Dominancia Relativa (DmR)
6.6. Identificación y Validación de Indicadores Ambientales
Theorem. (“algebraic version”; see [107]) Let 𝑓 : 𝑋 → 𝑌 be a projective morphism of Noetherian schemes (see“Schemes”). Then there exists a scheme 𝑍, a finite morphism 𝑔 : 𝑍 → 𝑌 and a projective morphism with connected fibres ℎ : 𝑋 → 𝑍 such that
𝑓 = 𝑔 ∘ ℎ .
Theorem. (“analytic version”); (see [241]). Let 𝑓 : 𝑋 → 𝑌 be a proper surjective mor-phism of reduced complex analytic spaces (see“Spaces, analytic -”). Then there exists a reduced complex analytic space 𝑍, a surjective morphism 𝑔 : 𝑍 → 𝑌 such that the fibres consist of a finite number of points and a surjective morphism with connected fibres ℎ : 𝑋 → 𝑍 such that
𝑓 = 𝑔 ∘ ℎ .
Subcanonical.
We say that a smooth subvariety 𝑍 of a smooth algebraic variety 𝑋 is subcanonical if there exists a line bundle 𝐿 on 𝑋 such that 𝐿|𝑍 = 𝐾𝑍, where 𝐾𝑍is the canonical bundle of 𝑍 (see“Canonical bundle, canonical sheaf”).Surfaces, algebraic -.
([18], [22], [25], [65], [93], [107], [253]). In the sequel the word surface will denote a smooth projective algebraic variety of dimension 2 over ℂ.We will denote the projective space ℙ𝑛ℂby ℙ𝑛.
We start by stating some basic theorems and notation.
Let 𝑆 be a surface; then 𝐻4(𝑆, ℤ) and 𝐻4(𝑆, ℤ) are isomorphic to ℤ, since 𝑆 is oriented (as it is a complex manifold), compact and of real dimension 4 (see“Singular homol-ogy and cohomolhomol-ogy”). We denote by [𝑆] the fundamental class of 𝑆, that is, the gener-ator of 𝐻4(𝑆, ℤ) giving the orientation of 𝑆. If 𝐿 and 𝐿are two holomorphic line bundles on 𝑆, we denote by 𝐿 ⋅ 𝐿the number obtained by evaluating 𝑐1(𝐿) ∪ 𝑐1(𝐿) ∈ 𝐻4(𝑆, ℤ) in [𝑆] (where ∪ is the cup product; see again“Singular homology and cohomology”).
For any divisors 𝐷 and 𝐷on 𝑆, let 𝐷 ⋅ 𝐷their intersection number (see“Intersection of cycles”).
Observe that, for any holomorphic line bundle 𝐿 on 𝑆, we have that 𝐻𝑖(O(𝐿)) = 0 for 𝑖 ≥ 3; to prove this, we can apply the Abstract de Rham’s theorem (see“Sheaves”) to the exact sequence
0 →O(𝐿) → 𝐶∞(𝐿)→ 𝐴𝜕 0,1(𝐿)→ 𝐴𝜕 0,2(𝐿) → 0,
where 𝐶∞(𝐿) is the sheaf of the 𝐶∞sections of 𝐿, 𝐴𝑝,𝑞(𝐿) is the sheaf of the 𝐶∞(𝑝, 𝑞)-forms with values in 𝐿 and the mapO(𝐿) → 𝐶∞(𝐿) is given by inclusion.
So, if we denote by 𝜒(O(𝐿)) the Euler characteristic of O(𝐿), i.e., the number
∑𝑖=0,...,∞(−1)𝑖ℎ𝑖(O(𝐿)), we have that, in this case (i.e., in the case of surfaces), 𝜒(O(𝐿)) = ℎ0(O(𝐿)) − ℎ1(O(𝐿)) + ℎ2(O(𝐿)),
Surfaces, algebraic - | 181 and so 𝜒(O𝑆) = ℎ0(O𝑆) − ℎ1(O𝑆) + ℎ2(O𝑆). Furthermore, we recall that 𝜒(𝑆) denotes the Euler-Poincaré characteristic of 𝑆, i.e., ∑𝑖=0,...,∞(−1)𝑖𝑏𝑖(𝑆), where 𝑏𝑖(𝑆) is the Betti number 𝑑𝑖𝑚 𝐻𝑖(𝑆, ℝ); in this case,
𝜒(𝑆) = 𝑏0(𝑆) − 𝑏1(𝑆) + 𝑏2(𝑆) − 𝑏3(𝑆) + 𝑏4(𝑆) = 2𝑏0(𝑆) − 2𝑏1(𝑆) + 𝑏2(𝑆) by Poincaré duality (see“Singular homology and cohomology”).
Index theorem. Let 𝑆 be a surface. The intersection form is negative definite on a sub-space of codimension 1 in 𝐻1,1(𝑆, ℂ), precisely is negative definite on the subspace of the primitive forms (see“Lefschetz decomposition and Hard Lefschetz theorem”for the definition of primitive).
Riemann–Roch theorem for surfaces. Let 𝐿 be a holomorphic line bundle on a sur-face 𝑆. We have:
𝜒(O(𝐿)) = 𝜒(O𝑆) +𝐿 ⋅ 𝐿 − 𝐿 ⋅ 𝐾𝑆
2 ,
where 𝐾𝑆 is the canonical bundle of 𝑆 (see “Canonical bundle, canonical sheaf”).
Noether’s theorem. Let 𝑆 be a surface. We denote by 𝑐𝑖(𝑆) the 𝑖-th Chern class of the holomorphic tangent bundle 𝑇1,0𝑆 and we denote by 𝑐1(𝑆)2the cup product 𝑐1(𝑆)∪𝑐1(𝑆).
Hopf theorem (see“Gauss–Bonnet–Hopf theorem”)𝑐2(𝑇1,0𝑆) = 𝜒(𝑆).
Castelnuovo–Enriques criterion. Let 𝑆 be a surface and 𝐶 a smooth rational curve in 𝑆 with 𝐶 ⋅ 𝐶 = −1. Then there exists a surface 𝑆and a map 𝑝 : 𝑆 → 𝑆such that 𝑝 is the blow-up of 𝑆in a point 𝑃 and 𝑝−1(𝑃) = 𝐶.
Structure of birational maps on surfaces. Let 𝑆1and 𝑆2be two surfaces and 𝑓 : 𝑆1→ 𝑆2be a birational map. Then there exists another surface ̃𝑆 and two morphisms, 𝑔1 and 𝑔2, both given by a sequence of blowing-up maps, such that the following diagram commutes:
182 | Surfaces, algebraic
-In other words, a birational map between surfaces is given by a sequence of blowing-ups followed by a sequence of blowing-downs.
Definition. We say that a surface 𝑆 is minimal if there does not exist another surface
̃𝑆 and a blowing-up 𝑆 → ̃𝑆. A minimal model for a surface 𝑆 is a minimal surface bira-tional to 𝑆.
We recall that if 𝑀 is a compact complex manifold of dimension 𝑛, we define – 𝑞(𝑀) = ℎ1,0(𝑀) (irregularity);
– 𝑃𝑟(𝑀) = ℎ0(O(𝐾𝑀𝑟 )) for any 𝑟 ∈ ℕ (plurigenera);
– 𝑝𝑔(𝑀) = ℎ𝑛,0(𝑀) = 𝑃1(𝑀) (geometric genus).
One can prove that they are birational invariants. With the notation above, we have 𝜒(O𝑆) = ℎ0(O𝑆) − ℎ1(O𝑆) + ℎ2(O𝑆)
= ℎ0,0(𝑆) − ℎ0,1(𝑆) + ℎ0,2(𝑆) = 1 − 𝑞(𝑆) + 𝑝𝑔(𝑆) by Dolbeault’s theorem and the Hodge theorem (see“Hodge theory”).
The most important tool to classify surfaces is Kodaira dimension 𝐾 (see“Kodaira dimension (or Kodaira number)”). For surfaces, it can be obviously only −∞, 0, 1, 2.
First, we will deal with rational and ruled surfaces.
Definition. We say that a surface is rational if it is birational to ℙ2.
Definition. We say that a surface is ruled over a compact Riemann surface 𝐶 if it is birational to 𝐶 × ℙ1.
Remark. Let 𝑆 be a surface. Then 𝑆 is ruled on ℙ1⇔ 𝑆 is rational (in fact it is birational to ℙ1× ℙ1if and only if it is birational to ℙ2).
Definition. We say that a surface 𝑆 is geometrically ruled over a compact Riemann surface 𝐶 if there exists a smooth morphism 𝑆 → 𝐶 such that the fibres are isomorphic to ℙ1.
(Please note that in some works the term “ruled” means “geometrically ruled”.) Noether–Enriques theorem. A geometrically ruled surface is equal to the projectivized ℙ(𝐸) of a vector bundle 𝐸 of rank 2.
Proposition 1. Let 𝑆 be a ruled surface on a compact Riemann surface 𝐶. Let 𝑔(𝐶) be the genus of 𝐶. Then
𝑞(𝑆) = 𝑔(𝐶), 𝑝𝑔(𝑆) = 0, 𝑃𝑛(𝑆) = 0 ∀𝑛 ≥ 2.
Moreover, if 𝑆 is geometrically ruled, then
𝐾𝑆2= 8(−𝑔(𝐶) + 1), 𝑏2(𝑆) = 2 .
Surfaces, algebraic - | 183
Definition. Let 𝑛 ∈ ℕ. The 𝑛-th Hirzebruch surface is 𝑆𝑛= ℙ(Oℙ1(𝑛) ⊕Oℙ1),
whereOℙ1is the trivial bundle on ℙ1andOℙ1(𝑛) is the 𝑛-th power of the hyperplane bundle on ℙ1; see“Hyperplane bundles, twisting sheaves”(we are making a slight abuse of the notation:Oℙ1(𝑛) generally denotes the sheaf of the holomorphic sections of the 𝑛-th power of the hyperplane bundle, but sometimes, as here, it denotes just the 𝑛-th power of the hyperplane bundle).
Remark. The Hirzebruch surfaces are exactly the surfaces that are geometrically ruled on ℙ1. In fact, by the Grothendieck–Segre theorem (see “Grothendieck–Segre theo-rem”) and by the Noether–Enriques theorem, we have that a surface that is geomet-rically ruled on ℙ1is equal to ℙ(𝐿1⊕ 𝐿2) = ℙ((𝐿1⊗ 𝐿−12 ) ⊕Oℙ1) for some 𝐿1and 𝐿2 holomorphic line bundles on ℙ1and ℙ((𝐿1⊗ 𝐿−12 ) ⊕Oℙ1) = ℙ(Oℙ1(𝑛) ⊕Oℙ1) for some 𝑛. Obviously we can suppose 𝑛 ∈ ℕ.
We can prove that for all 𝑛 ̸= 0 the surface 𝑆𝑛is the unique ℙ1-bundle on ℙ1with an irreducible curve 𝐸 with 𝐸 ⋅ 𝐸 = −𝑛 and that the blow-up of ℙ2in a point is 𝑆1. We can also prove that 𝑆𝑛−1can be obtained from 𝑆𝑛by a up followed by a blowing-down.
Let 𝐹 be a fibre of ℙ(Oℙ1(𝑛) ⊕OP1) and 𝑍 be the image of the section given by the zero section ofOℙ1(𝑛) and by the section 1 ofOP1. Let 𝑆𝑘,𝑛be the image of 𝑆𝑛by the map 𝜑(𝑍+𝑘𝐹)associated to the linear system |𝑍 + 𝑘𝐹| (see“Bundles fibre -”). We can prove that it is a surface in ℙ𝑛+2𝑘+1of degree 𝑛+2𝑘, which is the minimal achievable degree for a surface in ℙ𝑛+2𝑘+1(see“Minimal degree”). The surfaces 𝑆𝑘,𝑛are the rational normal scrolls of dimension 2 (see“Scrolls, rational normal -”).
Proposition 2. Every nondegenerate surface of minimal degree in ℙ𝑑, i.e., every non-degenerate surface of degree 𝑑 − 1 in ℙ𝑑, is either a rational normal scroll or the Veronese surface in ℙ5, i.e., the image of ℙ2embedded in ℙ5byOℙ2(2) (see“Veronese embedding”).
Another example of rational surfaces are Del Pezzo surfaces. Let 𝑟 ≤ 6. Let 𝑃1, . . . , 𝑃𝑟 be 𝑟 distinct points in ℙ2in general position (i.e., no 3 of them are collinear, and no 6 of them lie on a conic). Let 𝜋 : 𝑆 → ℙ2be the blowing-up of ℙ2in 𝑃1, . . . , 𝑃𝑟. Then
−𝐾𝑆defines an embedding 𝑖 : 𝑆 → ℙ9−𝑟, whose image has degree 9 − 𝑟 and is called a Del Pezzo surface.
If 𝑟 = 6, 𝑖(𝑆) is a smooth cubic in ℙ3; if 𝑟 = 5, 𝑖(𝑆) is a complete intersection of two quadrics in ℙ4.
One can show that 𝑖(𝑆) contains only a finite number of lines, precisely the images under 𝑖 of
(a) the exceptional curves;
(b) the strict transforms of the lines 𝑃𝑖𝑃𝑗for 𝑖 ̸= 𝑗;
184 | Surfaces, algebraic
-(c) the strict transforms of the conics through 5 of the 𝑃𝑖. In the case 𝑟 = 6 we have exactly 27 lines.
Theorem.
(i) The minimal ruled surfaces on ℙ1are isomorphic either to the Hirzebruch surfaces or to ℙ2.
(ii) The minimal ruled surfaces on a Riemann surface of genus ≥ 1 are the geometri-cally ruled ones and the minimal model is not unique.
Theorem. The nonruled surfaces have a unique minimal model (up to isomor-phisms).
Remark. If 𝑆 is a rational surface, then, for every 𝑛 ≥ 1, we have 𝑃𝑛(𝑆) = 𝑞(𝑆) = 𝑝𝑔(𝑆) = 0.
The following theorem tells us that also the converse is true.
Castelnuovo–Enriques theorem. Let 𝑆 be a surface such that 𝑞(𝑆) = 𝑃2(𝑆) = 0. Then 𝑆 is rational.
Theorem. Any unirational surface (see “Unirational, Lüroth problem”) is ratio-nal.
Theorem (Castelnuovo–De Franchis).
– Let 𝑆 be a minimal surface with 𝜒(𝑆) < 0; then 𝑆 is irrational ruled.
– Let 𝑆 be a minimal surface with 𝑐1(𝑆)2< 0; then 𝑆 is irrational ruled.
– Let 𝑆 be a minimal surface with 𝜒((O𝑆)) < 0; then 𝑆 is irrational ruled.
(Observe that the third statement follows trivially from the first two and Noether’s the-orem.)
Enriques’ theorem. A surface 𝑆 is ruled if and only if 𝑃12(𝑆) = 0.
As we have already said, the most important tool to classify surfaces is Kodaira dimen-sion.
Enriques’ theorem and Proposition 1 tell us that 𝐾(𝑆) = −∞ if and only if 𝑆 is ruled.
Definition. We say that a surface 𝑆 is hyperelliptic or bielliptic if it is equal to (𝐸 × 𝐹)/𝐺,
where 𝐸 and 𝐹 are elliptic curves and 𝐺 is a finite group of translations of 𝐸 acting on 𝐹 in such a way that 𝐹/𝐺 = ℙ1.
Definition. Let 𝐶 be a smooth curve. An elliptic surface 𝑆 with base 𝐶 is a surface such that there exists a surjective morphism 𝑆 → 𝐶 such that the generic fibre is an elliptic (irreducible) curve.
Surfaces, algebraic - | 185 Classification theorem (Enriques–Kodaira).
– A minimal surface 𝑆 with 𝐾(𝑆) = 0 is one of the following:
(a) a surface with 𝑞 = 0 and 𝑝𝑔= 1 (this implies 𝐾𝑆=O); such surfaces are called K3 surfaces;
(b) a surface with 𝑞 = 0 and 𝑝𝑔 = 0 (this implies 𝐾𝑆⊗2 = O); such surfaces are called Enriques surfaces;
(c) a hyperelliptic surface if 𝑞 = 1;
(d) an Abelian variety if 𝑞 = 2 (see“Tori, complex - and Abelian varieties”).
– A surface 𝑆 with 𝐾(𝑆) = 1 is elliptic.
A particular case of K3 surfaces are the Kummer surfaces. A Kummer surface 𝑋 is a surface obtained in the following way: let 𝐴 be an Abelian surface and let ̂𝐴 be the surface obtained from 𝐴 by blowing up the 16 points of order 2; define
𝑋 = ̂𝐴/⟨𝐼, 𝜎⟩,
where 𝜎 is the map induced on ̂𝐴 by the map 𝑎 → −𝑎 on 𝐴.
Proposition. The quotient of a K3 surface by a fixed-point free involution is an Enriques surface.
Conversely, let 𝑆 be an Enriques surface. As we have already said, 𝐾𝑆⊗2is the trivial bundle. Let ̃𝑆 = {𝑢 ∈ 𝐾𝑆| 𝑖(𝑢 ⊗ 𝑢) = 1} where 𝑖 is the isomorphism 𝐾𝑆⊗2 ≅ 𝑆 × ℂ. The surface ̃𝑆 is a (nonramified) double covering space of 𝑆, where the covering map is induced by the projection of 𝐾𝑆to 𝑆, and it is a K3 surface.
Definition. We say that a surface 𝑆 is of general type if 𝐾(𝑆) = 2.
Bogomolov–Miyaoka–Yau inequality. For a surface 𝑆 of general type, it holds that 𝑐12(𝑆) ≤ 3 𝑐2(𝑆),
where 𝑐𝑖(𝑆) = 𝑐𝑖(𝑇1,0𝑆).
Noether inequality. For a minimal surface of general type, 𝑆, it holds that 𝑝𝑔(𝑆) ≤ 12𝑐12(𝑆) + 2,
where 𝑐𝑖(𝑆) = 𝑐𝑖(𝑇1,0𝑆).
Bombieri and Mumford extended the classification of surfaces to arbitrary alge-braically closed fields (see [25]). The classification in the case where the characteristic is different from 2, 3 is analogous to the one over the complex numbers.
186 | Surfaces, algebraic
-Symmetric polynomials
([76], [164], [178], [181], [236], [238]). Let 𝑛 ∈ ℕ. Let 𝑃 ∈ ℤ[𝑥1, . . . , 𝑥𝑛]; we say that 𝑃 is symmetric if it is invariant for the action of the symmetric group 𝛴𝑛(in other words, if interchanging any of the variables does not modify the polynomial).We will denote by ℤ[𝑥1, . . . , 𝑥𝑛]𝛴𝑛 the set of the symmetric polynomials in 𝑥1, . . . , 𝑥𝑛 with coefficients in ℤ.
Let 𝐸𝑗(𝑥1, . . . , 𝑥𝑛) be the sum of the squarefree monomials of degree 𝑗 in 𝑥1, . . . , 𝑥𝑛 (squarefree means not divisible by the square of a variable). The polynomials 𝐸𝑗are symmetric and they are called elementary symmetric polynomials.
For example,
– 𝐸0(𝑥1, . . . , 𝑥𝑛) = 1;
– 𝐸1(𝑥1, . . . , 𝑥𝑛) = 𝑥1+ ⋅ ⋅ ⋅ + 𝑥𝑛; – 𝐸2(𝑥1, . . . , 𝑥𝑛) = ∑1≤𝑖<𝑗≤𝑛𝑥𝑖𝑥𝑗.
Observe that 𝐸𝑖(𝑥1, . . . , 𝑥𝑛) = 0 for any 𝑖 ≥ 𝑛 + 1.
The elementary symmetric polynomials can be defined also by the following formula:
𝛱𝑖=1,...,𝑛(1 + 𝑥𝑖𝑡) = ∑
𝑗∈ℕ𝐸𝑗(𝑥1, . . . , 𝑥𝑛)𝑡𝑗.
Gauss’ theorem. If 𝑃 ∈ ℤ[𝑥1, . . . , 𝑥𝑛]𝛴𝑛, then there exists a polynomial 𝐹 ∈ ℤ[𝑥1, . . . , 𝑥𝑛] such that 𝑃 = 𝐹(𝐸1, . . . , 𝐸𝑛).
Let 𝐶𝑗(𝑥1, . . . , 𝑥𝑛) be the sum of the monomials of degree 𝑗 in 𝑥1, . . . , 𝑥𝑛. The 𝐶𝑗 are called complete symmetric polynomials.
For example:
– 𝐶0(𝑥1, . . . , 𝑥𝑛) = 1;
– 𝐶1(𝑥1, . . . , 𝑥𝑛) = 𝑥1+ ⋅ ⋅ ⋅ + 𝑥𝑛; – 𝐶2(𝑥1, . . . , 𝑥𝑛) = ∑1≤𝑖≤𝑗≤𝑛𝑥𝑖𝑥𝑗.
The complete symmetric polynomials can be defined also by the following formula:
𝛱𝑖=1,...,𝑛 1 1 − 𝑥𝑖𝑡 = ∑
𝑗∈ℕ𝐶𝑗(𝑥1, . . . , 𝑥𝑛)𝑡𝑗. Remark. The following relations hold:
∑
𝑗∈ℕ
𝑥𝑛−𝑗𝑘 (−1)𝑗𝐸𝑗(𝑥1, . . . , 𝑥𝑛) = 0 ∀𝑘 ∈ {1, . . . , 𝑛},
∑
𝑗∈ℕ(−1)𝑗𝐸𝑗𝐶𝑝−𝑗= 0 ∀𝑝 ≥ 1.
Surfaces, algebraic - | 187 The first follows from the second definition of 𝐸𝑗taking 𝑡 = −1/𝑥𝑘. The second follows from the second definitions of 𝐸𝑗and 𝐶𝑗:
1 = 𝛱𝑖=1,...,𝑛(1 − 𝑥𝑖𝑡) 𝛱𝑖=1,...,𝑛 1 1 − 𝑥𝑖𝑡
= ( ∑
𝑗∈ℕ(−1)𝑗𝐸𝑗(𝑥1, . . . , 𝑥𝑛)𝑡𝑗)( ∑
𝑘∈ℕ𝐶𝑘(𝑥1, . . . , 𝑥𝑛)𝑡𝑘).
From the second relation we have ℤ[𝐶1, . . . , 𝐶𝑛] = ℤ[𝐸1, . . . , 𝐸𝑛], which, by Gauss’
theorem, is ℤ[𝑥1, . . . , 𝑥𝑛]𝛴𝑛. Thus both the elementary symmetric polynomials and the complete symmetric polynomials generate the algebra of the symmetric poly-nomials.
Now we will define another class of symmetric polynomials such that they generate ℤ[𝑥1, . . . , 𝑥𝑛]𝛴𝑛as ℤ-module: the Schur polynomials.
Let 𝜆 = (𝜆1, . . . , 𝜆𝑛) with 𝜆𝑖∈ ℕ and 𝜆1≥ ⋅ ⋅ ⋅ ≥ 𝜆𝑛(we call it a partition of 𝜆1+ ⋅ ⋅ ⋅ + 𝜆𝑛);
we can associate to 𝜆 a diagram, called a Young diagram, with 𝜆𝑖boxes in the 𝑖-th row for any 𝑖 ∈ {1, . . . , 𝑛} and the rows lined up on the left; see Figure19.
Fig. 19. Young diagram of (4, 3, 1).
Let 𝐴 be the matrix 𝑛 × ∞:
(
1 𝑥1 𝑥21 . . . .
. . . .
. . . .
. . . .
1 𝑥𝑛 𝑥2𝑛 . . . . ) .
Number the columns of 𝐴 beginning from 0. For distinct 𝑡1, . . . , 𝑡𝑛 ∈ ℕ, define 𝑎𝑡1,...,𝑡𝑛(𝑥1, . . . , 𝑥𝑛) to be the determinant of the matrix obtained by taking the columns 𝑡1, . . . , 𝑡𝑛of 𝐴.
For 𝜆 = (𝜆1, . . . , 𝜆𝑛) with 𝜆𝑖∈ ℕ and 𝜆1≥ ⋅ ⋅ ⋅ ≥ 𝜆𝑛, define 𝑆𝜆(𝑥1, . . . , 𝑥𝑛) = 𝑎𝜆+𝛿(𝑥1, . . . , 𝑥𝑛)
𝑎𝛿(𝑥1, . . . , 𝑥𝑛) ,
where 𝛿 = (𝑛 − 1, 𝑛 − 2, . . . , 0). The 𝑆𝜆are symmetric polynomials and they are called Schur polynomials.
For example: let 𝑛 = 3 and consider 𝜆 = (1, 1, 0), which we write (1, 1) (in general the zeroes at the end of a partition are omitted); we have
𝑆(1,1)= 𝑥23𝑥22(𝑥3− 𝑥2) − 𝑥21𝑥23(𝑥3− 𝑥1) + 𝑥22𝑥21(𝑥2− 𝑥1)
∏𝑖>𝑗(𝑥𝑖− 𝑥𝑗) = 𝑥1𝑥2+ 𝑥1𝑥3+ 𝑥2𝑥3.
188 | Syzygies
Theorem. The Schur polynomials are a basis of ℤ[𝑥1, . . . , 𝑥𝑛]𝛴𝑛as ℤ-module.
The following formulas express the Schur polynomials in terms of elementary sym-metric polynomials and in terms of complete symsym-metric polynomials.
Jacobi–Trudi–Giambelli formulas. For any 𝜆 = (𝜆1, . . . , 𝜆𝑛) with 𝜆𝑖∈ ℕ and 𝜆1≥ ⋅ ⋅ ⋅ ≥ column of the Young diagram of 𝜆 (in other words, the Young diagram of 𝛾 is obtained from the Young diagram of 𝜆 by interchanging rows and columns).
Littlewood–Richardson rule. For any 𝜆 = (𝜆1, . . . , 𝜆𝑛) with 𝜆𝑖 ∈ ℕ and 𝜆1 ≥ ⋅ ⋅ ⋅ ≥ 𝜆𝑛 and for any 𝜇 = (𝜇1, . . . , 𝜇𝑛) with 𝜇𝑖∈ ℕ and 𝜇1≥ ⋅ ⋅ ⋅ ≥ 𝜇𝑛, we have
𝑆𝜆𝑆𝜇= ∑
𝜈 𝑁𝜆,𝜇,𝜈𝑆𝜈,
where 𝑁𝜆,𝜇,𝜈is the number of the ways the Young diagram of 𝜆 can be expanded to the Young diagram of 𝜈 by a strict 𝜇-expansion, where
– a 𝜇 = (𝜇1, . . . , 𝜇𝑛)-expansion of the Young diagram of 𝜆 is a Young diagram ob-tained from the Young diagram of 𝜆 by adding 𝜇1boxes not two in the same col-umn, then 𝜇2boxes not two in the same column, and so on;
– a 𝜇-expansion is called strict if the following condition hold: put a 1 in each of the 𝜇1boxes, a 2 in each of the 𝜇2boxes, and so on; form a list reading the numbers in the boxes, reading from right to left and beginning from the top row; we must have that for every 𝑟 with 1 ≤ 𝑟 ≤ 𝜇1+ ⋅ ⋅ ⋅ + 𝜇𝑛, and for every 𝑝 with 1 ≤ 𝑝 ≤ 𝑛 − 1, in the first 𝑟 entries of the list, the number of the 𝑝’s is greater than or equal to the number of the (𝑝 + 1)’s.
Syzygies | 189 More generally, a syzygy among some 𝑚-uples of polynomials in 𝐾[𝑥1, . . . , 𝑥𝑛],
𝑃1= ( 𝑃11
...
𝑃1𝑚
) , . . . , 𝑃𝑟 = ( 𝑃𝑟1
...
𝑃𝑟𝑚 ) ,
is a 𝑟-uple of polynomials (𝑄1, . . . , 𝑄𝑟) with 𝑄1, . . . , 𝑄𝑟∈ 𝐾[𝑥1, . . . , 𝑥𝑛] such that
∑
𝑖=1,...,𝑟𝑄𝑖𝑃𝑖= 0.
Given a projective algebraic variety, one can study the syzygies among generators of the ideal of the variety and then the syzygies among these syzygies, and so on. In particular, one can study the degree of such syzygies. A definition which is often used is the following one; it is due to Green and Lazarsfeld (see [88], [89], [170]):
Let 𝑋 be a smooth complex projective algebraic variety of dimension 𝑛 and let 𝐿 be a holomorphic line bundle on 𝑋 defining an embedding 𝜑𝐿 : 𝑋 → ℙ, where ℙ = ℙ(𝐻0(𝑋, 𝐿)∨) (see“Bundles, fibre -”). Set 𝑆 = ⊕𝑑∈ℕ𝑆𝑦𝑚𝑑𝐻0(O(𝐿)), the homoge-neous coordinate ring of the projective space ℙ, and consider the graded 𝑆-module 𝐺 = ⊕𝑑∈ℕ𝐻0(𝑋,O(𝐿𝑑)). Let 𝐸∗
⋅ ⋅ ⋅ → 𝐸1→ 𝐸0→ 𝐺 → 0
be a minimal graded free resolution of 𝐺 (see“Minimal free resolutions”). For any 𝑝 ∈ ℕ, we say that the line bundle 𝐿 satisfies Property 𝑁𝑝if the two following condi-tions hold:
𝐸0= 𝑆,
𝐸𝑖= ⊕𝑆(−𝑖 − 1) for 1 ≤ 𝑖 ≤ 𝑝,
where the second condition means that 𝐸𝑖is the direct sum of some copies of 𝑆(−𝑖−1).
Observe that the kernel of the map 𝑆 → 𝐺 is the ideal of 𝜑𝐿(𝑋) (take the cohomology of the exact sequence 0 →I𝜑𝐿(𝑋)(𝑑) →Oℙ(𝑑) →O𝜑𝐿(𝑋)(𝑑) → 0, whereI𝜑𝐿(𝑋)is the ideal sheaf 𝜑𝐿(𝑋)).
Thus, 𝐿 satisfies Property 𝑁0if and only if 𝐿 is normally generated, 𝐿 satisfies Prop-erty 𝑁1if and only if it satisfies Property 𝑁0 and the ideal of 𝜑𝐿(𝑋) is generated by quadrics, and 𝐿 satisfies Property 𝑁2if and only if it satisfies Property 𝑁1and the syzygies among these quadrics are linear, and so on.
See also“Groebner bases”,“Hilbert syzygy theorem”.
190 | Tautological (or universal) bundle
T
Tautological (or universal) bundle.
([93], [188]).– Let 𝑉 be a vector space of dimension 𝑛 and let 𝑟 < 𝑛. The tautological (or uni-versal bundle) on the Grassmannian of 𝑟-subspaces in 𝑉, 𝐺(𝑟, 𝑉) (see “Grass-mannians”), is the bundle whose fibre on 𝑊 ∈ 𝐺(𝑟, 𝑉) is the 𝑟-subspace 𝑊. In particular the tautological bundle on the projective space ℙ(𝑉) is the line bundle whose fibre on 𝑙 is the line 𝑙; it is the dual of the hyperplane bundle (see “Hyper-plane bundles, twisting sheaves”).
– Let 𝐸 be a vector bundle on a manifold (or an algebraic variety) 𝑋 and let 𝜋 : ℙ(𝐸) → 𝑋 be the projectivized bundle, i.e., the bundle whose fibre on 𝑥 ∈ 𝑋 is the projectivized of 𝐸𝑥. The tautological (or universal) bundle 𝑈 on the projec-tivized bundle ℙ(𝐸) is the following bundle: the subbundle of 𝜋∗𝐸 whose fibre on a point 𝑙 of ℙ(𝐸) is the line represented by 𝑙.
Its dual is the line bundleOℙ(𝐸)(1), i.e., the line bundle whose restriction on ℙ(𝐸𝑥) isO(1) for all 𝑥 ∈ 𝑋.
– More generally: let 𝐸 be a vector bundle on a manifold (or an algebraic variety) 𝑋, and let 𝜋 : 𝐺(𝑟, 𝐸) → 𝑋 be the bundle whose fibre on 𝑥 ∈ 𝑋 is 𝐺(𝑟, 𝐸𝑥); the tau-tological (or universal) bundle 𝑈 on 𝐺(𝑟, 𝐸) is the following bundle on 𝐺(𝑟, 𝐸):
the subbundle of 𝜋∗𝐸 whose fibre on a point 𝑊 of 𝐺(𝑟, 𝐸) is the 𝑟-subspace of 𝐸𝜋(𝑊)= (𝜋∗𝐸)𝑊given by 𝑊:
𝑈 ⊂ 𝜋∗(𝐸)
𝐸
𝐺(𝑟, 𝐸) 𝜋 //𝑋
Obviously taking 𝑋 equal to a point, we get the notion of universal bundle on the Grassmannian and taking 𝑟 = 1 we get the notion of universal bundle on the projectivized of a bundle.
Tor, TOR.
([41], [62], [93], [116]). Let 𝑅 be a commutative ring with unity. Let 𝑀 be an 𝑅-module. Let⋅ ⋅ ⋅ → 𝑃𝑛→ 𝑃𝑛−1→ ⋅ ⋅ ⋅ → 𝑃0→ 𝑀 → 0
be a projective resolution of 𝑀 (see“Injective and projective resolutions”); we denote by 𝑃∗the complex
⋅ ⋅ ⋅ → 𝑃𝑛→ 𝑃𝑛−1→ ⋅ ⋅ ⋅ . → 𝑃0→ 0.
Let 𝑁 be another 𝑅-module. We can consider the complex 𝑃∗⊗𝑅𝑁:
⋅ ⋅ ⋅ → 𝑃𝑛⊗𝑅𝑁 → 𝑃𝑛−1⊗𝑅𝑁 → ⋅ ⋅ ⋅ → 𝑃0⊗𝑅𝑁 → 0.