \wedge G_4. \tag*{(13)} ∫ C 3 ∧ G 4 ∧ G 4 . (13)
Therefore, upon reduction on X ~ \widetilde{\mathcal X} X , its coefficients are the triple intersection numbers
κ A B C : = ∫ X ~ ω A ∧ ω B ∧ ω C = D A ⋅ D B ⋅ D C . (14) \kappa_{ABC} := \int_{\widetilde{\mathcal X}} \omega_A \wedge\omega_B \wedge\omega_C = D_A \cdot D_B \cdot D_C. \tag*{(14)} κ A B C := ∫ X ω A ∧ ω B ∧ ω C = D A ⋅ D B ⋅ D C . (14)
With the normalization in (3) , we obtain
c A B C = κ A B C , (15) c_{ABC} = \kappa_{ABC}, \tag*{(15)} c A B C = κ A B C , (15)
and so the cubic prepotential is
F ( t ) = 1 6 S 3 = 1 6 κ A B C t A t B t C . (16) \mathcal{F}(t) = \frac{1}{6}\mathcal{S}^3 = \frac{1}{6}\kappa_{ABC}t^At^Bt^C. \tag*{(16)} F ( t ) = 6 1 S 3 = 6 1 κ A B C t A t B t C . (16)
In particular,
∂ 3 F ∂ t A ∂ t B ∂ t C = κ A B C = D A ⋅ D B ⋅ D C . (17) \frac{\partial^3\mathcal{F}}{\partial t^A\partial t^B\partial t^C} = \kappa_{ABC} = D_A \cdot D_B \cdot D_C. \tag*{(17)} ∂ t A ∂ t B ∂ t C ∂ 3 F = κ A B C = D A ⋅ D B ⋅ D C . (17)
Only triple intersections containing at least one compact divisor enter the dynamical Coulomb-branch couplings. Purely non-compact triple intersections depend on choices at infinity and are not intrinsic to the local Calabi–Yau geometry. The intersection theory developed below will therefore extract the intrinsic triple products containing at least one compact divisor.
Toric Calabi–Yau Threefolds and Intersection Numbers
We now recall the toric case, which provides the basic combinatorial model for the intersection rules used later [52 , 53 , 54 ] , as well as the applications to the M-theory setting [3 , 26 , 27 ] .
Toric Polygon
Let Σ \Sigma Σ be a fan in N ≃ Z 3 N \simeq\mathbb{Z}^{3} N ≃ Z 3 defining a smooth toric Calabi–Yau threefold X Σ X_{\Sigma} X Σ . Its one-dimensional cones are generated by primitive lattice vectors v A ∈ N v_A \in N v A ∈ N and correspond to toric divisors D A D_A D A . A two-dimensional cone ⟨ v A , v B ⟩ \langle v_A, v_B \rangle ⟨ v A , v B ⟩ corresponds to the torus-invariant curve
C A B = D A ⋅ D B , (18) C_{AB} = D_A \cdot D_B, \tag*{(18)} C A B = D A ⋅ D B , (18)
and a three-dimensional cone ⟨ v A , v B , v C ⟩ \langle v_A, v_B, v_C \rangle ⟨ v A , v B , v C ⟩ determines the corresponding triple intersection. The canonical class is
K X Σ = − ∑ A D A . (19) K_{X_{\Sigma}} = -\sum_A D_A. \tag*{(19)} K X Σ = − A ∑ D A . (19)
The Calabi–Yau condition implies that, after an S L ( 3 , Z ) SL(3,\mathbb{Z}) S L ( 3 , Z ) transformation, the primitive generators may be written as
v A = ( v A 1 , v A 2 , 1 ) . (20) v_A = (v_A^1, v_A^2, 1). \tag*{(20)} v A = ( v A 1 , v A 2 , 1 ) . (20)
The fan can therefore be represented by a lattice polygon in Z 2 \mathbb{Z}^2 Z 2 . For a smooth crepant resolution the polygon is equipped with a regular unimodular triangulation. Interior lattice points correspond to compact divisors and boundary lattice points to non-compact divisors. The toric divisor classes obey the three linear relations
∑ A ( v A ) a D A = 0 , a = 1 , 2 , 3. (21) \sum_A (v_A)^a D_A = 0,\qquad a=1,2,3. \tag*{(21)} A ∑ ( v A ) a D A = 0 , a = 1 , 2 , 3. (21)
Triple Intersection Numbers
For three distinct divisors one has
D A ⋅ D B ⋅ D C = { 1 , ⟨ v A , v B , v C ⟩ is a maximal cone of Σ , 0 , otherwise. (22) D_A \cdot D_B \cdot D_C =
\begin{cases}
1, & \langle v_A,v_B,v_C\rangle\text{ is a maximal cone of }\Sigma,\\
0, & \text{otherwise.}
\end{cases} \tag*{(22)} D A ⋅ D B ⋅ D C = { 1 , 0 , ⟨ v A , v B , v C ⟩ is a maximal cone of Σ , otherwise. (22)
Mixed intersections containing a repeated divisor can be computed on the corresponding toric surface. Suppose that C A B = D A ∩ D B C_{AB}=D_A\cap D_B C A B = D A ∩ D B and that the two triangles adjacent to the edge [ v A , v B ] [v_A,v_B] [ v A , v B ] have third vertices v C v_C v C and v D v_D v D . The primitive rays in the fan of the surface D A D_A D A satisfy
( C A B ∣ D A ) 2 ( v B − v A ) + ( v C − v A ) + ( v D − v A ) = 0. (23) (C_{AB}|_{D_A})^2(v_B-v_A)+(v_C-v_A)+(v_D-v_A)=0. \tag*{(23)} ( C A B ∣ D A ) 2 ( v B − v A ) + ( v C − v A ) + ( v D − v A ) = 0. (23)
Since
( C A B ∣ D A ) 2 = D A ⋅ D B 2 , (24) (C_{AB}|_{D_A})^2=D_A\cdot D_B^2, \tag*{(24)} ( C A B ∣ D A ) 2 = D A ⋅ D B 2 , (24)
this determines intersections with two coincident divisor factors. If C A B ≃ P 1 C_{AB}\simeq\mathbb{P}^1 C A B ≃ P 1 , the Calabi–Yau condition on its normal bundle gives
( C A B ∣ D A ) 2 + ( C A B ∣ D B ) 2 = − 2. (25) (C_{AB}|_{D_A})^2+(C_{AB}|_{D_B})^2=-2. \tag*{(25)} ( C A B ∣ D A ) 2 + ( C A B ∣ D B ) 2 = − 2. (25)
Finally, let D i D_i D i be a smooth compact toric divisor and let N i N_i N i denote the number of vertices adjacent to the corresponding interior vertex of the triangulated polygon. Equivalently, $N_i is the number of rays in the two-dimensional fan of the smooth complete toric surface D i D_i D i . Ad-junction in the Calabi–Yau threefold gives
D i 3 = K D i 2 . (26) D_i^3 = K_{D_i}^2 . \tag*{(26)} D i 3 = K D i 2 . (26)
For a smooth complete toric rational surface,
χ ( O D i ) = 1 , c 2 ( D i ) = χ t o p ( D i ) = N i . (27) \chi(\mathcal{O}_{D_i}) = 1,\qquad c_2(D_i) = \chi_{\mathrm{top}}(D_i) = N_i . \tag*{(27)} χ ( O D i ) = 1 , c 2 ( D i ) = χ top ( D i ) = N i . (27)
Noether’s formula therefore gives
D i 3 = 12 − N i . (28) D_i^3 = 12 - N_i . \tag*{(28)} D i 3 = 12 − N i . (28)
Equivalently, h 1 , 1 ( D i ) = N i − 2 h^{1,1}(D_i) = N_i - 2 h 1 , 1 ( D i ) = N i − 2 and D i 3 = 10 − h 1 , 1 ( D i ) D_i^3 = 10 - h^{1,1}(D_i) D i 3 = 10 − h 1 , 1 ( D i ) .
Example
Consider the toric polygon
Figure 2.29.
There are two compact divisors, D 1 D_1 D 1 and D 2 D_2 D 2 . The five boundary vertices correspond to non-compact toric divisors. Because of the toric linear equivalences, only two independent non-compact divisor classes remain after the compact classes are fixed.
For three distinct divisors the non-zero intersections are
D 1 D 2 D 4 = D 1 D 4 D 5 = D 1 D 5 D 6 = D 1 D 2 D 6 = D 2 D 6 D 7 = D 2 D 7 D 3 = D 2 D 3 D 4 = 1. (29) D_1D_2D_4 = D_1D_4D_5 = D_1D_5D_6 = D_1D_2D_6 = D_2D_6D_7 = D_2D_7D_3 = D_2D_3D_4 = 1 . \tag*{(29)} D 1 D 2 D 4 = D 1 D 4 D 5 = D 1 D 5 D 6 = D 1 D 2 D 6 = D 2 D 6 D 7 = D 2 D 7 D 3 = D 2 D 3 D 4 = 1. (29)
For the curve C 12 = D 1 ∩ D 2 C_{12} = D_1 \cap D_2 C 12 = D 1 ∩ D 2 , the fan of D 1 D_1 D 1 gives
( C 12 ∣ D 1 ) 2 w 1 , 2 + w 1 , 4 + w 1 , 6 = 0. (30) (C_{12}|_{D_1})^2 w_{1,2} + w_{1,4} + w_{1,6} = 0 . \tag*{(30)} ( C 12 ∣ D 1 ) 2 w 1 , 2 + w 1 , 4 + w 1 , 6 = 0. (30)
Using
w 1 , 2 = ( 0 , 1 ) , w 1 , 4 = ( − 1 , − 1 ) , w 1 , 6 = ( 1 , − 1 ) , (31) w_{1,2} = (0,1),\qquad w_{1,4} = (-1,-1),\qquad w_{1,6} = (1,-1), \tag*{(31)} w 1 , 2 = ( 0 , 1 ) , w 1 , 4 = ( − 1 , − 1 ) , w 1 , 6 = ( 1 , − 1 ) , (31)
we obtain from (23) that
( C 12 ∣ D 1 ) 2 = D 1 D 2 2 = 2. (32) (C_{12}|_{D_1})^2 = D_1D_2^2 = 2 . \tag*{(32)} ( C 12 ∣ D 1 ) 2 = D 1 D 2 2 = 2. (32)
Similarly, the remaining repeated intersections are obtained from (23) and (25) . Finally, D 1 D_1 D 1 has four neighboring vertices and D 2 D_2 D 2 has five. Hence (28) gives
D 1 3 = 12 − 4 = 8 , D 2 3 = 12 − 5 = 7. (33) D_1^3 = 12 - 4 = 8,\qquad D_2^3 = 12 - 5 = 7 . \tag*{(33)} D 1 3 = 12 − 4 = 8 , D 2 3 = 12 − 5 = 7. (33)
Brane Webs and Generalized Toric Polygons
We begin by reviewing the realization of five-dimensional SCFTs through 5-brane webs and their correspondence with toric polygons. We then discuss 5-brane webs together with 7-branes, before introducing generalized toric polygons (GTPs), Symington polygons, and E E E -lines.
5-Brane Webs and 5d SCFTs
An alternative realization of a class of five-dimensional SCFTs is provided by webs of ( p , q ) (p,q) ( p , q ) 5-branes in Type IIB string theory [28 , 29 , 30 ] . We consider Type IIB string theory on R 1 , 9 \mathbb{R}^{1,9} R 1 , 9 , with coordinates x 0 , … , x 9 x^{0},\ldots,x^{9} x 0 , … , x 9 , where each ( p , q ) (p,q) ( p , q ) 5-brane extends along R 1 , 4 \mathbb{R}^{1,4} R 1 , 4 and a straight line in the ( x 5 , x 6 ) (x^{5},x^{6}) ( x 5 , x 6 ) plane, as summarized in Table 1 . The resulting brane web is represented by a planar graph whose edges have rational slopes and carry positive integral weights. Each edge e e e has a primitive direction u e = ( p , q ) ∈ Z 2 u_e=(p,q)\in\mathbb{Z}^{2} u e = ( p , q ) ∈ Z 2 , specifying the charge of the corresponding 5-brane, and a weight w e w_e w e giving its multiplicity. Supersymmetry requires charge conservation at every junction v v v :
Table 1. Spacetime orientations of ( p , q ) (p,q) ( p , q ) 5-branes in Type IIB string theory.
∑ e ∋ v w e u e = 0 , (34) \sum_{e\ni v} w_e u_e = 0, \tag*{(34)} e ∋ v ∑ w e u e = 0 , (34)
where the directions u e u_e u e are taken to be outgoing from v v v .
The 5-branes share the five-dimensional worldvolume R 1 , 4 \mathbb{R}^{1,4} R 1 , 4 . Contracting all finite edges produces the asymptotic web, whose common junction supports the 5d SCFT. Resolving the web by varying its finite segments describes Coulomb-branch deformations, while changing its asymptotic data describes mass deformations. We depict brane webs by drawing the ( x 5 , x 6 ) (x^{5},x^{6}) ( x 5 , x 6 ) plane.
Toric Polygons and Calabi–Yau Threefolds
5-brane webs admit a dual combinatorial description in terms of toric polygons. A toric polygon is a convex lattice polygon whose edges are perpendicular to the external 5-branes, with lattice lengths encoding their multiplicities. More generally, a polyhedral decomposition of the polygon determines a dual planar graph, in which vertices correspond to two-dimensional cells, edges to edges, and faces to vertices of the decomposition. This dual graph describes the associated 5-brane web, with internal edges corresponding to finite 5-brane segments. In particular, unimodular triangulations correspond to fully resolved webs, while coarser subdivisions describe partial resolutions. The correspondence between a toric polygon and its dual brane web is illustrated in Figure 3.1 .
Figure 3.1. (a) A toric polygon. (b) The corresponding dual planar graph, with the toric polygon shown in orange and the dual graph in blue. (c) The corresponding 5-brane web.
As reviewed in § 2.3.1 , toric polygons also encode toric Calabi–Yau threefolds, with their polyhedral decompositions describing partial or complete crepant resolutions. Combining the two descriptions gives a direct correspondence between the brane web and the geometry: bounded faces of the web correspond to compact divisors, while unbounded faces correspond to non-compact divisors. Finite 5-brane segments correspond to compact curves, whose contraction describes the shrinking of the corresponding curves in the Calabi–Yau threefold. Different unimodular triangulations are related by flops, realized in the web by contracting and re-expanding finite 5-brane segments, as illustrated in Figure 3.2 . The five-dimensional SCFT engineered by M-theory on the toric Calabi–Yau threefold coincides with the theory supported on the dual 5-brane web in Type IIB string theory [29 , 31 ] .
Figure 3.2. A flop transition in the toric polygon and its dual 5-brane web. (a) A fully triangulated toric polygon and its dual web, with the curve to be contracted highlighted in green. (b) The configuration after contraction of the green curve. (c) The resulting resolution after the flop, with the new curve highlighted in red.
5-Brane Webs with 7-Branes
More general Type IIB configurations allow the external ( p , q ) (p,q) ( p , q ) 5-branes to end on [ p , q ] [p,q] [ p , q ] 7-branes. A [ p , q ] [p,q] [ p , q ] 7-brane occupies the directions x 0 , … , x 4 , x 7 , x 8 , x 9 x^0,\ldots,x^4,x^7,x^8,x^9 x 0 , … , x 4 , x 7 , x 8 , x 9 , and is localized at a point in the ( x 5 , x 6 ) (x^5,x^6) ( x 5 , x 6 ) plane, as summarized in Table 2 . Supersymmetry requires a ( p , q ) (p,q) ( p , q ) 5-brane ending on a [ p , q ] [p,q] [ p , q ] 7-brane to have identical ( p , q ) (p,q) ( p , q ) charge. Several 5-branes of the same charge may end on the same 7-brane [22 , 32 , 55 , 56 ] .
Table 2. Spacetime orientations of ( p , q ) (p,q) ( p , q ) 5-branes and [ p , q ] [p,q] [ p , q ] 7-branes in Type IIB string theory.
A [ p , q ] [p,q] [ p , q ] 7-brane carries an S L ( 2 , Z ) SL(2,\mathbb{Z}) S L ( 2 , Z ) monodromy, represented by a cut with matrix
M [ p , q ] = ( 1 − p q p 2 − q 2 1 + p q ) , (35) M_{[p,q]} =
\begin{pmatrix}
1-pq & p^2 \\
-q^2 & 1+pq
\end{pmatrix}, \tag*{(35)} M [ p , q ] = ( 1 − pq − q 2 p 2 1 + pq ) , (35)
crossing the cut counterclockwise. Across the cut, the charge of a 5-brane is transformed by
the monodromy:
M [ p , q ] ⋅ ( r , s ) (36) M_{[p,q]}\cdot(r,s) \tag*{(36)} M [ p , q ] ⋅ ( r , s ) (36)
In figures, we generally suppress cuts that extend to infinity without crossing the 5-brane web.
In what follows, we consider 5-brane webs whose external legs can end on 7-branes, and assume that they are maximally resolved subject to these endpoint conditions. When several 5-branes end on the same 7-brane, such a configuration need not be maximally resolved as a web of 5-branes alone, since the endpoint constraints may obstruct certain resolutions. This distinction is illustrated in Figure 3.3 –Figure 3.4 .
Figure 3.3. A toric polygon and its dual 5-brane web, with all external 5-branes ending on 7-branes. This web contains in particular parallel external 5-branes ending on distinct 7-branes.
Figure 3.4. A toric polygon and its dual 5-brane web, with all external 5-branes ending on 7-branes. The white lattice point on the boundary of the polygon indicates that the dual web contains parallel external 5-branes ending on the same 7-brane.
In general, configurations of 5-branes ending on 7-branes need not preserve supersymmetry and therefore need not engineer 5d SCFTs. In addition to charge conservation at junctions, supersymmetry imposes further constraints, notably the s-rule [22 , 32 , 55 , 56 ] , illustrated in Figure 3.5 . Following [33 ] , we consider a stronger notion of consistency , formulated in terms of the successive pulling of 7-branes into the web, as discussed below. A local version of this has recently been discussed in [57 ] .
Figure 3.5. Two configurations of parallel 5-branes ending on a common 7-brane, violating and satisfying the s-rule, respectively. In the configuration on the left, pulling the 7-brane through the junction would require a negative number of 5-branes, whereas in the configuration on the right, the 7-brane can be pulled through the junction and becomes detached from the 5-branes.
Starting from a configuration in which the external 5-branes end on 7-branes, we move a given [ p , q ] [p,q] [ p , q ] 7-brane along the direction of the 5-branes attached to it, towards the interior of the web. When the 7-brane reaches a junction, it can be pulled through it by applying a local Hanany–Witten transition, as detailed in [33 §4.5] . This operation changes the local configuration of 5-branes, while preserving charge conservation. In this way, the 7-brane can pass through successive junctions, changing the combinatorial structure of the web. By repeating such operations, a 7-brane can be pushed through several junctions. Eventually, it may become detached from all 5-branes and lie in a bounded face of the web. Although it no longer appears as an endpoint of a 5-brane, the 7-brane still carries its S L ( 2 , Z ) SL(2,\mathbb{Z}) S L ( 2 , Z ) monodromy, represented by a cut emanating from its position, as illustrated in Figure 3.6 and Figure 3.7 .
Figure 3.6. The 5-brane web of Figure 3.3 after all the 7-branes have been pulled into bounded faces.
Figure 3.7. The 5-brane web of Figure 3.4 after the 7-branes have been pulled into bounded faces.
Following [33 ] , we say that a 5-brane web with 7-branes is consistent if there exists a sequence of pulling operations, such that the resulting configuration satisfies the following properties:
All 7-branes lie inside bounded faces of the web, with no 5-branes ending on them.
Every vertex of the resulting 5-brane web is trivalent and has multiplicity one, equivalently, it is dual to a lattice triangle of normalized area one.
For example, Figure 3.6 and Figure 3.7 show consistent 5-brane webs with pulled 7-branes satisfying the above properties.
Not every 5-brane web is consistent. Indeed, the possibility of pulling a 7-brane through a junction without creating negative numbers of 5-branes depends on the charges and multiplicities of the adjacent 5-branes, and pulling one 7-brane can affect the subsequent motion of the others. Consistency is therefore a global condition on the web, stronger than charge conservation at individual junctions or the local s-rule. In what follows, we restrict to consistent 5-brane webs.
Generalized Toric Polygons
We now explain how the correspondence between 5-brane webs and toric polygons extends to the case of 5-brane webs with 7-branes.
Consider a consistent 5-brane web with 7-branes, as reviewed above in § 3.3 . If we temporarily disregard the 7-brane endpoints and extend the attached external 5-branes to infinity, the resulting web is dual to a toric lattice polygon P P P with a regular polyhedral decomposition P \mathcal{P} P . A vertex of the web is dual to a two-dimensional cell of P \mathcal{P} P , and a 5-brane segment is dual to an edge of P \mathcal{P} P . In particular, the external 5-brane segments are dual to edges of P \mathcal{P} P contained in the boundary of P P P . The lattice length of each such edge is the number of parallel 5-branes represented by the corresponding segment of the web.
The endpoints on 7-branes supply additional data on these boundary edges. For each 7-brane α \alpha α , denote by L α L_\alpha L α the boundary edge of P \mathcal{P} P dual to the 5-branes ending on it. If L α L_\alpha L α has lattice length m α m_\alpha m α , it represents m α m_\alpha m α parallel 5-branes ending on the same 7-brane. We draw the m α − 1 m_{\alpha}-1 m α − 1 interior lattice points of L α L_{\alpha} L α in white, with its two endpoints in black. The other lattice points of the polygon are drawn in black – see Figure 3.4 for an example. This is the familiar white-dot convention for generalized toric polygons [32 , 37 ] . An edge L α L_{\alpha} L α of lattice length one has no interior lattice point to color.
Now fix a sequence of pullings of the 7-branes as in the consistency condition of § 3.3 . Label the 7-branes by α \alpha α and color by α \alpha α each cell of P P P dual to a vertex of the original web through which the 7-brane α \alpha α passes during the pulling. The possible colored cells are elementary lattice trapezoids and elementary lattice triangles, as defined in [33 §4.5] . An elementary trapezoid has two parallel edges of lattice lengths a a a and b b b , with a > b a>b a > b , separated by a lattice distance of a − b a-b a − b , and two remaining edges of lattice length one. An elementary triangle has a distinguished base of lattice length a a a , two remaining edges of lattice length one, and lattice height a a a above the base. These are the cells dual to the vertices across which a 7-brane can be pulled. We refer to [33 §§4.5–4.6] for a detailed description of the corresponding local operations on the web.
For each α \alpha α , the α \alpha α -colored cells form a consecutive strip beginning at L α L_{\alpha} L α . The strip consists of zero or more elementary trapezoids, followed by a single elementary triangle [33 §5.1] . The edge of the first colored cell facing the boundary is L α L_{\alpha} L α . In each trapezoid the edge opposite this incoming edge is shorter and serves as the incoming edge of the next cell. The final triangle has no opposite edge: the strip ends at its third vertex, which we denote by V α V_{\alpha} V α . Under duality with the web, V α V_{\alpha} V α corresponds to the bounded face in which the 7-brane α \alpha α lies after being fully pulled in. The entire α \alpha α -colored strip records the vertices of the initial web that have disappeared during this pulling. Different colored strips contain no common two-dimensional cells, although their tips V α V_{\alpha} V α may coincide.
All cells of P P P that are not colored are lattice triangles of normalized area one, or equivalently Euclidean area 1 / 2 1/2 1/2 in a lattice basis. These cells correspond to the trivalent vertices of multiplicity one that remain in the web with pulled-in 7-branes. In particular, the distinction is between the uncolored unit triangles, whose dual web vertices remain, and the elementary triangles and trapezoids colored by the 7-branes, whose dual vertices have been removed. Existence of a coloring with precisely these properties is the polygon counterpart of the consistency condition of § 3.3 . The precise polygon criterion is given in [33 §5.1] .
We call the resulting data—the toric polygon P P P , its chosen regular polyhedral decomposition P \mathcal{P} P , the boundary edges L α L_{\alpha} L α , and the coloring of its cells determined by a complete pulling sequence—a generalized toric polygon , or GTP . In [33 ] , this structure is called a decorated toric polygon equipped with a consistent regular polyhedral decomposition. In this paper, we use the shorter name GTP for this data. A simple example is given in Figure 3.8 .
Figure 3.8. A 7-brane pulled into a bounded face of the web (left), with its monodromy cut indicated. The dual GTP (right) shows the colored triangular region dual to the vertex of the original web that disappears after the 7-brane is pulled.
Symington Polygons and E E E -lines
Starting from a GTP ( P , P ) (P,\mathcal{P}) ( P , P ) as defined in § 3.4 , we now construct an integral-affine manifold with singularities, called the Symington polygon P ~ \widetilde{P} P associated with ( P , P ) (P,\mathcal{P}) ( P , P ) [33 §5.2] .
Recall that, for each 7-brane α \alpha α , the α \alpha α -colored cells form a strip beginning at the marked boundary edge L α L_{\alpha} L α : a possibly empty sequence of elementary trapezoids followed by a single elementary triangle with tip V α V_{\alpha} V α . We remove the interior of this strip, together with L α L_{\alpha} L α , and identify the two lateral edges of each colored cell, starting at L α L_{\alpha} L α and proceeding towards V α V_{\alpha} V α . The lateral edges are lattice segments of length one, so each pair can be identified while preserving its lattice subdivision. At the final triangle the two sides meeting at V α V_{\alpha} V α are identified, leaving the image of V α V_{\alpha} V α as a point of the resulting polygon. We carry out this operation for every colored strip. In the dual brane web, the colored cells correspond to the junctions through which the 7-branes passed during the pulling procedure. Removing them eliminates precisely those junctions, while the edge identifications reproduce the adjacencies of the fully pushed web.
The identifications also specify the integral-affine structure of the resulting polygon. On each remaining triangle we retain the usual lattice-affine coordinates. Across a pair of identified edges, these coordinates are related by an affine transformation whose linear part belongs to S L ( 2 , Z ) SL(2,\mathbb{Z}) S L ( 2 , Z ) . For each colored cell this transformation preserves the lattice direction parallel to its incoming edge and identifies the primitive directions of its two lateral edges. Along a strip the transformations are composed in the order of the successive gluings. Thus the remaining triangles fit together by integral-affine coordinate changes, except possibly at the images of the tips V α V_{\alpha} V α . Their union P ~ \widetilde{P} P with these identifications is an integral-affine manifold with singularities at the vertices V α V_{\alpha} V α , and is called a Symington polygon.
The uncolored cells of P \mathcal{P} P are triangles of normalized lattice area one. They survive the cutting and gluing, and so their images give a unimodular triangulation P ~ \widetilde{\mathcal{P}} P of P ~ \widetilde{P} P . This triangulation is dual to the web after all 7-branes have been pulled in: faces of the web correspond to vertices of P ~ \widetilde{\mathcal{P}} P , 5-brane segments to edges, and web junctions to triangles. By the consistency condition of § 3.3 , every junction of the fully pushed web is trivalent and has multiplicity one, that is, its dual cell is a triangle of normalized area one. We summarize the cut-and-glue construction by
( P , P ) ⟼ ( P ~ , P ~ ) . (37) (\mathcal{P},\mathcal{P}) \longmapsto(\widetilde{P},\widetilde{\mathcal{P}}). \tag*{(37)} ( P , P ) ⟼ ( P , P ) . (37)
If there are no 7-branes, nothing is removed or glued, and P ~ = P \widetilde{P}=P P = P with its ordinary affine structure and triangulation.
The simplest nontrivial cut-and-glue operation, involving a single colored triangle, is shown in Figure 3.9 . Its two lateral edges are glued after the shaded triangle is removed. To draw the result in the plane, one may introduce a monodromy cut starting at the image of the tip. This cut records the change of affine coordinates between the two sides. With the crossing orientation used in the figure, the lattice directions of the polygon transform according to M [ 0 , 1 ] T = ( 1 − 1 0 1 ) M^{T}_{[0,1]}=\begin{pmatrix}1&-1\\0&1\end{pmatrix} M [ 0 , 1 ] T = ( 1 0 − 1 1 ) . For example, the vectors ( 1 , 4 ) (1,4) ( 1 , 4 ) and ( 1 , 3 ) (1,3) ( 1 , 3 ) are represented on the other side by ( − 3 , 4 ) (-3,4) ( − 3 , 4 ) and ( − 2 , 3 ) (-2,3) ( − 2 , 3 ) , respectively. Crossing the cut in the opposite direction uses the inverse matrix. The transpose here follows the convention for polygon lattice directions dual to the 5-brane charges, whose monodromy M [ p , q ] M_{[p,q]} M [ p , q ] was introduced in § 3.3 .
Figure 3.9. Cutting and gluing a colored triangle. The shaded triangle is removed and its two lateral edges are identified. On the right, the resulting Symington polygon is displayed using a monodromy cut and the indicated lattice vectors are related across the cut by M [ 0 , 1 ] T M^{T}_{[0,1]} M [ 0 , 1 ] T . The dashed cut is not a boundary of the polygon.
The only points at which the integral-affine structure is singular are the images of the tips V α V_{\alpha} V α . If a vertex is the tip of a single colored strip, its image is a focus–focus singularity , as familiar from almost-toric geometry [58 , 59 ] and the theory of integral-affine manifolds with singularities [44 , 45 , 60 , 61 , 62 ] . Transporting a lattice direction around a small loop encircling this point gives a nontrivial integral shear. If two or more strips have the same tip, their images produce one singular point whose monodromy combines the contributions of the corresponding 7-branes. Such a point need not be a single focus–focus singularity. Away from the images of the tips, the coordinate changes constructed above provide an ordinary integral-affine structure.
It will often be convenient to separate any coincident singularities. This can be achieved by an arbitrarily small perturbation of the integral-affine structure near a common tip, replacing it by separate focus–focus singularities, one for each 7-brane. The perturbation is not unique: in particular, it involves a choice of the relative positions of the singularities. We fix such a choice and continue to write V α V_{\alpha} V α for the singularity associated with the 7-brane α \alpha α .
The monodromy around each V α V_{\alpha} V α fixes a primitive lattice direction. In local affine coordinates this direction is parallel to the marked boundary edge L α L_{\alpha} L α of the original GTP, transported to a neighborhood of V α V_{\alpha} V α through the gluings. Thus it is determined by the external 5-branes that originally ended on the 7-brane α \alpha α , to which L α L_\alpha L α is dual. The invariant direction is unoriented and determines two opposite rays issuing from V α V_\alpha V α . Either ray may be continued in the Symington polygon.
An E E E -line is the continuation of one of these rays as a straight line in the integral-affine structure of P ~ \widetilde{P} P . Away from the singularities, its direction is constant when expressed in any local affine coordinates. If a planar drawing uses a monodromy cut, the coordinates on its two sides differ by the corresponding S L ( 2 , Z ) SL(2,\mathbb{Z}) S L ( 2 , Z ) transformation. An E E E -line may therefore appear to bend upon crossing the cut, although it is straight in the integral-affine sense. The change in its drawn direction is the same change of coordinates illustrated in Figure 3.9 . Moreover, for a generic perturbation of all the singularities V α V_\alpha V α , an E E E -line will never cross a singularity, except possibly when the direction of the E E E -line coincides with the monodromy-invariant direction of the singularity. In this latter case, the E E E -line is extended across the singularity without ambiguity by following the monodromy-invariant direction.
For the Symington polygons arising from the consistent GTPs considered here, every E E E -line reaches the boundary of P ~ \widetilde{P} P rather than continuing indefinitely in its interior. This can be deduced from the existence of an associated Calabi–Yau geometry, as described in § 4 . A simple example of E E E -lines is illustrated in Figure 3.10 .
Figure 3.10. On the left: the E E E -lines E ~ 1 \widetilde{E}_1 E 1 and E ~ 2 \widetilde{E}_2 E 2 emerging from a singularity in a Symington polygon. The shaded triangle marks the region removed in the cut-and-glue construction; its lateral edges are identified. On the right: the dual paths starting from the 7-branes in the brane web.
An E E E -line also determines a dual path in the 5-brane web with pulled 7-branes. The point V α V_\alpha V α is dual to the bounded face containing the 7-brane α \alpha α , so the path starts at that 7-brane. When the E E E -line crosses an edge of P ~ \widetilde{P} P and enters an adjacent triangle, the dual path follows the corresponding 5-brane segment to the next web junction. In this way the path records the sequence of edges and triangles met by the E E E -line. However, the exact shape of the path is not fixed beyond these incidence conditions. In particular, it can be drawn with curved segments in the ( x 5 , x 6 ) (x^5,x^6) ( x 5 , x 6 ) plane, whereas the E E E -line is straight in the integral-affine structure of the Symington polygon. Examples illustrating these dual paths are given in Figure 3.10 and Figure 3.11 .
Figure 3.11. E E E -lines in the Symington polygon (left) and their dual paths in the 5-brane web with pulled 7-branes (right). The paths follow the dual combinatorial data and need not be straight in the web plane.
Generalized Toric Calabi–Yau Threefolds
In this section, we review the construction of Calabi–Yau threefolds from GTPs developed in [33 ] , and identify the divisors and curves relevant for the intersection theory studied in the following sections. In Section 4.1, we first recall the necessary properties of log Calabi–Yau surfaces, then explain how such surfaces are associated with the vertices of a Symington polygon and glued to form the central fiber of a one-parameter family of surfaces whose total space is a Calabi–Yau threefold. In Section 4.2, we describe the divisors and curves of the resulting threefold, distinguishing those arising from the central fiber from the additional non-compact divisors associated with E E E -lines, which we call exuberant divisors.
From Generalized Toric Polygons to Calabi–Yau Threefolds
We fix a GTP ( P , P ) (P,\mathcal{P}) ( P , P ) as in Section 3.4 and consider the associated Symington polygon P ~ \widetilde{P} P , equipped with its unimodular triangulation P ~ \widetilde{\mathcal{P}} P , as constructed in Section 3.5. We denote by P ~ [ 0 ] \widetilde{\mathcal{P}}^{[0]} P [ 0 ] , P ~ [ 1 ] \widetilde{\mathcal{P}}^{[1]} P [ 1 ] , and P ~ [ 2 ] \widetilde{\mathcal{P}}^{[2]} P [ 2 ] the sets of vertices, edges, and triangles of this triangulation, respectively. We first explain how to associate a surface to each vertex of P ~ \widetilde{\mathcal{P}} P and then how to glue these surfaces along curves corresponding to its edges. The resulting surface is the central fiber of a one-parameter family of surfaces whose total space will be the Calabi–Yau threefold associated to the GTP.
Surfaces from Vertices of the Symington Polygon
For each vertex v ∈ P ~ [ 0 ] v \in\widetilde{\mathcal{P}}^{[0]} v ∈ P [ 0 ] , we explain below how to associate a smooth surface X v X_v X v and a normal crossing divisor ∂ X v \partial X_v ∂ X v on X v X_v X v , referred to as the boundary of X v X_v X v . Every pair ( X v , ∂ X v ) (X_v,\partial X_v) ( X v , ∂ X v ) will be log Calabi–Yau, that is, satisfying
K X v + ∂ X v = 0 , (38) K_{X_v}+\partial X_v=0, \tag*{(38)} K X v + ∂ X v = 0 , (38)
where K X v K_{X_v} K X v denotes the canonical divisor of X v X_v X v .
If v v v is a smooth point of the integral-affine structure on P ~ \widetilde{P} P , that is v ≠ V α v\ne V_\alpha v = V α for every 7-brane α \alpha α , the primitive directions of the edges emanating from it form a fan Σ v \Sigma_v Σ v . The unimodularity of P ~ \widetilde{P} P ensures that adjacent rays form a basis of the local lattice. We take X v X_v X v to be the corresponding smooth toric surface and ∂ X v \partial X_v ∂ X v its toric boundary. The surface is compact when v v v is an interior vertex of P ~ \widetilde{P} P , and non-compact when v v v lies on its boundary. Equivalently, X v X_v X v is the toric surface with momentum polytope given by the face of the 5-brane web dual to v v v .
Now suppose that v v v is a singular point of the integral-affine structure on P ~ \widetilde{P} P , that is, the image in the Symington polygon of tips V α V_\alpha V α of the α \alpha α -colored strips associated to 7-branes α \alpha α . Before the corresponding cuts producing the Symington polygon, the star of the vertex in the polyhedral decomposition P \mathcal{P} P determines a toric surface ( X ‾ v , ∂ X ‾ v ) (\overline{X}_v,\partial\overline{X}_v) ( X v , ∂ X v ) , possibly singular. Equivalently, X ‾ v \overline{X}_v X v is the toric surface with momentum polytope given by the face of the 5-brane web dual to v v v before pulling the 7-brane. Each cone removed by a cut corresponds to a zero-dimensional stratum of the toric divisor ∂ X ‾ v \partial\overline{X}_v ∂ X v . The removed cone is the cone over an elementary lattice triangle whose base has the same lattice length n n n as its lattice height. For n > 1 n>1 n > 1 the associated affine toric surface has a Wahl singularity at the corresponding zero-dimensional toric stratum,
C 2 / ( Z / n 2 Z ) , (39) \mathbb{C}^2/(\mathbb{Z}/n^2\mathbb{Z}), \tag*{(39)} C 2 / ( Z / n 2 Z ) , (39)
where the cyclic group acts with weights ( 1 , a n − 1 ) (1,an-1) ( 1 , an − 1 ) for an integer a a a with gcd ( a , n ) = 1 \gcd(a,n)=1 g cd( a , n ) = 1 [63 ] . Equivalently, the singularity is the quotient of x y = z n xy=z^n x y = z n by Z / n Z \mathbb{Z}/n\mathbb{Z} Z / n Z , acting on ( x , y , z ) (x,y,z) ( x , y , z ) with weights ( 1 , − 1 , a ) (1,-1,a) ( 1 , − 1 , a ) . The surface X v X_v X v is obtained from X ‾ v \overline{X}_v X v by smoothing these Wahl singularities. Locally, these smoothings can be described as the quotient of
x y = z n + t (40) xy=z^n+t \tag*{(40)} x y = z n + t (40)
by the same Z / n Z \mathbb{Z}/n\mathbb{Z} Z / n Z -action. For t ≠ 0 t\ne0 t = 0 it smooths the surface singularity and the corresponding corner of ∂ X ‾ v \partial\overline{X}_v ∂ X v . For n = 1 n=1 n = 1 there is no surface singularity, and the corresponding operation replaces the normal-crossing divisor x y = 0 xy=0 x y = 0 by x y = t xy=t x y = t in a smooth surface. These local deformations can be carried out simultaneously for all cut cones [33 §6.2, Step I] . The result of this deformation is a smooth log Calabi–Yau pair ( X v , ∂ X v ) (X_v,\partial X_v) ( X v , ∂ X v ) , where the divisor ∂ X v \partial X_v ∂ X v is obtained by smoothing precisely the corners of the toric divisor ∂ X ‾ v \partial\overline{X}_v ∂ X v selected by the cuts. The deformation of the toric surface ( X ‾ v , ∂ X ‾ v ) (\overline{X}_v,\partial\overline{X}_v) ( X v , ∂ X v ) into the non-toric surface ( X v , ∂ X v ) (X_v,\partial X_v) ( X v , ∂ X v ) is the geometric counterpart of pulling 7-branes into the face of the 5-brane web dual to v v v . In particular, the face with 7-branes pulled in is a non-toric generalization of the toric momentum polytope, called a Symington polytope for the non-toric surface ( X v , ∂ X v ) (X_v,\partial X_v) ( X v , ∂ X v ) [58 , 59 ] . The complement X v ∖ ∂ X v X_v\setminus\partial X_v X v ∖ ∂ X v admits a torus fibration over this polytope, with nodal singular fibers over the 7-branes. In particular, the number of 7-branes in the face dual to v v v is equal to the charge Q X v Q_{X_v} Q X v of the log Calabi–Yau surface ( X v , ∂ X v ) (X_v,\partial X_v) ( X v , ∂ X v ) , that is, the topological Euler characteristic of the complement X v ∖ ∂ X v X_v\setminus\partial X_v X v ∖ ∂ X v .
Remark 4.1. Elementary triangles are also known as primitive T-cones , and some aspects of their relation with GTPs have also been explored in [39 , 40 , 41 , 42 , 57 ] . For the present paper, the important point is the observation made in [33 ] that the smoothing of the corresponding Wahl singularities is a basic operation in the construction of the Calabi–Yau threefold corresponding to a GTP.
Gluing, Smoothing, and the Generalized Toric Calabi–Yau Threefold
In § 4.1.1 , we constructed smooth surfaces X v X_v X v by deformation of possibly singular toric surfaces X ‾ v \overline{X}_v X v . We now explain how to glue these surfaces together to produce a normal crossing surface X ~ 0 \widetilde{\mathcal{X}}_0 X 0 .
For each edge e = [ v , w ] ∈ P ~ [ 1 ] e=[v,w]\in\widetilde{\mathcal{P}}^{[1]} e = [ v , w ] ∈ P [ 1 ] , the boundaries ∂ X v \partial X_v ∂ X v and ∂ X w \partial X_w ∂ X w contain curves corresponding to the primitive directions of e e e at v v v and w w w , respectively. We identify these curves according to the gluing data of the Symington triangulation. Performing these identifications for every edge produces a normal crossing surface
X ~ 0 = ⋃ v ∈ P ~ [ 0 ] X v . (41) \widetilde{\mathcal{X}}_0=\bigcup_{v\in\widetilde{\mathcal{P}}^{[0]}}X_v. \tag*{(41)} X 0 = v ∈ P [ 0 ] ⋃ X v . (41)
Each triangle of P ~ \widetilde{\mathcal{P}} P records a point where three components meet. Thus the vertices, edges, and triangles of P ~ \widetilde{\mathcal{P}} P correspond to the components, double curves, and triple points of the normal crossing surface X ~ 0 \widetilde{\mathcal{X}}_0 X 0 , as summarized in the table below:
Table 4.5.
The gluing can be chosen so that the normal crossing surface X ~ 0 \widetilde{\mathcal{X}}_0 X 0 admits a one-parameter smoothing [33 §6.2, Step III] . More precisely, there is a smooth non-compact Calabi–Yau threefold X ~ \widetilde{\mathcal{X}} X and a holomorphic map
π : X ~ ⟶ Δ , π − 1 ( 0 ) = X ~ 0 , (42) \pi:\widetilde{\mathcal{X}}\longrightarrow\Delta,\qquad\pi^{-1}(0)=\widetilde{\mathcal{X}}_0, \tag*{(42)} π : X ⟶ Δ , π − 1 ( 0 ) = X 0 , (42)
where Δ ⊂ C \Delta\subset\mathbb{C} Δ ⊂ C is a disk about the origin. We refer to the smooth non-compact Calabi–Yau threefold X ~ \widetilde{\mathcal{X}} X obtained by this construction as the generalized toric Calabi–Yau threefold associated with the given GTP.
Note that the deformation of the normal crossing surface X ~ 0 \widetilde{\mathcal{X}}_0 X 0 over Δ \Delta Δ to produce the threefold X ~ \widetilde{\mathcal{X}} X is distinct from the deformations considered in § 4.1.1 to construct the surfaces X v X_v X v . By construction, the surface components X v X_v X v of X ~ 0 \widetilde{\mathcal{X}}_0 X 0 are divisors in the threefold X ~ \widetilde{\mathcal{X}} X . Those associated with interior vertices are compact, whereas those associated with boundary vertices are non-compact. Contracting all compact curves and all compact divisors in X ~ \widetilde{\mathcal{X}} X yields the associated canonical Calabi–Yau threefold singularity
X ~ ⟶ X . (43) \widetilde{\mathcal{X}} \longrightarrow\mathcal{X}. \tag*{(43)} X ⟶ X . (43)
According to [33 ] , the 5d SCFT defined by a consistent 5-brane web with 7-branes in Type IIB string theory admits a dual description as M-theory on the canonical Calabi–Yau threefold singularity X \mathcal{X} X . The choice of unimodular triangulation of the Symington polygon specifies the resolved geometry and hence a chamber of the extended Coulomb branch. In the absence of 7-branes, the GTP is just a toric polygon, and the above construction recovers the associated toric Calabi–Yau threefold.
Remark 4.2. An alternative description of the threefolds X ~ \widetilde{\mathcal{X}} X and X \mathcal{X} X in terms of scattering diagrams and broken lines is given in [33 §6.4] and further developed in [46 ] . For our intersection-theoretic purposes, however, it is more convenient to realize the threefold as the total space of a smoothing of an explicit normal crossings surface, as reviewed in this section.
Divisors and Curves
We now identify the divisors and curves of the generalized toric Calabi–Yau threefold X ~ \widetilde{\mathcal{X}} X relevant for the intersection theory developed in § 5 . We first describe the divisors arising from the components of the central fiber X ~ 0 \widetilde{\mathcal{X}}_0 X 0 and the curves along which these components intersect. We then turn to the additional non-compact divisors associated with E E E -lines, which we call exuberant divisors , and describe their intersections with the central fiber. Finally, we introduce the notation for these divisors and curves and summarize the incidence data needed to compute their intersection products.
Divisors and Double Curves of the Central Fiber
For every vertex v i ∈ P ~ [ 0 ] v_i \in\widetilde{\mathcal{P}}^{[0]} v i ∈ P [ 0 ] , consider the divisor
D i : = X v i ⊂ X ~ (44) D_i := X_{v_i} \subset\widetilde{\mathcal{X}} \tag*{(44)} D i := X v i ⊂ X (44)
given by the corresponding irreducible component of the normal crossing surface X ~ 0 \widetilde{\mathcal{X}}_0 X 0 . We denote the index set of these divisors by I 0 I_0 I 0 and write
I 0 = I c ⊔ I 0 , ∂ , (45) I_0 = I_c \sqcup I_{0,\partial}, \tag*{(45)} I 0 = I c ⊔ I 0 , ∂ , (45)
where I c I_c I c indexes the interior vertices and the corresponding compact divisors, while I 0 , ∂ I_{0,\partial} I 0 , ∂ indexes boundary vertices and non-compact divisors.
An edge e a = [ v i , v j ] ∈ P ~ [ 1 ] e_a = [v_i,v_j] \in\widetilde{\mathcal{P}}^{[1]} e a = [ v i , v j ] ∈ P [ 1 ] specifies an irreducible double curve C a ⊂ D i ∩ D j C_a \subset D_i \cap D_j C a ⊂ D i ∩ D j . The full intersection D i ∩ D j D_i \cap D_j D i ∩ D j may contain several such curves if more than one edge joins v i v_i v i and v j v_j v j after the affine identifications. An explicit example where this happens is provided in § 7.1 . Let A 0 A_0 A 0 index the double curves and set
A 0 = A 0 , c ⊔ A 0 , ∂ . (46) A_0 = A_{0,c} \sqcup A_{0,\partial}. \tag*{(46)} A 0 = A 0 , c ⊔ A 0 , ∂ . (46)
with A 0 , c A_{0,c} A 0 , c corresponding to interior edges, whose curves are compact, and A 0 , ∂ A_{0,\partial} A 0 , ∂ to boundary edges, whose curves are non-compact. The incidence relation is
C a ⊂ D i ⟺ v i is an endpoint of e a . (47) C_a \subset D_i \Longleftrightarrow v_i\text{ is an endpoint of }e_a. \tag*{(47)} C a ⊂ D i ⟺ v i is an endpoint of e a . (47)
In particular, the irreducible curves C a C_a C a must be distinguished from the possibly reducible full intersection of two divisors.
Exuberant Divisors and E-lines
The components D i D_i D i are not all the divisors relevant to the intersection theory of X ~ \widetilde{X} X . In the toric presentation, a divisor is associated with a lattice vertex, or equivalently a face of the dual 5-brane web. The non-toric deformation changes this picture: the smoothing of a boundary corner associated with a pulled-in 7-brane can identify toric boundary pieces, while the resulting threefold has further non-compact divisors not contained in X ~ 0 \widetilde{X}_0 X 0 , that we call exuberant divisors .
Let I e I_e I e be an index set for the 7-branes. As in § 3.5 , we perturb infinitesimally the position of the focus–focus singularities V α V_\alpha V α in the Symington polygons for every α ∈ I e \alpha\in I_e α ∈ I e . Fix some α ∈ I e \alpha\in I_e α ∈ I e and denote by v v v the vertex of the Symington polygon at which the singularity V α V_\alpha V α is originally located before perturbation. As also described in § 3.5 , there are two E E E -lines emanating from V α V_\alpha V α . We denote by ℓ α \ell_\alpha ℓ α the E E E -line starting from V α V_\alpha V α and not passing through v v v , as illustrated in Figure 4.1 . As in the tropical correspondence theorem [64 ] , the E E E -line ℓ α \ell_\alpha ℓ α determines an algebraic curve L α L_\alpha L α in the normal crossings surface X ~ 0 \widetilde{X}_0 X 0 , whose tropicalization is ℓ α \ell_\alpha ℓ α . The irreducible components of L α L_\alpha L α correspond to the vertices of the triangulation traversed by ℓ α \ell_\alpha ℓ α before perturbation, with adjacent components meeting at nodes corresponding to the segments joining these vertices. In particular, the initial vertex v v v determines a distinguished irreducible component C α C_\alpha C α of L α L_\alpha L α , which we call the exuberant curve associated with ℓ α \ell_\alpha ℓ α .
Figure 4.1. The exuberant curve C α C_{\alpha} C α in X v X_v X v , image of an exceptional curve in X ~ v \widetilde{X}_v X v
The exuberant curve C α C_\alpha C α admits an explicit description in terms of a toric model of the log Calabi–Yau surface ( X v , ∂ X v ) (X_v,\partial X_v) ( X v , ∂ X v ) , that is, a presentation of ( X v , ∂ X v ) (X_v,\partial X_v) ( X v , ∂ X v ) as a blow-down of a non-toric blow-up of a toric surface. As explained in [65 ] , a generic perturbation separating the integral-affine singularities at v v v into distinct focus–focus singularities determines a choice of toric model for ( X v , ∂ X v ) (X_v,\partial X_v) ( X v , ∂ X v ) . Let Σ \Sigma Σ be the fan obtained from the primitive directions of the edges of P ~ \widetilde{P} P emanating from v v v by adding rays in the monodromy-invariant directions of the perturbed focus–focus singularities. Denote by X Σ X_\Sigma X Σ the corresponding toric surface. For each singularity V β V_\beta V β originally located at v v v , blow up a general point of the toric boundary component of X Σ X_\Sigma X Σ corresponding to its monodromy-invariant direction. These blow-ups are performed away from the toric fixed points. Let X ~ v \widetilde{X}_v X v be the resulting surface, with boundary ∂ X ~ v \partial\widetilde{X}_v ∂ X v given by the strict transform of the toric boundary of X ‾ Σ \overline{X}_\Sigma X Σ . The surface X v X_v X v is then obtained by contracting those irreducible components of ∂ X ~ v \partial\widetilde{X}_v ∂ X v corresponding to rays of Σ \Sigma Σ that are not directions of edges of P ~ \widetilde{P} P emanating from v v v . Under this contraction X ~ v → X v \widetilde{X}_v \to X_v X v → X v , the exceptional curve of the blow-up X ~ v → X ‾ Σ \widetilde{X}_v \to\overline{X}_\Sigma X v → X Σ associated with V α V_\alpha V α maps to the exuberant curve C α ⊂ X v C_\alpha\subset X_v C α ⊂ X v , as illustrated in Figure 4.1 . In particular, C α C_\alpha C α is a compact rational curve whose normalization is isomorphic to the projective line P 1 \mathbb{P}^1 P 1 . We denote by A e A_e A e the index set of these compact curves, with the natural bijection A e ≅ I e A_e \cong I_e A e ≅ I e .
Finally, the exuberant divisor E α E_{\alpha} E α is the total space of a one-parameter smoothing of the curve L α L_{\alpha} L α in X ~ 0 \widetilde{X}_0 X 0 with tropicalization the E E E -line ℓ α \ell_{\alpha} ℓ α . The existence of this smoothing follows as in the tropical correspondence theorem [64 ] . The exuberant divisor E α E_{\alpha} E α is a non-compact divisor in X ~ \widetilde{X} X , whose intersection with the central fiber X ~ 0 \widetilde{X}_0 X 0 is the reducible curve L α L_{\alpha} L α :
E α ∩ X ~ 0 = L α . (48) E_{\alpha} \cap\widetilde{X}_0 = L_{\alpha}. \tag*{(48)} E α ∩ X 0 = L α . (48)
The two types of divisors, components of the central fiber and exuberant divisors, may be summarized as follows:
Table 4.13.
Accordingly, the divisor index set used below is
I = I 0 ⊔ I e = I c ⊔ I 0 , ∂ ⊔ I e , (49) I = I_0 \sqcup I_e = I_c \sqcup I_{0,\partial} \sqcup I_e, \tag*{(49)} I = I 0 ⊔ I e = I c ⊔ I 0 , ∂ ⊔ I e , (49)
and for every i ∈ I i \in I i ∈ I , we will denote by D i D_i D i the corresponding divisor. In particular, for i = α ∈ I e i = \alpha\in I_e i = α ∈ I e , we will have D i = E α D_i = E_{\alpha} D i = E α . The curve index set used below is
A = A 0 ⊔ A e = A 0 , c ⊔ A 0 , ∂ ⊔ A e , A c = A 0 , c ⊔ A e , (50) A = A_0 \sqcup A_e = A_{0,c} \sqcup A_{0,\partial} \sqcup A_e,\qquad A_c = A_{0,c} \sqcup A_e, \tag*{(50)} A = A 0 ⊔ A e = A 0 , c ⊔ A 0 , ∂ ⊔ A e , A c = A 0 , c ⊔ A e , (50)
where A 0 A_0 A 0 is the index set of double curves in the central fiber, A e A_e A e is the index set of exuberant curves, A 0 , c A_{0,c} A 0 , c is the index set of compact double curves, A 0 , ∂ A_{0,\partial} A 0 , ∂ is the index set of non-compact double curves, and A c A_c A c is the index set of compact curves (either double curves or exuberant). For every a ∈ A a \in A a ∈ A , we will denote by C a C_a C a the corresponding curve.
Remark 4.3. Every divisor on X ~ \widetilde{X} X defines by Poincaré duality a class in H 2 ( X ~ , Z ) H^2(\widetilde{X},\mathbb{Z}) H 2 ( X , Z ) . It follows from [43 ] , Lemma 3.11-Corollary 8.14 that the classes of the divisors ( D i ) i ∈ I 0 (D_i)_{i\in I_0} ( D i ) i ∈ I 0 and ( E α ) α ∈ I e (E_\alpha)_{\alpha\in I_e} ( E α ) α ∈ I e span H 2 ( X ~ , Z ) H^2(\widetilde{X},\mathbb{Z}) H 2 ( X , Z ) and satisfy three independent linear relations, so that
rank H 2 ( X ~ , Z ) = ∣ I 0 ∣ + ∣ I e ∣ − 3 , \operatorname{rank} H^2(\widetilde{X},\mathbb{Z}) = |I_0| + |I_e| - 3, rank H 2 ( X , Z ) = ∣ I 0 ∣ + ∣ I e ∣ − 3 ,
where:
∣ I 0 ∣ |I_0| ∣ I 0 ∣ is the cardinality of I 0 I_0 I 0 , that is, the number of lattice points in the Symington polygon P ~ \widetilde{P} P ,
∣ I e ∣ |I_e| ∣ I e ∣ is the cardinality of I e I_e I e , that is, the total number of 7-branes in the web description, or equivalently the total number of focus–focus singularities in the Symington polygon after generic perturbation. We also have
∣ I e ∣ = ∑ v ∈ I 0 Q X v , |I_e| = \sum_{v \in I_0} Q_{X_v}, ∣ I e ∣ = v ∈ I 0 ∑ Q X v ,
where Q X v Q_{X_v} Q X v is the charge of the log Calabi–Yau surface ( X v , ∂ X v ) (X_v, \partial X_v) ( X v , ∂ X v ) defined at the end of Section 4.1.1.
The divisors ( D i ) i ∈ I 0 (D_i)_{i \in I_0} ( D i ) i ∈ I 0 always satisfy the relation
∑ i ∈ I 0 D i = 0 (51) \sum_{i \in I_0} D_i = 0 \tag*{(51)} i ∈ I 0 ∑ D i = 0 (51)
in H 2 ( X ~ , Z ) H^2(\widetilde{X},\mathbb{Z}) H 2 ( X , Z ) . Indeed, ∑ i ∈ I 0 D i \sum_{i \in I_0} D_i ∑ i ∈ I 0 D i is the class of the central fiber X ~ 0 \widetilde{X}_0 X 0 , which is always a principal divisor since it is the zero-locus of the projection π : X ~ → Δ \pi: \widetilde{X} \to\Delta π : X → Δ . The total number of independent linear relations among the divisors ( D i ) i ∈ I 0 (D_i)_{i \in I_0} ( D i ) i ∈ I 0 in H 2 ( X ~ , Z ) H^2(\widetilde{X},\mathbb{Z}) H 2 ( X , Z ) depends on the 7-brane configuration of the web underlying the GTP. If the web contains at least two 7-branes with distinct [ p , q ] [p,q] [ p , q ] charges, (4.16) is the unique such relation. This number increases to two when all 7-branes have identical charges, and to three in the absence of 7-branes, corresponding to the toric case.
Example
In this section, we illustrate the geometry of exuberant divisors in a particular example. Consider the GTP of Figure 3.8 , with E E E -lines represented in Figure 3.10 . We first consider the toric geometry, without 7-branes pulled in. The toric Calabi–Yau threefold can be visualized by thickening the web drawing in a third transverse direction, so that the faces of the web appear as surfaces in an ambient three-dimensional space. These faces give a representation of the toric divisors D i D_i D i in the corresponding toric Calabi–Yau threefold.
Figure 4.17.
After the [ 1 , − 1 ] [1,-1] [ 1 , − 1 ] 7-brane is pulled into a bounded face, the web and its monodromy cut take the form
Figure 4.18.
The central-fiber divisors, now denoted by D ~ i \widetilde{D}_i D i in the diagrams to distinguish them from the toric presentation, remain associated with faces:
Figure 4.19.
The cut identifies the pieces D ~ 4 ′ \widetilde{D}'_4 D 4 ′ and D ~ 4 ′ ′ \widetilde{D}''_4 D 4 ′′ of the face marked 4. Finally, we illustrate the exuberant divisors corresponding to the two E E E -lines: they are not represented by faces of the web but by surfaces transverse to the face presentation. The first is E ~ 1 \widetilde{E}_1 E 1 ,
Figure 4.20.
and the second is E ~ 2 \widetilde{E}_2 E 2 :
Intersection Theory
Intersection Products
In this section, we establish the intersection-theoretic framework used to compute triple intersection numbers on the generalized toric Calabi–Yau threefold X ~ \widetilde{X} X constructed in Section 4. Our ultimate goal is to determine the intrinsic triple intersection numbers of divisors, namely those involving at least one compact divisor. However, the reconstruction of these numbers will require a larger collection of intersection products, including curve-valued products of non-compact divisors and numerical intersections with non-compact curves whenever these are well-defined. We first recall the relevant intersection products on a non-compact threefold. We then introduce the divisor–curve intersection coefficients and characterize, in terms of the combinatorial data of the Symington polygon and its E E E -lines, the pairs for which numerical intersections are intrinsically defined.
Intersection Products on non-compact Calabi–Yau Threefolds
Let X ~ \widetilde{X} X be the generalized toric Calabi–Yau threefold associated with a GTP, as constructed in Section 4. We are interested in the triple intersection form
κ i j k : = D i ⋅ D j ⋅ D k , { i , j , k } ∩ I c ≠ ∅ , (52) \kappa_{ijk} := D_i \cdot D_j \cdot D_k,\qquad\{i,j,k\}\cap I_c \ne\varnothing, \tag*{(52)} κ ij k := D i ⋅ D j ⋅ D k , { i , j , k } ∩ I c = ∅ , (52)
where I c I_c I c denotes the set of compact divisors. On a non-compact threefold, it is important to distinguish between curve-valued intersections of divisors and numerical intersection products. In particular, while intersections of non-compact divisors may be well-defined as curve classes, their numerical triple intersections are not necessarily intrinsic. Compact divisors define Poincaré dual classes in the cohomology H c 2 ( X ~ , Z ) H^2_c(\widetilde{X},\mathbb{Z}) H c 2 ( X , Z ) with compact support, whereas possibly non-compact divisors define Poincaré dual classes in the cohomology H 2 ( X ~ , Z ) H^2(\widetilde{X},\mathbb{Z}) H 2 ( X , Z ) . The intersection of two possibly non-compact divisors produces a possibly non-compact curve class, which defines a Poincaré dual class in H 4 ( X ~ , Z ) H^4(\widetilde{X},\mathbb{Z}) H 4 ( X , Z ) . Equivalently, this intersection product for possibly non-compact divisors is given by the cup-product
H 2 ( X ~ , Z ) × H 2 ( X ~ , Z ) ⟶ H 4 ( X ~ , Z ) . (53) H^2(\widetilde{X},\mathbb{Z}) \times H^2(\widetilde{X},\mathbb{Z}) \longrightarrow H^4(\widetilde{X},\mathbb{Z}). \tag*{(53)} H 2 ( X , Z ) × H 2 ( X , Z ) ⟶ H 4 ( X , Z ) . (53)
Thus, neither divisor needs to be compact for the curve-valued product D i ⋅ D j D_i \cdot D_j D i ⋅ D j to be defined.
Now let C ⊂ X ~ C \subset\widetilde{X} C ⊂ X be a possibly non-compact curve. Whenever the geometric intersection D ∩ C D \cap C D ∩ C is compact, the corresponding zero-dimensional intersection cycle has a well-defined degree. In this case, we write
D ⋅ C ∈ Z . (54) D \cdot C \in\mathbb{Z}. \tag*{(54)} D ⋅ C ∈ Z . (54)
If instead D ∩ C D \cap C D ∩ C contains a non-compact component, no intrinsic numerical intersection is assigned without additional data at infinity.
Finally, if at least one of the three divisors is compact, their triple intersection is intrinsically defined through
H c 2 ( X ~ ) × H 2 ( X ~ ) × H 2 ( X ~ ) ⟶ H c 6 ( X ~ ) ⟶ ∼ Z . (55) H_c^2(\widetilde{X}) \times H^2(\widetilde{X}) \times H^2(\widetilde{X}) \longrightarrow H_c^6(\widetilde{X}) \overset{\sim}{\longrightarrow} \mathbb{Z}. \tag*{(55)} H c 2 ( X ) × H 2 ( X ) × H 2 ( X ) ⟶ H c 6 ( X ) ⟶ ∼ Z . (55)
In particular, if D i D_i D i is compact, then
D i ⋅ D j ⋅ D k = ∫ X ~ ω i ∧ ω j ∧ ω k (56) D_i \cdot D_j \cdot D_k = \int_{\widetilde{X}} \omega_i \wedge\omega_j \wedge\omega_k \tag*{(56)} D i ⋅ D j ⋅ D k = ∫ X ω i ∧ ω j ∧ ω k (56)
is well-defined for arbitrary divisors D j , D k D_j,D_k D j , D k .
These distinctions will play an essential role in what follows. We will use both curve-valued divisor products and admissible numerical divisor–curve intersections to obtain compatibility relations, including relations involving non-compact cycles. The final output, however, will consist only of the intrinsic triple intersection numbers containing at least one compact divisor. As reviewed in § 2 , these triple intersection numbers determine the prepotential of the 5d SCFT engineered by M-theory on X ~ \widetilde{X} X , that is, the 5d SCFT associated with the GTP.
Divisors, Curves, and Intersection Coefficients
We use the notation introduced in § 4.2 . Recall that the divisors are indexed by
I = I 0 ⊔ I e , I 0 = I c ⊔ I 0 , ∂ , (57) I = I_0 \sqcup I_e,\qquad I_0 = I_c \sqcup I_{0,\partial}, \tag*{(57)} I = I 0 ⊔ I e , I 0 = I c ⊔ I 0 , ∂ , (57)
where I 0 I_0 I 0 indexes the components of the central fiber, I c I_c I c the compact components, I 0 , ∂ I_{0,\partial} I 0 , ∂ the non-compact components, and I e I_e I e the exuberant divisors associated with E E E -lines.
Similarly, the curves are indexed by
A = A 0 ⊔ A e , A 0 = A 0 , c ⊔ A 0 , ∂ , (58) A = A_0 \sqcup A_e,\qquad A_0 = A_{0,c} \sqcup A_{0,\partial}, \tag*{(58)} A = A 0 ⊔ A e , A 0 = A 0 , c ⊔ A 0 , ∂ , (58)
where A 0 A_0 A 0 indexes the double curves of the central fiber and A e A_e A e the exuberant curves. We denote by A c ⊂ A A_c \subset A A c ⊂ A the set of compact curves.
We describe the relevant intersection theory in terms of two types of coefficients.
Curve-valued divisor products. For two central-fiber divisors D i D_i D i , D j D_j D j , their intersection is represented by the double curves of the central fiber. We write
D i ⋅ D j = ∑ a ∈ A 0 α i j a C a , ( i , j ) ∈ I 0 × I 0 , (59) D_i \cdot D_j = \sum_{a \in A_0} \alpha^a_{ij} C_a, \qquad(i,j) \in I_0 \times I_0, \tag*{(59)} D i ⋅ D j = a ∈ A 0 ∑ α ij a C a , ( i , j ) ∈ I 0 × I 0 , (59)
where α i j a ∈ Z \alpha^a_{ij} \in\mathbb{Z} α ij a ∈ Z and α i j a = α j i a \alpha^a_{ij} = \alpha^a_{ji} α ij a = α j i a . Products of the form E α ⋅ D j E_\alpha\cdot D_j E α ⋅ D j with ( α , j ) ∈ I e × I 0 (\alpha,j) \in I_e \times I_0 ( α , j ) ∈ I e × I 0 will likewise admit an expansion into curve classes:
E α ⋅ D j = ∑ a ∈ A α α j a C a . (60) E_\alpha\cdot D_j = \sum_{a \in A} \alpha^a_{\alpha j} C_a. \tag*{(60)} E α ⋅ D j = a ∈ A ∑ α α j a C a . (60)
We will not require a curve-class expansion of E α ⋅ E β E_\alpha\cdot E_\beta E α ⋅ E β .
Numerical divisor–curve intersections. For a divisor D i D_i D i and a curve C a C_a C a , we introduce the coefficients
β i a : = D i ⋅ C a (61) \beta_{ia} := D_i \cdot C_a \tag*{(61)} β ia := D i ⋅ C a (61)
whenever this numerical intersection is intrinsically defined. Unlike the curve-valued products above, the coefficients β i a \beta_{ia} β ia are not defined for every divisor–curve pair. Their domain of definition is determined by the compactness of the corresponding geometric intersection, which we now characterize combinatorially.
Admissible Divisor–Curve Intersections
Definition 5.1 (Admissible divisor–curve pairs). Let B ⊂ I × A \mathcal{B} \subset I \times A B ⊂ I × A be the set of pairs ( i , a ) (i,a) ( i , a ) for which the geometric intersection D i ∩ C a D_i \cap C_a D i ∩ C a is compact. For every ( i , a ) ∈ B (i,a) \in\mathcal{B} ( i , a ) ∈ B , we define the intrinsic numerical intersection coefficient
β i a : = D i ⋅ C a ∈ Z . (62) \beta_{ia} := D_i \cdot C_a \in\mathbb{Z}. \tag*{(62)} β ia := D i ⋅ C a ∈ Z . (62)
Every compact curve can be paired with every divisor. Consequently,
I × A c ⊂ B . (63) I \times A_c \subset\mathcal{B}. \tag*{(63)} I × A c ⊂ B . (63)
The only non-trivial admissibility conditions arise for non-compact boundary curves C a C_a C a , with a ∈ A 0 , ∂ a \in A_{0,\partial} a ∈ A 0 , ∂ . These conditions can be read directly from the Symington polygon and its E E E -lines.
Let e a e_a e a be the boundary edge corresponding to C a C_a C a , and denote its relative interior by e a ∘ e_a^\circ e a ∘ . First, consider a central-fiber divisor D i D_i D i , with i ∈ I 0 i \in I_0 i ∈ I 0 , corresponding to a vertex v i v_i v i of the Symington polygon. Then
D i ∩ C a is non-compact ⟺ v i ∈ e a . (64) D_i \cap C_a \text{ is non-compact} \quad\Longleftrightarrow\quad v_i \in e_a. \tag*{(64)} D i ∩ C a is non-compact ⟺ v i ∈ e a . (64)
Indeed, when v i ∈ e a v_i \in e_a v i ∈ e a , the curve C a C_a C a is contained in D i D_i D i , and hence their intersection is non-compact. Otherwise, the intersection consists of finitely many triple points or is empty, and therefore defines an intrinsic numerical intersection.
Next, consider an exuberant divisor E α E_\alpha E α , with α ∈ I e \alpha\in I_e α ∈ I e , and let ℓ α \ell_\alpha ℓ α be its associated E E E -line. In this case,
E α ∩ C a is non-compact ⟺ ℓ α ∩ e a ∘ ≠ ∅ . (65) E_\alpha\cap C_a \text{ is non-compact} \Longleftrightarrow\ell_\alpha\cap e_a^\circ\ne\varnothing. \tag*{(65)} E α ∩ C a is non-compact ⟺ ℓ α ∩ e a ∘ = ∅ . (65)
Thus, an intersection of the E E E -line with the relative interior of a boundary edge records a non-compact common locus of the corresponding exuberant divisor and boundary curve. An intersection occurring only at an endpoint of e a e_a e a does not exclude the pair from B \mathcal{B} B .
It is convenient to combine these two conditions into a single combinatorial criterion. For every a ∈ A 0 , ∂ a \in A_{0,\partial} a ∈ A 0 , ∂ , define
ϵ i a = { 1 , i ∈ I 0 and v i ∈ e a , 1 , i = α ∈ I e and ℓ α ∩ e a ∘ ≠ ∅ , 0 , otherwise . (66) \epsilon_{ia} =
\begin{cases}
1, & i \in I_0 \text{ and } v_i \in e_a,\\
1, & i=\alpha\in I_e \text{ and } \ell_\alpha\cap e_a^\circ\ne\varnothing,\\
0, & \text{otherwise}.
\end{cases} \tag*{(66)} ϵ ia = ⎩ ⎨ ⎧ 1 , 1 , 0 , i ∈ I 0 and v i ∈ e a , i = α ∈ I e and ℓ α ∩ e a ∘ = ∅ , otherwise . (66)
We obtain the following characterization of the admissibility domain:
B = ( I × A c ) ⊔ { ( i , a ) ∈ I × A 0 , ∂ : ϵ i a = 0 } . (67) \mathcal{B} = (I \times A_c) \sqcup\{(i,a) \in I \times A_{0,\partial} : \epsilon_{ia}=0\}. \tag*{(67)} B = ( I × A c ) ⊔ {( i , a ) ∈ I × A 0 , ∂ : ϵ ia = 0 } . (67)
Equivalently,
( i , a ) ∉ B ⟺ a ∈ A 0 , ∂ and { v i ∈ e a , i ∈ I 0 , ℓ i ∩ e a ∘ ≠ ∅ , i ∈ I e . (68) (i,a) \notin\mathcal{B} \Longleftrightarrow a \in A_{0,\partial} \text{ and }
\begin{cases}
v_i \in e_a, & i \in I_0,\\
\ell_i \cap e_a^\circ\ne\varnothing, & i \in I_e.
\end{cases} \tag*{(68)} ( i , a ) ∈ / B ⟺ a ∈ A 0 , ∂ and { v i ∈ e a , ℓ i ∩ e a ∘ = ∅ , i ∈ I 0 , i ∈ I e . (68)
In particular, the central-fiber and exuberant parts of the admissibility domain are
B 0 = ( I 0 × A 0 ) ∖ { ( i , a ) : a ∈ A 0 , ∂ , v i ∈ e a } , (69) \mathcal{B}_0 = (I_0 \times A_0) \setminus\{(i,a): a \in A_{0,\partial},\ v_i \in e_a\}, \tag*{(69)} B 0 = ( I 0 × A 0 ) ∖ {( i , a ) : a ∈ A 0 , ∂ , v i ∈ e a } , (69)
and
B e = ( I e × A c ) ⊔ { ( α , a ) ∈ I e × A 0 , ∂ : ℓ α ∩ e a ∘ = ∅ } . (70) \mathcal{B}_e = (I_e \times A_c) \sqcup\{(\alpha,a) \in I_e \times A_{0,\partial} : \ell_\alpha\cap e_a^\circ= \varnothing\}. \tag*{(70)} B e = ( I e × A c ) ⊔ {( α , a ) ∈ I e × A 0 , ∂ : ℓ α ∩ e a ∘ = ∅ } . (70)
Intersections among Central-Fiber Divisors
We now determine the intersection products involving the components of the central fiber X ~ 0 \widetilde{X}_0 X 0 from the combinatorial data of the unimodular subdivision P ~ \widetilde{\mathcal{P}} P of the Symington polygon.
Products of Central-Fiber Divisors
Let D i , D j D_i,D_j D i , D j be central-fiber divisors corresponding to vertices v i , v j v_i,v_j v i , v j of P ~ \widetilde{\mathcal{P}} P . Suppose first that i ≠ j i \ne j i = j . Their intersection is represented by the double curves corresponding to edges joining v i v_i v i and v j v_j v j . Thus
D i ⋅ D j = ∑ a ∈ A 0 C a ⊂ D i ∩ D j C a . (71) D_i \cdot D_j = \sum_{\substack{a \in A_0\\ C_a \subset D_i \cap D_j}} C_a. \tag*{(71)} D i ⋅ D j = a ∈ A 0 C a ⊂ D i ∩ D j ∑ C a . (71)
In terms of the coefficients introduced in Section 5.1.2, this gives
α i j a = { 1 , i ≠ j and e a joins v i and v j , 0 , otherwise, (72) \alpha^a_{ij} =
\begin{cases}
1, & i \ne j \text{ and } e_a \text{ joins } v_i \text{ and } v_j,\\
0, & \text{otherwise,}
\end{cases} \tag*{(72)} α ij a = { 1 , 0 , i = j and e a joins v i and v j , otherwise, (72)
with contributions added if several edge representatives are identified with the same curve class.
For the diagonal case i = j i=j i = j , note that the divisor X ~ 0 = ∑ j ∈ I 0 D j \widetilde{\mathcal{X}}_0=\sum_{j\in I_0}D_j X 0 = ∑ j ∈ I 0 D j is a fiber of X ~ → Δ \widetilde{\mathcal{X}}\to\Delta X → Δ and so is linearly equivalent to zero. Hence,
D i ∣ D i = − ∑ j ∈ I 0 j ≠ i D j ∣ D i . (73) D_i|_{D_i}=-\sum_{\substack{j\in I_0\\j\ne i}}D_j|_{D_i}. \tag*{(73)} D i ∣ D i = − j ∈ I 0 j = i ∑ D j ∣ D i . (73)
On the other hand, adjunction in the Calabi–Yau threefold gives D i ∣ D i = K D i D_i|_{D_i}=K_{D_i} D i ∣ D i = K D i . Since ( D i , ∂ D i ) (D_i,\partial D_i) ( D i , ∂ D i ) is log Calabi–Yau, we have D i ∣ D i = K D i = − ∂ D i D_i|_{D_i}=K_{D_i}=-\partial D_i D i ∣ D i = K D i = − ∂ D i . Consequently,
D i 2 = − ∑ a ∈ A 0 C a ⊂ ∂ D i C a . (74) D_i^2=-\sum_{\substack{a\in A_0\\C_a\subset\partial D_i}}C_a. \tag*{(74)} D i 2 = − a ∈ A 0 C a ⊂ ∂ D i ∑ C a . (74)
Thus
α i i a = { − 1 , e a is incident to v i , 0 , otherwise. (75) \alpha^a_{ii}=
\begin{cases}
-1, & e_a \text{ is incident to } v_i,\\
0, & \text{otherwise.}
\end{cases} \tag*{(75)} α ii a = { − 1 , 0 , e a is incident to v i , otherwise. (75)
These formulas apply to both compact and non-compact central-fiber divisors. In particular, the resulting curve-valued products may contain non-compact boundary curves C a C_a C a , with a ∈ A 0 , ∂ a\in A_{0,\partial} a ∈ A 0 , ∂ .
Intersections with Double Curves
We next compute the numerical intersection coefficients
β i a = D i ⋅ C a (76) \beta_{ia}=D_i\cdot C_a \tag*{(76)} β ia = D i ⋅ C a (76)
for central-fiber divisors D i D_i D i and double curves C a C_a C a , whenever ( i , a ) ∈ B 0 (i,a)\in B_0 ( i , a ) ∈ B 0 .
We distinguish intersections with the two divisors containing a double curve from transverse intersections with other central-fiber divisors.
Intersections with the two adjacent divisors. Let a ∈ A 0 , c a\in A_{0,c} a ∈ A 0 , c and let
e a = [ v , w ] (77) e_a=[v,w] \tag*{(77)} e a = [ v , w ] (77)
be the corresponding interior edge of P ~ \widetilde{\mathcal{P}} P . Denote its two adjacent unimodular triangles by
( v , w , x ) , ( v , w , y ) . (78) (v,w,x),\qquad(v,w,y). \tag*{(78)} ( v , w , x ) , ( v , w , y ) . (78)
The edge e a e_a e a corresponds to a compact double curve
C a ⊂ D v ∩ D w . (79) C_a \subset D_v \cap D_w. \tag*{(79)} C a ⊂ D v ∩ D w . (79)
We assume for simplicity that the two components are distinct and meet transversely along C a C_a C a , as in the normal-crossings central fiber of § 4.1.2 . The intersection numbers
β v a = D v ⋅ C a , β w a = D w ⋅ C a (80) \beta_{va} = D_v \cdot C_a,\qquad\beta_{wa} = D_w \cdot C_a \tag*{(80)} β v a = D v ⋅ C a , β w a = D w ⋅ C a (80)
are related to the self-intersections of C a C_a C a in the two surfaces. Indeed, since D w ∣ D v D_w|_{D_v} D w ∣ D v contains C a C_a C a with multiplicity one, and locally along C a C_a C a the two divisors meet transversely, we have
D w ⋅ C a = C a 2 ∣ D v , D v ⋅ C a = C a 2 ∣ D w . (81) D_w \cdot C_a = C_a^2|_{D_v},\qquad D_v \cdot C_a = C_a^2|_{D_w}. \tag*{(81)} D w ⋅ C a = C a 2 ∣ D v , D v ⋅ C a = C a 2 ∣ D w . (81)
We first compute C a 2 ∣ D v C_a^2|_{D_v} C a 2 ∣ D v . The primitive tangent vector
u = w − v (82) u = w-v \tag*{(82)} u = w − v (82)
determines the ray corresponding to C a C_a C a in the integral-affine fan at v v v . The two adjacent rays are obtained from the directions towards x x x and y y y .
These three rays must be expressed in a common integral-affine chart. If the two triangles are separated by an affine cut, we parallel transport the tangent vectors across the cut using the corresponding integral-affine transition. Choose the chart containing ( v , w , x ) (v,w,x) ( v , w , x ) , and let M ∈ S L ( 2 , Z ) M \in SL(2,\mathbb{Z}) M ∈ S L ( 2 , Z ) denote the linear part of the transition from the chart containing ( v , w , y ) (v,w,y) ( v , w , y ) to this chosen chart. Thus, as illustrated in Figure 5.1 , the three primitive rays at v v v are
Figure 5.1. (a) Two adjacent unimodular triangles sharing the edge e a = [ v , w ] e_a=[v,w] e a = [ v , w ] , corresponding to the double curve C a = D v ∩ D w C_a=D_v\cap D_w C a = D v ∩ D w . (b) The primitive directions in the fan at v v v satisfy u − + u + + β w a u = 0 u_-+u_++\beta_{wa}u=0 u − + u + + β w a u = 0 . (c) The corresponding boundary curves on D v D_v D v .
u = w − v , u − = x − v , u + = M ( y − v ) . (83) u = w-v,\qquad u_- = x-v,\qquad u_+ = M(y-v). \tag*{(83)} u = w − v , u − = x − v , u + = M ( y − v ) . (83)
The notation M ( y − v ) M(y-v) M ( y − v ) means that the tangent vector y − v y-v y − v is transported into the chosen chart. If no cut is crossed, M M M is the identity.
Since the two adjacent cones are unimodular, the primitive rays satisfy a unique integral relation
u − + u + + n v u = 0 (84) u_{-}+u_{+}+n_{v}u=0 \tag*{(84)} u − + u + + n v u = 0 (84)
where n v n_v n v is the self-intersection of the corresponding boundary curve in D v D_v D v :
n v = C a 2 ∣ D v . (85) n_v=C_a^2|_{D_v}. \tag*{(85)} n v = C a 2 ∣ D v . (85)
Using (81) , we obtain
n v = β w a . (86) n_v=\beta_{wa}. \tag*{(86)} n v = β w a . (86)
Therefore the integral-affine self-intersection rule at v v v is
( x − v ) + M ( y − v ) + β w a ( w − v ) = 0. (87) (x-v)+M(y-v)+\beta_{wa}(w-v)=0. \tag*{(87)} ( x − v ) + M ( y − v ) + β w a ( w − v ) = 0. (87)
This vector relation uniquely determines the integer β w a \beta_{wa} β w a .
Similarly, at the other endpoint w w w , the primitive ray corresponding to C a C_a C a is v − w v-w v − w . Using the same parallel-transport convention, the three consecutive rays are
v − w , x − w , M ( y − w ) . (88) v-w,\qquad x-w,\qquad M(y-w). \tag*{(88)} v − w , x − w , M ( y − w ) . (88)
The corresponding fan relation gives
( x − w ) + M ( y − w ) + β v a ( v − w ) = 0. (89) (x-w)+M(y-w)+\beta_{va}(v-w)=0. \tag*{(89)} ( x − w ) + M ( y − w ) + β v a ( v − w ) = 0. (89)
For computational purposes, one may extract the integer coefficients by taking an ordinary Euclidean inner product with the corresponding edge direction. This gives
β v a = − ( ( x − w ) + M ( y − w ) ) ⋅ ( v − w ) ∣ v − w ∣ 2 , β w a = − ( ( x − v ) + M ( y − v ) ) ⋅ ( w − v ) ∣ w − v ∣ 2 . (90) \begin{aligned}
\beta_{va} &= -\frac{\bigl((x-w)+M(y-w)\bigr)\cdot(v-w)}{|v-w|^2},\\
\beta_{wa} &= -\frac{\bigl((x-v)+M(y-v)\bigr)\cdot(w-v)}{|w-v|^2}. \tag*{(90)}
\end{aligned} β v a β w a = − ∣ v − w ∣ 2 ( ( x − w ) + M ( y − w ) ) ⋅ ( v − w ) , = − ∣ w − v ∣ 2 ( ( x − v ) + M ( y − v ) ) ⋅ ( w − v ) . (90)
The Euclidean inner product is used only to extract the integer coefficient from the vector relation. The underlying rule is integral-affine and does not depend on a choice of Euclidean metric.
When no affine cut is crossed, M = 1 M=1 M = 1 , and the formulas reduce to the standard toric self-intersection rules.
Transverse intersections with double curves.
Now let D i D_i D i be a central-fiber divisor whose corresponding vertex v i v_i v i is not an endpoint of e a e_a e a . The intersection of D i D_i D i with C a C_a C a occurs at the triple points corresponding to triangles containing both e a e_a e a and v i v_i v i . Thus, whenever ( i , a ) ∈ B 0 (i,a)\in B_0 ( i , a ) ∈ B 0 ,
β i a = # { σ ∈ P ~ [ 2 ] : e a ⊂ σ , v i ∈ σ ∖ e a } . (91) \beta_{ia}=\#\left\{\sigma\in\widetilde{\mathcal{P}}^{[2]}:e_a\subset\sigma,\ v_i\in\sigma\setminus e_a\right\}. \tag*{(91)} β ia = # { σ ∈ P [ 2 ] : e a ⊂ σ , v i ∈ σ ∖ e a } . (91)
This formula also applies to non-compact boundary curves C a C_a C a , with a ∈ A 0 , ∂ a\in A_{0,\partial} a ∈ A 0 , ∂ , provided v i ∉ e a v_i\notin e_a v i ∈ / e a . In that case the intersection is a finite collection of triple points and therefore has a well-defined degree. If instead v i ∈ e a v_i\in e_a v i ∈ e a , the non-compact curve C a C_a C a is contained in D i D_i D i . The pair ( i , a ) (i,a) ( i , a ) is then inadmissible, in agreement with § 5.1.3 .
Intersections involving Exuberant Divisors
We now turn to intersections involving the exuberant divisors E α E_\alpha E α , α ∈ I e \alpha\in I_e α ∈ I e , associated with the E E E -lines of the Symington polygon. Unlike intersections among central-fiber divisors, which are determined by the integral-affine triangulation, intersections involving exuberant divisors require additional geometric data. The primitive direction of an E E E -line determines the intersection numbers of its associated exuberant curve, while its incidence with the triangulation constrains the remaining divisor–curve intersections. These rules determine part of the intersection data directly. The remaining coefficients will be determined by the compatibility relations developed in § 5.5 .
Intersection Numbers of Exuberant Curves
Let E α E_\alpha E α , α ∈ I e \alpha\in I_e α ∈ I e , be an exuberant divisor whose E E E -line ℓ α \ell_\alpha ℓ α starts at the vertex v v v corresponding to a compact central-fiber divisor D D D . Let u u u be the primitive direction of ℓ α \ell_\alpha ℓ α at v v v . Choose a unimodular cone in the star of v v v , with primitive generators u 1 u_1 u 1 , u 2 u_2 u 2 , containing u u u . We may write
u = p u 1 + q u 2 , p , q ∈ Z ≥ 0 , gcd ( p , q ) = 1. (96) u = p u_1 + q u_2, \qquad p,q \in\mathbb{Z}_{\geq0}, \qquad\gcd(p,q)=1. \tag*{(96)} u = p u 1 + q u 2 , p , q ∈ Z ≥ 0 , g cd( p , q ) = 1. (96)
Let D 1 D_1 D 1 , D 2 D_2 D 2 be the central-fiber divisors corresponding to the far endpoints of the edges in the directions u 1 u_1 u 1 , u 2 u_2 u 2 , respectively. As reviewed in § 4.2.2 , the intersection of E α E_\alpha E α with D D D is the compact exuberant curve C α : = D ⋅ E α C_\alpha:= D \cdot E_\alpha C α := D ⋅ E α . The corresponding local geometry is illustrated in Figure 5.2 .
Figure 5.2. Local geometry of an exuberant curve C α C_{\alpha} C α at the initial vertex of its E E E -line. The primitive direction u = p u 1 + q u 2 u=pu_{1}+qu_{2} u = p u 1 + q u 2 lies in a unimodular cone of the Symington polygon, illustrated for ( p , q ) = ( 2 , 3 ) (p,q)=(2,3) ( p , q ) = ( 2 , 3 ) (left). The corresponding curve on D D D has local equation x p − y q = 0 x^{p}-y^{q}=0 x p − y q = 0 , with boundary components B 1 = D 1 ∣ D = { y = 0 } B_{1}=D_{1}|_{D}=\{y=0\} B 1 = D 1 ∣ D = { y = 0 } and B 2 = D 2 ∣ D = { x = 0 } B_{2}=D_{2}|_{D}=\{x=0\} B 2 = D 2 ∣ D = { x = 0 } (right).
Proposition 5.3. With the notation above,
D ⋅ C α = − p − q , D 1 ⋅ C α = p , D 2 ⋅ C α = q , E α ⋅ C α = p q − 1. (97) D \cdot C_\alpha= -p-q, \qquad D_1 \cdot C_\alpha= p, \qquad D_2 \cdot C_\alpha= q, \qquad E_\alpha\cdot C_\alpha= pq-1. \tag*{(97)} D ⋅ C α = − p − q , D 1 ⋅ C α = p , D 2 ⋅ C α = q , E α ⋅ C α = pq − 1. (97) All other central-fiber divisors have zero intersection with C α C_\alpha C α .
Proof. Near the point where C α C_{\alpha} C α meets the boundary of D D D , choose local coordinates x , y x,y x , y such that the boundary components corresponding to u 1 , u 2 u_1,u_2 u 1 , u 2 are given by y = 0 y=0 y = 0 and x = 0 x=0 x = 0 , respectively. Since u = p u 1 + q u 2 u=pu_1+qu_2 u = p u 1 + q u 2 , the curve C α C_{\alpha} C α has local equation
x p − y q = 0 , x^p-y^q=0, x p − y q = 0 ,
and hence admits the parametrization x = t q x=t^q x = t q , y = t p y=t^p y = t p . Its intersection multiplicities with the two boundary components are therefore p p p and q q q , respectively. As these boundary components are the restrictions of D 1 D_1 D 1 and D 2 D_2 D 2 to D D D , we obtain
D 1 ⋅ C α = p , D 2 ⋅ C α = q . D_1\cdot C_{\alpha}=p,\qquad D_2\cdot C_{\alpha}=q. D 1 ⋅ C α = p , D 2 ⋅ C α = q .
No other central-fiber divisor distinct from D D D meets C α C_{\alpha} C α .
To compute the intersection with D D D , recall that D D D is a log Calabi–Yau surface and that its normal bundle in X ~ \widetilde{\mathcal X} X is its canonical bundle. Thus D ∣ D = K D = − ∂ D D|_D=K_D=-\partial D D ∣ D = K D = − ∂ D . The curve C α C_{\alpha} C α meets the boundary ∂ D \partial D ∂ D at a single point, with total intersection multiplicity
C α ⋅ ∂ D = p + q . (98) C_{\alpha}\cdot\partial D=p+q. \tag*{(98)} C α ⋅ ∂ D = p + q . (98)
Consequently,
D ⋅ C α = K D ⋅ C α = − ( p + q ) . (99) D\cdot C_{\alpha}=K_D\cdot C_{\alpha}=-(p+q). \tag*{(99)} D ⋅ C α = K D ⋅ C α = − ( p + q ) . (99)
It remains to determine E α ⋅ C α E_{\alpha}\cdot C_{\alpha} E α ⋅ C α . Suppose first that p , q > 0 p,q>0 p , q > 0 . The local geometric construction describes C α C_{\alpha} C α as the image of an exceptional rational curve under the corresponding toric blow-down. Its normalization is therefore P 1 \mathbb{P}^1 P 1 , while its unique singular point has local analytic equation
f : = x p − y q = 0. (100) f:=x^p-y^q=0. \tag*{(100)} f := x p − y q = 0. (100)
Hence, the arithmetic genus of C α C_{\alpha} C α equals the delta invariant δ \delta δ of this singularity. Moreover, gcd ( p , q ) = 1 \gcd(p,q)=1 g cd( p , q ) = 1 implies that the singularity is unibranch. Its Milnor number is
μ = dim C C [ x , y ] ( ∂ x f , ∂ y f ) = ( p − 1 ) ( q − 1 ) . (101) \mu=\dim_{\mathbb{C}}\frac{\mathbb{C}[x,y]}{(\partial_x f,\partial_y f)}=(p-1)(q-1). \tag*{(101)} μ = dim C ( ∂ x f , ∂ y f ) C [ x , y ] = ( p − 1 ) ( q − 1 ) . (101)
Since the singularity is unibranch, the Milnor formula, [66 Theorem 10.5] , gives 2 δ = μ 2\delta=\mu 2 δ = μ . Hence
p a ( C α ) = δ = ( p − 1 ) ( q − 1 ) 2 . (102) p_a(C_{\alpha})=\delta=\frac{(p-1)(q-1)}{2}. \tag*{(102)} p a ( C α ) = δ = 2 ( p − 1 ) ( q − 1 ) . (102)
We now apply the adjunction formula for an integral curve on the smooth surface D D D :
2 p a ( C α ) − 2 = C α ∣ D 2 + K D ⋅ C α . (103) 2p_a(C_{\alpha})-2=C_{\alpha|D}^{2}+K_D\cdot C_{\alpha}. \tag*{(103)} 2 p a ( C α ) − 2 = C α ∣ D 2 + K D ⋅ C α . (103)
Using (5.51) and (5.48), we obtain
C α ∣ D 2 = 2 p a ( C α ) − 2 − K D ⋅ C α = ( p − 1 ) ( q − 1 ) − 2 + p + q = p q − 1. (104) \begin{aligned}
C_{\alpha|D}^{2}
&=2p_a(C_{\alpha})-2-K_D\cdot C_{\alpha}\\
&=(p-1)(q-1)-2+p+q\\
&=pq-1. \tag*{(104)}
\end{aligned} C α ∣ D 2 = 2 p a ( C α ) − 2 − K D ⋅ C α = ( p − 1 ) ( q − 1 ) − 2 + p + q = pq − 1. (104)
Since C α = D ∩ E α C_{\alpha}=D\cap E_{\alpha} C α = D ∩ E α , the restriction of E α E_{\alpha} E α to D D D is the divisor C α C_{\alpha} C α . Therefore
E α ⋅ C α = ( E α ∣ D ) ⋅ C α = C α ∣ D 2 = p q − 1. (105) E_{\alpha}\cdot C_{\alpha}=(E_{\alpha}|_{D})\cdot C_{\alpha}=C_{\alpha|D}^{2}=pq-1. \tag*{(105)} E α ⋅ C α = ( E α ∣ D ) ⋅ C α = C α ∣ D 2 = pq − 1. (105)
Finally, suppose that one of p , q p,q p , q vanishes. Primitivity forces the other to be one. Then C α C_{\alpha} C α is an exceptional curve, and hence
C α ∣ D 2 = − 1 = p q − 1. (106) C_{\alpha|D}^{2}=-1=pq-1. \tag*{(106)} C α ∣ D 2 = − 1 = pq − 1. (106)
Thus the formula holds in all cases.
Remark 5.4. The local intersection numbers satisfy
C α ⋅ ( D + D 1 + D 2 ) = − ( p + q ) + p + q = 0. (107) C_{\alpha}\cdot(D+D_{1}+D_{2})=-(p+q)+p+q=0. \tag*{(107)} C α ⋅ ( D + D 1 + D 2 ) = − ( p + q ) + p + q = 0. (107)
This is consistent with the fact that the central fiber X ~ 0 \widetilde{X}_{0} X 0 is linearly equivalent to zero, and therefore has zero intersection number with every compact curve.
Divisor Products determined by E E E -lines
Let j ∈ I 0 j\in I_{0} j ∈ I 0 . As introduced in § 5.1.2 , we write
E α ⋅ D j = ∑ a ∈ A α a j α C a . (108) E_{\alpha}\cdot D_{j}=\sum_{a\in A}\alpha_{aj}^{\alpha}C_{a}. \tag*{(108)} E α ⋅ D j = a ∈ A ∑ α aj α C a . (108)
To specify a unique curve expansion of E α ⋅ D j E_{\alpha}\cdot D_{j} E α ⋅ D j in (108) , we choose a small perturbation ℓ α ′ \ell_{\alpha}^{\prime} ℓ α ′ of the E E E -line ℓ α \ell_{\alpha} ℓ α that avoids all vertices of the triangulation except its initial vertex. At each vertex traversed by ℓ α ′ \ell'_{\alpha} ℓ α ′ , this amounts to choosing one of the two possible sides along which ℓ α ′ \ell'_{\alpha} ℓ α ′ bypasses the vertex, as illustrated in Figure 5.3 . This choice determines the set of curve classes allowed to appear in the expansion, as explained below.
Figure 5.3. Two possible infinitesimal deformations of an E -line ℓ α \ell_\alpha ℓ α passing through a vertex v v v of the triangulation P ~ \widetilde{\mathcal{P}} P . The dashed line represents ℓ α \ell_\alpha ℓ α . The deformed path ℓ α ′ \ell'_\alpha ℓ α ′ avoids v v v by passing on either side of it. The two choices intersect the interiors of different edges adjacent to v v v , and hence determine different allowed supports for the curve expansion of E α ⋅ D j E_\alpha\cdot D_j E α ⋅ D j .
First, denote by v j v_j v j the vertex of P ~ \widetilde{P} P corresponding to D j D_j D j . Define
A α j ⊂ A (109) A_{\alpha j} \subset A \tag*{(109)} A α j ⊂ A (109)
to be the set of a ∈ A a \in A a ∈ A such that either:
a ∈ A 0 a \in A_0 a ∈ A 0 , the corresponding edge e a e_a e a of P ~ \widetilde{P} P is incident to v j v_j v j , and ℓ α ′ \ell'_{\alpha} ℓ α ′ intersects the interior of e a e_a e a , or
a = α a = \alpha a = α , or
a = β ∈ A e a = \beta\in A_e a = β ∈ A e , the corresponding focus–focus singularity V β V_\beta V β is originally located before perturbation at v j v_j v j , and ℓ α ′ \ell'_{\alpha} ℓ α ′ passes between V β V_\beta V β and v j v_j v j .
Lemma 5.5. There exists a unique curve expansion of E α ⋅ D j E_\alpha\cdot D_j E α ⋅ D j such that
α α j α = 0 (110) \alpha^{\alpha}_{\alpha j} = 0 \tag*{(110)} α α j α = 0 (110) unless a ∈ A α j a \in A_{\alpha j} a ∈ A α j .
Proof. If v j v_j v j does not lie on ℓ α \ell_\alpha ℓ α , then every irreducible component of the geometric intersection E α ∩ D j E_\alpha\cap D_j E α ∩ D j is of the form C a C_a C a , where a ∈ A 0 a \in A_0 a ∈ A 0 is necessarily in A α j A_{\alpha j} A α j (Case 1.). If v j v_j v j is the initial point of ℓ α \ell_\alpha ℓ α , then E α ∩ D j = C α E_\alpha\cap D_j = C_\alpha E α ∩ D j = C α , with α ∈ A α j \alpha\in A_{\alpha j} α ∈ A α j (Case 2.).
The more interesting situation is when v j v_j v j lies on ℓ α \ell_\alpha ℓ α but is not the initial point. We focus on the case where v j v_j v j is a smooth point of the integral-affine structure, so that the corresponding surface X v j X_{v_j} X v j is toric. The case of a singular point can be treated similarly using a toric model, as in § 4.2.2 . Let u 1 , … , u N u_1,\ldots,u_N u 1 , … , u N be the primitive directions of the edges outgoing from v j v_j v j , spanning the fan of X v j X_{v_j} X v j and corresponding to toric divisors B 1 , … , B N B_1,\ldots,B_N B 1 , … , B N . Write the outgoing and incoming directions of ℓ α \ell_\alpha ℓ α in their respective unimodular cones as
u = p u k + q u k + 1 , − u = r u l + s u l + 1 . (111) u = p u_k + q u_{k+1}, \qquad-u = r u_l + s u_{l+1}. \tag*{(111)} u = p u k + q u k + 1 , − u = r u l + s u l + 1 . (111)
Then, arguing as in the beginning of the proof of Proposition 5.3 , the boundary intersection numbers of the curve C : = E α ∩ X v j C := E_\alpha\cap X_{v_j} C := E α ∩ X v j are determined by these decompositions:
B k ⋅ C = p , B k + 1 ⋅ C = q , B l ⋅ C = r , B l + 1 ⋅ C = s . (112) \begin{aligned}
B_k \cdot C &= p, \qquad B_{k+1} \cdot C = q, \\
B_l \cdot C &= r, \qquad B_{l+1} \cdot C = s. \tag*{(112)}
\end{aligned} B k ⋅ C B l ⋅ C = p , B k + 1 ⋅ C = q , = r , B l + 1 ⋅ C = s . (112)
All other boundary intersections vanish, so these intersection numbers uniquely determine the class of C C C in the toric surface X v j X_{v_j} X v j . Geometrically, C C C is the closure of a general translate of a one-parameter subgroup of the torus ( C ∗ ) 2 ⊂ X v j (\mathbb{C}^*)^2 \subset X_{v_j} ( C ∗ ) 2 ⊂ X v j with primitive direction u u u . The choice of how ℓ α ′ \ell'_{\alpha} ℓ α ′ bypasses v j v_j v j selects one of the two chains of toric boundary curves onto which C C C can degenerate. This yields a unique expansion of [ C ] [C] [ C ] in terms of the boundary curve classes belonging to A α j A_{\alpha j} A α j (Case 3.). More precisely, orient the fan so that the chosen chain lies counterclockwise from u u u to − u -u − u . Then
[ C ] = ∑ m = 1 N max { 0 , det ( u , u m ) } [ B m ] . (113) [C] = \sum_{m=1}^{N} \max\{0,\det(u,u_m)\}[B_m]. \tag*{(113)} [ C ] = m = 1 ∑ N max { 0 , det ( u , u m )} [ B m ] . (113)
Indeed, the right-hand side is the effective toric boundary divisor obtained by taking the positive part of the principal divisor of a character whose restriction to the one-parameter subgroup with direction u u u is trivial.
Uniqueness follows from the toric linear equivalences. Indeed, since the fan is complete, the rays in the complementary closed half-plane span the ambient two-dimensional vector space. Thus, any principal toric divisor supported on the selected chain must vanish.
For α ∈ I e \alpha\in I_e α ∈ I e , we denote by b ( α ) ∈ I c b(\alpha) \in I_c b ( α ) ∈ I c the index of the initial vertex v b ( α ) v_{b(\alpha)} v b ( α ) of ℓ α \ell_\alpha ℓ α , so that D b ( α ) D_{b(\alpha)} D b ( α ) is the compact divisor containing the exuberant curve C α C_\alpha C α . At the initial vertex v b ( α ) v_{b(\alpha)} v b ( α ) of ℓ α \ell_\alpha ℓ α , the intersection with the corresponding compact divisor contains the exuberant curve C α C_\alpha C α with multiplicity one. Thus
α α , b ( α ) α = 1. (114) \alpha^{\alpha}_{\alpha,b(\alpha)} = 1. \tag*{(114)} α α , b ( α ) α = 1. (114)
The E -line geometry also constrains the numerical divisor–curve intersections
β α a = E α ⋅ C a . (115) \beta_{\alpha a} = E_\alpha\cdot C_a. \tag*{(115)} β α a = E α ⋅ C a . (115)
For compact curves, these intersections are always admissible:
( α , a ) ∈ B , a ∈ A c . (116) (\alpha,a) \in\mathcal{B}, \qquad a \in A_c. \tag*{(116)} ( α , a ) ∈ B , a ∈ A c . (116)
For a non-compact boundary curve C a C_a C a , with a ∈ A 0 , ∂ a \in A_{0,\partial} a ∈ A 0 , ∂ , the admissibility criterion of § 5.1.3 gives
( α , a ) ∈ B ⟺ ℓ α ∩ e a ∘ = ∅ . (117) (\alpha,a)\in\mathcal{B}\iff\ell_\alpha\cap e_a^\circ=\varnothing. \tag*{(117)} ( α , a ) ∈ B ⟺ ℓ α ∩ e a ∘ = ∅ . (117)
Thus the numerical intersection β α a \beta_{\alpha a} β α a is intrinsically defined precisely when the corresponding E E E -line does not meet the relative interior of the boundary edge e a e_a e a .
Moreover, for every admissible pair ( α , a ) ∈ B (\alpha,a)\in\mathcal{B} ( α , a ) ∈ B , we have
β α a = 0 (118) \beta_{\alpha a}=0 \tag*{(118)} β α a = 0 (118)
unless the exuberant divisor E α E_\alpha E α and the curve C a C_a C a have a non-empty geometric intersection, that is, unless either
a ∈ A 0 a\in A_0 a ∈ A 0 and the corresponding edge e a e_a e a of P ~ \widetilde{P} P is an edge of a triangle traversed by ℓ α \ell_\alpha ℓ α , or
a = β ∈ A e a=\beta\in A_e a = β ∈ A e , and ℓ α \ell_\alpha ℓ α traverses a triangle with vertex v b ( β ) v_{b(\beta)} v b ( β ) and containing the initial direction of ℓ β \ell_\beta ℓ β .
The intersection data involving exuberant divisors are therefore partially determined by three types of geometric information:
The primitive direction of the E E E -line at its initial vertex determines the intersection numbers (97) .
The incidence of the E E E -line with the triangulation determines the possible support of the curve-valued products (108) .
The admissibility and disjointness conditions determine which numerical intersections β α a \beta_{\alpha a} β α a are defined and which of them vanish.
These rules determine part of the exuberant α \alpha α - and β \beta β -data directly. However, some coefficients allowed by the incidence conditions are not fixed yet. To determine the remaining coefficients, we impose compatibility relations arising from associativity of the intersection product, which we discuss in the following subsection.
Compatibility Relations from Associativity
The intersection rules derived in § 5.2 and § 5.4 determine part of the intersection data directly from the geometry of the Symington polygon and its E E E -lines. However, some coefficients in the curve-valued products E α ⋅ D j E_\alpha\cdot D_j E α ⋅ D j and the admissible numerical intersections E α ⋅ C a E_\alpha\cdot C_a E α ⋅ C a are not fixed by these local rules. We now explain how associativity of the intersection product provides compatibility relations among these coefficients.
Suppose that a curve-valued divisor product admits an expansion
D j ⋅ D k = ∑ a ∈ A α j k a C a . (119) D_j\cdot D_k=\sum_{a\in A}\alpha^a_{jk}C_a. \tag*{(119)} D j ⋅ D k = a ∈ A ∑ α j k a C a . (119)
We say that this expansion is admissible against D i D_i D i if one of the following holds:
( j , k ) ∈ I 0 × I 0 (j,k) \in I_0 \times I_0 ( j , k ) ∈ I 0 × I 0 , so that the coefficients α j k a \alpha^a_{jk} α j k a are known by (72) –(75) , and we require
α j k a ≠ 0 ⟹ ( i , a ) ∈ B . (120) \alpha^a_{jk} \ne0 \quad\Longrightarrow\quad(i,a) \in\mathcal{B}. \tag*{(120)} α j k a = 0 ⟹ ( i , a ) ∈ B . (120)
( j , k ) ∈ I e × I 0 (j,k) \in I_e \times I_0 ( j , k ) ∈ I e × I 0 , and we require
a ∈ A j k ⟹ ( i , a ) ∈ B , (121) a \in A_{jk} \quad\Longrightarrow\quad(i,a) \in\mathcal{B}, \tag*{(121)} a ∈ A j k ⟹ ( i , a ) ∈ B , (121)
where A j k ⊂ A A_{jk} \subset A A j k ⊂ A is defined in (109) .
Whenever this condition holds, we define the corresponding numerical iterated intersection by
T i ; j k : = D i ⋅ ( D j ⋅ D k ) = ∑ a ∈ A α j k a β i a . (122) T_{i;jk} := D_i \cdot(D_j \cdot D_k) = \sum_{a \in A} \alpha^a_{jk}\beta_{ia}. \tag*{(122)} T i ; j k := D i ⋅ ( D j ⋅ D k ) = a ∈ A ∑ α j k a β ia . (122)
The intersection product is associative and commutative. Consequently, different admissible presentations of the same triple intersection must agree. We therefore obtain the compatibility relation
T i ; j k = T j ; i k (123) T_{i;jk} = T_{j;ik} \tag*{(123)} T i ; j k = T j ; ik (123)
whenever both sides are defined. If all three presentations are admissible, then
T i ; j k = T j ; i k = T k ; i j . (124) T_{i;jk} = T_{j;ik} = T_{k;ij}. \tag*{(124)} T i ; j k = T j ; ik = T k ; ij . (124)
These equations provide constraints on the intersection coefficients that are not determined by the local geometric rules. An important feature is that the compatibility relations are not restricted to products containing a compact divisor. Indeed, a non-compact boundary curve C a C_a C a , with a ∈ A 0 , ∂ a \in A_{0,\partial} a ∈ A 0 , ∂ , may appear in the expansion (119) and still contribute to (122) , provided its intersection with D i D_i D i is admissible.
Triple Intersection Numbers
The compatibility relations of the preceding subsection determine the intersection coefficients needed to reconstruct the triple intersection numbers of divisors. Although these relations may involve non-compact curves and non-compact divisors, the intrinsic numerical triple intersections relevant for the five-dimensional prepotential are precisely those containing at least one compact divisor. We now explain how these intrinsic triple intersections are extracted from the intersection coefficients.
Let i ∈ I c i \in I_c i ∈ I c and let D j , D k D_j, D_k D j , D k be arbitrary divisors, with j , k ∈ I j,k \in I j , k ∈ I . Since D i D_i D i is compact, its intersection with any divisor D k D_k D k is a compact curve class. Hence it admits an expansion in compact curves:
D i ⋅ D k = ∑ a ∈ A c α i k a C a . (125) D_i \cdot D_k = \sum_{a \in A_c} \alpha^a_{ik} C_a. \tag*{(125)} D i ⋅ D k = a ∈ A c ∑ α ik a C a . (125)
Every compact curve can be paired numerically with every divisor. In particular,
( j , a ) ∈ B , j ∈ I , a ∈ A c . (126) (j,a) \in\mathcal{B}, \qquad j \in I, \qquad a \in A_c. \tag*{(126)} ( j , a ) ∈ B , j ∈ I , a ∈ A c . (126)
It follows that the intrinsic triple intersection number is
D i ⋅ D j ⋅ D k = ∑ a ∈ A c α i k a β j a , ( i , j , k ) ∈ I c × I × I . (127) D_i \cdot D_j \cdot D_k = \sum_{a \in A_c} \alpha_{ik}^{a}\beta_{ja}, \qquad(i,j,k) \in I_c \times I \times I. \tag*{(127)} D i ⋅ D j ⋅ D k = a ∈ A c ∑ α ik a β j a , ( i , j , k ) ∈ I c × I × I . (127)
This is the basic formula for extracting the intrinsic triple intersection form from the intersection coefficients.
For i ∈ I c i \in I_c i ∈ I c , the intrinsic triple intersection
D i ⋅ E α ⋅ E β (128) D_i \cdot E_\alpha\cdot E_\beta \tag*{(128)} D i ⋅ E α ⋅ E β (128)
can be computed by first expanding the compact curve class
D i ⋅ E β = ∑ a ∈ A c α i β a C a (129) D_i \cdot E_\beta= \sum_{a \in A_c} \alpha_{i\beta}^{a} C_a \tag*{(129)} D i ⋅ E β = a ∈ A c ∑ α i β a C a (129)
and then intersecting with E α E_\alpha E α :
D i ⋅ E α ⋅ E β = ∑ a ∈ A c α i β a β α a . (130) D_i \cdot E_\alpha\cdot E_\beta= \sum_{a \in A_c} \alpha_{i\beta}^{a}\beta_{\alpha a}. \tag*{(130)} D i ⋅ E α ⋅ E β = a ∈ A c ∑ α i β a β α a . (130)
Thus the curve-valued products involving at least one central-fiber divisor, together with the admissible divisor–curve intersection coefficients, suffice to determine all intrinsic triple intersection numbers.
Remark 5.6. For every compact component D i D_i D i of X ~ 0 \widetilde{X}_0 X 0 , the triple self-intersection number D i 3 D_i^3 D i 3 is given by
D i 3 = 12 − Q i − χ t o p ( ∂ D i ) , (131) D_i^3 = 12 - Q_i - \chi_{\mathrm{top}}(\partial D_i), \tag*{(131)} D i 3 = 12 − Q i − χ top ( ∂ D i ) , (131)
where Q i = χ t o p ( D i ∖ ∂ D i ) Q_i = \chi_{\mathrm{top}}(D_i \setminus\partial D_i) Q i = χ top ( D i ∖ ∂ D i ) is the charge of D i D_i D i . If ∂ D i \partial D_i ∂ D i is a cycle of r i r_i r i rational curves, then
D i 3 = 12 − Q i − r i . (132) D_i^3 = 12 - Q_i - r_i. \tag*{(132)} D i 3 = 12 − Q i − r i . (132)
Indeed, since X ~ \widetilde{X} X is Calabi–Yau, adjunction gives D i ∣ D i = K D i D_i|_{D_i} = K_{D_i} D i ∣ D i = K D i . Since ( D i , ∂ D i ) (D_i,\partial D_i) ( D i , ∂ D i ) is log Calabi–Yau, we have K D i = − ∂ D i K_{D_i} = -\partial D_i K D i = − ∂ D i . Therefore D i 3 = K D i 2 D_i^3 = K_{D_i}^2 D i 3 = K D i 2 . The surface D i D_i D i is rational, so Noether’s formula gives K D i 2 = 12 − χ t o p ( D i ) K_{D_i}^2 = 12 - \chi_{\mathrm{top}}(D_i) K D i 2 = 12 − χ top ( D i ) . By additivity of the topological Euler characteristic, χ t o p ( D i ) = χ t o p ( D i ∖ ∂ D i ) + χ t o p ( ∂ D i ) \chi_{\mathrm{top}}(D_i) = \chi_{\mathrm{top}}(D_i \setminus\partial D_i) + \chi_{\mathrm{top}}(\partial D_i) χ top ( D i ) = χ top ( D i ∖ ∂ D i ) + χ top ( ∂ D i ) . Using the definition of Q i Q_i Q i , we obtain χ t o p ( D i ) = Q i + χ t o p ( ∂ D i ) \chi_{\mathrm{top}}(D_i) = Q_i + \chi_{\mathrm{top}}(\partial D_i) χ top ( D i ) = Q i + χ top ( ∂ D i ) , which proves the formula (131) for D i 3 D_i^3 D i 3 .
This formula will not be needed as input to the reconstruction algorithm. Instead, it provides an independent consistency check on the intersection coefficients obtained from the combinatorial rules.
Algorithm for Triple Intersection Numbers
In this section, we show that the intersection rules developed in § 5 allow us to reconstruct the triple intersection numbers of the Calabi–Yau threefold associated with a GTP directly from its combinatorial data.
We fix a GTP, or equivalently the associated Symington polygon, together with a choice of perturbation of the integral-affine singularities, which induces a choice of exuberant divisors. The goal is to determine the triple intersection numbers
κ i j k : = D i ⋅ D j ⋅ D k , { i , j , k } ∩ I c ≠ ∅ (133) \kappa_{ijk} := D_i \cdot D_j \cdot D_k,\qquad\{i,j,k\}\cap I_c \ne\varnothing \tag*{(133)} κ ij k := D i ⋅ D j ⋅ D k , { i , j , k } ∩ I c = ∅ (133)
between divisors on the generalized toric Calabi–Yau threefold X ~ \widetilde{X} X associated to the GTP as in § 4 . As reviewed in § 2 , these triple intersection numbers determine the prepotential of the 5d SCFT associated with the GTP.
In § 6.1 , we formulate an explicit combinatorial algorithm. This algorithm involves solving a system of compatibility equations, expressing the associativity of the intersection product. In § 6.2 , we establish that this system always admits a unique solution, and therefore that the algorithm computes the desired triple intersection numbers.
Reconstruction Problem and Algorithm
We formulate the reconstruction of triple intersection numbers as a finite system of equations for the intersection coefficients introduced in § 5 . The combinatorial data determine some of these coefficients directly and restrict the support of the others. The remaining coefficients are determined by imposing the compatibility relations of § 5.5 . We first specify the input data and the unknowns of the reconstruction, and then give the algorithm.
Combinatorial Input
We fix a GTP ( P , P ) (P,\mathcal{P}) ( P , P ) , or equivalently the associated Symington polygon ( P ~ , P ~ ) (\widetilde{P},\widetilde{\mathcal{P}}) ( P , P ) , together with a choice of generic infinitesimal perturbation of the integral-affine singularities, which induces a choice of E E E -lines as in § 3.5 and of exuberant divisors as in § 4.2.2 . We record the index sets
I 0 = I c ⊔ I 0 , ∂ , A 0 = A 0 , c ⊔ A 0 , ∂ , (134) I_0 = I_c \sqcup I_{0,\partial},\qquad A_0 = A_{0,c} \sqcup A_{0,\partial}, \tag*{(134)} I 0 = I c ⊔ I 0 , ∂ , A 0 = A 0 , c ⊔ A 0 , ∂ , (134)
for vertices and edges of P ~ \widetilde{\mathcal{P}} P , together with the vertex–edge–triangle incidence relations.
For each E E E -line ℓ α \ell_\alpha ℓ α , we record its initial vertex v b ( α ) v_{b(\alpha)} v b ( α ) , corresponding to a compact divisor D b ( α ) D_{b(\alpha)} D b ( α ) , and the associated exuberant divisor E α E_\alpha E α and compact curve C α = D b ( α ) ⋅ E α C_\alpha= D_{b(\alpha)}\cdot E_\alpha C α = D b ( α ) ⋅ E α . As in § 5.4.2 , we introduce a small perturbation ℓ α ′ \ell'_\alpha ℓ α ′ of ℓ α \ell_\alpha ℓ α , containing the information on how to bypass vertices lying on ℓ α \ell_\alpha ℓ α . We also record the primitive direction of ℓ α \ell_\alpha ℓ α at its initial vertex, the unimodular cone containing that direction, and the sequence of edges and triangles met by ℓ α ′ \ell'_\alpha ℓ α ′ . Whenever an E E E -line crosses an affine cut, its direction is transported using the corresponding integral-affine transition map.
Together, these data determine the index sets
I = I 0 ⊔ I e , A = A 0 ⊔ A e , A c = A 0 , c ⊔ A e . (135) I = I_0 \sqcup I_e,\qquad A = A_0 \sqcup A_e,\qquad A_c = A_{0,c} \sqcup A_e. \tag*{(135)} I = I 0 ⊔ I e , A = A 0 ⊔ A e , A c = A 0 , c ⊔ A e . (135)
The reconstruction involves two families of intersection coefficients. The first consists of the coefficients α i j a \alpha_{ij}^{a} α ij a in the curve class expansions
D i ⋅ D j = ∑ a ∈ A α i j a C a , (136) D_i \cdot D_j = \sum_{a \in A} \alpha_{ij}^{a} C_a, \tag*{(136)} D i ⋅ D j = a ∈ A ∑ α ij a C a , (136)
where ( i , j ) ∈ ( I × I ) ∖ ( I e × I e ) (i,j) \in(I \times I) \setminus(I_e \times I_e) ( i , j ) ∈ ( I × I ) ∖ ( I e × I e ) , as in § 5.1.2 .
The second family consists of the numerical divisor–curve intersections
β i a = D i ⋅ C a , ( i , a ) ∈ B , (137) \beta_{ia} = D_i \cdot C_a,\qquad(i,a) \in\mathcal{B}, \tag*{(137)} β ia = D i ⋅ C a , ( i , a ) ∈ B , (137)
where B ⊂ I × A \mathcal{B} \subset I \times A B ⊂ I × A is the admissibility domain defined in § 5.1.3 .
The required output is the collection of triple intersection numbers
κ i j k = D i ⋅ D j ⋅ D k , { i , j , k } ∩ I c ≠ ∅ . (138) \kappa_{ijk} = D_i \cdot D_j \cdot D_k,\qquad\{i,j,k\} \cap I_c \ne\varnothing. \tag*{(138)} κ ij k = D i ⋅ D j ⋅ D k , { i , j , k } ∩ I c = ∅ . (138)
Reconstruction Algorithm
Starting from the combinatorial data specified above, we reconstruct the intersection coefficients and the triple intersection form in five steps.
Step 1. Central-fiber intersection data.
For every i , j ∈ I 0 i,j \in I_0 i , j ∈ I 0 , determine the curve-valued product
D i ⋅ D j = ∑ a ∈ A 0 α i j a C a (139) D_i \cdot D_j = \sum_{a \in A_0} \alpha_{ij}^{a} C_a \tag*{(139)} D i ⋅ D j = a ∈ A 0 ∑ α ij a C a (139)
using the incidence rules of § 5.2.1 :
α i j a = { 1 , i ≠ j and e a joins v i , v j , − 1 , i = j and e a is incident to v i , 0 , otherwise . (140) \alpha_{ij}^{a} =
\begin{cases}
1, & i \ne j \text{ and } e_a \text{ joins } v_i,v_j,\\
-1, & i=j \text{ and } e_a \text{ is incident to } v_i,\\
0, & \text{otherwise}.
\end{cases} \tag*{(140)} α ij a = ⎩ ⎨ ⎧ 1 , − 1 , 0 , i = j and e a joins v i , v j , i = j and e a is incident to v i , otherwise . (140)
Next, compute the numerical intersections of central-fiber divisors with double curves. For an interior edge e a = [ v , w ] e_a=[v,w] e a = [ v , w ] , a ∈ A 0 , c a \in A_{0,c} a ∈ A 0 , c , determine β v a \beta_{va} β v a and β w a \beta_{wa} β w a from the integral-affine fan relations of § 5.2.2 .
For a vertex v i v_i v i which is not an endpoint of e a e_a e a , the intersection is determined by the number of incident triangles:
β i a = # { σ ∈ P ~ [ 2 ] : e a ⊂ σ , v i ∈ σ ∖ e a } . (141) \beta_{ia} = \#\left\{\sigma\in\widetilde{\mathcal{P}}[2] : e_a \subset\sigma,\ v_i \in\sigma\setminus e_a\right\}. \tag*{(141)} β ia = # { σ ∈ P [ 2 ] : e a ⊂ σ , v i ∈ σ ∖ e a } . (141)
In particular, all central-fiber intersection coefficients are determined directly from the combinatorics of the unimodular triangulation and of the integral-affine structure on the Symington polygon.
Step 2. Admissibility conditions.
Construct the set B \mathcal{B} B of admissible divisor–curve pairs, that is, for which a numerical intersection is defined.
Every compact curve is admissible against every divisor:
I × A c ⊂ B . (142) I \times A_c \subset\mathcal{B}. \tag*{(142)} I × A c ⊂ B . (142)
For each non-compact boundary curve C a C_a C a , a ∈ A 0 , ∂ a \in A_{0,\partial} a ∈ A 0 , ∂ , exclude the pair ( i , a ) (i,a) ( i , a ) if
v i ∈ e a , i ∈ I 0 , (143) v_i \in e_a,\qquad i \in I_0, \tag*{(143)} v i ∈ e a , i ∈ I 0 , (143)
or if
ℓ α ∩ e a ∘ ≠ ∅ , i = α ∈ I e . (144) \ell_\alpha\cap e_a^\circ\ne\varnothing,\qquad i=\alpha\in I_e. \tag*{(144)} ℓ α ∩ e a ∘ = ∅ , i = α ∈ I e . (144)
Thus
B = ( I × A c ) ⊔ { ( i , a ) ∈ I × A 0 , ∂ : ϵ i a = 0 } , (145) \mathcal{B}=(I\times A_c)\sqcup\{(i,a)\in I\times A_{0,\partial}:\epsilon_{ia}=0\}, \tag*{(145)} B = ( I × A c ) ⊔ {( i , a ) ∈ I × A 0 , ∂ : ϵ ia = 0 } , (145)
where ϵ i a \epsilon_{ia} ϵ ia is the incidence function defined in (66) .
Only coefficients β i a \beta_{ia} β ia with ( i , a ) ∈ B (i,a)\in\mathcal{B} ( i , a ) ∈ B are considered.
Step 3. Exuberant intersection data.
For each E E E -line ℓ α \ell_\alpha ℓ α , let D = D b ( α ) D=D_{b(\alpha)} D = D b ( α ) be the compact divisor corresponding to its initial vertex. Write the primitive direction of ℓ α \ell_\alpha ℓ α in a unimodular cone as
u = p u 1 + q u 2 , p , q ∈ Z ≥ 0 , gcd ( p , q ) = 1. (146) u=pu_1+qu_2,\qquad p,q\in\mathbb{Z}_{\geq0},\qquad\gcd(p,q)=1. \tag*{(146)} u = p u 1 + q u 2 , p , q ∈ Z ≥ 0 , g cd( p , q ) = 1. (146)
Let D 1 , D 2 D_1,D_2 D 1 , D 2 be the central-fiber divisors corresponding to the two vertices adjacent to v b ( α ) v_{b(\alpha)} v b ( α ) in the directions u 1 u_1 u 1 and u 2 u_2 u 2 respectively. The local intersection rule (97) gives
D ⋅ C α = − p − q , D 1 ⋅ C α = p , D 2 ⋅ C α = q , E α ⋅ C α = p q − 1. (147) \begin{aligned}
D\cdot C_\alpha&=-p-q,\\
D_1\cdot C_\alpha&=p,\\
D_2\cdot C_\alpha&=q,\\
E_\alpha\cdot C_\alpha&=pq-1. \tag*{(147)}
\end{aligned} D ⋅ C α D 1 ⋅ C α D 2 ⋅ C α E α ⋅ C α = − p − q , = p , = q , = pq − 1. (147)
The intersections of C α C_\alpha C α with all other central-fiber divisors vanish.
For each central-fiber divisor D j D_j D j , j ∈ I 0 j\in I_0 j ∈ I 0 , consider the curve-valued product
E α ⋅ D j = ∑ a ∈ A α a j α C a . (148) E_\alpha\cdot D_j=\sum_{a\in A}\alpha^\alpha_{aj}C_a. \tag*{(148)} E α ⋅ D j = a ∈ A ∑ α aj α C a . (148)
Set to zero all coefficients with a ∉ A α j a\notin A_{\alpha j} a ∈ / A α j , where A α j ⊂ A A_{\alpha j}\subset A A α j ⊂ A is defined in (109) , and set
α a , b ( α ) α = 1. (149) \alpha^\alpha_{a,b(\alpha)}=1. \tag*{(149)} α a , b ( α ) α = 1. (149)
The remaining coefficients allowed by the incidence conditions are retained as unknown integers.
Similarly, consider the numerical intersections
β α a = E α ⋅ C a , ( α , a ) ∈ B . (150) \beta_{\alpha a}=E_\alpha\cdot C_a,\qquad(\alpha,a)\in\mathcal{B}. \tag*{(150)} β α a = E α ⋅ C a , ( α , a ) ∈ B . (150)
Insert the values fixed by the local E E E -line geometry (Proposition 5.3 ), and set to zero all coefficients excluded by the incidence conditions described immediately after (118) . All remaining admissible coefficients are retained as unknown integers.
At this stage, every coefficient entering the reconstruction is either known, forced to vanish, or represented by an integer unknown.
Step 4. Compatibility equations.
We now impose the compatibility relations of § 5.5 .
For each available curve-valued product
D j ⋅ D k = ∑ a ∈ A α j k a C a , (151) D_j\cdot D_k=\sum_{a\in A}\alpha_{jk}^{a}C_a, \tag*{(151)} D j ⋅ D k = a ∈ A ∑ α j k a C a , (151)
and each divisor D i D_i D i , determine whether the expansion is admissible against D i D_i D i , namely whether one of the following holds:
(a) ( j , k ) ∈ I 0 × I 0 (j,k)\in I_0\times I_0 ( j , k ) ∈ I 0 × I 0 , so that the coefficients α j k a \alpha_{jk}^{a} α j k a are known by (72) –(75) , and we require
α j k a ≠ 0 ⟹ ( i , a ) ∈ B . (152) \alpha_{jk}^{a}\ne0\quad\Longrightarrow\quad(i,a)\in\mathcal{B}. \tag*{(152)} α j k a = 0 ⟹ ( i , a ) ∈ B . (152)
(b) ( j , k ) ∈ I e × I 0 (j,k)\in I_e\times I_0 ( j , k ) ∈ I e × I 0 , and we require
a ∈ A j k ⟹ ( i , a ) ∈ B , (153) a\in A_{jk}\quad\Longrightarrow\quad(i,a)\in\mathcal{B}, \tag*{(153)} a ∈ A j k ⟹ ( i , a ) ∈ B , (153)
where A j k ⊂ A A_{jk}\subset A A j k ⊂ A is defined in (109) .
Whenever this condition holds, form
T i ; j k = ∑ a ∈ A α j k a β i a . (154) T_{i;jk}=\sum_{a\in A}\alpha_{jk}^{a}\beta_{ia}. \tag*{(154)} T i ; j k = a ∈ A ∑ α j k a β ia . (154)
For every pair of admissible presentations of the same triple intersection, impose
T i ; j k = T j ; i k . (155) T_{i;jk}=T_{j;ik}. \tag*{(155)} T i ; j k = T j ; ik . (155)
These equations form a compatibility system of quadratic equations.
Solve this system for the unknown intersection coefficients. The existence and uniqueness of the solution are established by Theorem 6.1 in § 6.2 .
Step 5. Triple intersection numbers.
Once the compatibility system has been solved, compute the triple intersection numbers containing at least one compact divisor.
Let i ∈ I c i \in I_c i ∈ I c and j , k ∈ I j,k \in I j , k ∈ I . Since D i D_i D i is compact, the curve-valued product D i ⋅ D k D_i \cdot D_k D i ⋅ D k is supported on compact curves:
D i ⋅ D k = ∑ a ∈ A c α i k a C a . (156) D_i \cdot D_k = \sum_{a \in A_c} \alpha_{ik}^{a} C_a. \tag*{(156)} D i ⋅ D k = a ∈ A c ∑ α ik a C a . (156)
Every compact curve can be paired with D j D_j D j , so the required triple intersection is
κ i j k = ∑ a ∈ A c α i k a β j a , i ∈ I c , j , k ∈ I . (157) \kappa_{ijk} = \sum_{a \in A_c} \alpha_{ik}^{a} \beta_{ja}, \qquad i \in I_c,\quad j,k \in I. \tag*{(157)} κ ij k = a ∈ A c ∑ α ik a β j a , i ∈ I c , j , k ∈ I . (157)
Existence and Uniqueness of the Reconstruction
In this section, we prove the existence and uniqueness of the solution to the compatibility system used in the algorithm in § 6.1 .
Theorem 6.1. Let ( P , P ) (P,\mathcal{P}) ( P , P ) be a GTP. Fix a generic perturbation of the focus–focus singularities of the associated Symington polygon, as in § 3.5 , and small deformations of the E E E -lines bypassing non-initial vertices, as in § 5.4.2 . Then, the compatibility system of § 6.1 has a unique solution.
Proof. The existence is guaranteed by the existence of a corresponding Calabi–Yau threefold X ~ \widetilde{X} X , established in § 4 . The geometric intersection coefficients provide a solution of the compatibility system by the intersection theory arguments presented in § 5 . Therefore, we only need to establish the uniqueness of a solution. The key observation is that the initially undetermined intersection coefficients can be determined recursively by applying the compatibility relations along each E E E -line, propagating from its initial vertex through successive crossings of edges and vertices.
Fix an exuberant divisor E = E α E = E_\alpha E = E α and orient its E E E -line ℓ α \ell_\alpha ℓ α from its initial vertex towards its terminal point. We want to determine the intersection coefficients α α j a \alpha_{\alpha j}^{a} α α j a in the expansions
E α ⋅ D j = ∑ a ∈ A α α j a C a (158) E_\alpha\cdot D_j = \sum_{a \in A} \alpha_{\alpha j}^{a} C_a \tag*{(158)} E α ⋅ D j = a ∈ A ∑ α α j a C a (158)
for j ∈ I 0 j \in I_0 j ∈ I 0 corresponding to a vertex of P ~ \widetilde{P} P on ℓ α \ell_\alpha ℓ α . At the initial vertex v b ( α ) v_{b(\alpha)} v b ( α ) , we have E α ⋅ D v b ( α ) = C α E_\alpha\cdot D_{v_{b(\alpha)}} = C_\alpha E α ⋅ D v b ( α ) = C α . Next, we explain how to uniquely propagate these curve expansions along the E E E -line ℓ α \ell_\alpha ℓ α , through intersection with edges and vertices of P ~ \widetilde{P} P .
We first consider an intersection with an interior edge of P ~ \widetilde{P} P . The case of a boundary edge can be treated similarly. Denote the interior edge by e = [ v , w ] e = [v,w] e = [ v , w ] , separating two unimodular triangles
σ − = ( v , w , x ) , σ + = ( v , w , y ) , (159) \sigma_- = (v,w,x), \qquad\sigma_+ = (v,w,y), \tag*{(159)} σ − = ( v , w , x ) , σ + = ( v , w , y ) , (159)
where σ − \sigma_{-} σ − lies before the crossing and σ + \sigma_{+} σ + lies after it. Let C e = D v ∩ D w C_e = D_v \cap D_w C e = D v ∩ D w be the corresponding compact double curve. Assume that the E E E -line ℓ α \ell_{\alpha} ℓ α goes out of the triangle σ + \sigma_{+} σ + through the edge [ w , y ] [w,y] [ w , y ] .
We first recall the relevant central-fiber intersection numbers. We express the two triangles in a common integral-affine chart, using parallel transport across any intervening affine cut. After an integral-affine change of coordinates, we may take
v = ( 0 , 0 ) , w = ( 1 , 0 ) , x = ( 0 , 1 ) , y = ( m , − 1 ) , (160) v = (0,0), \qquad w = (1,0), \qquad x = (0,1), \qquad y = (m,-1), \tag*{(160)} v = ( 0 , 0 ) , w = ( 1 , 0 ) , x = ( 0 , 1 ) , y = ( m , − 1 ) , (160)
for some m ∈ Z m \in\mathbb{Z} m ∈ Z . The integral-affine fan relations of § 5.2.2 give
D v ⋅ C e = m − 2 , D w ⋅ C e = − m . (161) D_v \cdot C_e = m - 2, \qquad D_w \cdot C_e = -m. \tag*{(161)} D v ⋅ C e = m − 2 , D w ⋅ C e = − m . (161)
The two remaining central-fiber divisors intersect C e C_e C e transversely at the triple points corresponding to σ − \sigma_{-} σ − and σ + \sigma_{+} σ + . Consequently, D x ⋅ C e = D y ⋅ C e = 1 D_x \cdot C_e = D_y \cdot C_e = 1 D x ⋅ C e = D y ⋅ C e = 1 .
By the assumption that ℓ α \ell_{\alpha} ℓ α goes out of σ + \sigma_{+} σ + through the edge [ w , y ] [w,y] [ w , y ] , we have
E α ⋅ D v = Γ o l d + n e C e , (162) E_{\alpha} \cdot D_v = \Gamma_{\mathrm{old}} + n_e C_e, \tag*{(162)} E α ⋅ D v = Γ old + n e C e , (162)
where Γ o l d \Gamma_{\mathrm{old}} Γ old is determined by the preceding data and n e n_e n e is the only new coefficient. Then, the compatibility relation
D x ⋅ ( E α ⋅ D v ) = ( E α ⋅ D x ) ⋅ D v (163) D_x \cdot(E_{\alpha} \cdot D_v) = (E_{\alpha} \cdot D_x) \cdot D_v \tag*{(163)} D x ⋅ ( E α ⋅ D v ) = ( E α ⋅ D x ) ⋅ D v (163)
uniquely determines n e n_e n e by
n e = ( E α ⋅ D x ) ⋅ D v − D x ⋅ Γ o l d , (164) n_e = (E_{\alpha} \cdot D_x) \cdot D_v - D_x \cdot\Gamma_{\mathrm{old}}, \tag*{(164)} n e = ( E α ⋅ D x ) ⋅ D v − D x ⋅ Γ old , (164)
where the right-hand side is already determined.
Finally, we consider an intersection of the E E E -line ℓ α \ell_{\alpha} ℓ α with a vertex v v v of P ~ \widetilde{\mathcal{P}} P . Then, the uniqueness of the curve expansion follows as in the proof of Lemma 5.5 .
Examples of Calabi–Yau Threefolds from GTPs
In this section, we work out explicitly a few examples, following the algorithm of § 6 . The first example is the one already discussed in § 3.3 , which we work out in full detail. We have chosen this example because it displays most of the subtleties of the problem. The second example is one for which a gauge theory phase is known, which allows us to compare our results with expectations from the gauge theoretic prepotential.
A fully worked out Example
Consider the GTP of § 3.3 , see Figure 7.1 . The GTP P \mathcal{P} P , with its triangulation P ~ \widetilde{\mathcal{P}} P , has 6 vertices, conveniently labelled by { 1 , 2 , 3 ′ , 3 ′ ′ , 4 , 5 } \{1,2,3',3'',4,5\} { 1 , 2 , 3 ′ , 3 ′′ , 4 , 5 } , which after identification correspond to the 5 vertices, labelled by { 1 , 2 , 3 , 4 , 5 } \{1,2,3,4,5\} { 1 , 2 , 3 , 4 , 5 } , of the Symington polygon P ~ \widetilde{P} P , drawn with a monodromy cut on the right of Figure 3.9 . The identification is the affine map fixing v 1 v_1 v 1 with linear part
Figure 7.1. Left: the GTP P P P , with the generalized face F F F shaded, the white dot on the decorated boundary segment [ v 3 ′ , v 3 ′ ′ ] [v_3^\prime,v_3^{\prime\prime}] [ v 3 ′ , v 3 ′′ ] , and the E E E -line ℓ 6 \ell_6 ℓ 6 of the exuberant divisor E 6 E_6 E 6 in red. Each edge is labelled by its double curve C a C_a C a . The two edges labelled C 2 C_2 C 2 are glued into a single edge of P ~ \widetilde{P} P . In both panels, the triangles are labelled σ 1 , … , σ 5 \sigma_1,\ldots,\sigma_5 σ 1 , … , σ 5 . Right: the Symington polygon P ~ \widetilde{P} P with its triangulation P ~ \widetilde{\mathcal{P}} P , drawn with curved edges. The vertex labelled i i i is v i v_i v i , dual to the component D i D_i D i of X ~ 0 \widetilde{X}_0 X 0 , and v 3 v_3 v 3 is the common image of v 3 ′ v_3^\prime v 3 ′ and v 3 ′ ′ v_3^{\prime\prime} v 3 ′′ . The focus–focus singularity at v 1 v_1 v 1 is marked in red. Each edge is labelled by its double curve C a C_a C a , and the circled number next to vertex i i i on the edge C a C_a C a is β i a \beta_{ia} β ia . The E E E -line ℓ 6 \ell_6 ℓ 6 starts at v 1 v_1 v 1 , where the exuberant curve C 7 = D 1 ⋅ E 6 C_7=D_1\cdot E_6 C 7 = D 1 ⋅ E 6 lies, crosses C 4 C_4 C 4 and C 5 C_5 C 5 , and ends on C 9 C_9 C 9 .
M = ( 1 − 1 0 1 ) = M [ 0 , 1 ] T , M ( − 1 − 2 ) = ( 1 − 2 ) , M ( 1 0 ) = ( 1 0 ) . (165) M=\begin{pmatrix}1&-1\\0&1\end{pmatrix}=M_{[0,1]}^{T},\qquad M\begin{pmatrix}-1\\-2\end{pmatrix}=\begin{pmatrix}1\\-2\end{pmatrix},\qquad M\begin{pmatrix}1\\0\end{pmatrix}=\begin{pmatrix}1\\0\end{pmatrix}. \tag*{(165)} M = ( 1 0 − 1 1 ) = M [ 0 , 1 ] T , M ( − 1 − 2 ) = ( 1 − 2 ) , M ( 1 0 ) = ( 1 0 ) . (165)
The Symington polygon thus has 5 vertices and 5 triangles. The vertex v 1 v_1 v 1 is the focus–focus singularity, with monodromy M M M and monodromy-invariant direction ( 1 , 0 ) (1,0) ( 1 , 0 ) . We denote by p σ p_\sigma p σ the triple point of X ~ 0 \widetilde{X}_0 X 0 dual to a triangle σ \sigma σ .
We label divisors and curves as in Figure 7.1 . The divisor index sets are
I 0 = { 1 , 2 , 3 , 4 , 5 } , I c = { 1 , 2 } , I 0 , ∂ = { 3 , 4 , 5 } , I e = { 6 } , I = { 1 , … , 6 } . (166) I_0=\{1,2,3,4,5\},\qquad I_c=\{1,2\},\qquad I_{0,\partial}=\{3,4,5\},\qquad I_e=\{6\},\qquad I=\{1,\ldots,6\}. \tag*{(166)} I 0 = { 1 , 2 , 3 , 4 , 5 } , I c = { 1 , 2 } , I 0 , ∂ = { 3 , 4 , 5 } , I e = { 6 } , I = { 1 , … , 6 } . (166)
with D i = X v i D_i=X_{v_i} D i = X v i for i ∈ I 0 i\in I_0 i ∈ I 0 , and E 6 E_6 E 6 the exuberant divisor of the single 7-brane. Of the two monodromy-invariant rays ± ( 1 , 0 ) \pm(1,0) ± ( 1 , 0 ) at v 1 v_1 v 1 , we take ℓ 6 \ell_6 ℓ 6 to start in the direction u = ( 1 , 0 ) u=(1,0) u = ( 1 , 0 ) , as in
Figure 7.1 . Its initial vertex is v b ( 6 ) = v 1 v_{b(6)}=v_1 v b ( 6 ) = v 1 , so b ( 6 ) = 1 b(6)=1 b ( 6 ) = 1 . The curve index sets are
A 0 , c = { 1 , … , 6 } , A 0 , ∂ = { 8 , 9 , 10 } , A e = { 7 } , A c = { 1 , … , 7 } , A = { 1 , … , 10 } , (167) A_{0,c}=\{1,\ldots,6\},\qquad A_{0,\partial}=\{8,9,10\},\qquad A_e=\{7\},\qquad A_c=\{1,\ldots,7\},\qquad A=\{1,\ldots,10\}, \tag*{(167)} A 0 , c = { 1 , … , 6 } , A 0 , ∂ = { 8 , 9 , 10 } , A e = { 7 } , A c = { 1 , … , 7 } , A = { 1 , … , 10 } , (167)
where C 7 = D 1 ⋅ E 6 C_7=D_1\cdot E_6 C 7 = D 1 ⋅ E 6 is the exuberant curve, so that the bijection A e ≃ I e A_e\simeq I_e A e ≃ I e of § 4.2.2 reads 7 ↔ 6 7\leftrightarrow6 7 ↔ 6 .
Step 1: central-fiber intersection data. The incidence rule (140) gives the curve-valued products
D 1 2 = − C 1 − C 2 , D 1 D 2 = C 1 , D 1 D 3 = C 2 , D 2 2 = − C 1 − C 3 − C 4 − C 5 − C 6 , D 2 D 3 = C 3 + C 4 , D 2 D 4 = C 5 , D 3 2 = − C 2 − C 3 − C 4 − C 8 − C 10 , D 2 D 5 = C 6 , D 3 D 4 = C 10 , D 4 2 = − C 5 − C 9 − C 10 , D 3 D 5 = C 8 , D 4 D 5 = C 9 , D 5 2 = − C 6 − C 8 − C 9 , D 1 D 4 = 0 , D 1 D 5 = 0. (168) \begin{aligned}
D_1^2&=-C_1-C_2, & D_1D_2&=C_1, & D_1D_3&=C_2,\\
D_2^2&=-C_1-C_3-C_4-C_5-C_6, & D_2D_3&=C_3+C_4, & D_2D_4&=C_5,\\
D_3^2&=-C_2-C_3-C_4-C_8-C_{10}, & D_2D_5&=C_6, & D_3D_4&=C_{10},\\
D_4^2&=-C_5-C_9-C_{10}, & D_3D_5&=C_8, & D_4D_5&=C_9,\\
D_5^2&=-C_6-C_8-C_9, & D_1D_4&=0, & D_1D_5&=0. \tag*{(168)}
\end{aligned} D 1 2 D 2 2 D 3 2 D 4 2 D 5 2 = − C 1 − C 2 , = − C 1 − C 3 − C 4 − C 5 − C 6 , = − C 2 − C 3 − C 4 − C 8 − C 10 , = − C 5 − C 9 − C 10 , = − C 6 − C 8 − C 9 , D 1 D 2 D 2 D 3 D 2 D 5 D 3 D 5 D 1 D 4 = C 1 , = C 3 + C 4 , = C 6 , = C 8 , = 0 , D 1 D 3 D 2 D 4 D 3 D 4 D 4 D 5 D 1 D 5 = C 2 , = C 5 , = C 10 , = C 9 , = 0. (168)
In particular, note that D 2 ⋅ D 3 D_2\cdot D_3 D 2 ⋅ D 3 has the two components C 3 C_3 C 3 and C 4 C_4 C 4 ! This feature is specific to generalized toric geometry.
To compute the coefficients β v a \beta_{va} β v a and β w a \beta_{wa} β w a for e a = [ v , w ] e_a=[v,w] e a = [ v , w ] , we use the fan relations (87) and (89) :
Table 7.1.
In every row β v a + β w a = − 2 \beta_{va}+\beta_{wa}=-2 β v a + β w a = − 2 , as required by (92) . This gives the circled numbers in Figure 7.1 .
The rows a = 1 , 2 a=1,2 a = 1 , 2 determine the surface D 1 D_1 D 1 . The rays of X ‾ v 1 \overline{X}_{v_1} X v 1 satisfy ( − 1 , − 2 ) + ( 1 , − 2 ) + 4 ( 0 , 1 ) = 0 (-1,-2)+(1,-2)+4(0,1)=0 ( − 1 , − 2 ) + ( 1 , − 2 ) + 4 ( 0 , 1 ) = 0 , so X ‾ v 1 ≃ P ( 1 , 1 , 4 ) \overline{X}_{v_1}\simeq\mathbb{P}(1,1,4) X v 1 ≃ P ( 1 , 1 , 4 ) , and the cut cone ⟨ ( − 1 , − 2 ) , ( 1 , − 2 ) ⟩ \langle(-1,-2),(1,-2)\rangle ⟨( − 1 , − 2 ) , ( 1 , − 2 )⟩ is the Wahl singularity C 2 / ( Z / 4 Z ) \mathbb{C}^2/(\mathbb{Z}/4\mathbb{Z}) C 2 / ( Z /4 Z ) with weights ( 1 , 1 ) (1,1) ( 1 , 1 ) , that is n = 2 n=2 n = 2 . Its smoothing is D 1 ≃ P 2 D_1\simeq\mathbb{P}^2 D 1 ≃ P 2 , with ∂ D 1 = C 1 + C 2 \partial D_1=C_1+C_2 ∂ D 1 = C 1 + C 2 . The curve C 1 C_1 C 1 of the ray ( 0 , 1 ) (0,1) ( 0 , 1 ) is a conic, ( C 1 ∣ D 1 ) 2 = β 21 = 4 (C_1|_{D_1})^2=\beta_{21}=4 ( C 1 ∣ D 1 ) 2 = β 21 = 4 . The two boundary curves of the rays ( ∓ 1 , − 2 ) (\mp1,-2) ( ∓ 1 , − 2 ) have self-intersection 1 4 \frac{1}{4} 4 1 and meet with intersection number 1 4 \frac{1}{4} 4 1 at the singular point. The smoothing merges them into the line C 2 C_2 C 2 , with ( C 2 ∣ D 1 ) 2 = 1 4 + 1 4 + 2 ⋅ 1 4 = 1 = β 32 (C_2|_{D_1})^2=\frac{1}{4}+\frac{1}{4}+2\cdot\frac{1}{4}=1=\beta_{32} ( C 2 ∣ D 1 ) 2 = 4 1 + 4 1 + 2 ⋅ 4 1 = 1 = β 32 .
For transverse intersections, (141) gives the non-zero values
β 31 = 2 , β 22 = 2 , β 13 = β 53 = 1 , β 14 = β 44 = 1 , β 35 = β 55 = 1 , β 36 = β 46 = 1 , \beta_{31}=2,\qquad\beta_{22}=2,\qquad\beta_{13}=\beta_{53}=1,\qquad\beta_{14}=\beta_{44}=1,\qquad\beta_{35}=\beta_{55}=1,\qquad\beta_{36}=\beta_{46}=1, β 31 = 2 , β 22 = 2 , β 13 = β 53 = 1 , β 14 = β 44 = 1 , β 35 = β 55 = 1 , β 36 = β 46 = 1 ,
β 28 = β 29 = β 2 , 10 = 1. (169) \beta_{28}=\beta_{29}=\beta_{2,10}=1. \tag*{(169)} β 28 = β 29 = β 2 , 10 = 1. (169)
Here note in particular that β 31 = 2 \beta_{31}=2 β 31 = 2 and β 22 = 2 \beta_{22}=2 β 22 = 2 because the two triangles containing e 1 e_1 e 1 , respectively e 2 e_2 e 2 , have the same third vertex v 3 v_3 v 3 , respectively v 2 v_2 v 2 , in P ~ \widetilde{\mathcal{P}} P . Again, this is specific to generalized toric geometry. All other transverse coefficients vanish.
Step 2: admissibility conditions. All pairs ( i , a ) (i,a) ( i , a ) with a ∈ A c a\in A_c a ∈ A c are admissible. For the boundary curves, (143) excludes the pairs with v i ∈ e a v_i\in e_a v i ∈ e a , namely ( 3 , 8 ) (3,8) ( 3 , 8 ) , ( 5 , 8 ) (5,8) ( 5 , 8 ) , ( 4 , 9 ) (4,9) ( 4 , 9 ) , ( 5 , 9 ) (5,9) ( 5 , 9 ) , ( 3 , 10 ) (3,10) ( 3 , 10 ) and ( 4 , 10 ) (4,10) ( 4 , 10 ) . The E E E -line ℓ 6 \ell_6 ℓ 6 meets the relative interior of e 9 e_9 e 9 and does not meet e 8 e_8 e 8 or e 10 e_{10} e 10 , so (144) excludes ( 6 , 9 ) (6,9) ( 6 , 9 ) only. Thus
B = ( I × A ) ∖ { ( 3 , 8 ) , ( 5 , 8 ) , ( 4 , 9 ) , ( 5 , 9 ) , ( 6 , 9 ) , ( 3 , 10 ) , ( 4 , 10 ) } . (170) \mathcal{B}=(I\times A)\setminus\{(3,8),(5,8),(4,9),(5,9),(6,9),(3,10),(4,10)\}. \tag*{(170)} B = ( I × A ) ∖ {( 3 , 8 ) , ( 5 , 8 ) , ( 4 , 9 ) , ( 5 , 9 ) , ( 6 , 9 ) , ( 3 , 10 ) , ( 4 , 10 )} . (170)
Collecting the results, with × \times × marking the pairs excluded in (170) , one finds the β \beta β -matrix
Table 7.8.
Each column indexed by a compact curve sums to zero, as it must since X ~ e 0 = ∑ i ∈ I 0 D i \widetilde{X}_{e_0}=\sum_{i\in I_0}D_i X e 0 = ∑ i ∈ I 0 D i is linearly equivalent to zero.
Step 3: exuberant intersection data. The direction u = ( 1 , 0 ) u=(1,0) u = ( 1 , 0 ) of ℓ 6 \ell_6 ℓ 6 lies in the unimodular cone at v 1 v_1 v 1 , generated by u 1 = v 3 ′ ′ − v 1 = ( 1 , − 2 ) u_1 = v_{3''} - v_1 = (1,-2) u 1 = v 3 ′′ − v 1 = ( 1 , − 2 ) and u 2 = v 2 − v 1 = ( 0 , 1 ) u_2 = v_2 - v_1 = (0,1) u 2 = v 2 − v 1 = ( 0 , 1 ) :
u = ( 1 , 0 ) = 1 ⋅ ( 1 , − 2 ) + 2 ⋅ ( 0 , 1 ) , ( p , q ) = ( 1 , 2 ) . (171) u = (1,0) = 1 \cdot(1,-2) + 2 \cdot(0,1), \qquad(p,q) = (1,2). \tag*{(171)} u = ( 1 , 0 ) = 1 ⋅ ( 1 , − 2 ) + 2 ⋅ ( 0 , 1 ) , ( p , q ) = ( 1 , 2 ) . (171)
The local rule (147) gives
β 17 = − p − q = − 3 , β 37 = p = 1 , β 27 = q = 2 , β 67 = p q − 1 = 1 , β 47 = β 57 = 0. (172) \beta_{17} = -p-q = -3,\qquad\beta_{37} = p = 1,\qquad\beta_{27} = q = 2,\qquad\beta_{67} = pq-1 = 1,\qquad\beta_{47} = \beta_{57} = 0. \tag*{(172)} β 17 = − p − q = − 3 , β 37 = p = 1 , β 27 = q = 2 , β 67 = pq − 1 = 1 , β 47 = β 57 = 0. (172)
In D 1 ≃ P 2 D_1 \simeq\mathbb{P}^2 D 1 ≃ P 2 , this says C 7 ⋅ C 1 = 2 C_7 \cdot C_1 = 2 C 7 ⋅ C 1 = 2 , C 7 ⋅ C 2 = 1 C_7 \cdot C_2 = 1 C 7 ⋅ C 2 = 1 and ( C 7 ∣ D 1 ) 2 = 1 (C_7|_{D_1})^2 = 1 ( C 7 ∣ D 1 ) 2 = 1 : the exuberant curve C 7 C_7 C 7 is a line, meeting ∂ D 1 \partial D_1 ∂ D 1 at the single point p σ 2 ∈ C 1 ∩ C 2 p_{\sigma_2} \in C_1 \cap C_2 p σ 2 ∈ C 1 ∩ C 2 with multiplicity p + q = 3 p+q = 3 p + q = 3 , i.e. the tangent line to the conic C 1 C_1 C 1 at p σ 2 p_{\sigma_2} p σ 2 , see Figure 7.2 .
Figure 7.2. The surface D 1 D_1 D 1 of Example 1. Each ray is labelled by its curve, with the self-intersection of that curve in the surface in brackets. (a) The star of v 1 v_1 v 1 in P \mathcal{P} P spans the fan of X ‾ v 1 ≃ P ( 1 , 1 , 4 ) \overline{X}_{v_1} \simeq\mathbb{P}(1,1,4) X v 1 ≃ P ( 1 , 1 , 4 ) , with ( − 1 , − 2 ) + ( 1 , − 2 ) + 4 ( 0 , 1 ) = 0 (-1,-2)+(1,-2)+4(0,1)=0 ( − 1 , − 2 ) + ( 1 , − 2 ) + 4 ( 0 , 1 ) = 0 . The shaded cut cone is the cone over F F F and gives the Wahl singularity 1 4 ( 1 , 1 ) \frac{1}{4}(1,1) 4 1 ( 1 , 1 ) . The boundary curves C 2 ′ C_2' C 2 ′ and C 2 ′ ′ C_2'' C 2 ′′ of the rays towards v 3 ′ v_3' v 3 ′ and v 3 ′ ′ v_3'' v 3 ′′ are merged into C 2 C_2 C 2 by the smoothing. (b) D 1 ≃ P 2 D_1 \simeq\mathbb{P}^2 D 1 ≃ P 2 , with ∂ D 1 = C 1 + C 2 \partial D_1=C_1+C_2 ∂ D 1 = C 1 + C 2 a conic and a line meeting at the triple points p σ 1 p_{\sigma_1} p σ 1 and p σ 2 p_{\sigma_2} p σ 2 . The exuberant curve C 7 C_7 C 7 of Step 3 is the tangent line to C 1 C_1 C 1 at p σ 2 p_{\sigma_2} p σ 2 .
Next, the incidence of ℓ 6 \ell_6 ℓ 6 restricts the curve-valued products E 6 ⋅ D j E_6 \cdot D_j E 6 ⋅ D j . By (114) , α 61 4 = 1 \alpha^4_{61} = 1 α 61 4 = 1 . Besides C 7 C_7 C 7 , the only curves that can appear are the double curves C 4 C_4 C 4 , C 5 C_5 C 5 , C 9 C_9 C 9 of the edges met by ℓ 6 \ell_6 ℓ 6 , in the products with the divisors attached to their endpoints. Hence
E 6 ⋅ D 1 = C 7 , E 6 ⋅ D 2 = α 62 4 C 4 + α 62 5 C 5 , E 6 ⋅ D 3 = α 63 4 C 4 , E 6 ⋅ D 4 = α 64 5 C 5 + α 64 9 C 9 , E 6 ⋅ D 5 = α 65 9 C 9 . (173) \begin{aligned}
E_6 \cdot D_1 &= C_7,\\
E_6 \cdot D_2 &= \alpha^4_{62}C_4 + \alpha^5_{62}C_5,\qquad E_6 \cdot D_3 = \alpha^4_{63}C_4,\\
E_6 \cdot D_4 &= \alpha^5_{64}C_5 + \alpha^9_{64}C_9,\qquad E_6 \cdot D_5 = \alpha^9_{65}C_9. \tag*{(173)}
\end{aligned} E 6 ⋅ D 1 E 6 ⋅ D 2 E 6 ⋅ D 4 = C 7 , = α 62 4 C 4 + α 62 5 C 5 , E 6 ⋅ D 3 = α 63 4 C 4 , = α 64 5 C 5 + α 64 9 C 9 , E 6 ⋅ D 5 = α 65 9 C 9 . (173)
with six unknown integers α 62 4 \alpha^4_{62} α 62 4 , α 62 5 \alpha^5_{62} α 62 5 , α 63 4 \alpha^4_{63} α 63 4 , α 64 5 \alpha^5_{64} α 64 5 , α 64 9 \alpha^9_{64} α 64 9 , α 65 9 \alpha^9_{65} α 65 9 .
For the numerical intersections β 6 a = E 6 ⋅ C a \beta_{6a} = E_6 \cdot C_a β 6 a = E 6 ⋅ C a , ( 6 , a ) ∈ B (6,a) \in\mathcal{B} ( 6 , a ) ∈ B , the value β 67 = 1 \beta_{67} = 1 β 67 = 1 is fixed by (172) . The curve E 6 ∩ X ~ 0 E_6 \cap\widetilde{X}_0 E 6 ∩ X 0 meets the double locus of X ~ 0 \widetilde{X}_0 X 0 only along C 4 ∪ C 5 ∪ C 9 C_4 \cup C_5 \cup C_9 C 4 ∪ C 5 ∪ C 9 , which contains the triple points p σ 2 , p σ 3 , p σ 4 p_{\sigma_2}, p_{\sigma_3}, p_{\sigma_4} p σ 2 , p σ 3 , p σ 4 of the triangles traversed by ℓ 6 \ell_6 ℓ 6 . Since two double curves meet only at the triple points of common triangles, a double curve C a C_a C a can meet E 6 E_6 E 6 only if e a e_a e a is an edge of σ 2 \sigma_2 σ 2 , σ 3 \sigma_3 σ 3 or σ 4 \sigma_4 σ 4 . The edges e 3 e_3 e 3 and e 8 e_8 e 8 belong only to σ 1 \sigma_1 σ 1 and σ 5 \sigma_5 σ 5 . Hence C 3 C_3 C 3 and C 8 C_8 C 8 are disjoint from E 6 E_6 E 6 , and by (118)
β 63 = β 68 = 0. (174) \beta_{63} = \beta_{68} = 0. \tag*{(174)} β 63 = β 68 = 0. (174)
The six remaining admissible coefficients β 61 \beta_{61} β 61 , β 62 \beta_{62} β 62 , β 64 \beta_{64} β 64 , β 65 \beta_{65} β 65 , β 66 \beta_{66} β 66 , β 6 , 10 \beta_{6,10} β 6 , 10 are unknown integers (recall that β 69 \beta_{69} β 69 is not defined). Altogether, the compatibility system has twelve unknowns.
Step 4: compatibility equations. For i , j , k ∈ I 0 i,j,k \in I_0 i , j , k ∈ I 0 , all admissible presentations T i ; j k T_{i;jk} T i ; j k follow from (168) and (7.8), and the relations T i ; j k = T j ; i k T_{i;jk} = T_{j;ik} T i ; j k = T j ; ik hold identically. For instance, T 1 ; 22 = T 2 ; 12 = 4 T_{1;22} = T_{2;12} = 4 T 1 ; 22 = T 2 ; 12 = 4 , T 3 ; 12 = T 1 ; 23 = T 2 ; 13 = 2 T_{3;12} = T_{1;23} = T_{2;13} = 2 T 3 ; 12 = T 1 ; 23 = T 2 ; 13 = 2 , T 2 ; 33 = T 3 ; 23 = − 1 T_{2;33} = T_{3;23} = -1 T 2 ; 33 = T 3 ; 23 = − 1 and T 5 ; 22 = T 2 ; 25 = − 7 T_{5;22} = T_{2;25} = -7 T 5 ; 22 = T 2 ; 25 = − 7 .
The non-trivial relations involve E 6 E_6 E 6 . For j , k ∈ I 0 j,k \in I_0 j , k ∈ I 0 , the triple product E 6 ⋅ D j ⋅ D k E_6 \cdot D_j \cdot D_k E 6 ⋅ D j ⋅ D k has the presentations T 6 ; j k T_{6;jk} T 6 ; j k , T j ; 6 k T_{j;6k} T j ; 6 k and T k ; 6 j T_{k;6j} T k ; 6 j of (154) , whenever these are admissible in the sense of (6.20). Using (7.4), (7.8), (7.10), (7.11) and (7.12), we find
( 6 , 1 , 1 ) : T 6 ; 11 = − β 61 − β 62 , T 1 ; 61 = β 17 = − 3 , ( 6 , 1 , 2 ) : T 6 ; 12 = β 61 , T 1 ; 62 = α 62 4 β 14 + α 62 5 β 15 = α 62 4 , T 2 ; 61 = β 27 = 2 , ( 6 , 1 , 3 ) : T 6 ; 13 = β 62 , T 1 ; 63 = α 63 4 β 14 = α 63 4 , T 3 ; 61 = β 37 = 1 , ( 6 , 1 , 4 ) : T 6 ; 14 = 0 , T 1 ; 64 = α 64 5 β 15 + α 64 9 β 19 = 0 , T 4 ; 61 = β 47 = 0 , ( 6 , 1 , 5 ) : T 6 ; 15 = 0 , T 1 ; 65 = α 65 9 β 19 = 0 , T 5 ; 61 = β 57 = 0 , ( 6 , 2 , 2 ) : T 6 ; 22 = − β 61 − β 63 − β 64 − β 65 − β 66 , T 2 ; 62 = α 62 4 β 24 + α 62 5 β 25 = − α 62 4 − α 62 5 , ( 6 , 2 , 3 ) : T 6 ; 23 = β 63 + β 64 , T 2 ; 63 = α 63 4 β 24 = − α 63 4 , T 3 ; 62 = α 62 4 β 34 + α 62 5 β 35 = − α 62 4 + α 62 5 , ( 6 , 2 , 4 ) : T 6 ; 24 = β 65 , T 2 ; 64 = α 64 5 β 25 + α 64 9 β 29 = − α 64 5 + α 64 9 , T 4 ; 62 = α 62 4 β 44 + α 62 5 β 45 = α 62 4 − α 62 5 , ( 6 , 2 , 5 ) : T 6 ; 25 = β 66 , T 2 ; 65 = α 65 9 β 29 = α 65 9 , T 5 ; 62 = α 62 4 β 54 + α 62 5 β 55 = α 62 5 , ( 6 , 3 , 3 ) : T 6 ; 33 = − β 62 − β 63 − β 64 − β 68 − β 6 , 10 , T 3 ; 63 = α 63 4 β 34 = − α 63 4 , ( 6 , 3 , 4 ) : T 6 ; 34 = β 6 , 10 , T 3 ; 64 = α 64 5 β 35 + α 64 9 β 39 = α 64 5 , T 4 ; 63 = α 63 4 β 44 = α 63 4 , ( 6 , 3 , 5 ) : T 6 ; 35 = β 68 , T 3 ; 65 = α 65 9 β 39 = 0 , T 5 ; 63 = α 63 4 β 54 = 0. (175) \begin{aligned}
(6,1,1):\quad&T_{6;11}=-\beta_{61}-\beta_{62},\quad T_{1;61}=\beta_{17}=-3,\\
(6,1,2):\quad&T_{6;12}=\beta_{61},\quad T_{1;62}=\alpha_{62}^{4}\beta_{14}+\alpha_{62}^{5}\beta_{15}=\alpha_{62}^{4},\quad T_{2;61}=\beta_{27}=2,\\
(6,1,3):\quad&T_{6;13}=\beta_{62},\quad T_{1;63}=\alpha_{63}^{4}\beta_{14}=\alpha_{63}^{4},\quad T_{3;61}=\beta_{37}=1,\\
(6,1,4):\quad&T_{6;14}=0,\quad T_{1;64}=\alpha_{64}^{5}\beta_{15}+\alpha_{64}^{9}\beta_{19}=0,\quad T_{4;61}=\beta_{47}=0,\\
(6,1,5):\quad&T_{6;15}=0,\quad T_{1;65}=\alpha_{65}^{9}\beta_{19}=0,\quad T_{5;61}=\beta_{57}=0,\\
(6,2,2):\quad&T_{6;22}=-\beta_{61}-\beta_{63}-\beta_{64}-\beta_{65}-\beta_{66},\\
&T_{2;62}=\alpha_{62}^{4}\beta_{24}+\alpha_{62}^{5}\beta_{25}=-\alpha_{62}^{4}-\alpha_{62}^{5},\\
(6,2,3):\quad&T_{6;23}=\beta_{63}+\beta_{64},\quad T_{2;63}=\alpha_{63}^{4}\beta_{24}=-\alpha_{63}^{4},\\
&T_{3;62}=\alpha_{62}^{4}\beta_{34}+\alpha_{62}^{5}\beta_{35}=-\alpha_{62}^{4}+\alpha_{62}^{5},\\
(6,2,4):\quad&T_{6;24}=\beta_{65},\quad T_{2;64}=\alpha_{64}^{5}\beta_{25}+\alpha_{64}^{9}\beta_{29}=-\alpha_{64}^{5}+\alpha_{64}^{9},\\
&T_{4;62}=\alpha_{62}^{4}\beta_{44}+\alpha_{62}^{5}\beta_{45}=\alpha_{62}^{4}-\alpha_{62}^{5},\\
(6,2,5):\quad&T_{6;25}=\beta_{66},\quad T_{2;65}=\alpha_{65}^{9}\beta_{29}=\alpha_{65}^{9},\quad T_{5;62}=\alpha_{62}^{4}\beta_{54}+\alpha_{62}^{5}\beta_{55}=\alpha_{62}^{5},\\
(6,3,3):\quad&T_{6;33}=-\beta_{62}-\beta_{63}-\beta_{64}-\beta_{68}-\beta_{6,10},\quad T_{3;63}=\alpha_{63}^{4}\beta_{34}=-\alpha_{63}^{4},\\
(6,3,4):\quad&T_{6;34}=\beta_{6,10},\quad T_{3;64}=\alpha_{64}^{5}\beta_{35}+\alpha_{64}^{9}\beta_{39}=\alpha_{64}^{5},\quad T_{4;63}=\alpha_{63}^{4}\beta_{44}=\alpha_{63}^{4},\\
(6,3,5):\quad&T_{6;35}=\beta_{68},\quad T_{3;65}=\alpha_{65}^{9}\beta_{39}=0,\quad T_{5;63}=\alpha_{63}^{4}\beta_{54}=0. \tag*{(175)}
\end{aligned} ( 6 , 1 , 1 ) : ( 6 , 1 , 2 ) : ( 6 , 1 , 3 ) : ( 6 , 1 , 4 ) : ( 6 , 1 , 5 ) : ( 6 , 2 , 2 ) : ( 6 , 2 , 3 ) : ( 6 , 2 , 4 ) : ( 6 , 2 , 5 ) : ( 6 , 3 , 3 ) : ( 6 , 3 , 4 ) : ( 6 , 3 , 5 ) : T 6 ; 11 = − β 61 − β 62 , T 1 ; 61 = β 17 = − 3 , T 6 ; 12 = β 61 , T 1 ; 62 = α 62 4 β 14 + α 62 5 β 15 = α 62 4 , T 2 ; 61 = β 27 = 2 , T 6 ; 13 = β 62 , T 1 ; 63 = α 63 4 β 14 = α 63 4 , T 3 ; 61 = β 37 = 1 , T 6 ; 14 = 0 , T 1 ; 64 = α 64 5 β 15 + α 64 9 β 19 = 0 , T 4 ; 61 = β 47 = 0 , T 6 ; 15 = 0 , T 1 ; 65 = α 65 9 β 19 = 0 , T 5 ; 61 = β 57 = 0 , T 6 ; 22 = − β 61 − β 63 − β 64 − β 65 − β 66 , T 2 ; 62 = α 62 4 β 24 + α 62 5 β 25 = − α 62 4 − α 62 5 , T 6 ; 23 = β 63 + β 64 , T 2 ; 63 = α 63 4 β 24 = − α 63 4 , T 3 ; 62 = α 62 4 β 34 + α 62 5 β 35 = − α 62 4 + α 62 5 , T 6 ; 24 = β 65 , T 2 ; 64 = α 64 5 β 25 + α 64 9 β 29 = − α 64 5 + α 64 9 , T 4 ; 62 = α 62 4 β 44 + α 62 5 β 45 = α 62 4 − α 62 5 , T 6 ; 25 = β 66 , T 2 ; 65 = α 65 9 β 29 = α 65 9 , T 5 ; 62 = α 62 4 β 54 + α 62 5 β 55 = α 62 5 , T 6 ; 33 = − β 62 − β 63 − β 64 − β 68 − β 6 , 10 , T 3 ; 63 = α 63 4 β 34 = − α 63 4 , T 6 ; 34 = β 6 , 10 , T 3 ; 64 = α 64 5 β 35 + α 64 9 β 39 = α 64 5 , T 4 ; 63 = α 63 4 β 44 = α 63 4 , T 6 ; 35 = β 68 , T 3 ; 65 = α 65 9 β 39 = 0 , T 5 ; 63 = α 63 4 β 54 = 0. (175)
The triples ( 6 , 4 , 4 ) (6,4,4) ( 6 , 4 , 4 ) , ( 6 , 4 , 5 ) (6,4,5) ( 6 , 4 , 5 ) and ( 6 , 5 , 5 ) (6,5,5) ( 6 , 5 , 5 ) give no relation: T 6 ; 44 T_{6;44} T 6 ; 44 , T 6 ; 45 T_{6;45} T 6 ; 45 , T 6 ; 55 T_{6;55} T 6 ; 55 are not admissible because D 4 2 D_{4}^{2} D 4 2 , D 4 D 5 D_{4}D_{5} D 4 D 5 , D 5 2 D_{5}^{2} D 5 2 contain C 9 C_{9} C 9 and ( 6 , 9 ) ∉ B (6,9) \notin\mathcal{B} ( 6 , 9 ) ∈ / B , and T 4 ; 64 T_{4;64} T 4 ; 64 , T 4 ; 65 T_{4;65} T 4 ; 65 , T 5 ; 64 T_{5;64} T 5 ; 64 , T 5 ; 65 T_{5;65} T 5 ; 65 are not admissible because E 6 D 4 E_{6}D_{4} E 6 D 4 , E 6 D 5 E_{6}D_{5} E 6 D 5 may contain C 9 C_{9} C 9 and ( 4 , 9 ) (4,9) ( 4 , 9 ) , ( 5 , 9 ) ∉ B (5,9) \notin\mathcal{B} ( 5 , 9 ) ∈ / B . The triples ( 6 , 6 , k ) (6,6,k) ( 6 , 6 , k ) give no relation either, since E 6 ⋅ E 6 E_{6}\cdot E_{6} E 6 ⋅ E 6 is not expanded and T 6 ; 6 k T_{6;6k} T 6 ; 6 k is the only presentation.
We impose that all the presentations of each triple in (175) agree, and solve in the order in which ℓ 6 \ell_{6} ℓ 6 traverses P ~ \widetilde{P} P , as in the propagation argument of § 6.2 .
At the initial vertex, E 6 ⋅ D 1 = C 7 E_{6}\cdot D_{1}=C_{7} E 6 ⋅ D 1 = C 7 . The relations ( 6 , 1 , 2 ) (6,1,2) ( 6 , 1 , 2 ) and ( 6 , 1 , 3 ) (6,1,3) ( 6 , 1 , 3 ) , that is D 1 ⋅ ( E 6 ⋅ D k ) = ( E 6 ⋅ D 1 ) ⋅ D k D_{1}\cdot(E_{6}\cdot D_{k})=(E_{6}\cdot D_{1})\cdot D_{k} D 1 ⋅ ( E 6 ⋅ D k ) = ( E 6 ⋅ D 1 ) ⋅ D k for k = 2 , 3 k=2,3 k = 2 , 3 , fix the coefficients of the curve C 4 C_{4} C 4 of the first crossed edge:
α 62 4 = β 61 = 2 , α 63 4 = β 62 = 1. (176) \alpha_{62}^{4}=\beta_{61}=2,\qquad\alpha_{63}^{4}=\beta_{62}=1. \tag*{(176)} α 62 4 = β 61 = 2 , α 63 4 = β 62 = 1. (176)
The relation ( 6 , 1 , 1 ) (6,1,1) ( 6 , 1 , 1 ) is then satisfied, − 2 − 1 = − 3 -2-1=-3 − 2 − 1 = − 3 .
Crossing e 5 e_{5} e 5 , the relations ( 6 , 2 , 3 ) (6,2,3) ( 6 , 2 , 3 ) and ( 6 , 3 , 4 ) (6,3,4) ( 6 , 3 , 4 ) give
β 64 = − α 63 4 = − 1 , α 62 5 = α 62 4 − α 63 4 = 1 , α 64 5 = β 6 , 10 = α 63 4 = 1. (177) \beta_{64}=-\alpha_{63}^{4}=-1,\qquad\alpha_{62}^{5}=\alpha_{62}^{4}-\alpha_{63}^{4}=1,\qquad\alpha_{64}^{5}=\beta_{6,10}=\alpha_{63}^{4}=1. \tag*{(177)} β 64 = − α 63 4 = − 1 , α 62 5 = α 62 4 − α 63 4 = 1 , α 64 5 = β 6 , 10 = α 63 4 = 1. (177)
At the boundary edge e 9 e_{9} e 9 , the relations ( 6 , 2 , 4 ) (6,2,4) ( 6 , 2 , 4 ) and ( 6 , 2 , 5 ) (6,2,5) ( 6 , 2 , 5 ) give
β 65 = α 62 4 − α 62 5 = 1 , α 64 9 = β 65 + α 64 5 = 2 , α 65 9 = β 66 = α 62 5 = 1. (178) \beta_{65}=\alpha_{62}^{4}-\alpha_{62}^{5}=1,\qquad\alpha_{64}^{9}=\beta_{65}+\alpha_{64}^{5}=2,\qquad\alpha_{65}^{9}=\beta_{66}=\alpha_{62}^{5}=1. \tag*{(178)} β 65 = α 62 4 − α 62 5 = 1 , α 64 9 = β 65 + α 64 5 = 2 , α 65 9 = β 66 = α 62 5 = 1. (178)
The remaining relations are now checks: ( 6 , 2 , 2 ) (6, 2, 2) ( 6 , 2 , 2 ) reads − ( 2 − 1 + 1 + 1 ) = − ( 2 + 1 ) -(2-1+1+1)=-(2+1) − ( 2 − 1 + 1 + 1 ) = − ( 2 + 1 ) , ( 6 , 3 , 3 ) (6, 3, 3) ( 6 , 3 , 3 ) reads − ( 1 − 1 + 1 ) = − 1 -(1-1+1)=-1 − ( 1 − 1 + 1 ) = − 1 , and ( 6 , 1 , 4 ) (6, 1, 4) ( 6 , 1 , 4 ) , ( 6 , 1 , 5 ) (6, 1, 5) ( 6 , 1 , 5 ) , ( 6 , 3 , 5 ) (6, 3, 5) ( 6 , 3 , 5 ) read 0 = 0 = 0 0=0=0 0 = 0 = 0 . The last one, T 6 ; 35 = β 68 = 0 T_{6;35}=\beta_{68}=0 T 6 ; 35 = β 68 = 0 , independently confirms (174) .
The compatibility system therefore has the unique solution
E 6 ⋅ D 1 = C 7 , E 6 ⋅ D 2 = 2 C 4 + C 5 , E 6 ⋅ D 3 = C 4 , E 6 ⋅ D 4 = C 5 + 2 C 9 , E 6 ⋅ D 5 = C 9 , (179) \begin{aligned}
E_{6}\cdot D_{1}&=C_{7}, & E_{6}\cdot D_{2}&=2C_{4}+C_{5}, & E_{6}\cdot D_{3}&=C_{4},\\
E_{6}\cdot D_{4}&=C_{5}+2C_{9}, & E_{6}\cdot D_{5}&=C_{9}, \tag*{(179)}
\end{aligned} E 6 ⋅ D 1 E 6 ⋅ D 4 = C 7 , = C 5 + 2 C 9 , E 6 ⋅ D 2 E 6 ⋅ D 5 = 2 C 4 + C 5 , = C 9 , E 6 ⋅ D 3 = C 4 , (179)
and the completed matrix of divisor–curve intersections is
( β i a ) i ∈ I , a ∈ A = C 1 C 2 C 3 C 4 C 5 C 6 C 7 C 8 C 9 C 10 D 1 − 6 − 3 1 1 0 0 − 3 0 0 0 D 2 4 2 − 2 − 1 − 1 − 7 2 1 1 1 D 3 2 1 0 − 1 1 1 1 × 0 × D 4 0 0 0 1 − 1 1 0 0 × × D 5 0 0 1 0 1 5 0 × × 0 E 6 2 1 0 − 1 1 1 1 0 × 1 (180) \left(\beta_{ia}\right)_{i\in I,\,a\in A}
=
\begin{array}{c|cccccccccc}
& C_{1} & C_{2} & C_{3} & C_{4} & C_{5} & C_{6} & C_{7} & C_{8} & C_{9} & C_{10}\\
D_{1} & -6 & -3 & 1 & 1 & 0 & 0 & -3 & 0 & 0 & 0\\
D_{2} & 4 & 2 & -2 & -1 & -1 & -7 & 2 & 1 & 1 & 1\\
D_{3} & 2 & 1 & 0 & -1 & 1 & 1 & 1 & \times& 0 & \times\\
D_{4} & 0 & 0 & 0 & 1 & -1 & 1 & 0 & 0 & \times& \times\\
D_{5} & 0 & 0 & 1 & 0 & 1 & 5 & 0 & \times& \times& 0\\
\hline
E_{6} & 2 & 1 & 0 & -1 & 1 & 1 & 1 & 0 & \times& 1
\end{array} \tag*{(180)} ( β ia ) i ∈ I , a ∈ A = D 1 D 2 D 3 D 4 D 5 E 6 C 1 − 6 4 2 0 0 2 C 2 − 3 2 1 0 0 1 C 3 1 − 2 0 0 1 0 C 4 1 − 1 − 1 1 0 − 1 C 5 0 − 1 1 − 1 1 1 C 6 0 − 7 1 1 5 1 C 7 − 3 2 1 0 0 1 C 8 0 1 × 0 × 0 C 9 0 1 0 × × × C 10 0 1 × × 0 1 (180)
The solution passes several checks. First, β 64 = − 1 < 0 \beta_{64}=-1<0 β 64 = − 1 < 0 forces C 4 ⊂ E 6 C_{4}\subset E_{6} C 4 ⊂ E 6 , in agreement with the incidence of ℓ 6 \ell_{6} ℓ 6 . Second, β 61 = 2 β 62 \beta_{61}=2\beta_{62} β 61 = 2 β 62 and, more generally, the columns of C 1 C_{1} C 1 , C 2 C_{2} C 2 , C 7 C_{7} C 7 in (180) satisfy β i 1 = 2 β i 2 = 2 β i 7 \beta_{i1}=2\beta_{i2}=2\beta_{i7} β i 1 = 2 β i 2 = 2 β i 7 , as they must since C 1 C_{1} C 1 is a conic and C 2 C_{2} C 2 , C 7 C_{7} C 7 are lines in D 1 ≃ P 2 D_{1}\simeq\mathbb{P}^{2} D 1 ≃ P 2 . Third, the restriction E 6 ∣ D 2 = 2 C 4 + C 5 E_{6}|_{D_{2}}=2C_{4}+C_{5} E 6 ∣ D 2 = 2 C 4 + C 5 reproduces, by intersections in D 2 D_{2} D 2 , the values ( β 61 , β 63 , β 64 , β 65 , β 66 ) = ( 2 , 0 , − 1 , 1 , 1 ) (\beta_{61},\beta_{63},\beta_{64},\beta_{65},\beta_{66})=(2,0,-1,1,1) ( β 61 , β 63 , β 64 , β 65 , β 66 ) = ( 2 , 0 , − 1 , 1 , 1 ) . Finally, ∑ j ∈ I 0 E 6 ⋅ D j = C 7 + 3 C 4 + 2 C 5 + 3 C 9 \sum_{j\in I_{0}}E_{6}\cdot D_{j}=C_{7}+3C_{4}+2C_{5}+3C_{9} ∑ j ∈ I 0 E 6 ⋅ D j = C 7 + 3 C 4 + 2 C 5 + 3 C 9 has vanishing intersection with both compact divisors, − 3 + 3 = 0 -3+3=0 − 3 + 3 = 0 and 2 − 3 − 2 + 3 = 0 2-3-2+3=0 2 − 3 − 2 + 3 = 0 , in agreement with E 6 ⋅ X ~ 0 = 0 E_{6}\cdot\widetilde{X}_{0}=0 E 6 ⋅ X 0 = 0 .
Step 5: triple intersection numbers. For i ∈ I c i\in I_{c} i ∈ I c , (157) uses the compact expansions of D i ⋅ D k D_{i}\cdot D_{k} D i ⋅ D k , which are read off from (168) and (179) . For every j ∈ I j\in I j ∈ I , with D j D_{j} D j standing for E 6 E_{6} E 6 when j = 6 j=6 j = 6 , they give
D 1 D j D 1 = − β j 1 − β j 2 , D 1 D j D 2 = β j 1 , D 1 D j D 3 = β j 2 , D 1 D j D 4 = D 1 D j D 5 = 0 , D 1 D j E 6 = β j 7 , D 2 D j D 1 = β j 1 , D 2 D j D 2 = − β j 1 − β j 3 − β j 4 − β j 5 − β j 6 , D 2 D j D 3 = β j 3 + β j 4 , D 2 D j D 4 = β j 5 , D 2 D j D 5 = β j 6 , D 2 D j E 6 = 2 β j 4 + β j 5 . (181) \begin{aligned}
D_{1}D_{j}D_{1}&=-\beta_{j1}-\beta_{j2}, & D_{1}D_{j}D_{2}&=\beta_{j1}, & D_{1}D_{j}D_{3}&=\beta_{j2},\\
D_{1}D_{j}D_{4}&=D_{1}D_{j}D_{5}=0, & D_{1}D_{j}E_{6}&=\beta_{j7},\\
D_{2}D_{j}D_{1}&=\beta_{j1}, & D_{2}D_{j}D_{2}&=-\beta_{j1}-\beta_{j3}-\beta_{j4}-\beta_{j5}-\beta_{j6}, & D_{2}D_{j}D_{3}&=\beta_{j3}+\beta_{j4},\\
D_{2}D_{j}D_{4}&=\beta_{j5}, & D_{2}D_{j}D_{5}&=\beta_{j6}, & D_{2}D_{j}E_{6}&=2\beta_{j4}+\beta_{j5}. \tag*{(181)}
\end{aligned} D 1 D j D 1 D 1 D j D 4 D 2 D j D 1 D 2 D j D 4 = − β j 1 − β j 2 , = D 1 D j D 5 = 0 , = β j 1 , = β j 5 , D 1 D j D 2 D 1 D j E 6 D 2 D j D 2 D 2 D j D 5 = β j 1 , = β j 7 , = − β j 1 − β j 3 − β j 4 − β j 5 − β j 6 , = β j 6 , D 1 D j D 3 D 2 D j D 3 D 2 D j E 6 = β j 2 , = β j 3 + β j 4 , = 2 β j 4 + β j 5 . (181)
Inserting (180) , we obtain
{#eq:7.20}
These are all the intrinsic triple intersection numbers (138) . The result passes the following checks.
Every column sums to zero over D 1 , … , D 5 D_1,\ldots,D_5 D 1 , … , D 5 , since X ~ 0 ∼ 0 \widetilde{X}_0 \sim0 X 0 ∼ 0 .
D 1 3 = 9 = 12 − Q 1 − r 1 D_1^3=9=12-Q_1-r_1 D 1 3 = 9 = 12 − Q 1 − r 1 with charge Q 1 = 1 Q_1=1 Q 1 = 1 and r 1 = 2 r_1=2 r 1 = 2 boundary curves, as in (132) , and D 1 3 = K P 2 2 D_1^3=K_{\mathbb{P}^2}^2 D 1 3 = K P 2 2 . Similarly, D 2 3 = 7 = 12 − 5 D_2^3=7=12-5 D 2 3 = 7 = 12 − 5 , as in (28) , and D 2 3 = K B l 1 F 6 2 D_2^3=K_{\mathrm{Bl}_1F_6}^2 D 2 3 = K Bl 1 F 6 2 .
The row of E 6 E_6 E 6 in (180) coincides with the row of D 3 D_3 D 3 on all compact curves, so E 6 E_6 E 6 and D 3 D_3 D 3 have the same triple intersections with the compact divisors, as seen in the tables. Writing ≡ \equiv ≡ for equality of intersection numbers with all compact curves, E 6 ≡ D 3 E_6\equiv D_3 E 6 ≡ D 3 . Together with ∑ i ∈ I 0 D i ∼ 0 \sum_{i\in I_0}D_i\sim0 ∑ i ∈ I 0 D i ∼ 0 and the relation D 2 − 2 D 3 − D 4 + 2 D 5 ≡ 0 D_2-2D_3-D_4+2D_5\equiv0 D 2 − 2 D 3 − D 4 + 2 D 5 ≡ 0 , which holds by (180) and corresponds to the M M M -invariant linear function ( x , y ) ↦ y (x,y)\mapsto y ( x , y ) ↦ y on P ~ \widetilde{P} P , the divisor classes modulo ≡ \equiv ≡ span a space of rank 3, corresponding to r = 2 r=2 r = 2 Coulomb parameters and one mass parameter.
S U ( 4 ) k + 1 A S + 7 F SU(4)_k+1AS+7F S U ( 4 ) k + 1 A S + 7 F
We close with a rank three example, the GTP for S U ( 4 ) k + 1 A S + 7 F SU(4)_k+1AS+7F S U ( 4 ) k + 1 A S + 7 F with ∣ k ∣ = 1 2 |k|=\frac{1}{2} ∣ k ∣ = 2 1 , taken from Table 6 of [24 ] , see Figure 7.3 . There are three compact divisors, and an additional 12 non-compact divisors, one of them being exuberant:
Figure 7.3. Left: the GTP P P P for S U ( 4 ) − 1 / 2 + 1 A S + 7 F SU(4)_{-1/2}+1\mathrm{AS}+7\mathrm{F} S U ( 4 ) − 1/2 + 1 AS + 7 F , with the generalized face F F F shaded, with the E E E -line ℓ 15 \ell_{15} ℓ 15 of the exuberant divisor E 15 E_{15} E 15 in red. The two edges labelled C 1 , 4 C_{1,4} C 1 , 4 are glued into a single edge of P ~ \widetilde{P} P . Right: the Symington polygon P ~ \widetilde{P} P with its triangulation P ~ \widetilde{\mathcal{P}} P . The monodromy cut runs along [ v 1 , v 7 ] [v_1,v_7] [ v 1 , v 7 ] . In both panels, the vertex labelled i i i is v i v_i v i , dual to the component D i D_i D i of X ~ 0 \widetilde{X}_0 X 0 , and the edge [ v i , v j ] [v_i,v_j] [ v i , v j ] carries the double curve C i , j C_{i,j} C i , j . The circled number next to vertex i i i on a compact edge [ v i , v j ] [v_i,v_j] [ v i , v j ] is β i a = D i ⋅ C a \beta_{ia}=D_i\cdot C_a β ia = D i ⋅ C a for C a = C i , j C_a=C_{i,j} C a = C i , j .
I c = { 1 , 2 , 3 } , I 0 , ∂ = { 4 , … , 14 } , I e = { 15 } . (182) I_c=\{1,2,3\},\qquad I_{0,\partial}=\{4,\ldots,14\},\qquad I_e=\{15\}. \tag*{(182)} I c = { 1 , 2 , 3 } , I 0 , ∂ = { 4 , … , 14 } , I e = { 15 } . (182)
In this case no two vertices of P ~ \widetilde{P} P are joined by more than one edge (as opposed to the previous example), and we write C i , j C_{i,j} C i , j for the double curve of the edge [ v i , v j ] [v_i,v_j] [ v i , v j ] . There are 17 compact double curves, C 1 , j C_{1,j} C 1 , j (2 ≤ j ≤ 9 2\le j\le9 2 ≤ j ≤ 9 ), C 2 , j C_{2,j} C 2 , j (j = 3 j=3 j = 3 and 9 ≤ j ≤ 13 9\le j\le13 9 ≤ j ≤ 13 ) and C 3 , j C_{3,j} C 3 , j (j = 4 , 13 , 14 j=4,13,14 j = 4 , 13 , 14 ), 11 non-compact double curves along ∂ P ~ \partial\widetilde{P} ∂ P , and the exuberant curve C E = D 1 ⋅ E 15 C_E=D_1\cdot E_{15} C E = D 1 ⋅ E 15 . We now proceed through the algorithm, giving only a few essential steps.
Steps 1–4. D 1 D_1 D 1 is the smoothing of the toric surface X ‾ v 1 \overline{X}_{v_1} X v 1 with the nine rays towards v 4 ′ ′ , v 5 , … , v 9 , v 2 , v 3 , v 4 ′ v_4^{\prime\prime},v_5,\ldots,v_9,v_2,v_3,v_4^{\prime} v 4 ′′ , v 5 , … , v 9 , v 2 , v 3 , v 4 ′ , which merges the curves of the first and last rays into C 1 , 4 C_{1,4} C 1 , 4 . The direction ( 1 , − 1 ) (1,-1) ( 1 , − 1 ) of ℓ 15 \ell_{15} ℓ 15 is the ray towards v 7 v_7 v 7 , so ( p , q ) = ( 1 , 0 ) (p,q)=(1,0) ( p , q ) = ( 1 , 0 ) and (147) gives
D 1 ⋅ C E = − 1 , D 7 ⋅ C E = 1 , E 15 ⋅ C E = − 1 (183) D_1\cdot C_E=-1,\qquad D_7\cdot C_E=1,\qquad E_{15}\cdot C_E=-1 \tag*{(183)} D 1 ⋅ C E = − 1 , D 7 ⋅ C E = 1 , E 15 ⋅ C E = − 1 (183)
C E C_E C E is a ( − 1 ) (-1) ( − 1 ) -curve in D 1 D_1 D 1 meeting ∂ D 1 \partial D_1 ∂ D 1 once, on C 1 , 7 C_{1,7} C 1 , 7 . The compatibility relations, for instance T 15 ; 17 = T 1 ; 15 7 = T 7 ; 15 1 T_{15;17}=T_{1;15}{}^{7}=T_{7;15}{}^{1} T 15 ; 17 = T 1 ; 15 7 = T 7 ; 15 1 , then give2
E 15 ⋅ D 7 = x C 6 , 7 + y C 7 , 8 , x + y = 1 , E 15 ⋅ C 1 , 7 = 1 , (184) E_{15}\cdot D_7=xC_{6,7}+yC_{7,8},\qquad x+y=1,\qquad E_{15}\cdot C_{1,7}=1, \tag*{(184)} E 15 ⋅ D 7 = x C 6 , 7 + y C 7 , 8 , x + y = 1 , E 15 ⋅ C 1 , 7 = 1 , (184)
with E 15 ⋅ D j = 0 E_{15}\cdot D_j=0 E 15 ⋅ D j = 0 for the other j ∈ I 0 j\in I_0 j ∈ I 0 and E 15 ⋅ C a = 0 E_{15}\cdot C_a=0 E 15 ⋅ C a = 0 for all other double curves. Only x + y x+y x + y is fixed, not x x x and y y y individually. In order to apply Theorem 6.1 which guarantees that all intersections are uniquely determined, one must fix a small deformation of the E E E -line, as shown in Figure 7.4 . However, no triple intersection number depends on the individual values of x x x and y y y , so in this particular situation the deformation of the line is not needed if one is just interested in triple intersections.
Figure 7.4. (a) Auxiliary endpoint on [ v 7 , v 8 ] [v_7,v_8] [ v 7 , v 8 ] (b) Auxiliary endpoint on [ v 6 , v 7 ] [v_6,v_7] [ v 6 , v 7 ] . Local enlargement near the edge [ v 1 , v 7 ] [v_1,v_7] [ v 1 , v 7 ] , showing the small deformations of the ℓ 15 \ell_{15} ℓ 15 line needed so that the unicity part of Theorem 6.1 applies. The dash-dotted paths illustrate two possible auxiliary support choices, selecting C 7 , 8 C_{7,8} C 7 , 8 or C 6 , 7 C_{6,7} C 6 , 7 . Hence in the first case ( x , y ) = ( 0 , 1 ) (x,y)=(0,1) ( x , y ) = ( 0 , 1 ) and in the second case ( x , y ) = ( 1 , 0 ) (x,y)=(1,0) ( x , y ) = ( 1 , 0 ) .
Step 5. The non-vanishing triple intersection numbers with at least one compact divisor are
D 1 D 2 D 3 D 4 D 5 D 6 D 7 D 8 D 9 E 15 D 1 3 − 1 0 − 1 0 0 0 0 − 1 − 1 D 2 ∗ − 1 1 0 0 0 0 0 1 0 D 3 ∗ ∗ − 2 1 0 0 0 0 0 0 D 4 ∗ ∗ ∗ − 1 1 0 0 0 0 0 D 5 ∗ ∗ ∗ ∗ − 2 1 0 0 0 0 D 6 ∗ ∗ ∗ ∗ ∗ − 2 1 0 0 0 D 7 ∗ ∗ ∗ ∗ ∗ ∗ − 2 1 0 1 D 8 ∗ ∗ ∗ ∗ ∗ ∗ ∗ − 2 1 0 D 9 ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ − 1 0 E 15 ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ − 1 (185) \begin{array}{c|rrrrrrrrrr}
& D_1 & D_2 & D_3 & D_4 & D_5 & D_6 & D_7 & D_8 & D_9 & E_{15} \\
\hline
D_1 & 3 & -1 & 0 & -1 & 0 & 0 & 0 & 0 & -1 & -1 \\
D_2 & * & -1 & 1 & 0 & 0 & 0 & 0 & 0 & 1 & 0 \\
D_3 & * & * & -2 & 1 & 0 & 0 & 0 & 0 & 0 & 0 \\
D_4 & * & * & * & -1 & 1 & 0 & 0 & 0 & 0 & 0 \\
D_5 & * & * & * & * & -2 & 1 & 0 & 0 & 0 & 0 \\
D_6 & * & * & * & * & * & -2 & 1 & 0 & 0 & 0 \\
D_7 & * & * & * & * & * & * & -2 & 1 & 0 & 1 \\
D_8 & * & * & * & * & * & * & * & -2 & 1 & 0 \\
D_9 & * & * & * & * & * & * & * & * & -1 & 0 \\
E_{15} & * & * & * & * & * & * & * & * & * & -1 \tag*{(185)}
\end{array} D 1 D 2 D 3 D 4 D 5 D 6 D 7 D 8 D 9 E 15 D 1 3 ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ D 2 − 1 − 1 ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ D 3 0 1 − 2 ∗ ∗ ∗ ∗ ∗ ∗ ∗ D 4 − 1 0 1 − 1 ∗ ∗ ∗ ∗ ∗ ∗ D 5 0 0 0 1 − 2 ∗ ∗ ∗ ∗ ∗ D 6 0 0 0 0 1 − 2 ∗ ∗ ∗ ∗ D 7 0 0 0 0 0 1 − 2 ∗ ∗ ∗ D 8 0 0 0 0 0 0 1 − 2 ∗ ∗ D 9 − 1 1 0 0 0 0 0 1 − 1 ∗ E 15 − 1 0 0 0 0 0 1 0 0 − 1 (185)
D 1 D 2 D 3 D 9 D 10 D 11 D 12 D 13 D 1 − 1 − 1 1 1 0 0 0 0 D 2 ∗ 5 − 1 − 1 0 − 1 0 − 1 D 3 ∗ ∗ − 1 0 0 0 0 1 D 9 ∗ ∗ ∗ − 1 1 0 0 0 D 10 ∗ ∗ ∗ ∗ − 2 1 0 0 D 11 ∗ ∗ ∗ ∗ ∗ − 1 1 0 D 12 ∗ ∗ ∗ ∗ ∗ ∗ − 2 1 D 13 ∗ ∗ ∗ ∗ ∗ ∗ ∗ − 1 D 1 D 2 D 3 D 4 D 13 D 14 D 1 0 1 − 2 1 0 0 D 2 ∗ − 1 − 1 0 1 0 D 3 ∗ ∗ 7 − 2 − 1 − 1 D 4 ∗ ∗ ∗ 0 0 1 D 13 ∗ ∗ ∗ ∗ − 1 1 D 14 ∗ ∗ ∗ ∗ ∗ − 1 (186) \begin{aligned}
\begin{array}{c|rrrrrrrr}
& D_1 & D_2 & D_3 & D_9 & D_{10} & D_{11} & D_{12} & D_{13} \\
\hline
D_1 & -1 & -1 & 1 & 1 & 0 & 0 & 0 & 0 \\
D_2 & * & 5 & -1 & -1 & 0 & -1 & 0 & -1 \\
D_3 & * & * & -1 & 0 & 0 & 0 & 0 & 1 \\
D_9 & * & * & * & -1 & 1 & 0 & 0 & 0 \\
D_{10} & * & * & * & * & -2 & 1 & 0 & 0 \\
D_{11} & * & * & * & * & * & -1 & 1 & 0 \\
D_{12} & * & * & * & * & * & * & -2 & 1 \\
D_{13} & * & * & * & * & * & * & * & -1
\end{array}
\qquad
\begin{array}{c|rrrrrr}
& D_1 & D_2 & D_3 & D_4 & D_{13} & D_{14} \\
\hline
D_1 & 0 & 1 & -2 & 1 & 0 & 0 \\
D_2 & * & -1 & -1 & 0 & 1 & 0 \\
D_3 & * & * & 7 & -2 & -1 & -1 \\
D_4 & * & * & * & 0 & 0 & 1 \\
D_{13} & * & * & * & * & -1 & 1 \\
D_{14} & * & * & * & * & * & -1
\end{array} \tag*{(186)}
\end{aligned} D 1 D 2 D 3 D 9 D 10 D 11 D 12 D 13 D 1 − 1 ∗ ∗ ∗ ∗ ∗ ∗ ∗ D 2 − 1 5 ∗ ∗ ∗ ∗ ∗ ∗ D 3 1 − 1 − 1 ∗ ∗ ∗ ∗ ∗ D 9 1 − 1 0 − 1 ∗ ∗ ∗ ∗ D 10 0 0 0 1 − 2 ∗ ∗ ∗ D 11 0 − 1 0 0 1 − 1 ∗ ∗ D 12 0 0 0 0 0 1 − 2 ∗ D 13 0 − 1 1 0 0 0 1 − 1 D 1 D 2 D 3 D 4 D 13 D 14 D 1 0 ∗ ∗ ∗ ∗ ∗ D 2 1 − 1 ∗ ∗ ∗ ∗ D 3 − 2 − 1 7 ∗ ∗ ∗ D 4 1 0 − 2 0 ∗ ∗ D 13 0 1 − 1 0 − 1 ∗ D 14 0 0 − 1 1 1 − 1 (186)
Match with the gauge theory. We now show that the triple intersection numbers above reproduce exactly the one-loop prepotential (6) of S U ( 4 ) − 1 / 2 + 1 A S + 7 F SU(4)_{-1/2}+1AS+7F S U ( 4 ) − 1/2 + 1 A S + 7 F . In particular, we show that one can read the Chern–Simons level k = − 1 2 k=-\frac{1}{2} k = − 2 1 from geometry.3
The three compact surfaces are ruled, with fiber classes
f 1 = C 2 , 9 + C 2 , 10 + C 2 , 11 ⊂ D 2 , f 2 = C 1 , 4 + C 1 , 5 + ⋯ + C 1 , 9 ⊂ D 1 , f 3 = C 3 , 4 ⊂ D 3 , (187) f_{1}=C_{2,9}+C_{2,10}+C_{2,11}\subset D_{2},\qquad f_{2}=C_{1,4}+C_{1,5}+\cdots+C_{1,9}\subset D_{1},\qquad f_{3}=C_{3,4}\subset D_{3}, \tag*{(187)} f 1 = C 2 , 9 + C 2 , 10 + C 2 , 11 ⊂ D 2 , f 2 = C 1 , 4 + C 1 , 5 + ⋯ + C 1 , 9 ⊂ D 1 , f 3 = C 3 , 4 ⊂ D 3 , (187)
and the tables give ( D 2 , D 1 , D 3 ) ⋅ f i = − C A 3 (D_{2},D_{1},D_{3})\cdot f_i=-C_{A_3} ( D 2 , D 1 , D 3 ) ⋅ f i = − C A 3 , minus the Cartan matrix of S U ( 4 ) SU(4) S U ( 4 ) . Hence the fibers are the W-bosons, and ( D 2 , D 1 , D 3 ) (D_{2},D_{1},D_{3}) ( D 2 , D 1 , D 3 ) correspond to the simple roots ( α 1 , α 2 , α 3 ) (\alpha_1,\alpha_2,\alpha_3) ( α 1 , α 2 , α 3 ) . With S = ∑ A t A D A S=\sum_A t^A D_A S = ∑ A t A D A and vol ( C ) = − S ⋅ C \operatorname{vol}(C)=-S\cdot C vol ( C ) = − S ⋅ C , we define the Coulomb parameters by vol ( f i ) = ∑ j ( C A 3 ) i j ψ j \operatorname{vol}(f_i)=\sum_j(C_{A_3})_{ij}\psi_j vol ( f i ) = ∑ j ( C A 3 ) ij ψ j and set ( a 1 , a 2 , a 3 , a 4 ) = ( ψ 1 , ψ 2 − ψ 1 , ψ 3 − ψ 2 , − ψ 3 ) (a_1,a_2,a_3,a_4)=(\psi_1,\psi_2-\psi_1,\psi_3-\psi_2,-\psi_3) ( a 1 , a 2 , a 3 , a 4 ) = ( ψ 1 , ψ 2 − ψ 1 , ψ 3 − ψ 2 , − ψ 3 ) .
The hypermultiplets sit in the degenerate fibers. The volumes of C 2 , 11 C_{2,11} C 2 , 11 , C 2 , 10 + C 2 , 11 C_{2,10}+C_{2,11} C 2 , 10 + C 2 , 11 , C 1 , 9 + ⋯ + C 1 , 9 − j C_{1,9}+\cdots+C_{1,9-j} C 1 , 9 + ⋯ + C 1 , 9 − j (0 ≤ j ≤ 4 0\leq j\leq4 0 ≤ j ≤ 4 ) and C 2 , 3 C_{2,3} C 2 , 3 define the masses m 1 , … , m 7 m_1,\ldots,m_7 m 1 , … , m 7 and m A S m_{\mathrm{AS}} m AS , the gauge-neutral curve C 2 , 12 C_{2,12} C 2 , 12 defines the inverse coupling h 0 h_0 h 0 , and the other entries of
Table 7.27.
are then consequences of the tables. See Figure 7.5 . Thus the flavors f = 1 , 2 f=1,2 f = 1 , 2 live in the fiber ( − 1 , − 2 , − 1 ) (-1,-2,-1) ( − 1 , − 2 , − 1 ) of D 2 D_2 D 2 , the flavors f = 3 , … , 7 f=3,\ldots,7 f = 3 , … , 7 in the fiber ( − 1 , − 2 , − 2 , − 2 , − 2 , − 1 ) (-1,-2,-2,-2,-2,-1) ( − 1 , − 2 , − 2 , − 2 , − 2 , − 1 ) of D 1 D_1 D 1 , and the antisymmetric in the fibers through the common curve C 2 , 3 C_{2,3} C 2 , 3 of D 2 D_2 D 2 and D 3 D_3 D 3 . Positivity of the volumes selects the chamber
Figure 7.5. The 5-brane web dual to the GTP of Figure 7.3 . It is labelled by vol ( C i , j ) \operatorname{vol}(C_{i,j}) vol ( C i , j ) , taken from (7.27), with m ˉ = 1 2 ∑ f m f \bar{m} = \frac{1}{2}\sum_f m_f m ˉ = 2 1 ∑ f m f . The [1 , 1 ] 7-brane lies in face 1, and its monodromy cut (dashed) points in the direction ( − 1 , − 1 ) (-1,-1) ( − 1 , − 1 ) of the three 5-branes that ended on it before it was pulled in. The red path from the 7-brane is dual to the E E E -line ℓ 15 \ell_{15} ℓ 15 . Its first segment, the ( 1 , 1 ) (1,1) ( 1 , 1 ) string ending on C 1 , 7 C_{1,7} C 1 , 7 , is the exuberant curve C E C_E C E , with vol ( C E ) = h 0 + a 1 + a 2 + m 1 + m 2 + m 3 + m 4 − m ˉ \operatorname{vol}(C_E) = h_0 + a_1 + a_2 + m_1 + m_2 + m_3 + m_4 - \bar{m} vol ( C E ) = h 0 + a 1 + a 2 + m 1 + m 2 + m 3 + m 4 − m ˉ . The path then continues into face 7, dual to the endpoint v 7 v_7 v 7 of ℓ 15 \ell_{15} ℓ 15 .
a 1 + m 1 , 2 > 0 > a 2 + m 1 , 2 , a 2 + m 3 , … , 7 > 0 > a 3 + m 3 , … , 7 , s i j ( a i + a j + m A S ) > 0 , (188) a_1+m_{1,2}>0>a_2+m_{1,2},\qquad a_2+m_{3,\ldots,7}>0>a_3+m_{3,\ldots,7},\qquad s_{ij}(a_i+a_j+m_{\mathrm{AS}})>0, \tag*{(188)} a 1 + m 1 , 2 > 0 > a 2 + m 1 , 2 , a 2 + m 3 , … , 7 > 0 > a 3 + m 3 , … , 7 , s ij ( a i + a j + m AS ) > 0 , (188)
with s 12 = s 13 = 1 s_{12}=s_{13}=1 s 12 = s 13 = 1 and s i j = − 1 s_{ij}=-1 s ij = − 1 otherwise. In this chamber, (6) reads
F I M S = h 0 2 ∑ i a i 2 + k 6 ∑ i a i 3 + 1 6 ∑ i < j ( a i − a j ) 3 − 1 12 ∑ f = 1 2 ( ( a 1 + m f ) 3 − ∑ i = 2 4 ( a i + m f ) 3 ) \mathcal{F}_{\mathrm{IMS}}=\frac{h_0}{2}\sum_i a_i^2+\frac{k}{6}\sum_i a_i^3+\frac{1}{6}\sum_{i<j}(a_i-a_j)^3-\frac{1}{12}\sum_{f=1}^{2}\left((a_1+m_f)^3-\sum_{i=2}^{4}(a_i+m_f)^3\right) F IMS = 2 h 0 i ∑ a i 2 + 6 k i ∑ a i 3 + 6 1 i < j ∑ ( a i − a j ) 3 − 12 1 f = 1 ∑ 2 ( ( a 1 + m f ) 3 − i = 2 ∑ 4 ( a i + m f ) 3 )
− 1 12 ∑ f = 3 7 ( ∑ i = 1 2 ( a i + m f ) 3 − ∑ i = 3 4 ( a i + m f ) 3 ) − 1 12 ∑ i < j s i j ( a i + a j + m A S ) 3 . (189) -\frac{1}{12}\sum_{f=3}^{7}\left(\sum_{i=1}^{2}(a_i+m_f)^3-\sum_{i=3}^{4}(a_i+m_f)^3\right)-\frac{1}{12}\sum_{i<j}s_{ij}(a_i+a_j+m_{\mathrm{AS}})^3. \tag*{(189)} − 12 1 f = 3 ∑ 7 ( i = 1 ∑ 2 ( a i + m f ) 3 − i = 3 ∑ 4 ( a i + m f ) 3 ) − 12 1 i < j ∑ s ij ( a i + a j + m AS ) 3 . (189)
On the geometric side, the ψ 3 \psi^3 ψ 3 terms of F = 1 6 S 3 \mathcal{F} = \frac{1}{6}S^3 F = 6 1 S 3 come from the entries of (185) –(186) with i , j , k ∈ I c i,j,k \in I_c i , j , k ∈ I c :
6 F ∣ ψ 3 = 5 ψ 1 3 + 3 ψ 2 3 + 7 ψ 3 3 − 3 ψ 1 2 ψ 2 − 3 ψ 1 ψ 2 2 − 3 ψ 1 2 ψ 3 − 3 ψ 1 ψ 3 2 − 6 ψ 2 ψ 3 2 + 6 ψ 1 ψ 2 ψ 3 . (190) 6\mathcal{F}\big|_{\psi^3}=5\psi_1^3+3\psi_2^3+7\psi_3^3-3\psi_1^2\psi_2-3\psi_1\psi_2^2-3\psi_1^2\psi_3-3\psi_1\psi_3^2-6\psi_2\psi_3^2+6\psi_1\psi_2\psi_3. \tag*{(190)} 6 F ψ 3 = 5 ψ 1 3 + 3 ψ 2 3 + 7 ψ 3 3 − 3 ψ 1 2 ψ 2 − 3 ψ 1 ψ 2 2 − 3 ψ 1 2 ψ 3 − 3 ψ 1 ψ 3 2 − 6 ψ 2 ψ 3 2 + 6 ψ 1 ψ 2 ψ 3 . (190)
This coincides with the ψ 3 \psi^3 ψ 3 part of (189) if and only if k = − 1 2 k=-\frac{1}{2} k = − 2 1 , so the geometry determines the Chern–Simons level. The agreement is complete. Expressing all t A t^A t A through ( ψ i , m f , m A S , h 0 ) (\psi_i,m_f,m_{\mathrm{AS}},h_0) ( ψ i , m f , m AS , h 0 ) via (7.27), 1 6 S 3 \frac{1}{6}S^3 6 1 S 3 coincides with (189) in every term of order ψ 3 \psi^3 ψ 3 , ψ 2 m \psi^2m ψ 2 m and ψ m 2 \psi m^2 ψ m 2 , where m m m is any of the nine mass parameters. It differs only by ψ \psi ψ -independent terms, which involve non-compact divisors only and are not intrinsic. Every intrinsic number in the tables, including those of E 15 E_{15} E 15 , enters this comparison.
Acknowledgements
We thank Cyril Closset, Andres Collinucci, Quentin Lamouret, and Marcus Sperling for discussions at various stages of this long journey. We thank the organizers of String Math at ICTP, where the start of this collaboration across the physics-math divide was enabled. PB acknowledges the support of a Sloan Research Fellowship from the Alfred P. Sloan Foundation. The work of SSN is supported by the EPSRC Open Fellowship EP/X01276X/1 (Schafer-Nameki). For the purpose of open access, the authors have applied a CC BY public copyright licence to any author accepted manuscript arising from this submission.
Anthropic’s Claude Opus 5.5 and OpenAI ChatGPT-6.1 Sol and Astra were used in the final stages of the project to improve the writing, draw figures, perform the gauge-parameter match in § 7.2 and polish the Mathematica code. The authors take full responsibility for the content of the paper.
Description of the Mathematica Code
For the reader’s convenience, we provide an implementation of the algorithm of § 6 in a Mathematica notebook, available at
https://github.com/AntoineBrgt/GTP-Intersections
The code follows closely the steps outlined there. The input can be given in two ways.
The first one uses text: one gives the list of edges of the GTP, followed by a list of pulled-in 7-branes. Each 7-brane is specified by its charge and the cut that gives rise to the Symington polygon. For instance, if one wants to check the computations of the example of § 7.1 , one can write
gtpEx1 = {{{0, 0}, {0, 1}}, {{0, 0}, {1, -2}}, {{0, 1}, {-1, -2}},
{{0, 1}, {1, -2}}, {{0, 1}, {1, -1}}, {{0, 1}, {0, 2}},
{{0, 0}, {-1, -2}}, {{-1, -2}, {0, 2}}, {{1, -1}, {0, 2}},
Figure A.1. User-friendly input method for GTPs. We illustrate with the example of § 7.1 .
{{1, -2}, {1, -1}}};
braneEx1 = {{{0, 1}, {{0, 0}, {1, -2}}}};
gtpRun[gtpEx1, braneEx1]
Here, gtpEx1 is the list of edges of the GTP, written using coordinates for the end points of each edge. And braneEx1 is the list of 7-branes. Here there is a single 7-brane, so the list contains a single element, which is a pair: the first entry {0, 1} is the charge of the brane, the second one {{0, 0}, {1, -2}} is the path followed by (one side of) the cut which defines the Symington polygon.
In either case, the output gives the intersection tables, as shown in Figure A.2.
Figure A.2. Output of the code on the example of § 7.1 .