Shadow of a Cubic Surface

Total Page:16

File Type:pdf, Size:1020Kb

Shadow of a Cubic Surface Faculteit B`etawetenschappen Shadow of a cubic surface Bachelor Thesis Rein Janssen Groesbeek Wiskunde en Natuurkunde Supervisors: Dr. Martijn Kool Departement Wiskunde Dr. Thomas Grimm ITF June 2020 Abstract 3 For a smooth cubic surface S in P we can cast a shadow from a point P 2 S that does not lie on one of the 27 lines of S onto a hyperplane H. The closure of this shadow is a smooth quartic curve. Conversely, from every smooth quartic curve we can reconstruct a smooth cubic surface whose closure of the shadow is this quartic curve. We will also present an algorithm to reconstruct the cubic surface from the bitangents of a quartic curve. The 27 lines of S together with the tangent space TP S at P are in correspondence with the 28 bitangents or hyperflexes of the smooth quartic shadow curve. Then a short discussion on F-theory is given to relate this geometry to physics. Acknowledgements I would like to thank Martijn Kool for suggesting the topic of the shadow of a cubic surface to me and for the discussions on this topic. Also I would like to thank Thomas Grimm for the suggestions on the applications in physics of these cubic surfaces. Finally I would like to thank the developers of Singular, Sagemath and PovRay for making their software available for free. i Contents 1 Introduction 1 2 The shadow of a smooth cubic surface 1 2.1 Projection of the first polar . .1 2.2 Reconstructing a cubic from the shadow . .5 3 The 27 lines and the 28 bitangents 9 3.1 Theorem of the apparent boundary . .9 3.2 The last bitangent . 11 3.3 Injectivity . 13 4 Recovering the cubic from a quartic with computer algebra 14 4.1 Calculating the bitangents . 15 4.2 Calculating the points of contact . 16 4.3 Finding a syzygetic quadruple of bitangents . 17 4.4 Finding a fourth bitangent . 18 4.5 Recovering the cubic surface . 18 5 Algebraic geometry in physics: F-theory 19 5.1 D7-branes . 19 5.2 Elliptic fibration . 19 5.3 Intersecting D7-branes . 20 5.4 Born infeld action . 20 6 Appendix: lemmas and notation 23 6.1 Notation . 23 6.2 Lemmas . 23 1 Introduction Two classical results in algebraic geometry are that a smooth cubic surface in P3 contains 27 lines and a smooth quartic curve in P2 contains 28 contact lines. Figure 2: A smooth quartic curve with 28 Figure 1: A smooth cubic surface containing contact lines. 27 lines. In the following paper we will show how by taking the shadow of the smooth cubic surface, also known as the projection from a point, one can obtain a smooth quartic curve and relate the 27 lines to the 28 contact lines. The paper is structured as follows: • In the second chapter we will show how the (closure of the) shadow of a smooth cubic surface is a smooth quartic curve. To show smoothness, we use ideas from section 5.8.12 of [Bel09]. Then for a smooth quartic curve we will give a construction of a smooth cubic surface that has this curve as its shadow. The idea of this construction is from exercise 5.7.11 of [Bel09] and the worked out details in coordinates is new. • In the third chapter we will show how the 27 lines get mapped to the 28 contact lines. Here we use ideas from section 5.8.9 of [Bel09]. • In the fourth chapter we will describe an algorithm that reconstructs a smooth cubic surface from a smooth quartic curve. This is largely based on the paper and code samples from [PSV11]. • In the fifth chapter we will give a short sketch of how algebraic surfaces are used in F-theory, this is unrelated to the previous chapters so can be read separately. The main reference for this chapter is [Wei18]. 2 The shadow of a smooth cubic surface 2.1 Projection of the first polar We will work in 3 = 3 , projective 3-dimensional space over , but the results also hold over P PC C any algebraically closed field of characteristic zero. 1 For a smooth cubic surface S in P3 we can create its shadow by projecting it from a point P 2 S. Because S is an algebraic variety, we expect that the shadow also is algebraic. In this section we will show that if we choose P to not lie on a line of S, then the shadow (or the closure thereof) is a smooth quartic curve. After that we will also show that any smooth quartic can be considered the projection of some smooth cubic surface. Definition 1. For S a smooth hypersurface in P3 we define the shadow Sh(S; P; H) from a point P 2 S onto a hyperplane H not containing P , to be the locus of points PQ \ H where PQ ⊂ TQS is a line tangent to S at some Q 2 S n fP g. ♦ This definition is close to our intuition of a shadow, but it is cumbersome to work with. We can reformulate it using the notion of a first polar. n Definition 2. Let X be a smooth hypersurface in P given by the polynomial f 2 k[x0; : : : ; xn] of degree r. Let P be a point in X. The first polar X1(P ) of X at P is defined as the hypersurface given by the degree r − 1 polynomial n X @f ∆1 f := P : (1) P i @x i=0 i So this is similar to the equation for the tangent space, except that we now leave the argument of the derivatives as variables, and we fix the point in the tangent space. This can be generalized to higher polars, for 0 ≤ k ≤ r we define the k-th polar Xk(P ) as the hypersurface given by the degree r − k polynomial k k X @ f ∆P f := Pi1 ····· Pik : (2) @xi1 : : : @xik i1;:::;ik=0;:::;n ♦ Now we will give an equivalent definition of a shadow using the first polar S1(P ). Note that S was chosen to be smooth, so this construction indeed works. Let π : P3 n fP g ! H be the projection that maps Q to the intersection PQ \ H. Let S1(P ) be the first polar of S at P . If S is given as the zero set of the cubic polynomial f, 1 P3 @f then S1(P ) is the quadric hypersurface with equation ∆ f := Pi . P i=0 @xi If a point Q 6= P lies in S and in S1(P ), then P lies in the tangent space at Q, which means the line PQ is tangent at Q. For every point R 2 Sh(S; P; H) ⊂ H there exists a point Q 2 S \S1(P ) with Q 6= P such that π(Q) = PQ \ H = R, which means we have Sh(S; P; H) ⊂ π(S \ S1(P ) n fP g). Conversely, for Q 2 S \ S1(P ) n fP g, the line PQ intersects H in π(Q), and this intersection lies in Sh(S; P; H), so π(S \ S1(P ) n fP g) ⊂ Sh(S; P; H). This means we get an alternative definition of Sh(S; P; H) = π(S \ S1(P ) n fP g) as the projection of the intersection of S with its first polar at P . Note that the previous discussion applies to any smooth hypersurface in Pn and not just for smooth cubic surfaces in P3. We have that S \ S1(P ) n fP g is not closed in S because it is missing the point P . Similarly Sh(S; P; H) is not closed in H because we are missing the limiting line obtained by taking the limit of the lines PQ as Q approaches P , thus the tangent lines of S at P . Thus if we take the closure of Sh(S; P; H) in H we add these missing points. In the following theorem we will prove that π induces an isomorphism of varieties between S \ S1(P ) n fP g and Sh(S; P; H) and that the closure Sh(S; P; H) in H is a smooth plane quartic. 2 Theorem 1. Let S be a non-singular cubic hypersurface in P3. Let P 2 S a point lying outside the 27 lines of S, and H a hyperplane not containing P . Then Sh(S; P; H), the closure of its shadow in H, is a smooth quartic curve. Proof. We first choose a change of coordinates g 2 GL(4; k) such that g([0 : 0 : 0 : 1]) = P and g(Z(x3)) = H, for the details on the construction of g see Lemma 6. From now on we will use these new coordinates. This means that S is given as the zero set of the homogeneous polynomial 2 f = x3f1 + x3f2 + f3 (3) with f1; f2; f3 homogeneous polynomials of degree 1,2,3. Also we can identify points in H with 2 points in P by the identification [x0 : x1 : x2 : 0] $ [x0 : x1 : x2], which means we can consider the shadow Sh(S; P; H) to be a subset of P2. 3 2 Let π : P n fP g ! P be the projection from a point defined as π([x0 : x1 : x2 : x3]) = [x0 : x1 : x2]. Let C := S \ S1(P ) be the intersection of S with S1(P ), its first polar at P . The first polar at P has equation 3 X @f @f ∆1 f = P = 1 · = 2x f + f : (4) P i @x @x 3 1 2 i=0 i 3 2 This means C = Z(x3f1 + x3f2 + f3; 2x3f1 + f2).
Recommended publications
  • Cubic Surfaces and Their Invariants: Some Memories of Raymond Stora
    Available online at www.sciencedirect.com ScienceDirect Nuclear Physics B 912 (2016) 374–425 www.elsevier.com/locate/nuclphysb Cubic surfaces and their invariants: Some memories of Raymond Stora Michel Bauer Service de Physique Theorique, Bat. 774, Gif-sur-Yvette Cedex, France Received 27 May 2016; accepted 28 May 2016 Available online 7 June 2016 Editor: Hubert Saleur Abstract Cubic surfaces embedded in complex projective 3-space are a classical illustration of the use of old and new methods in algebraic geometry. Recently, they made their appearance in physics, and in particular aroused the interest of Raymond Stora, to the memory of whom these notes are dedicated, and to whom I’m very much indebted. Each smooth cubic surface has a rich geometric structure, which I review briefly, with emphasis on the 27 lines and the combinatorics of their intersections. Only elementary methods are used, relying on first order perturbation/deformation theory. I then turn to the study of the family of cubic surfaces. They depend on 20 parameters, and the action of the 15 parameter group SL4(C) splits the family in orbits depending on 5 parameters. This takes us into the realm of (geometric) invariant theory. Ireview briefly the classical theorems on the structure of the ring of polynomial invariants and illustrate its many facets by looking at a simple example, before turning to the already involved case of cubic surfaces. The invariant ring was described in the 19th century. Ishow how to retrieve this description via counting/generating functions and character formulae. © 2016 The Author. Published by Elsevier B.V.
    [Show full text]
  • Classifying Smooth Cubic Surfaces up to Projective Linear Transformation
    Classifying Smooth Cubic Surfaces up to Projective Linear Transformation Noah Giansiracusa June 2006 Introduction We would like to study the space of smooth cubic surfaces in P3 when each surface is considered only up to projective linear transformation. Brundu and Logar ([1], [2]) de¯ne an action of the automorphism group of the 27 lines of a smooth cubic on a certain space of cubic surfaces parametrized by P4 in such a way that the orbits of this action correspond bijectively to the orbits of the projective linear group PGL4 acting on the space of all smooth cubic surfaces in the natural way. They prove several other results in their papers, but in this paper (the author's senior thesis at the University of Washington) we focus exclusively on presenting a reasonably self-contained and coherent exposition of this particular result. In doing so, we chose to slightly modify the action and ensuing proof, more aesthetically than substantially, in order to better reveal the intricate relation between combinatorics and geometry that underlies this problem. We would like to thank Professors Chuck Doran and Jim Morrow for much guidance and support. The Space of Cubic Surfaces Before proceeding, we need to de¯ne terms such as \the space of smooth cubic surfaces". Let W be a 4-dimensional vector-space over an algebraically closed ¯eld k of characteristic zero whose projectivization P(W ) = P3 is the ambient space in which the cubic surfaces we consider live. Choose a basis (x; y; z; w) for the dual vector-space W ¤. Then an arbitrary cubic surface is given by the zero locus V (F ) of an element F 2 S3W ¤ ½ k[x; y; z; w], where SnV denotes the nth symmetric power of a vector space V | which in this case simply means the set of degree three homogeneous polynomials.
    [Show full text]
  • [Math.AG] 1 Nov 2005 Where Uoopimgop Hscntuto Ensahlmrhcperio Holomorphic a Defines Construction This Group
    A COMPACTIFICATION OF M3 VIA K3 SURFACES MICHELA ARTEBANI Abstract. S. Kond¯odefined a birational period map from the moduli space of genus three curves to a moduli space of degree four polarized K3 surfaces. In this paper we extend the period map to a surjective morphism on a suitable compactification of M3 and describe its geometry. Introduction 14 Let V =| OP2 (4) |=∼ P be the space of plane quartics and V0 be the open subvariety of smooth curves. The degree four cyclic cover of the plane branched along a curve C ∈ V0 is a K3 surface equipped with an order four non-symplectic automorphism group. This construction defines a holomorphic period map: P0 : V0 −→ M, where V0 is the geometric quotient of V0 by the action of P GL(3) and M is a moduli space of polarized K3 surfaces. In [15] S. Kond¯oshows that P0 gives an isomorphism between V0 and the com- plement of two irreducible divisors Dn, Dh in M. Moreover, he proves that the generic points in Dn and Dh correspond to plane quartics with a node and to smooth hyperelliptic genus three curves respectively. The moduli space M is an arithmetic quotient of a six dimensional complex ball, hence a natural compactification is given by the Baily-Borel compactification M∗ (see [1]). On the other hand, geometric invariant theory provides a compact pro- jective variety V containing V0 as a dense subset, given by the categorical quotient of the semistable locus in V for the natural action of P GL(3). In this paper we prove that the map P0 can be extended to a holomorphic surjective map P : V −→M∗ on the blowing-up V of V in the point v0 corresponding to the orbit of double arXiv:math/0511031v1 [math.AG] 1 Nov 2005 e conics.
    [Show full text]
  • Arxiv:1712.01167V2 [Math.AG] 12 Oct 2018 12
    AUTOMORPHISMS OF CUBIC SURFACES IN POSITIVE CHARACTERISTIC IGOR DOLGACHEV AND ALEXANDER DUNCAN Abstract. We classify all possible automorphism groups of smooth cu- bic surfaces over an algebraically closed field of arbitrary characteristic. As an intermediate step we also classify automorphism groups of quar- tic del Pezzo surfaces. We show that the moduli space of smooth cubic surfaces is rational in every characteristic, determine the dimensions of the strata admitting each possible isomorphism class of automor- phism group, and find explicit normal forms in each case. Finally, we completely characterize when a smooth cubic surface in positive char- acteristic, together with a group action, can be lifted to characteristic zero. Contents 1. Introduction2 Acknowledgements8 2. Preliminaries8 3. Del Pezzo surfaces of degree 4 12 4. Differential structure in special characteristics 19 5. The Fermat cubic surface 24 6. General forms 28 7. Rationality of the moduli space 33 8. Conjugacy classes of automorphisms 36 9. Involutions 38 10. Automorphisms of order 3 46 11. Automorphisms of order 4 59 arXiv:1712.01167v2 [math.AG] 12 Oct 2018 12. Automorphisms of higher order 64 13. Collections of Eckardt points 69 14. Proof of the Main Theorem 72 15. Lifting to characteristic zero 73 Appendix 78 References 82 The second author was partially supported by National Security Agency grant H98230- 16-1-0309. 1 2 IGOR DOLGACHEV AND ALEXANDER DUNCAN 1. Introduction 3 Let X be a smooth cubic surface in P defined over an algebraically closed field | of arbitrary characteristic p. The primary purpose of this paper is to classify the possible automorphism groups of X.
    [Show full text]
  • UNIVERSAL TORSORS and COX RINGS Brendan Hassett and Yuri
    UNIVERSAL TORSORS AND COX RINGS by Brendan Hassett and Yuri Tschinkel Abstract. — We study the equations of universal torsors on rational surfaces. Contents Introduction . 1 1. Generalities on the Cox ring . 4 2. Generalities on toric varieties . 7 3. The E6 cubic surface . 12 4. D4 cubic surface . 23 References . 26 Introduction The study of surfaces over nonclosed fields k leads naturally to certain auxiliary varieties, called universal torsors. The proofs of the Hasse principle and weak approximation for certain Del Pezzo surfaces required a very detailed knowledge of the projective geometry, in fact, explicit equations, for these torsors [7], [9], [8], [22], [23], [24]. More recently, Salberger proposed using universal torsors to count rational points of The first author was partially supported by the Sloan Foundation and by NSF Grants 0196187 and 0134259. The second author was partially supported by NSF Grant 0100277. 2 BRENDAN HASSETT and YURI TSCHINKEL bounded height, obtaining the first sharp upper bounds on split Del Pezzo surfaces of degree 5 and asymptotics on split toric varieties over Q [21]. This approach was further developed in the work of Peyre, de la Bret`eche, and Heath-Brown [19], [20], [3], [14]. Colliot-Th´el`eneand Sansuc have given a general formalism for writing down equations for these torsors. We briefly sketch their method: Let X be a smooth projective variety and {Dj}j∈J a finite set of irreducible divisors on X such that U := X \ ∪j∈J Dj has trivial Picard group. In practice, one usually chooses generators of the effective cone of X, e.g., the lines on the Del Pezzo surface.
    [Show full text]
  • Combination of Cubic and Quartic Plane Curve
    IOSR Journal of Mathematics (IOSR-JM) e-ISSN: 2278-5728,p-ISSN: 2319-765X, Volume 6, Issue 2 (Mar. - Apr. 2013), PP 43-53 www.iosrjournals.org Combination of Cubic and Quartic Plane Curve C.Dayanithi Research Scholar, Cmj University, Megalaya Abstract The set of complex eigenvalues of unistochastic matrices of order three forms a deltoid. A cross-section of the set of unistochastic matrices of order three forms a deltoid. The set of possible traces of unitary matrices belonging to the group SU(3) forms a deltoid. The intersection of two deltoids parametrizes a family of Complex Hadamard matrices of order six. The set of all Simson lines of given triangle, form an envelope in the shape of a deltoid. This is known as the Steiner deltoid or Steiner's hypocycloid after Jakob Steiner who described the shape and symmetry of the curve in 1856. The envelope of the area bisectors of a triangle is a deltoid (in the broader sense defined above) with vertices at the midpoints of the medians. The sides of the deltoid are arcs of hyperbolas that are asymptotic to the triangle's sides. I. Introduction Various combinations of coefficients in the above equation give rise to various important families of curves as listed below. 1. Bicorn curve 2. Klein quartic 3. Bullet-nose curve 4. Lemniscate of Bernoulli 5. Cartesian oval 6. Lemniscate of Gerono 7. Cassini oval 8. Lüroth quartic 9. Deltoid curve 10. Spiric section 11. Hippopede 12. Toric section 13. Kampyle of Eudoxus 14. Trott curve II. Bicorn curve In geometry, the bicorn, also known as a cocked hat curve due to its resemblance to a bicorne, is a rational quartic curve defined by the equation It has two cusps and is symmetric about the y-axis.
    [Show full text]
  • A Mathematical Background to Cubic and Quartic Schilling Models
    Utrecht University A Mathematical Background to Cubic and Quartic Schilling Models Author: I.F.M.M.Nelen A thesis presented for the degree Master of Science Supervisor: Prof. Dr. F. Beukers Second Reader: Prof. Dr. C.F. Faber Masters Programme: Mathematical Sciences Department: Departement of Mathematics University: Utrecht University Abstract At the end of the twentieth century plaster models of algebraic surface were constructed by the company of Schilling. Many universities have some series of these models but a rigorous mathematical background to the theory is most often not given. In this thesis a mathematical background is given for the cubic surfaces and quartic ruled surfaces on which two series of Schilling models are based, series VII and XIII. The background consists of the classification of all complex cubic surface through the number and type of singularities lying on the surface. The real cubic sur- faces are classified by which of the singularities are real and the number and configuration of the lines lying on the cubic surface. The ruled cubic and quartic surfaces all have a singular curve lying on them and they are classified by the degree of this curve. Acknowledgements Multiple people have made a contribution to this thesis and I want to extend my graditute here. First of all I want to thank Prof. Dr. Frits Beukers for being my super- visor, giving me an interesting subject to write about and helping me get the answers needed to finish this thesis. By his comments he gave me a broader un- derstanding of the topic and all the information I needed to complete this thesis.
    [Show full text]
  • Moduli Spaces and Invariant Theory
    MODULI SPACES AND INVARIANT THEORY JENIA TEVELEV CONTENTS §1. Syllabus 3 §1.1. Prerequisites 3 §1.2. Course description 3 §1.3. Course grading and expectations 4 §1.4. Tentative topics 4 §1.5. Textbooks 4 References 4 §2. Geometry of lines 5 §2.1. Grassmannian as a complex manifold. 5 §2.2. Moduli space or a parameter space? 7 §2.3. Stiefel coordinates. 8 §2.4. Complete system of (semi-)invariants. 8 §2.5. Plücker coordinates. 9 §2.6. First Fundamental Theorem 10 §2.7. Equations of the Grassmannian 11 §2.8. Homogeneous ideal 13 §2.9. Hilbert polynomial 15 §2.10. Enumerative geometry 17 §2.11. Transversality. 19 §2.12. Homework 1 21 §3. Fine moduli spaces 23 §3.1. Categories 23 §3.2. Functors 25 §3.3. Equivalence of Categories 26 §3.4. Representable Functors 28 §3.5. Natural Transformations 28 §3.6. Yoneda’s Lemma 29 §3.7. Grassmannian as a fine moduli space 31 §4. Algebraic curves and Riemann surfaces 37 §4.1. Elliptic and Abelian integrals 37 §4.2. Finitely generated fields of transcendence degree 1 38 §4.3. Analytic approach 40 §4.4. Genus and meromorphic forms 41 §4.5. Divisors and linear equivalence 42 §4.6. Branched covers and Riemann–Hurwitz formula 43 §4.7. Riemann–Roch formula 45 §4.8. Linear systems 45 §5. Moduli of elliptic curves 47 1 2 JENIA TEVELEV §5.1. Curves of genus 1. 47 §5.2. J-invariant 50 §5.3. Monstrous Moonshine 52 §5.4. Families of elliptic curves 53 §5.5. The j-line is a coarse moduli space 54 §5.6.
    [Show full text]
  • Polynomial Curves and Surfaces
    Polynomial Curves and Surfaces Chandrajit Bajaj and Andrew Gillette September 8, 2010 Contents 1 What is an Algebraic Curve or Surface? 2 1.1 Algebraic Curves . .3 1.2 Algebraic Surfaces . .3 2 Singularities and Extreme Points 4 2.1 Singularities and Genus . .4 2.2 Parameterizing with a Pencil of Lines . .6 2.3 Parameterizing with a Pencil of Curves . .7 2.4 Algebraic Space Curves . .8 2.5 Faithful Parameterizations . .9 3 Triangulation and Display 10 4 Polynomial and Power Basis 10 5 Power Series and Puiseux Expansions 11 5.1 Weierstrass Factorization . 11 5.2 Hensel Lifting . 11 6 Derivatives, Tangents, Curvatures 12 6.1 Curvature Computations . 12 6.1.1 Curvature Formulas . 12 6.1.2 Derivation . 13 7 Converting Between Implicit and Parametric Forms 20 7.1 Parameterization of Curves . 21 7.1.1 Parameterizing with lines . 24 7.1.2 Parameterizing with Higher Degree Curves . 26 7.1.3 Parameterization of conic, cubic plane curves . 30 7.2 Parameterization of Algebraic Space Curves . 30 7.3 Automatic Parametrization of Degree 2 Curves and Surfaces . 33 7.3.1 Conics . 34 7.3.2 Rational Fields . 36 7.4 Automatic Parametrization of Degree 3 Curves and Surfaces . 37 7.4.1 Cubics . 38 7.4.2 Cubicoids . 40 7.5 Parameterizations of Real Cubic Surfaces . 42 7.5.1 Real and Rational Points on Cubic Surfaces . 44 7.5.2 Algebraic Reduction . 45 1 7.5.3 Parameterizations without Real Skew Lines . 49 7.5.4 Classification and Straight Lines from Parametric Equations . 52 7.5.5 Parameterization of general algebraic plane curves by A-splines .
    [Show full text]
  • Universal Unramified Cohomology of Cubic Fourfolds Containing a Plane
    UNIVERSAL UNRAMIFIED COHOMOLOGY OF CUBIC FOURFOLDS CONTAINING A PLANE ASHER AUEL, JEAN-LOUIS COLLIOT-THEL´ ENE,` R. PARIMALA Abstract. We prove the universal triviality of the third unramified cohomology group of a very general complex cubic fourfold containing a plane. The proof uses results on the unramified cohomology of quadrics due to Kahn, Rost, and Sujatha. Introduction 5 Let X be a smooth cubic fourfold, i.e., a smooth cubic hypersurface in P .A well-known problem in algebraic geometry concerns the rationality of X over C. Expectation. The very general cubic fourfold over C is irrational. Here, \very general" is usually taken to mean \in the complement of a countable union of Zariski closed subsets" in the moduli space of cubic fourfolds. At present, however, not a single cubic fourfold is provably irrational, though many families of rational cubic fourfolds are known. 5 If X contains a plane P (i.e., a linear two dimensional subvariety of P ), then X 2 is birational to the total space of a quadric surface bundle Xe ! P by projecting 2 from P . Its discriminant divisor D ⊂ P is a sextic curve. The rationality of a cubic fourfold containing a plane over C is also a well-known problem. Expectation. The very general cubic fourfold containing a plane over C is irra- tional. Assuming that the discriminant divisor D is smooth, the discriminant double 2 cover S ! P branched along D is then a K3 surface of degree 2 and the even Clifford 2 algebra of the quadric fibration Xe ! P gives rise to a Brauer class β 2 Br(S), called the Clifford invariant of X.
    [Show full text]
  • Tropical Curves
    Tropical Curves Nathan Pflueger 24 February 2011 Abstract A tropical curve is a graph with specified edge lengths, some of which may be infinite. Various facts and attributes about algebraic curves have analogs for tropical curves. In this article, we focus on divisors and linear series, and prove the Riemann-Roch formula for divisors on tropical curves. We describe two ways in which algebraic curves may be transformed into tropical curves: by aboemas and by specialization on arithmetic surfaces. We discuss how the study of linear series on tropical curves can be used to obtain results about linear series on algebraic curves, and summarize several recent applications. Contents 1 Introduction 2 2 From curves to graphs 3 2.1 Amoebas of plane curves . .3 2.2 Curves over the field of Puiseux series . .4 2.3 Specialization . .5 3 Metric graphs and tropical curves 6 4 Divisors and linear equivalence on tropical curves 9 4.1 The Riemann-Roch criterion . 12 4.2 Tropical Riemann-Roch . 14 5 Tropical plane curves 17 5.1 Tropical algebra and tropical projective space . 17 5.2 Tropical curves in R2 ................................... 18 5.3 Calculation of the genus . 23 5.4 Stable intersection and the tropical B´ezouttheorem . 24 5.5 Classical B´ezoutfrom tropical B´ezout . 29 5.6 Enumerative geometry of tropical plane curves . 31 6 Tropical curves via specialization 32 6.1 The specialization map and specialization lemma . 32 6.2 The canonical divisor of a graph is canonical . 34 6.3 A tropical proof of the Brill-Noether theorem . 34 1 1 Introduction The origins of tropical geometry lie in the study of tropical algebra, whose basic object is the set R [ {−∞} equipped with the operations x ⊕ y = max(x; y) and x ⊗ y = x + y.
    [Show full text]
  • The Galois Group of the 27 Lines on a Rational Cubic Surface
    The Galois group of the 27 lines on a rational cubic surface Arav Karighattam Mentor – Yongyi Chen May 18, 2019 MIT-PRIMES Conference Cubic surfaces and lines A cubic surface is the set of all points (x, y, z) satisfying a cubic polynomial equation. In this talk we will be considering smooth cubic surfaces. Example. The Fermat cubic is the set of all points (x, y, z) with + + = 1. 3 3 3 − Cubic surfaces and lines The lines in the Fermat cubic + + = 1 are: 3 3 3 (-ζ, t, -ωt),− (-ζt, t, -ω), (-ζt, -ω, t), where t varies over , and ζ and ω are cube roots of 1. Notice that there areℂ 27 lines on this cubic surface. Theorem. (Cayley, 1849) Every smooth cubic surface contains exactly 27 lines. Cubic surfaces and lines This is an example of a cubic surface. Source: 27 Lines on a Cubic Surface by Greg Egan http://www.gregegan.net/ Projective space Define n-dimensional projective space as the space of nonzero (n + 1)-tuples of complex numbers … up to scaling. We may embed n-dimensional space ℙ into projective space by setting one of the 0 ∶ 1 ∶ 2 ∶ ∶ projective coordinates to 1. Projective space can be considered as ℂ with points at infinity. A line in is a linear embedding . We can parametrize a general line in ℂ as [s : t : as3 + bt : cs + dt] for some fixed1 complex3 numbers a, b, c, and d. 3 For the restℙ of the talk, we will be consideringℙ ↪ ℙ cubic surfaces in projective space. ℙ The blow-up of at six points 2 ℙ The blow-up of at the origin is \ {0} with a line at 0, with each point on the line corresponding to a2 tangent direction at2 the point.
    [Show full text]