Criteria of Separatedness and Properness

Total Page:16

File Type:pdf, Size:1020Kb

Criteria of Separatedness and Properness VALUATIVE CRITERIA OF SEPARATEDNESS AND PROPERNESS STERGIOS ANTONAKOUDIS 1. Introduction We discuss the valuative criteria of separatedness and properness for Noetherian rings using discrete valuation rings. We consider schemes of finite type over an algebraically closed field and see how to substitute DVR's with smooth one-dimensional k-algebras. These arguments easily generalize to a less restrictive setting [GD67, II.7.2 and II.7.3]. Our exposition follows [Har77] closely. Acknowledgments. We would like to thank Dennis Gaitsgory, Bao Le Hung and Thanos Papa¨ıoannoufor fruitful conversations and help. 2. Valuative criterion of separatedness 2.1. Motivation. Here is the rough idea: begin with a curve C missing a point P , and a scheme X over a field. We would like to say that X is separated if given any morphism from C − P to X, there is at most one way that we can extend it to a morphism from all of C into X. Since the above question is local, we should replace the curve by its local ring at P , which is a discrete valuation ring (we will see that in the case of schemes of finite type over a field this is not necessary). Theorem 1. Let f : X ! Y be a morphism of Noetherian schemes. Then f is separated if and only if the following condition holds. Given any DVR R with field of fractions K, with i : Spec(K) ! Spec(R) induced from the inclusion of R in K, and given any two morphisms Spec(K) ! X and Spec(R) ! Y such that the diagram Spec(K) / X ; i f Spec(R) / Y commutes, there is at most one morhism from Spec(R) to X making the whole diagram commutative. Before we prove the theorem we need to prove three lemmas. Date: December 3, 2009. 1 2 STERGIOS ANTONAKOUDIS Lemma 1. Let R be a valuation ring of a field K. Let T = Spec(R) and let U = Spec(K). To give a morphism of U to a scheme X is equivalent to giving a point x1 2 X and an inclusion of fields k(x1) ⊆ K. To give a morphism from T to X is equivalent to giving two points x0; x1 2 X, with x0 a specialization of x1, and an inclusion of fields k(x1) ⊆ K, such that R dominates the local ring O of x0 on the subscheme Z = fx1g of X with its reduced induced structure. Proof. [Har77, Section 2.4, Lemma 4.4]. Lemma 2. Let f : X ! Y be a quasi-compact morphism of schemes. Then the subset f(X) of Y is closed if and only if it is stable under specialization. Proof. [Har77, Section 2.4, Lemma 4.5.]. Lemma 3. Let O be a noetherian local domain with field of fractions K, and let L be a finitely-generated field extension of K, then there exists a discrete valuation ring R of L dominating O, i.e. the inclusion O ! R is a local ring homomorphism. Proof. Let y1; ··· ; yn be transcendental elements of L over K such that L is finite over K(y1; ··· ; yn). Then the extension of the maximal ideal m of O in O[y1; ··· ; yn] is not the unit ideal and therefore by localizing at a prime ideal above it, it suffices to find a DVR in L dominating that localization. Hence we can assume without loss of generality that L is a finite field extension of K. Next, let x1; ··· ; xn be a system of parameters for m, then x1; ··· ; xn are algebraically independent over K. Therefore the extension of m in 0 O = O[x2=x1; ··· ; xn=x1] is (x1), which is not the unit ideal. (An alternative way to see this is to choose x1 a non-nilpotent element of degree 1 in gr(m), which exists since the latter has the same dimension with O and hence it is not Artinian). Let p be a minimal 0 prime above (x1). Then p has height 1 by Krull's theorem. Let B be the localization of O at p. Then B is a noetherian local domain of dimension 1 and hence by the Krull-Akizuki theorem (See Appendix, Lemma 7) the integral closure of B in L, call it C, is noetherian of dimension 1. In particular if we localize C at a maximal ideal we get a DVR in L, which dominates O, as desired. Proof of theorem. First suppose that f is separated, and suppose given a diagram as above, with T = Spec(R) and U = Spec(K), where there are two morphisms h; h0 of T to X making the whole diagram commutative. Spec(K) ; / X www; h wwww i wwww f wwwwh0 www Spec(R) / Y 00 Then we obtain a morphism h : T ! X ×Y X. We observe that the restrictions of h and h0 to U are the same and the closure of U is all of T . Since ∆(X) is closed, the image of T lies in the diagonal (in the set-theoric sense). But since T is reduced it factors through the reduced scheme structure of the diagonal. But that implies that h and h0 are equal, by the universal property of h00. VALUATIVE CRITERIA OF SEPARATEDNESS AND PROPERNESS 3 Conversely, let us suppose that the condition of the theorem is satisfied. We need to show that f is separated and hence it will be enough to show that ∆(X) is closed. Since X is Noetherian, the diagonal morphism is quasi-compact and hence by the previous lemma it suffices to show that ∆(X) is stable under specialization. Assume that there exists p1 in ∆(X) with p0 2 fp1g. Let's assume that p0 is not in the diagonal. Let K = k(p1) and O be the local ring at p0 of fp1g, with its reduced induced structure. Then O is a local noetherian domain with field of fractions K. Hence by the lemma we proved above there is a DVR R of K that dominates O. The field of fractions of R is K and if we let T = Spec(R) and U = Spec(K) then by the lemma above we get a morphism T ! X ×Y X, sending the generic point t1 of T to p1 and the closed point t0 of T to p0. Now, we are almost done, since composing with the two projections π1 X ×Y X //X π2 gives two morphisms of T to X which give the same morhism to Y, and moreover whose restrictions to U are the same. So by the conditions of the theorem these two morphisms are the same. That means that the morphim T ! X ×Y X factors through the diagonal morhism and hence the point p0 belongs to the diagonal ∆(X). That finishes the proof. 3. Valuative criterion of properness 3.1. Motivation. The rough idea is similar to that of the valuative criterion of separat- edness in Section 2.1, namely given a curve C missing a point P , and a scheme X over a field, we would like to say that X is proper if given any morphism from C − P to X, there is precisely one way to we extend it to a morphism from all of C into X. Since the above question is local, again, we should replace the curve by its local ring at P , which is a discrete valuation ring. Theorem 2. Let f : X ! Y be a morphism of finite type of Noetherian schemes. Then f is proper if and only if the following condition holds. For any DVR R with field of fractions K, with i : Spec(K) ! Spec(R) induced from the inclusion of R in K, for any two morphisms Spec(K) ! X and Spec(R) ! Y such that the following diagram commutes. U / X > i f T / Y Then there is a unique morhism of Spec(R) to X making the whole diagram commutative. Proof. First assume that f is proper. Then f is also separated and by the previous theorem if such a morphism T ! X exists, it has to be unique. So we only need to prove that it exists. Consider the base change T ! Y and let XT be X ×Y T . Then we get a map 4 STERGIOS ANTONAKOUDIS U ! XT by the universality of fiber product, such that the following diagram commutes. U / XT / X f 0 f T / Y Let p1 2 XT be the image of the unique point t1 of U. Let Z = fp1g. Then Z is a closed 0 subset of XT . Since f is proper, it is also universally closed and hence f is a closed map. In particular the image of Z under f 0 is a closed subset of T . But by the commutativity 0 of the diagram above we have that f (p1) = t1, which is the generic point of T , hence 0 0 f (Z) = T . In particular there is a point p0 2 Z such that f (p0) = t0, the closed point of T . That gives us a homomorphism of local rings R ! OZ;p0 corresponding to the morphism 0 f . The function field of Z is k(p1), which is contained in K. But R is a discrete valuation ring, dominated by a local ring inside K, therefore R is isomorphic to OZ;p0 . Here I use the fact that valuation rings of a field K are maximal subrings of K with respect to the ordering given by domination. Therefore by the lemma in the previous section we get a morphism from T to XT sending t0; t1 to p0; p1, respectively.
Recommended publications
  • Arxiv:1708.06494V1 [Math.AG] 22 Aug 2017 Proof
    CLOSED POINTS ON SCHEMES JUSTIN CHEN Abstract. This brief note gives a survey on results relating to existence of closed points on schemes, including an elementary topological characterization of the schemes with (at least one) closed point. X Let X be a topological space. For a subset S ⊆ X, let S = S denote the closure of S in X. Recall that a topological space is sober if every irreducible closed subset has a unique generic point. The following is well-known: Proposition 1. Let X be a Noetherian sober topological space, and x ∈ X. Then {x} contains a closed point of X. Proof. If {x} = {x} then x is a closed point. Otherwise there exists x1 ∈ {x}\{x}, so {x} ⊇ {x1}. If x1 is not a closed point, then continuing in this way gives a descending chain of closed subsets {x} ⊇ {x1} ⊇ {x2} ⊇ ... which stabilizes to a closed subset Y since X is Noetherian. Then Y is the closure of any of its points, i.e. every point of Y is generic, so Y is irreducible. Since X is sober, Y is a singleton consisting of a closed point. Since schemes are sober, this shows in particular that any scheme whose under- lying topological space is Noetherian (e.g. any Noetherian scheme) has a closed point. In general, it is of basic importance to know that a scheme has closed points (or not). For instance, recall that every affine scheme has a closed point (indeed, this is equivalent to the axiom of choice). In this direction, one can give a simple topological characterization of the schemes with closed points.
    [Show full text]
  • 3 Lecture 3: Spectral Spaces and Constructible Sets
    3 Lecture 3: Spectral spaces and constructible sets 3.1 Introduction We want to analyze quasi-compactness properties of the valuation spectrum of a commutative ring, and to do so a digression on constructible sets is needed, especially to define the notion of constructibility in the absence of noetherian hypotheses. (This is crucial, since perfectoid spaces will not satisfy any kind of noetherian condition in general.) The reason that this generality is introduced in [EGA] is for the purpose of proving openness and closedness results on the locus of fibers satisfying reasonable properties, without imposing noetherian assumptions on the base. One first proves constructibility results on the base, often by deducing it from constructibility on the source and applying Chevalley’s theorem on images of constructible sets (which is valid for finitely presented morphisms), and then uses specialization criteria for constructible sets to be open. For our purposes, the role of constructibility will be quite different, resting on the interesting “constructible topology” that is introduced in [EGA, IV1, 1.9.11, 1.9.12] but not actually used later in [EGA]. This lecture is organized as follows. We first deal with the constructible topology on topological spaces. We discuss useful characterizations of constructibility in the case of spectral spaces, aiming for a criterion of Hochster (see Theorem 3.3.9) which will be our tool to show that Spv(A) is spectral, our ultimate goal. Notational convention. From now on, we shall write everywhere (except in some definitions) “qc” for “quasi-compact”, “qs” for “quasi-separated”, and “qcqs” for “quasi-compact and quasi-separated”.
    [Show full text]
  • What Is a Generic Point?
    Generic Point. Eric Brussel, Emory University We define and prove the existence of generic points of schemes, and prove that the irreducible components of any scheme correspond bijectively to the scheme's generic points, and every open subset of an irreducible scheme contains that scheme's unique generic point. All of this material is standard, and [Liu] is a great reference. Let X be a scheme. Recall X is irreducible if its underlying topological space is irre- ducible. A (nonempty) topological space is irreducible if it is not the union of two proper distinct closed subsets. Equivalently, if the intersection of any two nonempty open subsets is nonempty. Equivalently, if every nonempty open subset is dense. Since X is a scheme, there can exist points that are not closed. If x 2 X, we write fxg for the closure of x in X. This scheme is irreducible, since an open subset of fxg that doesn't contain x also doesn't contain any point of the closure of x, since the compliment of an open set is closed. Therefore every open subset of fxg contains x, and is (therefore) dense in fxg. Definition. ([Liu, 2.4.10]) A point x of X specializes to a point y of X if y 2 fxg. A point ξ 2 X is a generic point of X if ξ is the only point of X that specializes to ξ. Ring theoretic interpretation. If X = Spec A is an affine scheme for a ring A, so that every point x corresponds to a unique prime ideal px ⊂ A, then x specializes to y if and only if px ⊂ py, and a point ξ is generic if and only if pξ is minimal among prime ideals of A.
    [Show full text]
  • DENINGER COHOMOLOGY THEORIES Readers Who Know What the Standard Conjectures Are Should Skip to Section 0.6. 0.1. Schemes. We
    DENINGER COHOMOLOGY THEORIES TAYLOR DUPUY Abstract. A brief explanation of Denninger's cohomological formalism which gives a conditional proof Riemann Hypothesis. These notes are based on a talk given in the University of New Mexico Geometry Seminar in Spring 2012. The notes are in the same spirit of Osserman and Ile's surveys of the Weil conjectures [Oss08] [Ile04]. Readers who know what the standard conjectures are should skip to section 0.6. 0.1. Schemes. We will use the following notation: CRing = Category of Commutative Rings with Unit; SchZ = Category of Schemes over Z; 2 Recall that there is a contravariant functor which assigns to every ring a space (scheme) CRing Sch A Spec A 2 Where Spec(A) = f primes ideals of A not including A where the closed sets are generated by the sets of the form V (f) = fP 2 Spec(A) : f(P) = 0g; f 2 A: By \f(P ) = 000 we means f ≡ 0 mod P . If X = Spec(A) we let jXj := closed points of X = maximal ideals of A i.e. x 2 jXj if and only if fxg = fxg. The overline here denote the closure of the set in the topology and a singleton in Spec(A) being closed is equivalent to x being a maximal ideal. 1 Another word for a closed point is a geometric point. If a point is not closed it is called generic, and the set of generic points are in one-to-one correspondence with closed subspaces where the associated closed subspace associated to a generic point x is fxg.
    [Show full text]
  • Chapter 1: Lecture 11
    Chapter 1: Lecture 11 1. The Spectrum of a Ring & the Zariski Topology Definition 1.1. Let A be a ring. For I an ideal of A, define V (I)={P ∈ Spec(A) | I ⊆ P }. Proposition 1.2. Let A be a ring. Let Λ be a set of indices and let Il denote ideals of A. Then (1) V (0) = Spec(R),V(R)=∅; (2) ∩l∈ΛV (Il)=V (Pl∈Λ Il); k k (3) ∪l=1V (Il)=V (∩l=1Il); Then the family of all sets of the form V (I) with I ideal in A defines a topology on Spec(A) where, by definition, each V (I) is a closed set. We will call this topology the Zariski topology on Spec(A). Let Max(A) be the set of maximal ideals in A. Since Max(A) ⊆ Spec(A) we see that Max(A) inherits the Zariski topology. n Now, let k denote an algebraically closed fied. Let Y be an algebraic set in Ak .Ifwe n consider the Zariski topology on Y ⊆ Ak , then points of Y correspond to maximal ideals in A that contain I(Y ). That is, there is a natural homeomorphism between Y and Max(k[Y ]). Example 1.3. Let Y = {(x, y) | x2 = y3}⊂k2 with k algebraically closed. Then points in Y k[x, y] k[x, y] correspond to Max( ) ⊆ Spec( ). The latter set has a Zariski topology, and (x2 − y3) (x2 − y3) the restriction of the Zariski topology to the set of maximal ideals gives a topological space homeomorphic to Y (with the Zariski topology).
    [Show full text]
  • 18.726 Algebraic Geometry Spring 2009
    MIT OpenCourseWare http://ocw.mit.edu 18.726 Algebraic Geometry Spring 2009 For information about citing these materials or our Terms of Use, visit: http://ocw.mit.edu/terms. 18.726: Algebraic Geometry (K.S. Kedlaya, MIT, Spring 2009) More properties of schemes (updated 9 Mar 09) I’ve now spent a fair bit of time discussing properties of morphisms of schemes. How­ ever, there are a few properties of individual schemes themselves that merit some discussion (especially for those of you interested in arithmetic applications); here are some of them. 1 Reduced schemes I already mentioned the notion of a reduced scheme. An affine scheme X = Spec(A) is reduced if A is a reduced ring (i.e., A has no nonzero nilpotent elements). This occurs if and only if each stalk Ap is reduced. We say X is reduced if it is covered by reduced affine schemes. Lemma. Let X be a scheme. The following are equivalent. (a) X is reduced. (b) For every open affine subsheme U = Spec(R) of X, R is reduced. (c) For each x 2 X, OX;x is reduced. Proof. A previous exercise. Recall that any closed subset Z of a scheme X supports a unique reduced closed sub- scheme, defined by the ideal sheaf I which on an open affine U = Spec(A) is defined by the intersection of the prime ideals p 2 Z \ U. See Hartshorne, Example 3.2.6. 2 Connected schemes A nonempty scheme is connected if its underlying topological space is connected, i.e., cannot be written as a disjoint union of two open sets.
    [Show full text]
  • ALGEBRAIC GEOMETRY NOTES 1. Conventions and Notation Fix A
    ALGEBRAIC GEOMETRY NOTES E. FRIEDLANDER J. WARNER 1. Conventions and Notation Fix a field k. At times we will require k to be algebraically closed, have a certain charac- teristic or cardinality, or some combination of these. AN and PN are affine and projective spaces in N variables over k. That is, AN is the set of N-tuples of elements of k, and PN N+1 is the set of equivalence classes of A − 0 under the relation (a0; : : : ; aN ) ∼ (b0; : : : ; bN ) if and only if there exists c 2 k with cai = bi for i = 0;:::;N. 2. Monday, January 14th Theorem 2.1 (Bezout). Let X and Y be curves in P2 of degrees d and e respectively. Let fP1;:::;Png be the points of intersection of X and Y . Then we have: n X (1) i(X; Y ; Pi) = de i=1 3. Wednesday, January 16th - Projective Geometry, Algebraic Varieties, and Regular Functions We first introduce the notion of projective space required to make Bezout's Theorem work. Notice in A2 we can have two lines that don't intersect, violating (1). The idea is that these two lines intersect at a "point at infinity". Projective space includes these points. 3.1. The Projective Plane. Define the projective plane, P2, as a set, as equivalence classes 3 of A under the relation (a0; a1; a2) ∼ (b0; b1; b2) if and only if there exists c 2 k with cai = bi for i = 0; 1; 2. It is standard to write [a0 : a1 : a2] for the equivalence class represented by 3 (a0; a1; a2) 2 A .
    [Show full text]
  • AN INTRODUCTION to AFFINE SCHEMES Contents 1. Sheaves in General 1 2. the Structure Sheaf and Affine Schemes 3 3. Affine N-Space
    AN INTRODUCTION TO AFFINE SCHEMES BROOKE ULLERY Abstract. This paper gives a basic introduction to modern algebraic geom- etry. The goal of this paper is to present the basic concepts of algebraic geometry, in particular affine schemes and sheaf theory, in such a way that they are more accessible to a student with a background in commutative al- gebra and basic algebraic curves or classical algebraic geometry. This paper is based on introductions to the subject by Robin Hartshorne, Qing Liu, and David Eisenbud and Joe Harris, but provides more rudimentary explanations as well as original proofs and numerous original examples. Contents 1. Sheaves in General 1 2. The Structure Sheaf and Affine Schemes 3 3. Affine n-Space Over Algebraically Closed Fields 5 4. Affine n-space Over Non-Algebraically Closed Fields 8 5. The Gluing Construction 9 6. Conclusion 11 7. Acknowledgements 11 References 11 1. Sheaves in General Before we discuss schemes, we must introduce the notion of a sheaf, without which we could not even define a scheme. Definitions 1.1. Let X be a topological space. A presheaf F of commutative rings on X has the following properties: (1) For each open set U ⊆ X, F (U) is a commutative ring whose elements are called the sections of F over U, (2) F (;) is the zero ring, and (3) for every inclusion U ⊆ V ⊆ X such that U and V are open in X, there is a restriction map resV;U : F (V ) ! F (U) such that (a) resV;U is a homomorphism of rings, (b) resU;U is the identity map, and (c) for all open U ⊆ V ⊆ W ⊆ X; resV;U ◦ resW;V = resW;U : Date: July 26, 2009.
    [Show full text]
  • Monodromy Map for Tropical Dolbeault Cohomology
    Algebraic Geometry 6 (4) (2019) 384{409 doi:10.14231/AG-2019-018 Monodromy map for tropical Dolbeault cohomology Yifeng Liu Abstract We define monodromy maps for the tropical Dolbeault cohomology of algebraic vari- eties over non-Archimedean fields. We propose a conjecture of Hodge isomorphisms via monodromy maps and provide some evidence. 1. Introduction Let K be a complete non-Archimedean field. For a separated scheme X of finite type over K, one an Q T p has the associated K-analytic space X in the sense of Berkovich [Ber93]. Using the -sheaf Xan an p;q on X constructed in [Liu17], we obtain a new cohomology theory Htrop(X), functorial in X. p;q 6 6 They are rational vector spaces satisfying Htrop(X) 6= 0 only when 0 p; q dim X. Denote by p;q p;q htrop(X) the dimension of Htrop(X), which might be infinity. A •;• R an Using a bicomplex Xan of -sheaves of real forms on X , constructed by Chambert-Loir and p;q p;q R Ducros [CD12], we can recover Htrop(X)R := Htrop(X) ⊗Q via certain Dolbeault cohomology groups. In this article, we consider a natural question, motivated by complex geometry, about p;q q;p the Hodge relation: Do we have the equality htrop(X) = htrop(X)? However, the perspective we take is very different from the one in complex geometry. In fact, we will construct, for p > 1 and q > 0, a functorial map p;q p−1;q+1 NX :Htrop(X)R ! Htrop (X)R ; A p;q A p−1;q+1 which we call the monodromy map, via constructing a canonical map Xan ! Xan on the level of sheaves.
    [Show full text]
  • Arxiv:Math/0509587V6
    THE SPECIALIZATIONS IN A SCHEME FENG-WEN AN Abstract. In this paper we will obtain some further properties for specializations in a scheme. Using these results, we can take a picture for a scheme and a picture for a morphism of schemes. In particular, we will prove that every morphism of schemes is specialization-preserving and of norm not greater than one (under some condition); a necessary and sufficient condition will be given for an injective morphism between irreducible schemes. Introduction Specializations are concrete and intuitive for one to study classical varieties[6]. The results on this topic relating to schemes are mainly presented in Grothendieck’s EGA. In this paper we try to obtain some properties for the specializations in a scheme such as the lengths of specializations. Together with specializations, a scheme can be regarded as a partially ordered set. In §1 we will prove that every morphism of schemes is specialization-preserving (Proposition 1.3 ) and that every specilization in a scheme is contained in an affine open subset (Proposition 1.9 ). Using those results, we can take a picture for a scheme (Remark 1.10 ): A scheme can be described to be a number of trees standing on the ground such that arXiv:math/0509587v6 [math.AG] 5 Jun 2007 each irreducible component is a tree; the generic point of an irreducible component is the root of the cor- responding tree; the closed points of an irreducible component are the top leaves of the corresponding tree; each specialization in an irreducible component are the branches of the corresponding tree.
    [Show full text]
  • A Site-Theoretic Characterization of Points in a Topological Space
    A SITE-THEORETIC CHARACTERIZATION OF POINTS IN A TOPOLOGICAL SPACE ANDREW ARCHIBALD, PETE L. CLARK, AND DAVID SAVITT Abstract. We study points in the context of Grothendieck topologies (sites), especially the relationship between \honest" points of a topological space X and site-theoretic points of Xtop. We o®er an elementary and self-contained proof of a theorem of Grothendieck-Hakim which gives necessary and su±cient conditions for these two notions to coincide. We also explore analogues of this result for other sites encountered in geometry by giving axioms on a site su±cient to ensure the correspondence between site-theoretic points and the points of a certain topological space. These axioms are satis¯ed for the ¶etale, fppf, fpqc and crystalline sites. They are not satis¯ed by the large ¶etalesite of Spec C, in which (as we show) the site-theoretic points form a proper class. 1. Introduction Recall that a site, or Grothendieck topology (C; T ), consists of (i) a category C possessing a ¯nal object and closed under ¯ber products, together with (ii) for each object U in C, a distinguished collection of families of morphisms fUi ! Ugi2I , called the covering families of U. We require the following properties: (GT1) If U 0 ! U is an isomorphism, then fU 0 ! Ug is a covering family. (GT2) If fUi ! Ugi2I is a covering family and V ! U is a morphism, then fUi £U V ! V gi2I is a covering family of V . (GT3) If fUi ! Ugi2I is a covering family and for each i 2 I, fVi;j ! Uigj2J is a covering family of Ui, then the composite family fVi;j ! Ui ! Ug(i;j)2I£J is a covering family of U.
    [Show full text]
  • The Cohomology of Coherent Sheaves
    CHAPTER VII The cohomology of coherent sheaves 1. Basic Cechˇ cohomology We begin with the general set-up. (i) X any topological space = U an open covering of X U { α}α∈S a presheaf of abelian groups on X. F Define: (ii) Ci( , ) = group of i-cochains with values in U F F = (U U ). F α0 ∩···∩ αi α0,...,αYi∈S We will write an i-cochain s = s(α0,...,αi), i.e., s(α ,...,α ) = the component of s in (U U ). 0 i F α0 ∩··· αi (iii) δ : Ci( , ) Ci+1( , ) by U F → U F i+1 δs(α ,...,α )= ( 1)j res s(α ,..., α ,...,α ), 0 i+1 − 0 j i+1 Xj=0 b where res is the restriction map (U U ) (U U ) F α ∩···∩ Uαj ∩···∩ αi+1 −→ F α0 ∩··· αi+1 and means “omit”. Forb i = 0, 1, 2, this comes out as δs(cα , α )= s(α ) s(α ) if s C0 0 1 1 − 0 ∈ δs(α , α , α )= s(α , α ) s(α , α )+ s(α , α ) if s C1 0 1 2 1 2 − 0 2 0 1 ∈ δs(α , α , α , α )= s(α , α , α ) s(α , α , α )+ s(α , α , α ) s(α , α , α ) if s C2. 0 1 2 3 1 2 3 − 0 2 3 0 1 3 − 0 1 2 ∈ One checks very easily that the composition δ2: Ci( , ) δ Ci+1( , ) δ Ci+2( , ) U F −→ U F −→ U F is 0. Hence we define: 211 212 VII.THECOHOMOLOGYOFCOHERENTSHEAVES s(σβ0, σβ1) defined here U σβ0 Uσβ1 Vβ1 Vβ0 ref s(β0, β1) defined here Figure VII.1 (iv) Zi( , ) = Ker δ : Ci( , ) Ci+1( , ) U F U F −→ U F = group of i-cocycles, Bi( , ) = Image δ : Ci−1( , ) Ci( , ) U F U F −→ U F = group of i-coboundaries Hi( , )= Zi( , )/Bi( , ) U F U F U F = i-th Cech-cohomologyˇ group with respect to .
    [Show full text]