Computing Node Polynomials for Plane Curves Florian Block - - PowerPoint PPT Presentation

computing node polynomials for plane curves
SMART_READER_LITE
LIVE PREVIEW

Computing Node Polynomials for Plane Curves Florian Block - - PowerPoint PPT Presentation

Computing Node Polynomials for Plane Curves Florian Block University of Michigan FPSAC 2010 August 5, 2010 http://www-personal.umich.edu/ blockf/NodePolyTalk.pdf arXiv:1006.0218 Florian Block (U of Michigan) Computing Node Polynomials


slide-1
SLIDE 1

Computing Node Polynomials for Plane Curves

Florian Block University of Michigan FPSAC 2010 August 5, 2010 http://www-personal.umich.edu/∼blockf/NodePolyTalk.pdf arXiv:1006.0218

Florian Block (U of Michigan) Computing Node Polynomials August 5, 2010 1 / 14

slide-2
SLIDE 2

Combinatorial Rules in Enumerative Geometry

Enumerative Geometry: counting geometric objects with certain properties. Strategy: reduce the problem to enumeration of combinatorial gadgets. Example: the Littlewood - Richardson - Rule. Why are such rules needed? This talk: Enumeration of plane curves via (marked) floor diagrams. #                       = #

✉ ✉ ✉ ✉ ✉ ✉ ✉ ❡ ❡ ❡ ❡ ❡ ❡ ❡ ❡ ✲ ✲ ✲ ✲

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

  • Florian Block (U of Michigan)

Computing Node Polynomials August 5, 2010 2 / 14

slide-3
SLIDE 3

Counting plane curves

Question

How many (possibly reducible) nodal algebraic curves in CP2 have degree d, δ nodes, and pass through (d+3)d

2

− δ generic points? Number of such curves is the Severi degree Nd,δ. N1,0 = #{lines through 2 points} = 1. N2,1 = #{1-nodal conics through 4 points} = 3 N4,4 = #{4-nodal quartics through 10 points} = 666. If d ≥ δ + 2, a curve is irreducible by B´ ezout’s Theorem, so Nd,δ = Gromov-Witten invariant Nd,g, with g = (d−1)(d−2)

2

− δ.

Florian Block (U of Michigan) Computing Node Polynomials August 5, 2010 3 / 14

slide-4
SLIDE 4

Node Polynomials

Theorem (Fomin–Mikhalkin, 2009)

For a fixed δ, we have Nd,δ = Nδ(d), for a combinatorially defined polynomial Nδ(d) ∈ Q[d], provided d ≥ 2δ. Polynomiality was conjectured by P. Di Francesco–C. Itzykson (1994) and by L. G¨

  • ttsche (1997).

The Nδ(d) are called node polynomials. We have deg(Nδ(d)) = 2δ.

Florian Block (U of Michigan) Computing Node Polynomials August 5, 2010 4 / 14

slide-5
SLIDE 5

Chronology of Node Polynomials

  • J. Steiner (1848):

Nd,1 = 3(d − 1)2 (d ≥ 1)

  • A. Cayley (1863):

Nd,2 = 3

2(d − 1)(d − 2)(3d2 − 3d − 11)

(d ≥ 1)

  • S. Roberts (1867):

Nd,3 = 9

2d6 − 27d5 + 9 2d4 + 423 2 d3 − 229d2 − 829 2 d + 525

(d ≥ 3)

  • I. Vainsencher (1995):

δ = 4, 5, 6

  • S. Kleiman–R. Piene (2001):

δ = 7, 8

Florian Block (U of Michigan) Computing Node Polynomials August 5, 2010 5 / 14

slide-6
SLIDE 6

New Node Polynomials

Theorem (B.)

The node polynomials Nδ(d), for δ = 9, 10, 11, 12, 13, 14, are given by . . . .

Florian Block (U of Michigan) Computing Node Polynomials August 5, 2010 6 / 14

slide-7
SLIDE 7

New Node Polynomials

Theorem (B.)

The node polynomials Nδ(d), for δ = 9, 10, 11, 12, 13, 14, are given by . . . .

N14(d) = 19683 358758400 d28 − 19683 12812800 d27 − 6561 2562560 d26 + 1751787 3942400 d25 − 4529277 1971200 d24 − 562059 9856 d23 + 398785599 788480 d22 + 5214288411 1254400 d21 − 4860008991 89600 d20 − 63174295089 358400 d19 + 332872084467 89600 d18 + 3103879378581 985600 d17 − 4913807521304691 27596800 d16 + 899178800016807 8968960 d15 + 279086438050359453 44844800 d14 − 468967272863997483 51251200 d13 − 318443311640108577 1971200 d12 + 328351365725506869 985600 d11 + 1120586814080571923 358400 d10 − 9448861028448843949 1254400 d9 − 30880785216736406143 689920 d8 + 444525313669622586903 3942400 d7 + 11429038221675466251 24640 d6 − 269709254062572016617 246400 d5 − 74660630664748878665353 22422400 d4 + 140531359469510983018159 22422400 d3 + 16863931195154225977601 1121120 d2 − 64314454486825349085 4004 d − 32644422296329680.

Based on an algorithm of S. Fomin–G. Mikhalkin, with improvements. Maple calculation of N14(d) took 70 days.

Florian Block (U of Michigan) Computing Node Polynomials August 5, 2010 6 / 14

slide-8
SLIDE 8

Polynomiality Threshold

Define d∗(δ) = smallest d∗ such that d ≥ d∗(δ) implies Nd,δ = Nδ(d).

  • S. Fomin–G. Mikhalkin: d∗(δ) ≤ 2δ for δ ≥ 1.

Theorem (B.)

For δ ≥ 1, we have d∗(δ) ≤ δ.

  • L. G¨
  • ttsche (1997) conjectured: d∗(δ) ≤ ⌈ δ

2⌉ + 1.

Theorem (B.)

For 3 ≤ δ ≤ 14, we have d∗(δ) = ⌈ δ

2⌉ + 1.

For δ ≤ 8, this was established by S. Kleiman–R. Piene (2001).

Florian Block (U of Michigan) Computing Node Polynomials August 5, 2010 7 / 14

slide-9
SLIDE 9

Leading terms of the node polynomials

Theorem (B.)

For any δ, the nine leading terms of Nδ(d) are:

Nδ(d) = 3δ δ! » d2δ − 2δd2δ−1 − δ(δ − 4) 3 d2δ−2 + δ(δ − 1)(20δ − 13) 6 d2δ−3+ − δ(δ − 1)(69δ2 − 85δ + 92) 54 d2δ−4 − δ(δ − 1)(δ − 2)(702δ2 − 629δ − 286) 270 d2δ−5+ + δ(δ − 1)(δ − 2)(6028δ3 − 15476δ2 + 11701δ + 4425) 3240 d2δ−6+ + δ(δ − 1)(δ − 2)(δ − 3)(13628δ3 − 6089δ2 − 29572δ − 24485) 11340 d2δ−7+ − δ(δ − 1)(δ − 2)(δ − 3)(282855δ4 − 931146δ3 + 417490δ2 + 425202δ + 1141616) 204120 d2δ−8 + · · · # .

The first 7 terms were conjectured by P. Di Francesco–C. Itzykson (1994).

Florian Block (U of Michigan) Computing Node Polynomials August 5, 2010 8 / 14

slide-10
SLIDE 10

Main Technique: (Marked) Floor Diagrams

Mikhalkin’s Correspondence Theorem (2005) replaces enumeration of algebraic plane curves by weighted enumeration of tropical plane curves.

  • E. Brugall´

e–S. Fomin–G. Mikhalkin (2007, 2009) reduced the latter to enumeration of combinatorial gadgets called (marked) floor diagrams. #                       = #                                               = # {

✉ ✉ ✉ ✲ ✲ ✲ ✲ ❡ ❡ ❡ ❡ ❡

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

}

Florian Block (U of Michigan) Computing Node Polynomials August 5, 2010 9 / 14

slide-11
SLIDE 11

Main Technique: (Marked) Floor Diagrams

Mikhalkin’s Correspondence Theorem (2005) replaces enumeration of algebraic plane curves by weighted enumeration of tropical plane curves.

  • E. Brugall´

e–S. Fomin–G. Mikhalkin (2007, 2009) reduced the latter to enumeration of combinatorial gadgets called (marked) floor diagrams. #                       = #                                               = #                       

✉ ✉ ✉ ❡ ❡ ❡ ✲ ✲ ✲ ✲ ✉ ❡ ✉ ❡ ✉ ❡ ✉ ❡ ✉ ❡

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

                      

Florian Block (U of Michigan) Computing Node Polynomials August 5, 2010 10 / 14

slide-12
SLIDE 12

Definition of Marked Floor Diagrams

Marked floor diagram: a directed, weighted graph on linearly ordered, bipartite vertex set ({ ✇

, . . . , ✇

}, { ✇

, . . . , ✇

,

✇ ❣

, . . . , ✇

}) with: all edges go from smaller to larger vertices, all ✇

’s are sinks, each ✇

❣has one incoming and one outgoing edge,

edge weights are positive integers, div( ✇

) = 1, div( ✇

) = 0, div( ✇

) = −1. 2 2

✇ ✇ ✇ ✇ ✇ ✇ ✇ ✇ ✇ ✇ ✇ ❣ ❣ ❣ ❣ ❣ ❣ ❣ ❣ ❣ ❣ ❣ ❣ ❣ ❣ ❣ ✲ ✲ ✲ ✲

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

✲ ✲

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

✲ ✲

.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Florian Block (U of Michigan) Computing Node Polynomials August 5, 2010 11 / 14

slide-13
SLIDE 13

Combinatorial Correspondence Theorem

Theorem (Brugall´ e–Mikhalkin 2007)

Nd,δ =

  • M

wt(M), the sum over all marked floor diagrams M with d red vertices, and with

d(d−3) 2

+ b0(M) − b1(M) = δ. Here wt(M) = product of all edge weights of M. b0(M) = number of connected components of M. b1(M) = sum of the genera of the connected components of M.

Florian Block (U of Michigan) Computing Node Polynomials August 5, 2010 12 / 14

slide-14
SLIDE 14

Template Decomposition

First step: Decomposition of marked floor diagrams into templates.

✇ ❣ ✇ ❣ ✇ ❣ ✇ ❣ ✇ ❣ ✇ ❣ ✇ ❣ ✇ ❣ ✇ ❣ ✇ ❣ ✇ ❣ ✇ ❣ ✇ ❣ ✇ ❣ ✇ ❣ ✲ ✲

2 2

✲ ✲ ✲

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

✲ ✲ ✲

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

3 3

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

✇ ❣ ✇ ❣ ✇ ❣ ✇ ❣ ✇ ❣ ✇ ❣

2✲ 3✲

Computation of N14(d) required enumeration of 220006996 templates.

Florian Block (U of Michigan) Computing Node Polynomials August 5, 2010 13 / 14

slide-15
SLIDE 15

References and Work in Progress

  • F. Block, Computing Node Polynomials for Plane Curves,

arXiv:1006.0218. Relative case (several tangency conditions to one line):

  • F. Block, Relative Node Polynomials, in preparation.

Floor diagrams for “Psi-classes” (many lines, one tangency each):

  • F. Block, A. Gathmann, and H. Markwig, Psi-floor diagrams and a

Caporaso-Harris type recursion, arXiv:1003.2067. Other toric surfaces:

  • F. Ardila and F. Block, Curve Enumeration in Toric Surfaces and

Discrete Measures of Polytopes, in preparation. Real curves.

Florian Block (U of Michigan) Computing Node Polynomials August 5, 2010 14 / 14

slide-16
SLIDE 16

References and Work in Progress

  • F. Block, Computing Node Polynomials for Plane Curves,

arXiv:1006.0218. Relative case (several tangency conditions to one line):

  • F. Block, Relative Node Polynomials, in preparation.

Floor diagrams for “Psi-classes” (many lines, one tangency each):

  • F. Block, A. Gathmann, and H. Markwig, Psi-floor diagrams and a

Caporaso-Harris type recursion, arXiv:1003.2067. Other toric surfaces:

  • F. Ardila and F. Block, Curve Enumeration in Toric Surfaces and

Discrete Measures of Polytopes, in preparation. Real curves. Thank you!

Florian Block (U of Michigan) Computing Node Polynomials August 5, 2010 14 / 14