1  

A note on Veronese varieties

Margherita Barile † † Partially supported by PRIN Algebra Commutativa e Computazionale, Italian Ministry of Education,University and Research. Dipartimento di Matematica, Università di Bari, Via E. Orabona 4, 70125 Bari, Italy e-mail:barile@dm.uniba.it

November 27, 2006

Abstract
We show that for every prime p   , there is a class of Veronese varieties which are set-theoretic complete intersections if and only if the ground field has characteristic p   .
Introduction An affine or projective variety V   is called a set-theoretic complete intersection if it can be defined by the least possible number of equations, i.e., by codim V   equations. This property can depend upon the characteristic of the ground field, but not many examples of this kind have been discovered so far. In [2] the authors describe a class of codimension 2 toric varieties which are set-theoretic complete intersections only in one positive characteristic. In this paper we show that there is a class of Veronese varieties, of arbitrarily high codimension, having the same property. We shall present them by means of their parametrizations as toric varieties, which will enable us to apply the criteria on the associated semigroup developed in [3] . We shall also use cohomological methods to provide lower bounds for the minimum number of defining equations.

1 Preliminary results

Let K   be an algebraically closed field. Let p   be a prime number, h   a positive integer and n ≥ 3   an integer. Consider the following subset of N n   :
T = { ∑ i = 1 n a i e i | ∑ i = 1 n a i = p h , a i ∈ N for all i = 1 , . . . , n } ,   where { e 1 , . . . , e n }   is the standard basis of Z n   . With every ∑ i = 1 n a i e i ∈ T   we can associate the p h   -uple ( i 1 , . . . , i p h ) = ( 1 , . . . , 1 ︸ a 1 t i m e s , 2 , . . . , 2 ︸ a 2 t i m e s , . . . , n , . . . , n ︸ a n t i m e s ) ,   and this defines a bijection between T   and the set
P = { ( i 1 , . . . , i p h ) | 1 ≤ i 1 ≤ ⋯ ≤ i p h ≤ n } . (1)
We know that | T | = ( n + p h − 1 p h )   . In the affine space K | T |   we fix the coordinates x i 1 . . . i p h   , ( i 1 , . . . , i p h ) ∈ P   . With T   we can associate the simplicial affine toric variety V ⊂ K | T |   , defined by the following parametrization:
V = V p h n : { x 1 . . . 1 = u 1 p h x 2 . . . 2 = u 2 p h . . . x n . . . n = u n p h x 11 . . . 112 = u 1 p h − 1 u 2 x 11 . . . 122 = u 1 p h − 2 u 2 2 . . . x 12 . . . 222 = u 1 u 2 p h − 1 x 11 . . . 123 = u 1 p h − 2 u 2 u 3 . . . x n − 1 n . . . n = u n − 1 u n p h − 1   It has codimension N = | T | − n   . In the sequel, for the sake of simplicity, we shall set p h = q   . Note that V   is the affine cone over the projective toric variety of P | T | − 1   having the same parametrization; the latter is called Veronese variety, and was extensively studied by Gröbner [6] .
We consider the polynomial ring R = K [ x i 1 . . . i q | ( i 1 , . . . , i q ) ∈ P ]   , and, in view of ( 1 ), we define the content of the indeterminate x i 1 . . . i q   as the set (with repeated elements):
{ u 1 , . . . , u 1 ︸ a 1 t i m e s , u 2 , . . . , u 2 ︸ a 2 t i m e s , . . . , u n , . . . , u n ︸ a n t i m e s } .   More generally, the content of a product of indeterminates will be the (disjoint) union of the contents of its factors. Let I ( V )   be the defining ideal of V   in R   . Then I ( V )   is generated by binomials. Clearly, given two monomials M , N ∈ R   , we have that M − N ∈ I ( V )   if and only if M   and N   have the same content. We have just established the following
Proposition 1 The binomials in I ( V )   are the (non zero) differences (*) F =   x i 1 . . . i q x i q + 1 . . . i 2 q ⋯ x i ( s − 1 ) q + 1 . . . i s q − x i σ ( 1 ) . . . i σ ( q ) x i σ ( q + 1 ) . . . i σ ( 2 q ) ⋯ x i σ ( ( s − 1 ) q + 1 ) . . . i σ ( s q )   , where s   is a positive integer, and σ   any element of the symmetric group S s q   .
Remark 1 We have that F = 0   if and only if σ ( { i ( k − 1 ) q + 1 , . . . , i k q } ) = { i ( k − 1 ) q + 1 , . . . , i k q }   for some k ∈ { 1 , . . . , s }   . This occurs if and only if F   has a monomial factor.
Example 1 For p = 2   , h = 1   , and n = 3   the variety V   admits the following parametrization V : { x 11 = u 1 2 x 22 = u 2 2 x 33 = u 3 2 x 12 = u 1 u 2 x 13 = u 1 u 3 x 23 = u 2 u 3   The ideal I ( V )   is minimally generated by the following six quadratic binomials:
x 12 2 − x 11 x 22 , x 13 2 − x 11 x 33 , x 23 2 − x 22 x 33 ,   x 12 x 33 − x 13 x 23 , x 13 x 22 − x 12 x 23 , x 23 x 11 − x 12 x 13 .  
In general we have
Theorem 1 I ( V )   is generated by its binomials of degree 2.
Proof .-Let B   the set of binomials of type (*) having degree 2. Let F   be any non zero binomial of type (*), and suppose that σ = τ 1 ⋯ τ m   , where, for all i = 1 , . . . , m   , τ i   is a 2-cycle. In view of Remark 1, we can skip all τ i   which involve two indices belonging to the same indeterminate. Up to a change of indices we may thus assume that τ m = ( 1 q + 1 )   . We prove that F ∈ ( B )   by induction on m ≥ 1   . First suppose that m = 1   . Then σ = ( 1 q + 1 )   , so that
F = x i 1 i 2 . . . i q x i q + 1 i q + 2 . . . i 2 q x i 2 q + 1 . . . i 3 q ⋯ x i ( s − 1 ) q + 1 . . . i s q
− x i q + 1 i 2 . . . i q x i 1 i q + 2 . . . i 2 q x i 2 q + 1 . . . i 3 q ⋯ x i ( s − 1 ) q + 1 . . . i s q
= ( x i 1 i 2 . . . i q x i q + 1 i q + 2 . . . i 2 q − x i q + 1 i 2 . . . i q x i 1 i q + 2 . . . i 2 q ) x i 2 q + 1 . . . i 3 q ⋯ x i ( s − 1 ) q + 1 . . . i s q ,
where the term in brackets belongs to B   . Hence F ∈ ( B )   . Now assume that m > 1   and suppose the claim true for m − 1   . We have:
F = x i 1 i 2 . . . i q x i q + 1 i q + 2 . . . i 2 q x i 2 q + 1 . . . i 3 q ⋯ x i ( s − 1 ) q + 1 . . . i s q
− x i q + 1 i 2 . . . i q x i 1 i q + 2 . . . i 2 q x i 2 q + 1 . . . i 3 q ⋯ x i ( s − 1 ) q + 1 . . . i s q
+ x i q + 1 i 2 . . . i q x i 1 i q + 2 . . . i 2 q x i 2 q + 1 . . . i 3 q ⋯ x i ( s − 1 ) q + 1 . . . i s q
− x i σ ( 1 ) i σ ( 2 ) . . . i σ ( q ) x i σ ( q + 1 ) i σ ( q + 2 ) . . . i σ ( 2 q ) x i σ ( 2 q + 1 ) . . . i σ ( 3 q ) ⋯ x i σ ( ( s − 1 ) q + 1 ) . . . i σ ( s q )
= ( x i 1 i 2 . . . i q x i q + 1 i q + 2 . . . i 2 q − x i q + 1 i 2 . . . i q x i 1 i q + 2 . . . i 2 q ) x i 2 q + 1 . . . i 3 q ⋯ x i ( s − 1 ) q + 1 . . . i s q
+ x i q + 1 i 2 . . . i q x i 1 i q + 2 . . . i 2 q x i 2 q + 1 . . . i 3 q ⋯ x i ( s − 1 ) q + 1 . . . i s q
− x i σ τ m ( q + 1 ) i σ τ m ( 2 ) . . . i σ τ m ( q ) x i σ τ m ( 1 ) i σ τ m ( q + 2 ) . . . i σ τ m ( 2 q )
x i σ τ m ( 2 q + 1 ) . . . i σ τ m ( 3 q ) ⋯ x i σ τ m ( ( s − 1 ) q + 1 ) . . . i σ τ m ( s q ) .
Induction applies to σ τ m = τ 1 ⋯ τ m − 1   , so that F ∈ ( B )   in this case, too.
This completes the proof.
The previous lemma was already shown in [6] , but our approach is simpler and emphasizes the combinatorial aspect. In [6] one can also find a proof of the next result, for which we follow a more direct method.
Let { e i 1 . . . i q | ( i 1 , . . . , i q ) ∈ P }   be the standard basis of Z | T | .   Here we assume that the indices ( i 1 , . . . , i q )   are arranged in the ascending lexicographic order.
Lemma 1 The origin is the only singular point of V   .
Proof .-We prove the claim using the Jacobian criterion. Let
w = ( x ¯ 1 . . . 1 , . . . , x ¯ n . . . n , x ¯ 1 . . . 12 , . . . , x ¯ n − 1 n . . . n )
= ( u ¯ 1 q , . . . , u ¯ n q , u ¯ 1 q − 1 u ¯ 2 , . . . , u ¯ n − 1 u ¯ n q − 1 ) ∈ V ,
and let J ( w )   be a Jacobian matrix associated with the set B   of generators, evaluated at w   . Its rows are ∑ ( i 1 , . . . , i q ) ∈ P ( ∂ F ∂ x i 1 . . . i q ) ( w ) e i 1 . . . i q ,   where F ∈ B   . Clearly J ( 0 )   is the zero matrix, so that 0   is a singular point for V   . Now suppose that w ≠ 0   . Up to a change of indices we may assume that u ¯ 1 ≠ 0   . We prove that rank J ( w ) ≥ N   . Set F i 1 . . . i q = x 1 . . . 1 x i 1 . . . i q − x 1 . . . 1 i q x 1 i 1 . . . i q − 1 ,   for all ( i 1 , . . . , i q ) ∈ P ′ = P \ { ( 1 , . . . , 1 , i ) | i = 1 , . . . , n } .   These are N = | T | − n   elements of B   . Let J ′   be the N × N   -submatrix of J ( w )   formed by the entries ( ∂ F i 1 . . . i q ∂ x j 1 . . . j q ) ( w ) for all ( i 1 , . . . , i q ) , ( j i , . . . , j q ) ∈ P ′ .   The rows of J ′   are x ¯ 1 . . . 1 e i 1 . . . i q − x ¯ 1 . . . 1 i q e 1 i 1 . . . i q − 1 ,   where the second summand is missing whenever i 1 = ⋯ = i q − 2 = 1   .
Otherwise ( 1 , i 1 , . . . , i q − 1 )   is lexicographically smaller than ( i 1 , . . . , i q )   . Hence J ′   is a lower triangular matrix whose diagonal entries are all equal to x ¯ 1 . . . 1 ≠ 0   . Thus J ′   is invertible. It follows that rank J ( w ) ≥ N   , and w   is not a singular point for V   . This completes the proof.
The following result contains the notion of Galois covering: we refer to [8] , p. 43 for a definition.
Lemma 2 Suppose that char K ≠ p   . The map φ : K n \ { 0 } ⟶ V \ { 0 }   ( u 1 , . . . , u n ) ⟶ ( u 1 q , . . . , u n q , u 1 q − 1 u 2 , . . . , u n − 1 u n q − 1 )   is a finite Galois covering with cyclic Galois group of order q   .
Proof .-We show that φ   is an étale finite surjective map and that the multiplicative group C q   of the complex q   -th roots of unity acts transitively and faithfully on every fiber of φ   . Surjectivity and finiteness are clear. The coordinate ring of V   is A = K [ u 1 q , . . . , u n q , u 1 q − 1 u 2 , . . . , u n − 1 u n q − 1 ]   , and the coordinate ring of K n   is K [ u 1 , . . . , u n ] = A [ T 1 , . . . , T n ] / ( T 1 − u 1 , . . . , T n − u n ) .   Set P i = T i − u i   for all i = 1 , . . . , n   . Then the Jacobian matrix ( ∂ P i ∂ T j )   is the identity matrix of order n   . According to [8] , Corollary 3.16, p. 27, this implies that φ   is étale.
Let w = ( u ¯ 1 q , . . . , u ¯ n q , u ¯ 1 q − 1 u ¯ 2 , . . . , u ¯ n − 1 u ¯ n q − 1 ) ∈ V \ { 0 }   . Up to a permutation of indices we may assume that u ¯ 1 = ⋯ = u ¯ k = 0   and u ¯ k + 1 , . . . , u ¯ n ≠ 0   , for some k   , 0 ≤ k ≤ n − 1   . Let v = ( 0 , . . . , 0 , u ¯ k + 1 , . . . , u ¯ n ) ∈ φ − 1 ( w )   .
Then for all v ′ ∈ φ − 1 ( w )   , v ′ = ( 0 , . . . , 0 , g k + 1 u ¯ k + 1 , . . . , g n u ¯ n )   , where, for all i = k + 1 , . . . , n   , g i ∈ C q   . We are going to show that
g i = g j for all indices i , j such that k + 1 ≤ i < j ≤ n . (2)
Let k + 1 ≤ i < j ≤ n   . First assume that p = 2   , and h = 1   . In this case g i ∈ { − 1 , 1 }   for all i = k + 1 , . . . , n   . Equating the entries of φ ( v ′ )   and φ ( v )   of index i j   we have g i u ¯ i g j u ¯ j = u ¯ i u ¯ j   from which we deduce that g i g j = 1 ,   and this implies ( 2 ). Now suppose that p = 2   and h > 1   . Equating the entries of φ ( v ′ )   and φ ( v )   of index i . . . i ︸ 2 h − 1 j . . . j ︸ 2 h − 1   we get g i 2 h − 1 u ¯ i 2 h − 1 g j 2 h − 1 u ¯ j 2 h − 1 = u ¯ i 2 h − 1 u ¯ j 2 h − 1   , whence
g i 2 h − 1 g j 2 h − 1 = 1 , (3)
and equating their entries of index i . . . i ︸ 2 h − 1 − 1 j . . . j ︸ 2 h − 1 + 1   we get, similarly,
g i 2 h − 1 − 1 g j 2 h − 1 + 1 = 1 . (4)
If we divide ( 3 ) by ( 4 ) we obtain g i g j − 1 = 1   , which implies ( 2 ). Finally suppose that p   is odd. Equating the entries of φ ( v ′ )   and φ ( v )   of index i . . . i ︸ q + 1 2 j . . . j ︸ q − 1 2   we deduce that
g i q + 1 2 g j q − 1 2 = 1 , (5)
and equating their entries of index i . . . i ︸ q − 1 2 j . . . j ︸ q + 1 2   we get
g i q − 1 2 g j q + 1 2 = 1 . (6)
From ( 5 ) and ( 6 ) we again deduce ( 2 ). It follows that, in all cases v ′ = g v for some g ∈ C q .   This shows that φ − 1 ( v ) ⊂ { g v | g ∈ C q } ,   The opposite inclusion is obvious. This completes the proof.

2 The set-theoretic complete intersection property

In this section we show that V = V q n   is a set-theoretic complete intersection (i.e., it is set-theoretically defined by N   equations) if and only if char K = p   .
We shall consider the subsemigroup N T   of N n   and the subgroup Z T   of Z n   generated by T   . We need to recall the following two definitions, both quoted from [3] , pp. 1894–1895.
Definition 1 Let p   be a prime number and let T 1   and T 2   be non-empty subsets of T   such that T = T 1 ∪ T 2   and T 1 ∩ T 2 = ∅   . Then T   is called a p-gluing of T 1   and T 2   if there is s ∈ N   and a nonzero element α ∈ N n   such that Z T 1 ∩ Z T 2 = Z α   and p s α ∈ N T 1 ∩ N T 2   .
Definition 2 An affine semigroup N T   is called completely p   -glued if T   is the p   -gluing of T 1   and T 2   , where each of the semigroups N T 1 , N T 2   is completely p   -glued or a free abelian semigroup.
There is a large class of completely p   -glued semigroups:
Lemma 3 Let p   be a prime number, h   a positive integer, and let T 0 = { p h e 1 . . . 1 , . . . , p h e n . . . n } .   If T 0 ⊂ T ⊂ N n   , where T   is finite, then the semigroup N T   is completely p   -glued.
Proof .-We proceed by induction on | T | = t ≥ n   . Since T 0   is free, for t = n   there is nothing to prove. Suppose t > n   and the claim true for all smaller t   . Pick any α ∈ T \ T 0   and set T 1 = T \ { α }   , T 2 = { α }   .
Then, being T 0 ⊂ T 1   , N T 1   is completely p   -glued by induction and, moreover, p h α ∈ N T 0 ⊂ N T 1   , so that p h α ∈ N T 1 ∩ N T 2   . Hence T   is the p   -gluing of T 1   ad T 2   . This proves that N T   is completely p   -glued.
In the same paper (see [3] , Theorem 5) we find the following characterization of set-theoretic complete intersections:
Theorem 2 Suppose that char K = p > 0   . Then a toric variety in over K   is a set-theoretic complete intersection on binomials if and only if the associated semigroup is completely p   -glued.
We shall also need the following criterion, cited from [4] , Lemma 3 ′   .
Lemma 4 Let W ⊂ W ~   be affine varieties. Let d = dim W ~ \ W   . If there are s   equations F 1 , . . . , F s   such that W = W ~ ∩ V ( F 1 , . . . , F s )   , then H e t d + i ( W ~ \ W , Z / r Z ) = 0 for all i ≥ s   and for all r ∈ Z   which are prime to char K   .
Finally we recall one result from group cohomology.
Lemma 5 Let p   be a prime, and h   be a positive integer. Let G   be a cyclic group of order p h   . Then H i ( G , Z / p h Z ) ≠ 0 for all i ≥ 0 ,   with respect to any G   -action on Z / p h Z   .
Proof .-We consider G   as a multiplicative group. Let g   be a generator of G   , and let g [ 1 ] = [ a ]   , where a ∈ Z   and [] denotes the residue class mod p h   . Then the action of g   on A = Z / p h Z   is multiplication by a   . Moreover, g p h = 1   , so that [ 1 ] = g p h [ 1 ] = [ a ] p h   .
Hence a p h ≡ 1   (mod p   ). By Fermat's Little Theorem it follows that a ≡ 1   (mod p   ). Consequently, p h − 1 a ≡ p h − 1   (mod p h   ). Thus g [ p h − 1 ] = [ p h − 1 a ] = [ p h − 1 ] .   This shows that [ p h − 1 ] ∈ A G   , the submodule of g   -invariants of A   . Now, by definition of group cohomology, H 0 ( G ; A ) = A G   , so that H 0 ( G ; A ) ≠ 0 .   By a well-known property (see, e.g., [11] , Proposition 3-2-3), all cohomology groups of a cyclic group with respect to a finite G   -module A   have the same order.
This completes the proof.
We are now ready to prove:
Theorem 3 V   is a set-theoretic complete intersection if and only if char K = p   .
Proof .-Let T   be the subset of N n   associated with V   . Then N T   is completely p   -glued by virtue of Lemma  3 . By Theorem  2 it follows that, if char K = p   , V   is a set-theoretic complete intersection. Now suppose that char K ≠ p   . According to [8] , Theorem 2.20, p. 105, the map φ   in Lemma  2 gives rise to the following Hochschild-Serre spectral sequence:
H i ( Z / q Z , H e t j ( K n \ { 0 } , Z / q Z ) ) ⟹ H e t i + j ( V \ { 0 } , Z / q Z ) ,   where H ⋅   and H e t ⋅   denote group cohomology and étale cohomology respectively.
Recall that
H e t j ( K n \ { 0 } , Z / q Z ) ≃ Z / q Z if j = 0 , 2 n − 1 0 else (7)
and that, according to Lemma  5 ,
H i ( Z / q Z , Z / q Z ) ≠ 0 for all i ≥ 0 . (8)
Since 2 n − 1 > 2   , ( 7 ) and ( 8 ) imply that
H e t n + 1 ( V \ { 0 } , Z / q Z ) ≃ H n + 1 ( Z / q Z , H e t 0 ( K n \ { 0 } , Z / q Z ) ) ≠ 0 . (9)
Since, by Lemma  1 , V \ { 0 }   is a smooth n   -dimensional variety, we can apply Poincaré Duality to it (see [8] , Theorem 1.11, p. 276), so that
H c n − 1 ( V \ { 0 } , Z / q Z ) ≃ H e t n + 1 ( V \ { 0 } , Z / q Z ) , (10)
where H c ⋅   denotes cohomology with compact support. In the sequel we shall omit the coefficient group Z / q Z   for the sake of simplicity. From ( 9 ) and ( 10 ) we conclude that
H c n − 1 ( V \ { 0 } ) ≠ 0 . (11)
We have a long exact sequence of cohomology with compact support ⋯ ⟶ H c n − 2 ( 0 ) ⟶ H c n − 1 ( V \ { 0 } ) ⟶ H c n − 1 ( V ) ⟶ H c n − 1 ( 0 ) ⟶ ⋯ ,   where, being n ≥ 3   , H c n − 2 ( 0 ) = H c n − 1 ( 0 ) = 0   . Hence H c n − 1 ( V ) ≃ H c n − 1 ( V \ { 0 } ) ,   so that, by ( 11 ),
H c n − 1 ( V ) ≠ 0 . (12)
We also have a long exact sequence ⋯ ⟶ H c n − 1 ( K N + n ) ⟶ H c n − 1 ( V ) ⟶ H c n ( K N + n \ V ) ⟶ H c n ( K N + n ) ⟶ ⋯ ,   where H c n − 1 ( K N + n ) = H c n ( K N + n ) = 0   , so that H c n ( K N + n \ V ) ≃ H c n − 1 ( V )   , and thus, in view of ( 12 ), H c n ( K N + n \ V ) ≠ 0 .   If we apply Poincaré Duality, we finally obtain H e t N + n + N ( K N + n \ V ) ≠ 0 .   By Lemma  4 it follows that, under our present assumption that char K ≠ p   , V   is not set-theoretically defined by N   equations, i.e. it is not a set-theoretic complete intersection. This completes the proof.
According to the methods developed in [3] , V q n   is, in characteristic p   , a set-theoretic complete intersection on the following set of N   binomials x i 1 . . . i q q − x 1 . . . 1 i 1 ⋯ x q . . . q i q , ( i 1 , . . . , i p ) ∈ P \ { ( 1 . . . 1 ) , . . . , ( n . . . n ) } .   The same set of N   binomials was found by Gattazzo [5] by direct computations.
Remark 2 The variety V 2 n ( n + 1 ) 2   (of codimension n ( n − 1 ) 2   ) is the determinantal variety defined by the vanishing of the 2-minors of an n × n   symmetric variety of indeterminates over K   . Theorem  2 generalizes part of the results in [1] : there it was shown that the minimum number of equations required to define V 2 n ( n + 1 ) 2   set-theoretically is n ( n + 1 ) 2 − 2 if char K ≠ 2 n ( n − 1 ) 2 if char K = 2   which means that, for n ≥ 3   , V 2 n ( n + 1 ) 2   is a set-theoretic complete intersection if and only if char K = 2   .
Remark 3 A general lower bound for the minimum number of equations which define a variety set-theoretically (the so-called arithmetical rank, ara) is given by the local cohomological dimension: if I = I ( V )   , this number is c d I = max { n ∈ N | H I n ( R ) ≠ 0 } ,   where H I ⋅   denotes local cohomology with respect to I   . In the case of the Veronese variety V = V p h n   , the ideal I   is perfect (see [6] , p. 259), so that, according to [9] , Prop. 4.1, c d I = h t I = N .   However, it follows from Theorem  3 that N = c d I < a r a I if char K ≠ p ,   so that the lower bound for the arithmetical rank provided by the local cohomological dimension is almost always non-sharp. Varieties where the local cohomological dimension and the arithmetical rank differ are particularly sought. The only previously known examples are the Pfaffian ideals of an alternating matrix of indeterminates, in all positive characteristics (see [1] , Remarks 6.2), and Reisner's variety, in all characteristics different from 2 (see [10] , p. 250, [7] , Example 1, and [12] , Example 2).
References

  1. Barile, M.: Arithmetical ranks of ideals associated to symmetric and alternating matrices. J. Algebra, 176, 59–82 (1995)
  2. Barile, M., Lyubeznik, G.: Set-theoretic complete intersections in characteristic p   . Preprint (2004). To appear in: Proceedings of the American Mathematical Society.
  3. Barile, M., Morales, M., Thoma, A.: Set-Theoretic Complete Intersections on Binomials. Proc. Amer. Soc., 130, 1893–1903 (2002)
  4. Bruns, W., Schwänzl: The number of equations defining a determinantal variety. Bull. London Math. Soc., 22, 439–445 (1990)
  5. Gattazzo, R.: In characteristic p = 2   the Veronese variety V m ⊂ P m ( m + 3 ) / 2   and each of its generic projection is set-theoretic complete intersection, in: Greco, S., Strano, R. (eds.), Complete Intersections, Acireale 1983, Lecture Notes in Mathematics 1092, 221–228, Springer, Berlin-Heidelberg (1984)
  6. Gröbner, W.: Über Veronesesche Varietäten und deren Projektionen. Arch. Math., 16, 257–264 (1965)
  7. Lyubeznik, G.: On the local cohomology modules H A i ( R )   for ideals A   generated by monomials in an R   -sequence, in: Greco, S., Strano, R. (eds.), Complete Intersections, Acireale 1983, Lecture Notes in Mathematics 1092, 221–228, Springer, Berlin-Heidelberg (1984)
  8. Milne, J.: Étale Cohomology. Princeton University Press, Princeton (1980)
  9. Peskine, C., Szpiro, L.: Dimension finie et cohomologie locale. Inst. Hautes Études Sci. Publ. Math. 42, 47–119 (1973)
  10. Schmitt, Th., Vogel, W.: Note on Set-Theoretic Intersections of Subvarieties of Projective Space. Math. Ann. 245, 247–253 (1979)
  11. Weiss, E.: Cohomology of Groups. Academic Press, New York, London (1969)
  12. Yan, Z.: An étale analog of the Goresky-MacPherson formula for subspace arrangements. J. Pure Appl. Algebra 146, 305–318 (2000)

1   MSC 2000: 14M25, 14M10, 14M12, 14F20, 20J06