Veronese Varieties over Fields with non-zero Characteristic: A - - PDF document

veronese varieties over fields with non zero
SMART_READER_LITE
LIVE PREVIEW

Veronese Varieties over Fields with non-zero Characteristic: A - - PDF document

Veronese Varieties over Fields with non-zero Characteristic: A Survey Hans Havlicek Vienna University of Technology Combinatorics 2000 Introduction If F is a (commutative) field of characteristic 2 then all tangent lines of a conic in PG(2 , F


slide-1
SLIDE 1

Veronese Varieties over Fields with non-zero Characteristic: A Survey

Hans Havlicek Vienna University of Technology

Combinatorics 2000

slide-2
SLIDE 2

Introduction

If F is a (commutative) field of characteristic 2 then all tangent lines of a conic in PG(2, F) are concurrent at a point called nucleus. What happens in higher dimensions?

  • Normal rational curves
  • Veronese varieties

J.A. Thas, 1969. H. Timmermann, 1977, 1978.

  • A. Herzer, 1982. H. Karzel, 1987.

Combinatorics 2000 1

slide-3
SLIDE 3

Part 1 Pascal’s triangle modulo a prime

Combinatorics 2000 2

slide-4
SLIDE 4

Representations in base p

Let p be a fixed prime. The representation of a non- negative integer n ∈ N := {0, 1, 2, . . . } in base p has the form n =

  • σ=0

nσpσ =: nσ with only finitely many digits nσ ∈ {0, 1, . . . , p − 1} different from 0.

Combinatorics 2000 3

slide-5
SLIDE 5

A theorem of Lucas

Let nσ and jσ be the representations of non- negative integers n and j in base p. Then n j

  • σ=0

nσ jσ

  • (mod p).

Combinatorics 2000 4

slide-6
SLIDE 6

Pascal’s triangle modulo 3

Combinatorics 2000 5

slide-7
SLIDE 7

Pascal’s triangle modulo p

∆ denotes Pascal’s triangle modulo p and ∆i is the subtriangle of ∆ that is formed by the rows 0, 1, . . . , pi − 1. Each triangle ∆i+1 (i ≥ 0) has the following form, with products taken modulo p:

  • ∆i

1

  • ∆i ∇i 1

1

  • ∆i

2

  • ∆i ∇i 2

1

  • ∆i ∇i 2

2

  • ∆i

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . p−1

  • ∆i ∇i

. . . ∇i p−1

p−1

  • ∆i

Here the ∇i’s are subtriangles with all entries equal to zero.

  • E. Hexel and H. Sachs, 1978.

C.T. Long, 1981. N.A. Volodin, 1994.

Combinatorics 2000 6

slide-8
SLIDE 8

A partition of zero entries

The zero entries of Pascal’s triangle modulo p fall into (disjoint) maximal subtriangles ∇i (i ∈ N+). We get a partition of all zero entries of ∆ by gluing together all triangles ∇i of same size to one class i, say. Example. Pascal’s triangle modulo 2

Combinatorics 2000 7

slide-9
SLIDE 9

Counting zero entries

Then the number of entries in row n of ∆ belonging to class i equals Φ(i, n) :=

  • pi − 1 −

i−1

  • µ=0

nµpµ · ni ·

  • σ=i+1

(nσ + 1). The number of entries in row n of ∆ belonging to classes i, (i + 1), . . . is Σ(i, n) :=

  • η=i

Φ(η, n) = n + 1 −

  • 1 +

i−1

  • µ=0

nµpµ ∞

  • σ=i

(nσ + 1). N.J. Fine, 1947. J. Gmainer, 1999.

Combinatorics 2000 8

slide-10
SLIDE 10

The top line function T(R, b)

Given b ∈ N+ and R ∈ N then let T(R, b) :=

  • σ=R

bσpσ. This function has the following property: If (n, j) ∈ i and b := n + 1 then T(i, b) gives the top line of the triangle ∇i containing the (n, j)-entry of ∆, i.e. 0 ≡ n

j

n−1

j

  • ≡ . . . ≡

T (i,b)

j

  • (mod p),

0 ≡ T (i,b)−1

j

  • (mod p).

Combinatorics 2000 9

slide-11
SLIDE 11

Part 2 Nuclei of normal rational curves

Combinatorics 2000 10

slide-12
SLIDE 12

Normal rational curves

In terms of (adequately chosen) coordinates and an inhomogeneous parameter a normal rational curve Vn

1

in PG(n, F) is the point set {F(1, x, . . . , xn) | x ∈ F ∪ {∞}}. Example. Twisted cubic P0 P1 P2 P3

Combinatorics 2000 11

slide-13
SLIDE 13

Osculating subspaces

If we fix one u ∈ F then columns of the regular matrix          

  • . . .

1

  • u

1

1

  • . . .

2

  • u2

2

1

  • u

2

2

  • . . .

. . . ... . . . n

  • un

n

1

  • un−1

n

2

  • un−2

. . . n

n

         give, respectively, a point of the NRC and its derivative points. The k-osculating subspace (k ∈ {−1, 0, . . . , n−1}) of Vn

1 at the given point is the k-dimensional projective

subspace spanned by the first k + 1 columns of the matrix.

Combinatorics 2000 12

slide-14
SLIDE 14

Nuclei

The k-nucleus N (k)Vn

1

(k ∈ {−1, 0, . . . , n − 1})

  • f a normal rational curve Vn

1 is the intersection of all

its k-osculating subspaces.

Combinatorics 2000 13

slide-15
SLIDE 15

The main theorem

Theorem 1. If #F ≥ k + 1, then the k-nucleus N (k)Vn

1 is spanned by those base points Pj, where

j ∈ {0, 1, . . . , n} is subject to n j

n − 1 j

  • ≡ . . . ≡

k + 1 j

  • ≡ 0 (mod char F).

Combinatorics 2000 14

slide-16
SLIDE 16

Example

Let n = 14 and p = 2, whence n + 1 = b = 15.

1 1 1 1 0 1 1 1 1 1 1 0 0 0 1 1 1 0 0 1 1 1 0 1 0 1 0 1 1 1 1 1 1 1 1 1 T (3, 15) = 1, 0, 0, 0 → 1 0 0 0 0 0 0 0 1 1 1 0 0 0 0 0 0 1 1 1 0 1 0 0 0 0 0 1 0 1 1 1 1 1 0 0 0 0 1 1 1 1 T (2, 15) = 1, 1, 0, 0 → 1 0 0 0 1 0 0 0 1 0 0 0 1 1 1 0 0 1 1 0 0 1 1 0 0 1 1 T (1, 15) = 1, 1, 1, 0 → 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 T (0, 15) = 1, 1, 1, 1 → k −1, 0, . . . , 6 7, 8, 9, 10 11, 12 13 dim −1 2 6

Combinatorics 2000 15

slide-17
SLIDE 17

A dimension formula

Theorem 2. If #F ≥ k + 1, char F = p > 0, and T(R, b) =

  • µ=R

bµpµ ≤ k + 1 <

  • σ=Q

bσpσ = T(Q, b) with at most one bσ = 0 for σ ∈ {Q, Q+1, . . . , R−1}, then the k-nucleus of Vn

1 has dimension

n −

  • 1 +

R−1

  • µ=0

nµpµ

  • σ=R

(nσ + 1) = Σ(R, n) − 1.

  • H. Timmermann, 1978. J. Gmainer, 1999.

Combinatorics 2000 16

slide-18
SLIDE 18

Number of nuclei

Theorem 3. If #F ≥ n, then there are as many distinct nuclei of Vn

1

as non-zero digits in the representation of b = n + 1 in base p.

  • J. Gmainer, 1999.

Example. p = 2, n ∈ {2, 3, . . . , 31}

2 4 6 8 10 12 14 16 18 20 5 10 15 20 25 30

Combinatorics 2000 17

slide-19
SLIDE 19

Other topics

  • Geometric meaning of nuclei
  • One-point nuclei
  • Lattice of invariant subspaces
  • Projection of a NRC from a nucleus or an invariant

subspace

Combinatorics 2000 18

slide-20
SLIDE 20

Part 3 Pascal’s simplex modulo a prime

Combinatorics 2000 19

slide-21
SLIDE 21

Multinomial coefficients

Let t, e0, e1, . . . , em ∈ N. If (e0, e1, . . . , em) ∈ Et

m, i.e.,

e0 + e1 + . . . + em = t then

  • t

e0, e1, . . . , em

  • :=

t! e0! e1! · · · em!

  • therwise
  • t

e0, e1, . . . , em

  • := 0.

Combinatorics 2000 20

slide-22
SLIDE 22

Lucas revisited

If p is a prime then, in terms of digits in base p,

  • t

e0, . . . , em

  • σ∈N

e0,σ, . . . , em,σ

  • (mod p).

Combinatorics 2000 21

slide-23
SLIDE 23

Pascal’s pyramid modulo 2

  • P. Hilton and J. Pedersen, 1999. H. Walser.

Combinatorics 2000 22

slide-24
SLIDE 24

Counting zero entries

Let p be a prime. The number of (m + 1)–tuples (e0, e1, . . . , em) ∈ Et

m

such that

  • t

e0, e1, . . . , em

  • ≡ 0 (mod p)

equals m + t t

  • σ∈N

m + tσ tσ

  • .

F.T. Howard, 1974. N.A. Volodin, 1989.

Combinatorics 2000 23

slide-25
SLIDE 25

Part 4 Nuclei of Veronese varieties

Combinatorics 2000 24

slide-26
SLIDE 26

Veronese varieties

In terms of (adequately chosen) coordinates the Veronese mapping is given by F(x0, x1, . . . , xm) → F(. . . , xe0

0 xe1 1 . . . xem m , . . . )

where xi ∈ F and (e0, e1, . . . , em) ∈ Et

m.

Its image is a Veronese variety Vt

  • m. Its ambient space

has dimension m + t t

  • − 1.

(By putting m := 1 and n := t a NRC Vn

1 is obtained.)

Combinatorics 2000 25

slide-27
SLIDE 27

Osculating subspaces

The Veronese image of each r-dimensional subspace of the parameter space (0 ≤ r < m) is a sub-Veronesean Vt

r of Vt m.

There exists a k-osculating subspace of Vt

m along Vt r

for each k ∈ {−1, 0, . . . , t − 1}. We call it an (r, k)-

  • sculating subspace of Vt
  • m. Its dimension equals

t

  • i=t−k

r + i i m + t − r − i − 1 t − i

  • − 1.

In particular, each (t−1, m−1)-osculating subspace of a Veronese variety Vt

m is a hyperplane of the ambient

space; it is called an osculating hyperplane or a contact hyperplane.

Combinatorics 2000 26

slide-28
SLIDE 28

Nuclei

The (r, k)-nucleus of a Veronese variety Vt

m is the

intersection of all its (r, k)-osculating subspaces.

Combinatorics 2000 27

slide-29
SLIDE 29

Intersection of osculating hyperplanes

Theorem 4. If #F ≥ t then the (m − 1, t − 1)- nucleus of a Veronese variety Vt

m is spanned by those

base points Pe0,e1,... ,em satisfying

  • t

e0, e1, . . . , em

  • ≡ 0

mod char F. Theorem 5. Let

σ∈N tσpσ be the representation of

t in base p = char F > 0. If #F ≥ t, then the (m−1, t−1)-nucleus of a Veronese variety Vt

m has dimension

m + t t

  • σ∈N

m + tσ tσ

  • − 1.
  • J. Gmainer, H. H., 2000.

Combinatorics 2000 28

slide-30
SLIDE 30

Remarks and open problems

  • Connection to symmetric powers
  • Coordinate-free definitions
  • Nuclei of sub-Veroneseans
  • A general dimension formula for nuclei ?
  • Geometric meaning of nuclei ?
  • Invariant subspaces?

Combinatorics 2000 29

slide-31
SLIDE 31

References

  • E. Bertini, 1907, 1924.
  • H. Brauner, 1976.

A.E. Brouwer and H.A. Wilbrink, 1995.

  • W. Burau, 1961, 1974.

N.J. Fine, 1947. D.G. Glynn, 1986.

  • J. Gmainer, 1999, 2000.

V.D. Goppa, 1988.

  • H. Harborth, 1975.
  • H. Hasse, 1937.
  • A. Herzer, 1982.
  • E. Hexel and H. Sachs, 1978.

Combinatorics 2000 30

slide-32
SLIDE 32
  • P. Hilton and J. Pedersen, 1999.

J.W.P. Hirschfeld, 1985, 1998. J.W.P. Hirschfeld and L. Storme, 1998. J.W.P. Hirschfeld and J.A. Thas, 1991. F.T. Howard, 1974. V.V. Karachik, 1996.

  • H. Karzel, 1987.

C.T. Long, 1981.

  • N. Melone, 1983.
  • R. Riesinger, 1981.

J.B. Roberts, 1957.

  • L. Storme and J.A. Thas, 1994.

J.A. Thas, 1969, 1992.

  • H. Timmermann, 1977, 1978.

Combinatorics 2000 31

slide-33
SLIDE 33

N.A. Volodin, 1989, 1994.

  • S. Wolfram, 1984.
  • C. Zanella, 1997, 1998.
  • J. Zeuge, 1972, 1977.

Combinatorics 2000 32