Geometry of Discrete Painlev e Equations and Applications Anton - - PowerPoint PPT Presentation

geometry of discrete painlev e equations and applications
SMART_READER_LITE
LIVE PREVIEW

Geometry of Discrete Painlev e Equations and Applications Anton - - PowerPoint PPT Presentation

Geometry of Discrete Painlev e Equations and Applications Anton Dzhamay School of Mathematical Sciences, University of Northern Colorado, Greeley, CO Based on the joint work with Tomoyuki Takenawa Tokyo University of Marine Science and


slide-1
SLIDE 1

Geometry of Discrete Painlev´ e Equations and Applications Anton Dzhamay

School of Mathematical Sciences, University of Northern Colorado, Greeley, CO Based on the joint work with Tomoyuki Takenawa Tokyo University of Marine Science and Technology Painlev´ e Equations and Applications: A Workshop in Memory of A. A. Kapaev Michigan Center for Applied Interdisciplinary Mathematics (MCAIM) The University of Michigan Ann Arbor, MI, August 25–29, 2017

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 1 / 41

slide-2
SLIDE 2

Prologue

Are the following two equations: Same? Different? Equivalent? Related?

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 2 / 41

slide-3
SLIDE 3

Prologue

Are the following two equations: Same? Different? Equivalent? Related?          ¯ x = (α − β)(αx(θ1

1 − θ2 1) + (1 + θ2 0)(x(y − θ2 1) + y(θ1 0 − θ2 0)))

(α − β)(x(y − θ2

1) + (θ1 0 − θ2 0)y) − α(θ1 1 + 1)(θ1 0 − θ2 0)

¯ y = (α − β)(y(x + θ1

0 − θ2 0) − θ2 1x)

α(θ1

0 − θ2 0)

, (1) where θj

i and κi are some parameters and

α(x, y) =

  • yr1 + x(θ2

0r1+r2)

x+θ1

0−θ2

  • (x + y)(θ1

1 − θ2 1) ,

β(x, y) =

  • (y + θ2

0)r1 + r2

  • (x + y)(θ1

1 − θ2 1) ,

r1(x, y) = κ1κ2 + κ2κ3 + κ3κ1 − (y − θ2

1)(x − θ2 0) − θ1 0(y + θ2 0) − θ1 1(θ1 0 + θ2 0 + θ2 1),

r2(x, y) = κ1κ2κ3 + θ1

1((y − θ2 1)(x − θ2 0) + θ1 0(y + θ2 0)).

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 2 / 41

slide-4
SLIDE 4

Prologue

Are the following two equations: Same? Different? Equivalent? Related?          ¯ x = (α − β)(αx(θ1

1 − θ2 1) + (1 + θ2 0)(x(y − θ2 1) + y(θ1 0 − θ2 0)))

(α − β)(x(y − θ2

1) + (θ1 0 − θ2 0)y) − α(θ1 1 + 1)(θ1 0 − θ2 0)

¯ y = (α − β)(y(x + θ1

0 − θ2 0) − θ2 1x)

α(θ1

0 − θ2 0)

, (1) where θj

i and κi are some parameters and

α(x, y) =

  • yr1 + x(θ2

0r1+r2)

x+θ1

0−θ2

  • (x + y)(θ1

1 − θ2 1) ,

β(x, y) =

  • (y + θ2

0)r1 + r2

  • (x + y)(θ1

1 − θ2 1) ,

r1(x, y) = κ1κ2 + κ2κ3 + κ3κ1 − (y − θ2

1)(x − θ2 0) − θ1 0(y + θ2 0) − θ1 1(θ1 0 + θ2 0 + θ2 1),

r2(x, y) = κ1κ2κ3 + θ1

1((y − θ2 1)(x − θ2 0) + θ1 0(y + θ2 0)).

         (f + g)(¯ f + g) = (g + b1)(g + b2)(g + b3)(g + b4) (g − b5 − δ)(g − b6 − δ) (¯ f + g)(¯ f + ¯ g) = (¯ f − b1)(¯ f − b2)(¯ f − b3)(¯ f − b4) (¯ f + b7 − δ)(¯ f − b8 − δ) , (2) where b1, . . . , b8 are some parameters and δ = b1 + · · · + b8.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 2 / 41

slide-5
SLIDE 5

Both equations are in fact very natural expressions (in their respective settings, of course) of difference Painlev´ e equations of type d-P

  • A(1)∗

2

  • with symmetry

W

  • E (1)

6

  • , and so a question

about the relationship between the them is a very reasonable one.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 3 / 41

slide-6
SLIDE 6

Both equations are in fact very natural expressions (in their respective settings, of course) of difference Painlev´ e equations of type d-P

  • A(1)∗

2

  • with symmetry

W

  • E (1)

6

  • , and so a question

about the relationship between the them is a very reasonable one. The first equation describes one of the simplest elementary Schlesinger transformation of a Fuchsian system (T. Takenawa, A.D).

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 3 / 41

slide-7
SLIDE 7

Both equations are in fact very natural expressions (in their respective settings, of course) of difference Painlev´ e equations of type d-P

  • A(1)∗

2

  • with symmetry

W

  • E (1)

6

  • , and so a question

about the relationship between the them is a very reasonable one. The first equation describes one of the simplest elementary Schlesinger transformation of a Fuchsian system (T. Takenawa, A.D). The second equation is obtained by the application of the singularity confinement criterion to a deautonomization of a QRT map (B. Grammaticos, A. Ramani, Y. Ohta).

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 3 / 41

slide-8
SLIDE 8

Both equations are in fact very natural expressions (in their respective settings, of course) of difference Painlev´ e equations of type d-P

  • A(1)∗

2

  • with symmetry

W

  • E (1)

6

  • , and so a question

about the relationship between the them is a very reasonable one. The first equation describes one of the simplest elementary Schlesinger transformation of a Fuchsian system (T. Takenawa, A.D). The second equation is obtained by the application of the singularity confinement criterion to a deautonomization of a QRT map (B. Grammaticos, A. Ramani, Y. Ohta). There are infinitely many discrete Painlev´ e equations of the same type, but some of those equations are simpler and more “natural” than others, it’s important to identify such equations.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 3 / 41

slide-9
SLIDE 9

Both equations are in fact very natural expressions (in their respective settings, of course) of difference Painlev´ e equations of type d-P

  • A(1)∗

2

  • with symmetry

W

  • E (1)

6

  • , and so a question

about the relationship between the them is a very reasonable one. The first equation describes one of the simplest elementary Schlesinger transformation of a Fuchsian system (T. Takenawa, A.D). The second equation is obtained by the application of the singularity confinement criterion to a deautonomization of a QRT map (B. Grammaticos, A. Ramani, Y. Ohta). There are infinitely many discrete Painlev´ e equations of the same type, but some of those equations are simpler and more “natural” than others, it’s important to identify such equations. According to the Sakai’s classification scheme, a discrete Painlev´ e equation is a birational map of a complex projective plane that corresponds to a translation element in the symmetry sub-lattice

  • f a Picard lattice of a certain rational algebraic surface, known as the Okamoto Space of Initial

Conditions, that is obtained when we resolve the indeterminacies of the equation by using a blowup procedure. Our approach is to exploit the structure of the extended affine Weyl symmetry group W

  • E (1)

6

  • f the surface.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 3 / 41

slide-10
SLIDE 10

Both equations are in fact very natural expressions (in their respective settings, of course) of difference Painlev´ e equations of type d-P

  • A(1)∗

2

  • with symmetry

W

  • E (1)

6

  • , and so a question

about the relationship between the them is a very reasonable one. The first equation describes one of the simplest elementary Schlesinger transformation of a Fuchsian system (T. Takenawa, A.D). The second equation is obtained by the application of the singularity confinement criterion to a deautonomization of a QRT map (B. Grammaticos, A. Ramani, Y. Ohta). There are infinitely many discrete Painlev´ e equations of the same type, but some of those equations are simpler and more “natural” than others, it’s important to identify such equations. According to the Sakai’s classification scheme, a discrete Painlev´ e equation is a birational map of a complex projective plane that corresponds to a translation element in the symmetry sub-lattice

  • f a Picard lattice of a certain rational algebraic surface, known as the Okamoto Space of Initial

Conditions, that is obtained when we resolve the indeterminacies of the equation by using a blowup procedure. Our approach is to exploit the structure of the extended affine Weyl symmetry group W

  • E (1)

6

  • f the surface.

Main result: These two equations are equivalent through an explicit change of variables transforming one equation into the other: f = x(y − θ1

1) + y(θ1 0 + κ1) + (θ2 0 + κ1)(θ1 0 + θ2 0 + θ1 1 + 2κ1)

y + θ2

0 + κ1

g = x(y − θ2

0 − θ1 1 − κ1) + y(θ1 0 − θ2 0) + (θ2 0 + κ1)(θ1 0 + θ2 0 + 2κ1)

x − θ2

0 − κ1

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 3 / 41

slide-11
SLIDE 11

Classification Scheme for Painlev´ e Equations

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 4 / 41

slide-12
SLIDE 12

Classification Scheme for Painlev´ e Equations

n=1: L. Fuchs, H. Poincar´ e

dy dt 2 = 4y 3 − g2y − g3, g2, g3 ∈ C Weierstrass ℘(t|g2, g3) dy dt = a(t)y 2 + b(t)y + c(t), (Riccati equation)

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 4 / 41

slide-13
SLIDE 13

Classification Scheme for Painlev´ e Equations

n=1: L. Fuchs, H. Poincar´ e

dy dt 2 = 4y 3 − g2y − g3, g2, g3 ∈ C Weierstrass ℘(t|g2, g3) dy dt = a(t)y 2 + b(t)y + c(t), (Riccati equation)

n=2: P. Painlev´ e, B. Gambier — Painlev´ e equations and Painlev´ e transcendents:

(P-I) d2y dt2 = 6y 2 + t; (P-II) d2y dt2 = 2y 3 + ty + α; (P-III) d2y dt2 = 1 y dy dt 2 − 1 t dy dt + 1 t (αy 2 + β) + γy 3 + δ y ; (P-IV) d2y dt2 = 1 2y dy dt 2 + 3 2 y 3 + 4ty 2 + 2(t2 − α)y + β y ; (P-V) d2y dt2 = 1 2y + 1 y − 1 dy dt 2 − 1 t dy dt + (y − 1)2 t2

  • αy + β

y

  • + γ y

t + δ y(y + 1) y − 1 ; (P-VI) d2y dt2 = 1 2 1 y + 1 y − 1 + 1 y − t dy dt 2 − 1 t + 1 t − 1 + 1 y − t dy dt + y(y − 1)(y − t) t2(t − 1)2

  • α + β t

y 2 + γ t − 1 (y − 1)2 + δ t(t − 1) (y − t)2

  • .

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 4 / 41

slide-14
SLIDE 14

Classification Scheme for Painlev´ e Equations

n=1: L. Fuchs, H. Poincar´ e

dy dt 2 = 4y 3 − g2y − g3, g2, g3 ∈ C Weierstrass ℘(t|g2, g3) dy dt = a(t)y 2 + b(t)y + c(t), (Riccati equation)

n=2: P. Painlev´ e, B. Gambier — Painlev´ e equations and Painlev´ e transcendents:

(P-I) d2y dt2 = 6y 2 + t; Painlev´ e equations have parameters! (P-II) d2y dt2 = 2y 3 + ty + α; (P-III) d2y dt2 = 1 y dy dt 2 − 1 t dy dt + 1 t (αy 2 + β) + γy 3 + δ y ; (P-IV) d2y dt2 = 1 2y dy dt 2 + 3 2 y 3 + 4ty 2 + 2(t2 − α)y + β y ; (P-V) d2y dt2 = 1 2y + 1 y − 1 dy dt 2 − 1 t dy dt + (y − 1)2 t2

  • αy + β

y

  • + γ y

t + δ y(y + 1) y − 1 ; (P-VI) d2y dt2 = 1 2 1 y + 1 y − 1 + 1 y − t dy dt 2 − 1 t + 1 t − 1 + 1 y − t dy dt + y(y − 1)(y − t) t2(t − 1)2

  • α + β t

y 2 + γ t − 1 (y − 1)2 + δ t(t − 1) (y − t)2

  • .

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 4 / 41

slide-15
SLIDE 15

Classification Scheme for Painlev´ e Equations

n=1: L. Fuchs, H. Poincar´ e

dy dt 2 = 4y 3 − g2y − g3, g2, g3 ∈ C Weierstrass ℘(t|g2, g3) dy dt = a(t)y 2 + b(t)y + c(t), (Riccati equation)

n=2: P. Painlev´ e, B. Gambier — Painlev´ e equations and Painlev´ e transcendents:

(P-I) d2y dt2 = 6y 2 + t; Painlev´ e equations have parameters! (P-II) d2y dt2 = 2y 3 + ty + α; PVI PV PIV PIII PII PI (P-III) d2y dt2 = 1 y dy dt 2 − 1 t dy dt + 1 t (αy 2 + β) + γy 3 + δ y ; (P-IV) d2y dt2 = 1 2y dy dt 2 + 3 2 y 3 + 4ty 2 + 2(t2 − α)y + β y ; (P-V) d2y dt2 = 1 2y + 1 y − 1 dy dt 2 − 1 t dy dt + (y − 1)2 t2

  • αy + β

y

  • + γ y

t + δ y(y + 1) y − 1 ; (P-VI) d2y dt2 = 1 2 1 y + 1 y − 1 + 1 y − t dy dt 2 − 1 t + 1 t − 1 + 1 y − t dy dt + y(y − 1)(y − t) t2(t − 1)2

  • α + β t

y 2 + γ t − 1 (y − 1)2 + δ t(t − 1) (y − t)2

  • .

n ≥ 3: Still open.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 4 / 41

slide-16
SLIDE 16

Classification Scheme for Painlev´ e Equations

n=1: L. Fuchs, H. Poincar´ e

dy dt 2 = 4y 3 − g2y − g3, g2, g3 ∈ C Weierstrass ℘(t|g2, g3) dy dt = a(t)y 2 + b(t)y + c(t), (Riccati equation)

n=2: P. Painlev´ e, B. Gambier — Painlev´ e equations and Painlev´ e transcendents:

(P-I) d2y dt2 = 6y 2 + t; Painlev´ e equations have parameters! (P-II) d2y dt2 = 2y 3 + ty + α; PVI PV PIV PIII PII PI (P-III) d2y dt2 = 1 y dy dt 2 − 1 t dy dt + 1 t (αy 2 + β) + γy 3 + δ y ; (P-IV) d2y dt2 = 1 2y dy dt 2 + 3 2 y 3 + 4ty 2 + 2(t2 − α)y + β y ; Gauss Kummer Hermite Bessel Airy (P-V) d2y dt2 = 1 2y + 1 y − 1 dy dt 2 − 1 t dy dt + (y − 1)2 t2

  • αy + β

y

  • + γ y

t + δ y(y + 1) y − 1 ; (P-VI) d2y dt2 = 1 2 1 y + 1 y − 1 + 1 y − t dy dt 2 − 1 t + 1 t − 1 + 1 y − t dy dt + y(y − 1)(y − t) t2(t − 1)2

  • α + β t

y 2 + γ t − 1 (y − 1)2 + δ t(t − 1) (y − t)2

  • .

n ≥ 3: Still open.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 4 / 41

slide-17
SLIDE 17

Classification Scheme for Discrete Painlev´ e Equations

In 2001 H. Sakai, developing the ideas of K. Okamoto in the differential case, proposed a classification scheme for Painlev´ e equations based on algebraic geometry. To each equation corresponds a pair of orthogonal sub-lattices (Π(R), Π(R⊥)) — the surface and the symmetry sub-lattice in the E (1)

8

lattice, and a translation element in ˜ W (R⊥).

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 5 / 41

slide-18
SLIDE 18

Classification Scheme for Discrete Painlev´ e Equations

In 2001 H. Sakai, developing the ideas of K. Okamoto in the differential case, proposed a classification scheme for Painlev´ e equations based on algebraic geometry. To each equation corresponds a pair of orthogonal sub-lattices (Π(R), Π(R⊥)) — the surface and the symmetry sub-lattice in the E (1)

8

lattice, and a translation element in ˜ W (R⊥).

  • E(1)

8 e

  • E(1)

8 q

  • E(1)

7 q

  • E(1)

6 q

  • D(1)

5 q

  • A(1)

4 q

  • (A2 + A1)(1)q
  • (A1 + A1)(1)q

|α|2 = 14

  • A(1)

1 q |α|2 = 4

  • A(1)

q

  • A(1)

1 q |α|2=8

  • E(1)

8 δ

  • E(1)

7 δ

  • E(1)

6 δ

  • D(1)

4 c,δ

  • A(1)

3 c,δ

  • 2A(1)

1 c,δ

  • A(1)

1 c,δ

  • A(1)

c

  • 2A(1)

1 c,δ

  • A(1)

1 c,δ

  • A(1)

c q-PVI q-PV q-PIV, q-PIII q-PI PVI, d-PV PV, d-PIV d-PIII PIII

  • alt. d-PII

PIII PIII PIV, d-PII PII, alt.d-PI PI

Symmetry-type classification scheme for Painlev´ e equations

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 5 / 41

slide-19
SLIDE 19

Classification Scheme for Discrete Painlev´ e Equations

In 2001 H. Sakai, developing the ideas of K. Okamoto in the differential case, proposed a classification scheme for Painlev´ e equations based on algebraic geometry. To each equation corresponds a pair of orthogonal sub-lattices (Π(R), Π(R⊥)) — the surface and the symmetry sub-lattice in the E (1)

8

lattice, and a translation element in ˜ W (R⊥).

  • E(1)

8 e

  • E(1)

8 q

  • E(1)

7 q

  • E(1)

6 q

  • D(1)

5 q

  • A(1)

4 q

  • (A2 + A1)(1)q
  • (A1 + A1)(1)q

|α|2 = 14

  • A(1)

1 q |α|2 = 4

  • A(1)

q

  • A(1)

1 q |α|2=8

  • E(1)

8 δ

  • E(1)

7 δ

  • E(1)

6 δ

  • D(1)

4 c,δ

  • A(1)

3 c,δ

  • 2A(1)

1 c,δ

  • A(1)

1 c,δ

  • A(1)

c

  • 2A(1)

1 c,δ

  • A(1)

1 c,δ

  • A(1)

c q-PVI q-PV q-PIV, q-PIII q-PI PVI, d-PV PV, d-PIV d-PIII PIII

  • alt. d-PII

PIII PIII PIV, d-PII PII, alt.d-PI PI

Symmetry-type classification scheme for Painlev´ e equations So we see that in the discrete case the classification scheme is very rich, there are twenty-two different cases, and moreover, there are no generic expressions for equations of each type, the classification scheme is very algebraic.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 5 / 41

slide-20
SLIDE 20

Classification Scheme for Discrete Painlev´ e Equations

In 2001 H. Sakai, developing the ideas of K. Okamoto in the differential case, proposed a classification scheme for Painlev´ e equations based on algebraic geometry. To each equation corresponds a pair of orthogonal sub-lattices (Π(R), Π(R⊥)) — the surface and the symmetry sub-lattice in the E (1)

8

lattice, and a translation element in ˜ W (R⊥).

  • E(1)

8 e

  • E(1)

8 q

  • E(1)

7 q

  • E(1)

6 q

  • D(1)

5 q

  • A(1)

4 q

  • (A2 + A1)(1)q
  • (A1 + A1)(1)q

|α|2 = 14

  • A(1)

1 q |α|2 = 4

  • A(1)

q

  • A(1)

1 q |α|2=8

  • E(1)

8 δ

  • E(1)

7 δ

  • E(1)

6 δ

  • D(1)

4 c,δ

  • A(1)

3 c,δ

  • 2A(1)

1 c,δ

  • A(1)

1 c,δ

  • A(1)

c

  • 2A(1)

1 c,δ

  • A(1)

1 c,δ

  • A(1)

c q-PVI q-PV q-PIV, q-PIII q-PI PVI, d-PV PV, d-PIV d-PIII PIII

  • alt. d-PII

PIII PIII PIV, d-PII PII, alt.d-PI PI

Symmetry-type classification scheme for Painlev´ e equations So we see that in the discrete case the classification scheme is very rich, there are twenty-two different cases, and moreover, there are no generic expressions for equations of each type, the classification scheme is very algebraic. One of my goals for today is to explain the main ideas behind this scheme.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 5 / 41

slide-21
SLIDE 21

Classification Scheme for Discrete Painlev´ e Equations

First, we note that although the previous classification scheme according to symmetries is more traditional, for the geometric approach, it is more natural to look at the classification based on the point-configuration or surface type:

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 6 / 41

slide-22
SLIDE 22

Classification Scheme for Discrete Painlev´ e Equations

First, we note that although the previous classification scheme according to symmetries is more traditional, for the geometric approach, it is more natural to look at the classification based on the point-configuration or surface type:

A(1) A(1)∗ A(1) 1 A(1) 2 A(1) 3 A(1) 4 A(1) 5 A(1) 6 A(1) 7 A(1) 8 A(1) 7 A(1)∗∗ A(1)∗ 1 A(1)∗ 2 D(1) 4 D(1) 5 D(1) 6 D(1) 7 D(1) 8 E(1) 6 E(1) 7 E(1) 8

Surface-type classification scheme for Painlev´ e equations

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 6 / 41

slide-23
SLIDE 23

Classification Scheme for Discrete Painlev´ e Equations

First, we note that although the previous classification scheme according to symmetries is more traditional, for the geometric approach, it is more natural to look at the classification based on the point-configuration or surface type:

A(1) A(1)∗ A(1) 1 A(1) 2 A(1) 3 A(1) 4 A(1) 5 A(1) 6 A(1) 7 A(1) 8 A(1) 7 A(1)∗∗ A(1)∗ 1 A(1)∗ 2 D(1) 4 D(1) 5 D(1) 6 D(1) 7 D(1) 8 E(1) 6 E(1) 7 E(1) 8

Surface-type classification scheme for Painlev´ e equations In this description each letter stands for a Dynkin diagram describing intersection configuration of the irreducible components of an anti-canonical divisor of a certain algebraic surface, known as the generalized Halphen surface, that is obtained by blowing up P1 × P1 at eight (possibly infinitely-close) points.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 6 / 41

slide-24
SLIDE 24

Classification Scheme for Discrete Painlev´ e Equations

First, we note that although the previous classification scheme according to symmetries is more traditional, for the geometric approach, it is more natural to look at the classification based on the point-configuration or surface type:

A(1) A(1)∗ A(1) 1 A(1) 2 A(1) 3 A(1) 4 A(1) 5 A(1) 6 A(1) 7 A(1) 8 A(1) 7 A(1)∗∗ A(1)∗ 1 A(1)∗ 2 D(1) 4 D(1) 5 D(1) 6 D(1) 7 D(1) 8 E(1) 6 E(1) 7 E(1) 8

Surface-type classification scheme for Painlev´ e equations In this description each letter stands for a Dynkin diagram describing intersection configuration of the irreducible components of an anti-canonical divisor of a certain algebraic surface, known as the generalized Halphen surface, that is obtained by blowing up P1 × P1 at eight (possibly infinitely-close) points. The very top A(1) node corresponds to the the points lying on a smooth elliptic curves, A(1)∗ is a nodal curve (hence the multiplicative dynamic on the parameters), A(1)∗∗ is a cusp curve (hence the additive dynamic on the parameters), and others are various degenerations into rational components.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 6 / 41

slide-25
SLIDE 25

Origins of Discrete Painlev´ e Equations

Discrete Painlev´ e Equations appear in a variety of contexts:

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 7 / 41

slide-26
SLIDE 26

Origins of Discrete Painlev´ e Equations

Discrete Painlev´ e Equations appear in a variety of contexts: As certain second-order (or two-dimensional) non-autonomous nonlinear recurrence relations that has the usual differential Painlev´ e equations as some continuous limits (Shohat, Br´ ezin-Kazakov, Gross-Migdal).

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 7 / 41

slide-27
SLIDE 27

Origins of Discrete Painlev´ e Equations

Discrete Painlev´ e Equations appear in a variety of contexts: As certain second-order (or two-dimensional) non-autonomous nonlinear recurrence relations that has the usual differential Painlev´ e equations as some continuous limits (Shohat, Br´ ezin-Kazakov, Gross-Migdal). As symmetries (B¨ acklund transformations) of continous Painlev´ e equations. Geometric theory of B¨ acklund transformations of continous Painlev´ e equations was developed in the works of Okamoto, who also introduced the important notion of the space of initial conditions, which eventually led to Sakai’s theory.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 7 / 41

slide-28
SLIDE 28

Origins of Discrete Painlev´ e Equations

Discrete Painlev´ e Equations appear in a variety of contexts: As certain second-order (or two-dimensional) non-autonomous nonlinear recurrence relations that has the usual differential Painlev´ e equations as some continuous limits (Shohat, Br´ ezin-Kazakov, Gross-Migdal). As symmetries (B¨ acklund transformations) of continous Painlev´ e equations. Geometric theory of B¨ acklund transformations of continous Painlev´ e equations was developed in the works of Okamoto, who also introduced the important notion of the space of initial conditions, which eventually led to Sakai’s theory. As deautonomizations of different autonomous systems, such as QRT maps, together with requiring the singularity confinement condition (Grammaticos, Ramani, Ohta, Papageorgiu, Nijhoff).

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 7 / 41

slide-29
SLIDE 29

Origins of Discrete Painlev´ e Equations

Discrete Painlev´ e Equations appear in a variety of contexts: As certain second-order (or two-dimensional) non-autonomous nonlinear recurrence relations that has the usual differential Painlev´ e equations as some continuous limits (Shohat, Br´ ezin-Kazakov, Gross-Migdal). As symmetries (B¨ acklund transformations) of continous Painlev´ e equations. Geometric theory of B¨ acklund transformations of continous Painlev´ e equations was developed in the works of Okamoto, who also introduced the important notion of the space of initial conditions, which eventually led to Sakai’s theory. As deautonomizations of different autonomous systems, such as QRT maps, together with requiring the singularity confinement condition (Grammaticos, Ramani, Ohta, Papageorgiu, Nijhoff). As birational representations of affine Weyl groups (Noumi, Yamada, Kajiwara, Ohta).

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 7 / 41

slide-30
SLIDE 30

Origins of Discrete Painlev´ e Equations

Discrete Painlev´ e Equations appear in a variety of contexts: As certain second-order (or two-dimensional) non-autonomous nonlinear recurrence relations that has the usual differential Painlev´ e equations as some continuous limits (Shohat, Br´ ezin-Kazakov, Gross-Migdal). As symmetries (B¨ acklund transformations) of continous Painlev´ e equations. Geometric theory of B¨ acklund transformations of continous Painlev´ e equations was developed in the works of Okamoto, who also introduced the important notion of the space of initial conditions, which eventually led to Sakai’s theory. As deautonomizations of different autonomous systems, such as QRT maps, together with requiring the singularity confinement condition (Grammaticos, Ramani, Ohta, Papageorgiu, Nijhoff). As birational representations of affine Weyl groups (Noumi, Yamada, Kajiwara, Ohta). As reductions of Schlesinger transformations of Fuchsian systems (Jimbo, Sakai, Grammaticos, Ramani)

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 7 / 41

slide-31
SLIDE 31

Origins of Discrete Painlev´ e Equations

Discrete Painlev´ e Equations appear in a variety of contexts: As certain second-order (or two-dimensional) non-autonomous nonlinear recurrence relations that has the usual differential Painlev´ e equations as some continuous limits (Shohat, Br´ ezin-Kazakov, Gross-Migdal). As symmetries (B¨ acklund transformations) of continous Painlev´ e equations. Geometric theory of B¨ acklund transformations of continous Painlev´ e equations was developed in the works of Okamoto, who also introduced the important notion of the space of initial conditions, which eventually led to Sakai’s theory. As deautonomizations of different autonomous systems, such as QRT maps, together with requiring the singularity confinement condition (Grammaticos, Ramani, Ohta, Papageorgiu, Nijhoff). As birational representations of affine Weyl groups (Noumi, Yamada, Kajiwara, Ohta). As reductions of Schlesinger transformations of Fuchsian systems (Jimbo, Sakai, Grammaticos, Ramani) (the list of names above is, of course, far from complete).

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 7 / 41

slide-32
SLIDE 32

Origins of Discrete Painlev´ e Equations

Discrete Painlev´ e Equations appear in a variety of contexts: As certain second-order (or two-dimensional) non-autonomous nonlinear recurrence relations that has the usual differential Painlev´ e equations as some continuous limits (Shohat, Br´ ezin-Kazakov, Gross-Migdal). As symmetries (B¨ acklund transformations) of continous Painlev´ e equations. Geometric theory of B¨ acklund transformations of continous Painlev´ e equations was developed in the works of Okamoto, who also introduced the important notion of the space of initial conditions, which eventually led to Sakai’s theory. As deautonomizations of different autonomous systems, such as QRT maps, together with requiring the singularity confinement condition (Grammaticos, Ramani, Ohta, Papageorgiu, Nijhoff). As birational representations of affine Weyl groups (Noumi, Yamada, Kajiwara, Ohta). As reductions of Schlesinger transformations of Fuchsian systems (Jimbo, Sakai, Grammaticos, Ramani) (the list of names above is, of course, far from complete). For the geometric description, the most natural approach is via the deautonomization of QRT maps, so we briefly recall this construction.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 7 / 41

slide-33
SLIDE 33

Geometry of a QRT Map

The Quispel-Roberts-Thompson (or QRT) mapping can be describing geometrically as follows (following T. Tsuda).

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 8 / 41

slide-34
SLIDE 34

Geometry of a QRT Map

The Quispel-Roberts-Thompson (or QRT) mapping can be describing geometrically as follows (following T. Tsuda). Consider a bi-quadractic curve Γ on P1 × P1. In an affine C2-chart Γ is given by a bi-degree (2, 2) polynomial equation a00x2y2 + a01x2y + a02x2 + a10xy2 + a11xy + a12x + a20y2 + a21y + a22 = 0.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 8 / 41

slide-35
SLIDE 35

Geometry of a QRT Map

The Quispel-Roberts-Thompson (or QRT) mapping can be describing geometrically as follows (following T. Tsuda). Consider a bi-quadractic curve Γ on P1 × P1. In an affine C2-chart Γ is given by a bi-degree (2, 2) polynomial equation a00x2y2 + a01x2y + a02x2 + a10xy2 + a11xy + a12x + a20y2 + a21y + a22 = 0. This equation can be written as xT Ay = x2 x 1   a00 a01 a02 a10 a11 a12 a20 a21 a22     y2 y 1   =

2

  • i,j=0

aijx2−iy2−j = 0, where x = x2, x, 1, y = y2, y, 1, A ∈ Mat3×3(C).

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 8 / 41

slide-36
SLIDE 36

Geometry of a QRT Map

The Quispel-Roberts-Thompson (or QRT) mapping can be describing geometrically as follows (following T. Tsuda). Consider a bi-quadractic curve Γ on P1 × P1. In an affine C2-chart Γ is given by a bi-degree (2, 2) polynomial equation a00x2y2 + a01x2y + a02x2 + a10xy2 + a11xy + a12x + a20y2 + a21y + a22 = 0. This equation can be written as xT Ay = x2 x 1   a00 a01 a02 a10 a11 a12 a20 a21 a22     y2 y 1   =

2

  • i,j=0

aijx2−iy2−j = 0, where x = x2, x, 1, y = y2, y, 1, A ∈ Mat3×3(C). In general, Γ is an elliptic curve that can be rewritten in a Weierstrass normal form Y 2 = 4X 3 − g2X − g3.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 8 / 41

slide-37
SLIDE 37

Geometry of a QRT Map

Since Γ has bi-degree (2, 2), we can define two involutions, rx : (x, y) → (x′, y) and ry : (x, y) → (x, y′)

(x, y) (x, y0) (x0, y) (¯ x, ¯ y) rx ry rx Γ

as well as their composition rx ◦ ry : (x, y) → (¯ x, ¯ y).

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 9 / 41

slide-38
SLIDE 38

Geometry of a QRT Map

Since Γ has bi-degree (2, 2), we can define two involutions, rx : (x, y) → (x′, y) and ry : (x, y) → (x, y′)

(x, y) (x, y0) (x0, y) (¯ x, ¯ y) rx ry rx Γ

as well as their composition rx ◦ ry : (x, y) → (¯ x, ¯ y). The main idea of the QRT map is to extend rx ◦ ry to all of the P1 × P1.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 9 / 41

slide-39
SLIDE 39

Geometry of a QRT Map

For that, take two matrices A, B ∈ Mat3×3(C) and consider a pencil (i.e., a one-dimensional family) of such curves Γ[α:β] : αxT Ay + βxT By = 0, [α : β] ∈ P1

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 10 / 41

slide-40
SLIDE 40

Geometry of a QRT Map

For that, take two matrices A, B ∈ Mat3×3(C) and consider a pencil (i.e., a one-dimensional family) of such curves Γ[α:β] : αxT Ay + βxT By = 0, [α : β] ∈ P1

Γ[α:β] xT By = 0 xT Ay = 0

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 10 / 41

slide-41
SLIDE 41

Geometry of a QRT Map

For that, take two matrices A, B ∈ Mat3×3(C) and consider a pencil (i.e., a one-dimensional family) of such curves Γ[α:β] : αxT Ay + βxT By = 0, [α : β] ∈ P1

Γ[α:β] xT By = 0 xT Ay = 0

Then, given a point (x∗, y∗), there is only one curve from a family with the parameter [α : β] = [−xT

∗ By∗, xT ∗ Ay∗], except for the eight base points xT ∗ Ay∗ = xT ∗ By∗ = 0.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 10 / 41

slide-42
SLIDE 42

Geometry of a QRT Map

For that, take two matrices A, B ∈ Mat3×3(C) and consider a pencil (i.e., a one-dimensional family) of such curves Γ[α:β] : αxT Ay + βxT By = 0, [α : β] ∈ P1

Γ[α:β] xT By = 0 xT Ay = 0

Then, given a point (x∗, y∗), there is only one curve from a family with the parameter [α : β] = [−xT

∗ By∗, xT ∗ Ay∗], except for the eight base points xT ∗ Ay∗ = xT ∗ By∗ = 0.

Resolving these points using the blowup, we get a rational elliptic surface X with the QRT automorphism rx ◦ ry preserving the elliptic fibration π : X → P1, and π−1([α : β]) is an elliptic curve except for 12 points corresponding to singular fibers (classified by K. Kodaira into 22 types).

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 10 / 41

slide-43
SLIDE 43

Geometry of a QRT Map

For that, take two matrices A, B ∈ Mat3×3(C) and consider a pencil (i.e., a one-dimensional family) of such curves Γ[α:β] : αxT Ay + βxT By = 0, [α : β] ∈ P1

Γ[α:β] xT By = 0 xT Ay = 0

Then, given a point (x∗, y∗), there is only one curve from a family with the parameter [α : β] = [−xT

∗ By∗, xT ∗ Ay∗], except for the eight base points xT ∗ Ay∗ = xT ∗ By∗ = 0.

Resolving these points using the blowup, we get a rational elliptic surface X with the QRT automorphism rx ◦ ry preserving the elliptic fibration π : X → P1, and π−1([α : β]) is an elliptic curve except for 12 points corresponding to singular fibers (classified by K. Kodaira into 22 types). Deautonomization is performed with different choices of a fiber on which the blowup points lie (that fiber is exactly the anti-canonical divisor, −KX that is the key object of the geometric theory). Allowing the points move along a particular fiber, either smooth or singular, breaks down the elliptic surface structure and the dynamic becomes non-autonomus.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 10 / 41

slide-44
SLIDE 44

The A(1)∗

2

Deautonomization Example

For the example that we are interested in, we can take the matrices A, B as A =   1 1   , B =   1 −(a + a−1) 1 −(a + a−1) (b + b−1)2 1 1   , where a = b = 0, ±1.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 11 / 41

slide-45
SLIDE 45

The A(1)∗

2

Deautonomization Example

For the example that we are interested in, we can take the matrices A, B as A =   1 1   , B =   1 −(a + a−1) 1 −(a + a−1) (b + b−1)2 1 1   , where a = b = 0, ±1. Here is an example of two fibers in this family: a smooth elliptic A(1)

0 -fiber

and a singular A(1)∗

2

  • fiber:

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 11 / 41

slide-46
SLIDE 46

The A(1)∗

2

Deautonomization Example

For the example that we are interested in, we can take the matrices A, B as A =   1 1   , B =   1 −(a + a−1) 1 −(a + a−1) (b + b−1)2 1 1   , where a = b = 0, ±1. Here is an example of two fibers in this family: a smooth elliptic A(1)

0 -fiber

and a singular A(1)∗

2

  • fiber:

The base points of the map are shown in red, allowing them to move along the A(1)∗

2

  • fiber (points

shown in blue) resulted in the deautonomization example of Grammaticos-Ramani-Ohta.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 11 / 41

slide-47
SLIDE 47

The A(1)∗

2

Deautonomization Example

For the example that we are interested in, we can take the matrices A, B as A =   1 1   , B =   1 −(a + a−1) 1 −(a + a−1) (b + b−1)2 1 1   , where a = b = 0, ±1. Here is an example of two fibers in this family: a smooth elliptic A(1)

0 -fiber

and a singular A(1)∗

2

  • fiber:

The base points of the map are shown in red, allowing them to move along the A(1)∗

2

  • fiber (points

shown in blue) resulted in the deautonomization example of Grammaticos-Ramani-Ohta. However, we can in fact create the mapping starting just from the Dynkin diagrams and a choice

  • f a translation element. We explain how to do that next.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 11 / 41

slide-48
SLIDE 48

Canonical Model of the Okamoto Surface of Type A(1)∗

2 Let us start by understanding the structure of a generalized Halphen surface of type A(1)∗

2

.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 12 / 41

slide-49
SLIDE 49

Canonical Model of the Okamoto Surface of Type A(1)∗

2 Let us start by understanding the structure of a generalized Halphen surface of type A(1)∗

2

. Such a surface X is obtained by blowing up P1 × P1 at 8 points, and the type corresponds to the intersection structure of the irreducible components of the anti-canonical divisor −KX in the Picard Lattice Pic(X) = ZHf

  • ZHg
  • 8

i=1 ZEi,

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 12 / 41

slide-50
SLIDE 50

Canonical Model of the Okamoto Surface of Type A(1)∗

2 Let us start by understanding the structure of a generalized Halphen surface of type A(1)∗

2

. Such a surface X is obtained by blowing up P1 × P1 at 8 points, and the type corresponds to the intersection structure of the irreducible components of the anti-canonical divisor −KX in the Picard Lattice Pic(X) = ZHf

  • ZHg
  • 8

i=1 ZEi,

−KX = 2Hf + 2Hg − E1 − · · · − E8 =

  • i

miDi.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 12 / 41

slide-51
SLIDE 51

Canonical Model of the Okamoto Surface of Type A(1)∗

2 Let us start by understanding the structure of a generalized Halphen surface of type A(1)∗

2

. Such a surface X is obtained by blowing up P1 × P1 at 8 points, and the type corresponds to the intersection structure of the irreducible components of the anti-canonical divisor −KX in the Picard Lattice Pic(X) = ZHf

  • ZHg
  • 8

i=1 ZEi,

−KX = 2Hf + 2Hg − E1 − · · · − E8 =

  • i

miDi. Dynkin diagram A(1)

2

and the anti-canonical divisor decomposition D0 D1 D2   −2 1 1 1 −2 1 1 1 −2   −KX = D0 + D1 + D2 Dynkin diagram A(1)

2

its Cartan matrix −KX decomposition

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 12 / 41

slide-52
SLIDE 52

Canonical Model of the Okamoto Surface of Type A(1)∗

2 Let us start by understanding the structure of a generalized Halphen surface of type A(1)∗

2

. Such a surface X is obtained by blowing up P1 × P1 at 8 points, and the type corresponds to the intersection structure of the irreducible components of the anti-canonical divisor −KX in the Picard Lattice Pic(X) = ZHf

  • ZHg
  • 8

i=1 ZEi,

−KX = 2Hf + 2Hg − E1 − · · · − E8 =

  • i

miDi. Dynkin diagram A(1)

2

and the anti-canonical divisor decomposition D0 D1 D2   −2 1 1 1 −2 1 1 1 −2   −KX = D0 + D1 + D2 Dynkin diagram A(1)

2

its Cartan matrix −KX decomposition Without loss of generality, we can put

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 12 / 41

slide-53
SLIDE 53

Canonical Model of the Okamoto Surface of Type A(1)∗

2 Let us start by understanding the structure of a generalized Halphen surface of type A(1)∗

2

. Such a surface X is obtained by blowing up P1 × P1 at 8 points, and the type corresponds to the intersection structure of the irreducible components of the anti-canonical divisor −KX in the Picard Lattice Pic(X) = ZHf

  • ZHg
  • 8

i=1 ZEi,

−KX = 2Hf + 2Hg − E1 − · · · − E8 =

  • i

miDi. Dynkin diagram A(1)

2

and the anti-canonical divisor decomposition D0 D1 D2   −2 1 1 1 −2 1 1 1 −2   −KX = D0 + D1 + D2 Dynkin diagram A(1)

2

its Cartan matrix −KX decomposition Without loss of generality, we can put D0 = Hf + Hg − E1 − E2 − E3 − E4 D1 = Hf − E5 − E6 D2 = Hg − E7 − E8.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 12 / 41

slide-54
SLIDE 54

Canonical Model of the Okamoto Surface of Type A(1)∗

2 There are, however, two different geometric configurations related to the algebraic intersection structure given by this Dynkin diagram:

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 13 / 41

slide-55
SLIDE 55

Canonical Model of the Okamoto Surface of Type A(1)∗

2 There are, however, two different geometric configurations related to the algebraic intersection structure given by this Dynkin diagram: D0 D1 D2 Dynkin diagram A(1)

2

A(1)

2

surface (multiplicative) A(1)∗

2

surface (additive)

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 13 / 41

slide-56
SLIDE 56

Canonical Model of the Okamoto Surface of Type A(1)∗

2 There are, however, two different geometric configurations related to the algebraic intersection structure given by this Dynkin diagram: D0 D1 D2 Dynkin diagram A(1)

2

A(1)

2

surface (multiplicative) A(1)∗

2

surface (additive) We are interested in the additive dynamic given by A(1)∗

2

, so we want all of the irreducible components of the anti-canonical divisor to intersect at one point.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 13 / 41

slide-57
SLIDE 57

Canonical Model of the Okamoto Surface of Type A(1)∗

2 There are, however, two different geometric configurations related to the algebraic intersection structure given by this Dynkin diagram: D0 D1 D2 Dynkin diagram A(1)

2

A(1)

2

surface (multiplicative) A(1)∗

2

surface (additive) We are interested in the additive dynamic given by A(1)∗

2

, so we want all of the irreducible components of the anti-canonical divisor to intersect at one point. Again, without the loss of generality (i.e., acting by affine transformations on each of the two P1 factors) we can assume that the component D1 = Hf − E5 − E6 under the blowing down map projects to the line f = ∞ (and so there are two blowup points p5(∞, b5) and p6(∞, b6) on that line), the component D2 = Hg − E7 − E8 projects to the line g = ∞ with points p7(−b6, ∞) and p8(−b8, ∞), and the component D0 = Hf + Hg − E1 − E2 − E3 − E4 projects to the line f + g = 0.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 13 / 41

slide-58
SLIDE 58

Canonical Model of the Okamoto Surface of Type A(1)∗

2 Thus, we get the following geometric realization of a (family of) surface(s) Xb of type A(1)∗

2

:

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 14 / 41

slide-59
SLIDE 59

Canonical Model of the Okamoto Surface of Type A(1)∗

2 Thus, we get the following geometric realization of a (family of) surface(s) Xb of type A(1)∗

2

:

p7(−b7, ∞) p8(−b8, ∞) p5(∞, b5) p6(∞, b6) p1(b1, −b1) p2(b2, −b2) p3(b3, −b3) p4(b4, −b4) g = ∞ Hg g = 0 Hg f = 0 f = ∞ Hf Hf Hf + Hg h = f + g = 0 Hg − E7 − E8 Hf − E5 − E6 Hf + Hg − E1 − E2 − E3 − E4 E1 E2 E3 E4 E5 E6 E7 E8

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 14 / 41

slide-60
SLIDE 60

Canonical Model of the Okamoto Surface of Type A(1)∗

2 Thus, we get the following geometric realization of a (family of) surface(s) Xb of type A(1)∗

2

:

p7(−b7, ∞) p8(−b8, ∞) p5(∞, b5) p6(∞, b6) p1(b1, −b1) p2(b2, −b2) p3(b3, −b3) p4(b4, −b4) g = ∞ Hg g = 0 Hg f = 0 f = ∞ Hf Hf Hf + Hg h = f + g = 0 Hg − E7 − E8 Hf − E5 − E6 Hf + Hg − E1 − E2 − E3 − E4 E1 E2 E3 E4 E5 E6 E7 E8

Note that the lines in the above configuration form a pole divisor of the symplectic form ω = df ∧ dg (f + g) = − dF ∧ dg F(1 + Fg) = − df ∧ dG G(fG + 1) = dF ∧ dG (F + G) = dh ∧ dg h = · · ·

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 14 / 41

slide-61
SLIDE 61

Canonical Model of the Okamoto Surface of Type A(1)∗

2 Thus, we get the following geometric realization of a (family of) surface(s) Xb of type A(1)∗

2

:

p7(−b7, ∞) p8(−b8, ∞) p5(∞, b5) p6(∞, b6) p1(b1, −b1) p2(b2, −b2) p3(b3, −b3) p4(b4, −b4) g = ∞ Hg g = 0 Hg f = 0 f = ∞ Hf Hf Hf + Hg h = f + g = 0 Hg − E7 − E8 Hf − E5 − E6 Hf + Hg − E1 − E2 − E3 − E4 E1 E2 E3 E4 E5 E6 E7 E8

Note that the lines in the above configuration form a pole divisor of the symplectic form ω = df ∧ dg (f + g) = − dF ∧ dg F(1 + Fg) = − df ∧ dG G(fG + 1) = dF ∧ dG (F + G) = dh ∧ dg h = · · · However, there is still a two-parameter family of transformations preserving this configuration: b1 b2 b3 b4 b5 b6 b7 b8; f , g

αb1 + β αb2 + β αb3 + β αb4 + β αb5 − β αb6 − β αb7 − β αb8 − β; αf + β, αg − β

  • , α = 0.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 14 / 41

slide-62
SLIDE 62

The Symmetry Group and the Symmetry Sub-Lattice

A more invariant way to parameterize the surface is to use the so-called Period Map. For that we first need to define the symmetry sublattice.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 15 / 41

slide-63
SLIDE 63

The Symmetry Group and the Symmetry Sub-Lattice

A more invariant way to parameterize the surface is to use the so-called Period Map. For that we first need to define the symmetry sublattice. Symmetry sublattice Q ⊳ Pic(X) Q = (SpanZ{D0, D1, D2})⊥ = Q

  • (A(1)

2 )⊥

= SpanZ{α0, α1, α2, α3, α4, α5, α6} = Q

  • E (1)

6

  • ,

where the simple roots αi are given by α0 α1 α2 α3 α4 α5 α6 α0 = E3 − E4, α4 = E7 − E8, α1 = E2 − E3, α5 = Hg − E1 − E5, α2 = E1 − E2, α6 = E5 − E6. α3 = Hf − E1 − E7, Note also that δ = −KX = α0 + 2α1 + 3α2 + 2α3 + α4 + 2α5 + α6.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 15 / 41

slide-64
SLIDE 64

The Symmetry Group and the Symmetry Sub-Lattice

A more invariant way to parameterize the surface is to use the so-called Period Map. For that we first need to define the symmetry sublattice. Symmetry sublattice Q ⊳ Pic(X) Q = (SpanZ{D0, D1, D2})⊥ = Q

  • (A(1)

2 )⊥

= SpanZ{α0, α1, α2, α3, α4, α5, α6} = Q

  • E (1)

6

  • ,

where the simple roots αi are given by α0 α1 α2 α3 α4 α5 α6 α0 = E3 − E4, α4 = E7 − E8, α1 = E2 − E3, α5 = Hg − E1 − E5, α2 = E1 − E2, α6 = E5 − E6. α3 = Hf − E1 − E7, Note also that δ = −KX = α0 + 2α1 + 3α2 + 2α3 + α4 + 2α5 + α6. The period mapping is the map χ : Q → C, χ(αi) = ai defined on the simple roots and then extended by the linearity.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 15 / 41

slide-65
SLIDE 65

The Period Map

χ(αi) = χ

  • [C 1

i ] − [C 0 1 ]

  • =

Qi

Pi

1 2πi

  • Dk

ω = Qi

Pi

resDk ω, ω = df ∧ dg f + g Pi Qi Dk C 0

i

C 1

i

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 16 / 41

slide-66
SLIDE 66

The Period Map

χ(αi) = χ

  • [C 1

i ] − [C 0 1 ]

  • =

Qi

Pi

1 2πi

  • Dk

ω = Qi

Pi

resDk ω, ω = df ∧ dg f + g Pi Qi Dk C 0

i

C 1

i

Examples of the Period Map computations p1(b1, −b1) p2(b2, −b2) p3(b3, −b3) p4(b4, −b4) p5(∞, b5) p6(∞, b6) p7(−b7, ∞) p8(−b8, ∞)

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 16 / 41

slide-67
SLIDE 67

The Period Map

χ(αi) = χ

  • [C 1

i ] − [C 0 1 ]

  • =

Qi

Pi

1 2πi

  • Dk

ω = Qi

Pi

resDk ω, ω = df ∧ dg f + g Pi Qi Dk C 0

i

C 1

i

Examples of the Period Map computations

  • α0 = E3 − E4 = [E3] − [E4],

Dk = D0 = {h = f + g = 0} ω = df ∧ dg f + g = dh ∧ dg h , resh=0 ω = dg χ(α0) = −b3

−b4

dg = b4 − b3 = a0 p1(b1, −b1) p2(b2, −b2) p3(b3, −b3) p4(b4, −b4) p5(∞, b5) p6(∞, b6) p7(−b7, ∞) p8(−b8, ∞)

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 16 / 41

slide-68
SLIDE 68

The Period Map

χ(αi) = χ

  • [C 1

i ] − [C 0 1 ]

  • =

Qi

Pi

1 2πi

  • Dk

ω = Qi

Pi

resDk ω, ω = df ∧ dg f + g Pi Qi Dk C 0

i

C 1

i

Examples of the Period Map computations

  • α0 = E3 − E4 = [E3] − [E4],

Dk = D0 = {h = f + g = 0} ω = df ∧ dg f + g = dh ∧ dg h , resh=0 ω = dg χ(α0) = −b3

−b4

dg = b4 − b3 = a0

  • α3 = Hf − E1 − E7 = [Hf − E1] − [E7],

Dk = D2 = {g = ∞} = {G = 0} ω = df ∧ dg f + g = − df ∧ dG G(fG + 1), resG=0 ω = df χ(α3) = b1

−b7

df = b1 + b7 = a3 Hf − E1 f = b1 p1(b1, −b1) p2(b2, −b2) p3(b3, −b3) p4(b4, −b4) p5(∞, b5) p6(∞, b6) p7(−b7, ∞) p8(−b8, ∞)

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 16 / 41

slide-69
SLIDE 69

The Period Map

The Period Map, ai = χ(αi) are the root variables a0 = b4 − b3, a3 = b1 + b7, a6 = b6 − b5, a1 = b3 − b2, a4 = b8 − b7, a2 = b2 − b1, a5 = b1 + b5.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 17 / 41

slide-70
SLIDE 70

The Period Map

The Period Map, ai = χ(αi) are the root variables a0 = b4 − b3, a3 = b1 + b7, a6 = b6 − b5, a1 = b3 − b2, a4 = b8 − b7, a2 = b2 − b1, a5 = b1 + b5. Parameterization by the root variables ai b1 b2 b3 b4 b5 b6 b7 b8; f , g

  • =
  • b1

b1 + a2 b1 + a1 + a2 b1 + a0 + a1 + a2 a5 − b1 a5 + a6 − b1 a3 − b1 a3 + a4 − b1 ; f , g

  • ,

and so we see that b1 is one free parameter (translation of the origin). To fix the global scaling parameter we usually normalize χ(δ) = χ(−KX ) = χ(a0 + 2a1 + 3a2 + 2a3 + a4 + 2a5 + a6) = b1 + b2 + b3 + b4 + b5 + b6 + b7 + b8.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 17 / 41

slide-71
SLIDE 71

The Period Map

The Period Map, ai = χ(αi) are the root variables a0 = b4 − b3, a3 = b1 + b7, a6 = b6 − b5, a1 = b3 − b2, a4 = b8 − b7, a2 = b2 − b1, a5 = b1 + b5. Parameterization by the root variables ai b1 b2 b3 b4 b5 b6 b7 b8; f , g

  • =
  • b1

b1 + a2 b1 + a1 + a2 b1 + a0 + a1 + a2 a5 − b1 a5 + a6 − b1 a3 − b1 a3 + a4 − b1 ; f , g

  • ,

and so we see that b1 is one free parameter (translation of the origin). To fix the global scaling parameter we usually normalize χ(δ) = χ(−KX ) = χ(a0 + 2a1 + 3a2 + 2a3 + a4 + 2a5 + a6) = b1 + b2 + b3 + b4 + b5 + b6 + b7 + b8. The usual normalization is to put χ(δ) = 1, and one can also ask the same for b1. We will not do that, but we will require that, when resolving the normalization ambiguity, both χ(δ) and b1 are fixed — this ensures the group structure on the level of elementary birational maps.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 17 / 41

slide-72
SLIDE 72

The Extended Affine Weyl Symmetry Group W

  • E (1)

6

  • The next step in understanding the structure of difference Painlev´

e equations of type d-P

  • A(1)∗

2

  • is to describe the realization of the symmetry group in terms of elementary bilinear maps.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 18 / 41

slide-73
SLIDE 73

The Extended Affine Weyl Symmetry Group W

  • E (1)

6

  • The next step in understanding the structure of difference Painlev´

e equations of type d-P

  • A(1)∗

2

  • is to describe the realization of the symmetry group in terms of elementary bilinear maps.
  • W
  • E (1)

6

  • = Aut(E (1)

6

) ⋉ W (E (1)

6

) The full extended Weyl symmetry group W

  • E (1)

6

  • is a semi-direct product of

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 18 / 41

slide-74
SLIDE 74

The Extended Affine Weyl Symmetry Group W

  • E (1)

6

  • The next step in understanding the structure of difference Painlev´

e equations of type d-P

  • A(1)∗

2

  • is to describe the realization of the symmetry group in terms of elementary bilinear maps.
  • W
  • E (1)

6

  • = Aut(E (1)

6

) ⋉ W (E (1)

6

) The full extended Weyl symmetry group W

  • E (1)

6

  • is a semi-direct product of

The affine Weyl symmetry group of reflections wi = wαi W (E (1)

6

) =

  • w0, . . . , w6
  • w2

i = e

wi ◦ wj = wj ◦ wi when αi αj wi ◦ wj ◦ wi = wj ◦ wi ◦ wj when αi αj

  • α0 α1 α2 α3 α4

α5 α6

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 18 / 41

slide-75
SLIDE 75

The Extended Affine Weyl Symmetry Group W

  • E (1)

6

  • The next step in understanding the structure of difference Painlev´

e equations of type d-P

  • A(1)∗

2

  • is to describe the realization of the symmetry group in terms of elementary bilinear maps.
  • W
  • E (1)

6

  • = Aut(E (1)

6

) ⋉ W (E (1)

6

) The full extended Weyl symmetry group W

  • E (1)

6

  • is a semi-direct product of

The affine Weyl symmetry group of reflections wi = wαi W (E (1)

6

) =

  • w0, . . . , w6
  • w2

i = e

wi ◦ wj = wj ◦ wi when αi αj wi ◦ wj ◦ wi = wj ◦ wi ◦ wj when αi αj

  • α0 α1 α2 α3 α4

α5 α6 The finite group of Dynkin diagram automorphisms Aut

  • E (1)

6

  • ≃ Aut
  • A(1)

2

  • ≃ D3,

where D3 = {e, m0, m1, m2, r, r2} = m0, r | m2

0 = r3 = e, m0r = r2m0 is the usual dihedral

group of the symmetries of a triangle.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 18 / 41

slide-76
SLIDE 76

The Affine Weyl Group W (E (1)

6 ) Theorem Reflections wi are induced by the following elementary birational mappings (also denoted by wi)

  • n the family Xb fixing b1 and χ(δ) (we put bi···k = bi + · · · + bk, e.g., b12 = b1 + b2 and so on)

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 19 / 41

slide-77
SLIDE 77

The Affine Weyl Group W (E (1)

6 ) Theorem Reflections wi are induced by the following elementary birational mappings (also denoted by wi)

  • n the family Xb fixing b1 and χ(δ) (we put bi···k = bi + · · · + bk, e.g., b12 = b1 + b2 and so on)

b1 b2 b3 b4 b5 b6 b7 b8; f g

  • w0

− → b1 b2 b4 b3 b5 b6 b7 b8; f g

  • ,

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 19 / 41

slide-78
SLIDE 78

The Affine Weyl Group W (E (1)

6 ) Theorem Reflections wi are induced by the following elementary birational mappings (also denoted by wi)

  • n the family Xb fixing b1 and χ(δ) (we put bi···k = bi + · · · + bk, e.g., b12 = b1 + b2 and so on)

b1 b2 b3 b4 b5 b6 b7 b8; f g

  • w0

− → b1 b2 b4 b3 b5 b6 b7 b8; f g

  • ,

b1 b2 b3 b4 b5 b6 b7 b8; f g

  • w1

− → b1 b3 b2 b4 b5 b6 b7 b8; f g

  • ,

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 19 / 41

slide-79
SLIDE 79

The Affine Weyl Group W (E (1)

6 ) Theorem Reflections wi are induced by the following elementary birational mappings (also denoted by wi)

  • n the family Xb fixing b1 and χ(δ) (we put bi···k = bi + · · · + bk, e.g., b12 = b1 + b2 and so on)

b1 b2 b3 b4 b5 b6 b7 b8; f g

  • w0

− → b1 b2 b4 b3 b5 b6 b7 b8; f g

  • ,

b1 b2 b3 b4 b5 b6 b7 b8; f g

  • w1

− → b1 b3 b2 b4 b5 b6 b7 b8; f g

  • ,

b1 b2 b3 b4 b5 b6 b7 b8; f g

  • w2

− →

  • b1

b11 − b2 b13 − b2 b14 − b2 b52 − b1 b62 − b1 b72 − b1 b82 − b1; f + b1 − b2 g − b1 + b2,

  • ,

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 19 / 41

slide-80
SLIDE 80

The Affine Weyl Group W (E (1)

6 ) Theorem Reflections wi are induced by the following elementary birational mappings (also denoted by wi)

  • n the family Xb fixing b1 and χ(δ) (we put bi···k = bi + · · · + bk, e.g., b12 = b1 + b2 and so on)

b1 b2 b3 b4 b5 b6 b7 b8; f g

  • w0

− → b1 b2 b4 b3 b5 b6 b7 b8; f g

  • ,

b1 b2 b3 b4 b5 b6 b7 b8; f g

  • w1

− → b1 b3 b2 b4 b5 b6 b7 b8; f g

  • ,

b1 b2 b3 b4 b5 b6 b7 b8; f g

  • w2

− →

  • b1

b11 − b2 b13 − b2 b14 − b2 b52 − b1 b62 − b1 b72 − b1 b82 − b1; f + b1 − b2 g − b1 + b2,

  • ,

b1 b2 b3 b4 b5 b6 b7 b8; f g

  • w3

− →

  • b1

b217 b317 b417 b5 b6 −b117 b8 − b17; f + b17

(g+b1)(f +b7) f −b1

− b1

  • ,

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 19 / 41

slide-81
SLIDE 81

The Affine Weyl Group W (E (1)

6 ) Theorem Reflections wi are induced by the following elementary birational mappings (also denoted by wi)

  • n the family Xb fixing b1 and χ(δ) (we put bi···k = bi + · · · + bk, e.g., b12 = b1 + b2 and so on)

b1 b2 b3 b4 b5 b6 b7 b8; f g

  • w0

− → b1 b2 b4 b3 b5 b6 b7 b8; f g

  • ,

b1 b2 b3 b4 b5 b6 b7 b8; f g

  • w1

− → b1 b3 b2 b4 b5 b6 b7 b8; f g

  • ,

b1 b2 b3 b4 b5 b6 b7 b8; f g

  • w2

− →

  • b1

b11 − b2 b13 − b2 b14 − b2 b52 − b1 b62 − b1 b72 − b1 b82 − b1; f + b1 − b2 g − b1 + b2,

  • ,

b1 b2 b3 b4 b5 b6 b7 b8; f g

  • w3

− →

  • b1

b217 b317 b417 b5 b6 −b117 b8 − b17; f + b17

(g+b1)(f +b7) f −b1

− b1

  • ,

b1 b2 b3 b4 b5 b6 b7 b8; f g

  • w4

− → b1 b2 b3 b4 b5 b6 b8 b7; f g

  • ,

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 19 / 41

slide-82
SLIDE 82

The Affine Weyl Group W (E (1)

6 ) Theorem Reflections wi are induced by the following elementary birational mappings (also denoted by wi)

  • n the family Xb fixing b1 and χ(δ) (we put bi···k = bi + · · · + bk, e.g., b12 = b1 + b2 and so on)

b1 b2 b3 b4 b5 b6 b7 b8; f g

  • w0

− → b1 b2 b4 b3 b5 b6 b7 b8; f g

  • ,

b1 b2 b3 b4 b5 b6 b7 b8; f g

  • w1

− → b1 b3 b2 b4 b5 b6 b7 b8; f g

  • ,

b1 b2 b3 b4 b5 b6 b7 b8; f g

  • w2

− →

  • b1

b11 − b2 b13 − b2 b14 − b2 b52 − b1 b62 − b1 b72 − b1 b82 − b1; f + b1 − b2 g − b1 + b2,

  • ,

b1 b2 b3 b4 b5 b6 b7 b8; f g

  • w3

− →

  • b1

b217 b317 b417 b5 b6 −b117 b8 − b17; f + b17

(g+b1)(f +b7) f −b1

− b1

  • ,

b1 b2 b3 b4 b5 b6 b7 b8; f g

  • w4

− → b1 b2 b3 b4 b5 b6 b8 b7; f g

  • ,

b1 b2 b3 b4 b5 b6 b7 b8; f g

  • w5

− →

  • b1

b215 b315 b415 −b115 b6 − b15 b7 b8 ;

(f −b1)(g−b5) g+b1

+ b1 g − b15

  • ,

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 19 / 41

slide-83
SLIDE 83

The Affine Weyl Group W (E (1)

6 ) Theorem Reflections wi are induced by the following elementary birational mappings (also denoted by wi)

  • n the family Xb fixing b1 and χ(δ) (we put bi···k = bi + · · · + bk, e.g., b12 = b1 + b2 and so on)

b1 b2 b3 b4 b5 b6 b7 b8; f g

  • w0

− → b1 b2 b4 b3 b5 b6 b7 b8; f g

  • ,

b1 b2 b3 b4 b5 b6 b7 b8; f g

  • w1

− → b1 b3 b2 b4 b5 b6 b7 b8; f g

  • ,

b1 b2 b3 b4 b5 b6 b7 b8; f g

  • w2

− →

  • b1

b11 − b2 b13 − b2 b14 − b2 b52 − b1 b62 − b1 b72 − b1 b82 − b1; f + b1 − b2 g − b1 + b2,

  • ,

b1 b2 b3 b4 b5 b6 b7 b8; f g

  • w3

− →

  • b1

b217 b317 b417 b5 b6 −b117 b8 − b17; f + b17

(g+b1)(f +b7) f −b1

− b1

  • ,

b1 b2 b3 b4 b5 b6 b7 b8; f g

  • w4

− → b1 b2 b3 b4 b5 b6 b8 b7; f g

  • ,

b1 b2 b3 b4 b5 b6 b7 b8; f g

  • w5

− →

  • b1

b215 b315 b415 −b115 b6 − b15 b7 b8 ;

(f −b1)(g−b5) g+b1

+ b1 g − b15

  • ,

b1 b2 b3 b4 b5 b6 b7 b8; f g

  • w6

− → b1 b2 b3 b4 b6 b5 b7 b8; f g

  • .

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 19 / 41

slide-84
SLIDE 84

Sketch of the proof

Since α0 = E2 − E3, w0 : E2 ↔ E3, which just swaps the parameters b2 and b3.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 20 / 41

slide-85
SLIDE 85

Sketch of the proof

Since α0 = E2 − E3, w0 : E2 ↔ E3, which just swaps the parameters b2 and b3. Similarly, since α2 = E1 − E2, w2 : E1 ↔ E2, which swaps b1 and b2, but then we need to use the normalization freedom to ensure that b1 is fixed, b2 b1 b4 b3 b5 b6 b7 b8; f g

  • b1

b11 − b2 b13 − b2 b14 − b2 b52 − b1 b62 − b1 b72 − b1 b82 − b1; f + b1 − b2 g − b1 + b2,

  • .

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 20 / 41

slide-86
SLIDE 86

Sketch of the proof

Since α0 = E2 − E3, w0 : E2 ↔ E3, which just swaps the parameters b2 and b3. Similarly, since α2 = E1 − E2, w2 : E1 ↔ E2, which swaps b1 and b2, but then we need to use the normalization freedom to ensure that b1 is fixed, b2 b1 b4 b3 b5 b6 b7 b8; f g

  • b1

b11 − b2 b13 − b2 b14 − b2 b52 − b1 b62 − b1 b72 − b1 b82 − b1; f + b1 − b2 g − b1 + b2,

  • .

Consider now α3 = Hf − E1 − E7. Then w3(Hf ) = Hf , w3(Hg) = Hf + Hg − E1 − E7, w3(E1) = Hf − E7, w3(E7) = Hf − E1, and w3(Ei) = Ei otherwise.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 20 / 41

slide-87
SLIDE 87

Sketch of the proof

Since α0 = E2 − E3, w0 : E2 ↔ E3, which just swaps the parameters b2 and b3. Similarly, since α2 = E1 − E2, w2 : E1 ↔ E2, which swaps b1 and b2, but then we need to use the normalization freedom to ensure that b1 is fixed, b2 b1 b4 b3 b5 b6 b7 b8; f g

  • b1

b11 − b2 b13 − b2 b14 − b2 b52 − b1 b62 − b1 b72 − b1 b82 − b1; f + b1 − b2 g − b1 + b2,

  • .

Consider now α3 = Hf − E1 − E7. Then w3(Hf ) = Hf , w3(Hg) = Hf + Hg − E1 − E7, w3(E1) = Hf − E7, w3(E7) = Hf − E1, and w3(Ei) = Ei otherwise. Thus, w−1

3

(H¯

g) = Hf + Hg − E1 − E7, i.e., ¯

g is a coordinate on a pencil of (1, 1) curves passing through p1(b1, −b1) and p7(−b7, ∞): |H¯

g| = {Afg + Bf + Cg + D = 0 | −Ab2 1 + (B − C)b1 + D = −Ab7 + C = 0}

= {A(fg + b7g + b2

1 + b1b7) + b(f − b1 = 0} =

⇒ ¯ g = P(fg + b7g + b2

1 + b1b7) + Q(f − b1)

R(fg + b7g + b2

1 + b1b7) + S(f − b1) ,

¯ f = Lf + M Nf + T .

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 20 / 41

slide-88
SLIDE 88

Sketch of the proof

Since α0 = E2 − E3, w0 : E2 ↔ E3, which just swaps the parameters b2 and b3. Similarly, since α2 = E1 − E2, w2 : E1 ↔ E2, which swaps b1 and b2, but then we need to use the normalization freedom to ensure that b1 is fixed, b2 b1 b4 b3 b5 b6 b7 b8; f g

  • b1

b11 − b2 b13 − b2 b14 − b2 b52 − b1 b62 − b1 b72 − b1 b82 − b1; f + b1 − b2 g − b1 + b2,

  • .

Consider now α3 = Hf − E1 − E7. Then w3(Hf ) = Hf , w3(Hg) = Hf + Hg − E1 − E7, w3(E1) = Hf − E7, w3(E7) = Hf − E1, and w3(Ei) = Ei otherwise. Thus, w−1

3

(H¯

g) = Hf + Hg − E1 − E7, i.e., ¯

g is a coordinate on a pencil of (1, 1) curves passing through p1(b1, −b1) and p7(−b7, ∞): |H¯

g| = {Afg + Bf + Cg + D = 0 | −Ab2 1 + (B − C)b1 + D = −Ab7 + C = 0}

= {A(fg + b7g + b2

1 + b1b7) + b(f − b1 = 0} =

⇒ ¯ g = P(fg + b7g + b2

1 + b1b7) + Q(f − b1)

R(fg + b7g + b2

1 + b1b7) + S(f − b1) ,

¯ f = Lf + M Nf + T . Since ¯ g(−b8, ∞) = ∞, R = 0, and since ¯ f (∞, b5) = ∞, N = 0.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 20 / 41

slide-89
SLIDE 89

Sketch of the proof (cont.)

So we have ¯ f = Lf + M, ¯ g = P fg + b7g + b2

1 + b1b7

f − b1 + Q.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 21 / 41

slide-90
SLIDE 90

Sketch of the proof (cont.)

So we have ¯ f = Lf + M, ¯ g = P fg + b7g + b2

1 + b1b7

f − b1 + Q. Then we have (¯ b1, − ¯ b1) = (¯ f , ¯ g)(f = −b7) = (−Lb7 + M, −Pb1 + Q), (¯ b2, − ¯ b2) = (¯ f , ¯ g)(b2, −b2) = (Lb2 + M, −P(b1 + b2 + b7) + Q), (¯ b3, − ¯ b3) = (¯ f , ¯ g)(b3, −b3) = (Lb3 + M, −P(b1 + b3 + b7) + Q), (¯ b4, − ¯ b4) = (¯ f , ¯ g)(b4, −b4) = (Lb4 + M, −P(b1 + b4 + b7) + Q), (∞, ¯ b5) = (¯ f , ¯ g)(∞, b5) = (∞, Pb5 + Q), (∞, ¯ b6) = (¯ f , ¯ g)(∞, b6) = (∞, Pb6 + Q), (− ¯ b7, ∞) = (¯ f , ¯ g)(f = b1) = (Lb1 + M, ∞), (− ¯ b8, ∞) = (¯ f , ¯ g)(−b8, ∞) = (−Lb8 + M, ∞).

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 21 / 41

slide-91
SLIDE 91

Sketch of the proof (cont.)

So we have ¯ f = Lf + M, ¯ g = P fg + b7g + b2

1 + b1b7

f − b1 + Q. Then we have (¯ b1, − ¯ b1) = (¯ f , ¯ g)(f = −b7) = (−Lb7 + M, −Pb1 + Q), (¯ b2, − ¯ b2) = (¯ f , ¯ g)(b2, −b2) = (Lb2 + M, −P(b1 + b2 + b7) + Q), (¯ b3, − ¯ b3) = (¯ f , ¯ g)(b3, −b3) = (Lb3 + M, −P(b1 + b3 + b7) + Q), (¯ b4, − ¯ b4) = (¯ f , ¯ g)(b4, −b4) = (Lb4 + M, −P(b1 + b4 + b7) + Q), (∞, ¯ b5) = (¯ f , ¯ g)(∞, b5) = (∞, Pb5 + Q), (∞, ¯ b6) = (¯ f , ¯ g)(∞, b6) = (∞, Pb6 + Q), (− ¯ b7, ∞) = (¯ f , ¯ g)(f = b1) = (Lb1 + M, ∞), (− ¯ b8, ∞) = (¯ f , ¯ g)(−b8, ∞) = (−Lb8 + M, ∞). The first four equations give L = P and then to preserve χ(δ) we must have L = P = 1. Then, to fix b1, M = b1 + b7 and Q = 0, which gives the required map. Other cases are similar to the ones considered.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 21 / 41

slide-92
SLIDE 92

The Automorphism Group Aut(A(1)

2 ) ≃ Aut(E (1) 6 ) ≃ D3 Theorem The acton of the automorphisms on the Picard lattice Pic(X), the symmetry sub-lattice SpanZ{αi} and the surface sub-lattice SpanZ{Di} is given by: m0 = (D1D2) = (α3α5)(α4α6), Hf → Hg, E1 → E1, E3 → E3, E5 → E7, E7 → E5, Hg → Hf , E2 → E2, E4 → E4, E6 → E8, E8 → E6; m1 = (D0D2) = (α0α4)(α1α3), Hf → Hf , E1 → Hf − E2, E3 → E7, E5 → E5, E7 → E3, Hg → Hf + Hg − E1 − E2, E2 → Hf − E1, E4 → E8, E6 → E6, E8 → E4; m2 = (D0D1) = (α0α6)(α1α5), Hf → Hf + Hg − E1 − E2, E1 → Hg − E2, E3 → E5, E5 → E3, E7 → E7, Hg → Hg, E2 → Hg − E1, E4 → E6, E6 → E4, E8 → E8; r = (D0D1D2) = (α0α6α4)(α1α5α3), Hf → Hg, E1 → Hg − E2, E3 → E5, E5 → E7, E7 → E3, Hg → Hf + Hg − E1 − E2, E2 → Hg − E1, E4 → E6, E6 → E8, E8 → E4; r2 = (D0D2D1) = (α0α4α6)(α1α3α5), Hf → Hf + Hg − E1 − E2, E1 → Hf − E2, E3 → E7, E5 → E3, E7 → E5, Hg → Hf , E2 → Hf − E1, E4 → E8, E6 → E4, E8 → E6.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 22 / 41

slide-93
SLIDE 93

Sketch of the proof

This is almost obvious from looking at the diagrams. For example, for m2 we have

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 23 / 41

slide-94
SLIDE 94

Sketch of the proof

This is almost obvious from looking at the diagrams. For example, for m2 we have D0 = Hf + Hg − E1 − E2 − E3 − E4 D1 = Hf − E5 − E6 D2 = Hg − E7 − E8 α6 = E5 − E6 α5 = Hg − E1 − E5 α2 = E1 − E2 α1 = E2 − E3 α0 = E3 − E4 α3 = Hf − E1 − E7 α4 = E7 − E8

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 23 / 41

slide-95
SLIDE 95

Sketch of the proof

This is almost obvious from looking at the diagrams. For example, for m2 we have D0 = Hf + Hg − E1 − E2 − E3 − E4 D1 = Hf − E5 − E6 D2 = Hg − E7 − E8 α6 = E5 − E6 α5 = Hg − E1 − E5 α2 = E1 − E2 α1 = E2 − E3 α0 = E3 − E4 α3 = Hf − E1 − E7 α4 = E7 − E8 Hence, m2 is given by Hf → Hf + Hg − E1 − E2, E1 → Hg − E2, E3 → E5, E5 → E3, E7 → E7, Hg → Hg, E2 → Hg − E1, E4 → E6, E6 → E4, E8 → E8;

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 23 / 41

slide-96
SLIDE 96

The Automorphism Group Aut(A(1)

2 ) ≃ Aut(E (1) 6 ) ≃ D3 Theorem The automorphisms are given by the following elementary birational maps on the family Xb fixing b1 and χ(δ) b1 b2 b3 b4 b5 b6 b7 b8; f g

  • m0

− → b1 b2 b4 b3 b7 b8 b5 b6; −f −g

  • ,

b1 b2 b3 b4 b5 b6 b7 b8; f g

  • m1

− →

  • b1

b2 b127 b128 b5 b6 b3 − b12 b4 − b12; b12 − f

g(f −b12)−b1b2 f +g

  • ,

b1 b2 b3 b4 b5 b6 b7 b8; f g

  • m2

− →

  • b1

b2 b125 b126 b3 − b12 b4 − b12 b7 b8 ;

f (g+b12)−b1b2 f +g

−g − b12

  • ,

b1 b2 b3 b4 b5 b6 b7 b8; f g

  • r

− →

  • b1

b2 b127 b128 b3 − b12 b4 − b12 b5 b6 ; − g(f −b12)−b1b2

f +g

f − b12

  • ,

b1 b2 b3 b4 b5 b6 b7 b8; f g

  • r2

− →

  • b1

b2 b125 b126 b7 b8 b3 − b12 b4 − b12; g + b12 − f (g+b12)−b1b2

f +g

  • .

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 24 / 41

slide-97
SLIDE 97

The Automorphism Group Aut(A(1)

2 ) ≃ Aut(E (1) 6 ) ≃ D3 Theorem The automorphisms are given by the following elementary birational maps on the family Xb fixing b1 and χ(δ) b1 b2 b3 b4 b5 b6 b7 b8; f g

  • m0

− → b1 b2 b4 b3 b7 b8 b5 b6; −f −g

  • ,

b1 b2 b3 b4 b5 b6 b7 b8; f g

  • m1

− →

  • b1

b2 b127 b128 b5 b6 b3 − b12 b4 − b12; b12 − f

g(f −b12)−b1b2 f +g

  • ,

b1 b2 b3 b4 b5 b6 b7 b8; f g

  • m2

− →

  • b1

b2 b125 b126 b3 − b12 b4 − b12 b7 b8 ;

f (g+b12)−b1b2 f +g

−g − b12

  • ,

b1 b2 b3 b4 b5 b6 b7 b8; f g

  • r

− →

  • b1

b2 b127 b128 b3 − b12 b4 − b12 b5 b6 ; − g(f −b12)−b1b2

f +g

f − b12

  • ,

b1 b2 b3 b4 b5 b6 b7 b8; f g

  • r2

− →

  • b1

b2 b125 b126 b7 b8 b3 − b12 b4 − b12; g + b12 − f (g+b12)−b1b2

f +g

  • .

Proof is similar to the previous theorem. Notice that the group structure is preserved on the level

  • f the maps.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 24 / 41

slide-98
SLIDE 98

The Semi-Direct Product Structure

The extended affine Weyl group W (E (1)

6

) is a semi-direct product of its normal subgroup W (E (1)

6

) ⊳ W (E (1)

6

) and the subgroup of the diagram automorphisms Aut(E (1)

6

),

  • W (E (1)

6

) = Aut(D(1)

6 ) ⋉ W (D(1) 6 ).

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 25 / 41

slide-99
SLIDE 99

The Semi-Direct Product Structure

The extended affine Weyl group W (E (1)

6

) is a semi-direct product of its normal subgroup W (E (1)

6

) ⊳ W (E (1)

6

) and the subgroup of the diagram automorphisms Aut(E (1)

6

),

  • W (E (1)

6

) = Aut(D(1)

6 ) ⋉ W (D(1) 6 ).

We have just described the group structure of W (E (1)

6

) and Aut(E (1)

6

) using generators and relations, so it remains to give the action of Aut(E (1)

6

) on W (E (1)

6

).

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 25 / 41

slide-100
SLIDE 100

The Semi-Direct Product Structure

The extended affine Weyl group W (E (1)

6

) is a semi-direct product of its normal subgroup W (E (1)

6

) ⊳ W (E (1)

6

) and the subgroup of the diagram automorphisms Aut(E (1)

6

),

  • W (E (1)

6

) = Aut(D(1)

6 ) ⋉ W (D(1) 6 ).

We have just described the group structure of W (E (1)

6

) and Aut(E (1)

6

) using generators and relations, so it remains to give the action of Aut(E (1)

6

) on W (E (1)

6

). But elements of Aut(E (1)

6

) act as permutations of the simple roots αi, and so the action is just the corresponding permutation of the corresponding reflections, σtwαi σ−1

t

= wt(αi ), where t is the permutation of αi’s corresponding to σt.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 25 / 41

slide-101
SLIDE 101

The Semi-Direct Product Structure

The extended affine Weyl group W (E (1)

6

) is a semi-direct product of its normal subgroup W (E (1)

6

) ⊳ W (E (1)

6

) and the subgroup of the diagram automorphisms Aut(E (1)

6

),

  • W (E (1)

6

) = Aut(D(1)

6 ) ⋉ W (D(1) 6 ).

We have just described the group structure of W (E (1)

6

) and Aut(E (1)

6

) using generators and relations, so it remains to give the action of Aut(E (1)

6

) on W (E (1)

6

). But elements of Aut(E (1)

6

) act as permutations of the simple roots αi, and so the action is just the corresponding permutation of the corresponding reflections, σtwαi σ−1

t

= wt(αi ), where t is the permutation of αi’s corresponding to σt. Example: σ1 = σm1 = (α0α4)(α1α3) acts as σ1w0σ1 = w4, σ1w4σ1 = w0, σ1w1σ1 = w3, σ1w3σ1 = w1, σ1wiσ1 = wi

  • therwise .

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 25 / 41

slide-102
SLIDE 102

Decomposition of Translation Elements

Finally, we need an algorithm for representing a translation element of W (E 1

6 ) as a composition

  • f the generators of the group, then the corresponding discrete Painlev´

e equation can be written as a composition of elementary birational maps.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 26 / 41

slide-103
SLIDE 103

Decomposition of Translation Elements

Finally, we need an algorithm for representing a translation element of W (E 1

6 ) as a composition

  • f the generators of the group, then the corresponding discrete Painlev´

e equation can be written as a composition of elementary birational maps. For this, we use the following Lemma:

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 26 / 41

slide-104
SLIDE 104

Decomposition of Translation Elements

Finally, we need an algorithm for representing a translation element of W (E 1

6 ) as a composition

  • f the generators of the group, then the corresponding discrete Painlev´

e equation can be written as a composition of elementary birational maps. For this, we use the following Lemma: Reduction Lemma (V. Kac, Infinite dimensional Lie algebras, Lemma 3.11) If w(αi) < 0, then l(w ◦ wi) < l(w), where l(w) is length of w ∈ W , and αi is a simple root.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 26 / 41

slide-105
SLIDE 105

Decomposition of Translation Elements

Finally, we need an algorithm for representing a translation element of W (E 1

6 ) as a composition

  • f the generators of the group, then the corresponding discrete Painlev´

e equation can be written as a composition of elementary birational maps. For this, we use the following Lemma: Reduction Lemma (V. Kac, Infinite dimensional Lie algebras, Lemma 3.11) If w(αi) < 0, then l(w ◦ wi) < l(w), where l(w) is length of w ∈ W , and αi is a simple root. As an example, consider the following translational mapping: ϕ∗ : (α0, α1, α2, α3, α4, α5, α6) → (α0, α1, α2, α3 + δ, α4, α5 − δ, α6), where δ = α0 + 2α1 + 3α2 + 2α3 + α4 + 2α5 + α6 as usual.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 26 / 41

slide-106
SLIDE 106

Decomposition of Translation Elements

Finally, we need an algorithm for representing a translation element of W (E 1

6 ) as a composition

  • f the generators of the group, then the corresponding discrete Painlev´

e equation can be written as a composition of elementary birational maps. For this, we use the following Lemma: Reduction Lemma (V. Kac, Infinite dimensional Lie algebras, Lemma 3.11) If w(αi) < 0, then l(w ◦ wi) < l(w), where l(w) is length of w ∈ W , and αi is a simple root. As an example, consider the following translational mapping: ϕ∗ : (α0, α1, α2, α3, α4, α5, α6) → (α0, α1, α2, α3 + δ, α4, α5 − δ, α6), where δ = α0 + 2α1 + 3α2 + 2α3 + α4 + 2α5 + α6 as usual. Put α = (α0, α1, α2, α3, α4, α5, α6). Then the algorithm works as follows:

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 26 / 41

slide-107
SLIDE 107

α0 α1 α2 α3 α4 α5 α6

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 27 / 41

slide-108
SLIDE 108

α0 α1 α2 α3 α4 α5 α6 ϕ∗(α) = (α0, α1, α2, α3 + δ, α4, α5 − δ, α6),

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 27 / 41

slide-109
SLIDE 109

α0 α1 α2 α3 α4 α5 α6 ϕ∗(α) = (α0, α1, α2, α3 + δ, α4, α5 − δ, α6),

  • ϕ(1)

= ϕ∗ ◦ w5

  • (α) = (α0, α1, α25 − δ, α3 + δ, α4, δ − α5, α56 − δ),

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 27 / 41

slide-110
SLIDE 110

α0 α1 α2 α3 α4 α5 α6 ϕ∗(α) = (α0, α1, α2, α3 + δ, α4, α5 − δ, α6),

  • ϕ(1)

= ϕ∗ ◦ w5

  • (α) = (α0, α1, α25 − δ, α3 + δ, α4, δ − α5, α56 − δ),
  • ϕ(2)

= ϕ(1)

  • w6
  • (α) = (α0, α1, α25 − δ, α3 + δ, α4, α6, δ − α56),

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 27 / 41

slide-111
SLIDE 111

α0 α1 α2 α3 α4 α5 α6 ϕ∗(α) = (α0, α1, α2, α3 + δ, α4, α5 − δ, α6),

  • ϕ(1)

= ϕ∗ ◦ w5

  • (α) = (α0, α1, α25 − δ, α3 + δ, α4, δ − α5, α56 − δ),
  • ϕ(2)

= ϕ(1)

  • w6
  • (α) = (α0, α1, α25 − δ, α3 + δ, α4, α6, δ − α56),
  • ϕ(3)

= ϕ(2)

  • w2
  • (α) = (α0, α125 − δ, δ − α25, α235, α4, α256 − δ, δ − α56),

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 27 / 41

slide-112
SLIDE 112

α0 α1 α2 α3 α4 α5 α6 ϕ∗(α) = (α0, α1, α2, α3 + δ, α4, α5 − δ, α6),

  • ϕ(1)

= ϕ∗ ◦ w5

  • (α) = (α0, α1, α25 − δ, α3 + δ, α4, δ − α5, α56 − δ),
  • ϕ(2)

= ϕ(1)

  • w6
  • (α) = (α0, α1, α25 − δ, α3 + δ, α4, α6, δ − α56),
  • ϕ(3)

= ϕ(2)

  • w2
  • (α) = (α0, α125 − δ, δ − α25, α235, α4, α256 − δ, δ − α56),
  • ϕ(4)

= ϕ(3)

  • w1
  • (α) = (α0125 − δ, δ − α125, α1, α235, α4, α256 − δ, δ − α56),

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 27 / 41

slide-113
SLIDE 113

α0 α1 α2 α3 α4 α5 α6 ϕ∗(α) = (α0, α1, α2, α3 + δ, α4, α5 − δ, α6),

  • ϕ(1)

= ϕ∗ ◦ w5

  • (α) = (α0, α1, α25 − δ, α3 + δ, α4, δ − α5, α56 − δ),
  • ϕ(2)

= ϕ(1)

  • w6
  • (α) = (α0, α1, α25 − δ, α3 + δ, α4, α6, δ − α56),
  • ϕ(3)

= ϕ(2)

  • w2
  • (α) = (α0, α125 − δ, δ − α25, α235, α4, α256 − δ, δ − α56),
  • ϕ(4)

= ϕ(3)

  • w1
  • (α) = (α0125 − δ, δ − α125, α1, α235, α4, α256 − δ, δ − α56),
  • ϕ(5)

= ϕ(4)

  • w0
  • (α) = (δ − α0125, α0, α1, α235, α4, α256 − δ, δ − α56),

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 27 / 41

slide-114
SLIDE 114

α0 α1 α2 α3 α4 α5 α6 ϕ∗(α) = (α0, α1, α2, α3 + δ, α4, α5 − δ, α6),

  • ϕ(1)

= ϕ∗ ◦ w5

  • (α) = (α0, α1, α25 − δ, α3 + δ, α4, δ − α5, α56 − δ),
  • ϕ(2)

= ϕ(1)

  • w6
  • (α) = (α0, α1, α25 − δ, α3 + δ, α4, α6, δ − α56),
  • ϕ(3)

= ϕ(2)

  • w2
  • (α) = (α0, α125 − δ, δ − α25, α235, α4, α256 − δ, δ − α56),
  • ϕ(4)

= ϕ(3)

  • w1
  • (α) = (α0125 − δ, δ − α125, α1, α235, α4, α256 − δ, δ − α56),
  • ϕ(5)

= ϕ(4)

  • w0
  • (α) = (δ − α0125, α0, α1, α235, α4, α256 − δ, δ − α56),
  • ϕ(6)

= ϕ(5)

  • w5
  • (α) = (δ − α0125, α0, α1256 − δ, α235, α4, δ − α256, α2),

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 27 / 41

slide-115
SLIDE 115

α0 α1 α2 α3 α4 α5 α6 ϕ∗(α) = (α0, α1, α2, α3 + δ, α4, α5 − δ, α6),

  • ϕ(1)

= ϕ∗ ◦ w5

  • (α) = (α0, α1, α25 − δ, α3 + δ, α4, δ − α5, α56 − δ),
  • ϕ(2)

= ϕ(1)

  • w6
  • (α) = (α0, α1, α25 − δ, α3 + δ, α4, α6, δ − α56),
  • ϕ(3)

= ϕ(2)

  • w2
  • (α) = (α0, α125 − δ, δ − α25, α235, α4, α256 − δ, δ − α56),
  • ϕ(4)

= ϕ(3)

  • w1
  • (α) = (α0125 − δ, δ − α125, α1, α235, α4, α256 − δ, δ − α56),
  • ϕ(5)

= ϕ(4)

  • w0
  • (α) = (δ − α0125, α0, α1, α235, α4, α256 − δ, δ − α56),
  • ϕ(6)

= ϕ(5)

  • w5
  • (α) = (δ − α0125, α0, α1256 − δ, α235, α4, δ − α256, α2),
  • ϕ(7)

= ϕ(6)

  • w2
  • (α) = (α12233456, −α1223345, α01223345, −α01234, α4, α1, α2),

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 27 / 41

slide-116
SLIDE 116

α0 α1 α2 α3 α4 α5 α6 ϕ∗(α) = (α0, α1, α2, α3 + δ, α4, α5 − δ, α6),

  • ϕ(1)

= ϕ∗ ◦ w5

  • (α) = (α0, α1, α25 − δ, α3 + δ, α4, δ − α5, α56 − δ),
  • ϕ(2)

= ϕ(1)

  • w6
  • (α) = (α0, α1, α25 − δ, α3 + δ, α4, α6, δ − α56),
  • ϕ(3)

= ϕ(2)

  • w2
  • (α) = (α0, α125 − δ, δ − α25, α235, α4, α256 − δ, δ − α56),
  • ϕ(4)

= ϕ(3)

  • w1
  • (α) = (α0125 − δ, δ − α125, α1, α235, α4, α256 − δ, δ − α56),
  • ϕ(5)

= ϕ(4)

  • w0
  • (α) = (δ − α0125, α0, α1, α235, α4, α256 − δ, δ − α56),
  • ϕ(6)

= ϕ(5)

  • w5
  • (α) = (δ − α0125, α0, α1256 − δ, α235, α4, δ − α256, α2),
  • ϕ(7)

= ϕ(6)

  • w2
  • (α) = (α12233456, −α1223345, α01223345, −α01234, α4, α1, α2),
  • ϕ(8)

= ϕ(7)

  • w1
  • (α) = (α6, α1223345, α0, −α01234, α4, α1, α2),

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 27 / 41

slide-117
SLIDE 117

α0 α1 α2 α3 α4 α5 α6 ϕ∗(α) = (α0, α1, α2, α3 + δ, α4, α5 − δ, α6),

  • ϕ(1)

= ϕ∗ ◦ w5

  • (α) = (α0, α1, α25 − δ, α3 + δ, α4, δ − α5, α56 − δ),
  • ϕ(2)

= ϕ(1)

  • w6
  • (α) = (α0, α1, α25 − δ, α3 + δ, α4, α6, δ − α56),
  • ϕ(3)

= ϕ(2)

  • w2
  • (α) = (α0, α125 − δ, δ − α25, α235, α4, α256 − δ, δ − α56),
  • ϕ(4)

= ϕ(3)

  • w1
  • (α) = (α0125 − δ, δ − α125, α1, α235, α4, α256 − δ, δ − α56),
  • ϕ(5)

= ϕ(4)

  • w0
  • (α) = (δ − α0125, α0, α1, α235, α4, α256 − δ, δ − α56),
  • ϕ(6)

= ϕ(5)

  • w5
  • (α) = (δ − α0125, α0, α1256 − δ, α235, α4, δ − α256, α2),
  • ϕ(7)

= ϕ(6)

  • w2
  • (α) = (α12233456, −α1223345, α01223345, −α01234, α4, α1, α2),
  • ϕ(8)

= ϕ(7)

  • w1
  • (α) = (α6, α1223345, α0, −α01234, α4, α1, α2),
  • ϕ(9)

= ϕ(8)

  • w3
  • (α) = (α6, α1223345, −α1234, α01234, −α0123, α1, α2),

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 27 / 41

slide-118
SLIDE 118

α0 α1 α2 α3 α4 α5 α6 ϕ∗(α) = (α0, α1, α2, α3 + δ, α4, α5 − δ, α6),

  • ϕ(1)

= ϕ∗ ◦ w5

  • (α) = (α0, α1, α25 − δ, α3 + δ, α4, δ − α5, α56 − δ),
  • ϕ(2)

= ϕ(1)

  • w6
  • (α) = (α0, α1, α25 − δ, α3 + δ, α4, α6, δ − α56),
  • ϕ(3)

= ϕ(2)

  • w2
  • (α) = (α0, α125 − δ, δ − α25, α235, α4, α256 − δ, δ − α56),
  • ϕ(4)

= ϕ(3)

  • w1
  • (α) = (α0125 − δ, δ − α125, α1, α235, α4, α256 − δ, δ − α56),
  • ϕ(5)

= ϕ(4)

  • w0
  • (α) = (δ − α0125, α0, α1, α235, α4, α256 − δ, δ − α56),
  • ϕ(6)

= ϕ(5)

  • w5
  • (α) = (δ − α0125, α0, α1256 − δ, α235, α4, δ − α256, α2),
  • ϕ(7)

= ϕ(6)

  • w2
  • (α) = (α12233456, −α1223345, α01223345, −α01234, α4, α1, α2),
  • ϕ(8)

= ϕ(7)

  • w1
  • (α) = (α6, α1223345, α0, −α01234, α4, α1, α2),
  • ϕ(9)

= ϕ(8)

  • w3
  • (α) = (α6, α1223345, −α1234, α01234, −α0123, α1, α2),
  • ϕ(10)

= ϕ(9)

  • w4
  • (α) = (α6, α1223345, −α1234, α4, α0123, α1, α2),

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 27 / 41

slide-119
SLIDE 119

α0 α1 α2 α3 α4 α5 α6 ϕ∗(α) = (α0, α1, α2, α3 + δ, α4, α5 − δ, α6),

  • ϕ(1)

= ϕ∗ ◦ w5

  • (α) = (α0, α1, α25 − δ, α3 + δ, α4, δ − α5, α56 − δ),
  • ϕ(2)

= ϕ(1)

  • w6
  • (α) = (α0, α1, α25 − δ, α3 + δ, α4, α6, δ − α56),
  • ϕ(3)

= ϕ(2)

  • w2
  • (α) = (α0, α125 − δ, δ − α25, α235, α4, α256 − δ, δ − α56),
  • ϕ(4)

= ϕ(3)

  • w1
  • (α) = (α0125 − δ, δ − α125, α1, α235, α4, α256 − δ, δ − α56),
  • ϕ(5)

= ϕ(4)

  • w0
  • (α) = (δ − α0125, α0, α1, α235, α4, α256 − δ, δ − α56),
  • ϕ(6)

= ϕ(5)

  • w5
  • (α) = (δ − α0125, α0, α1256 − δ, α235, α4, δ − α256, α2),
  • ϕ(7)

= ϕ(6)

  • w2
  • (α) = (α12233456, −α1223345, α01223345, −α01234, α4, α1, α2),
  • ϕ(8)

= ϕ(7)

  • w1
  • (α) = (α6, α1223345, α0, −α01234, α4, α1, α2),
  • ϕ(9)

= ϕ(8)

  • w3
  • (α) = (α6, α1223345, −α1234, α01234, −α0123, α1, α2),
  • ϕ(10)

= ϕ(9)

  • w4
  • (α) = (α6, α1223345, −α1234, α4, α0123, α1, α2),
  • ϕ(11)

= ϕ(10)

  • w2
  • (α) = (α6, α235, α1234, −α123, α0123, −α234, α2),

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 27 / 41

slide-120
SLIDE 120

α0 α1 α2 α3 α4 α5 α6

  • ϕ(10)

= ϕ(9)

  • w4
  • (α) = (α6, α1223345, −α1234, α4, α0123, α1, α2),

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 28 / 41

slide-121
SLIDE 121

α0 α1 α2 α3 α4 α5 α6

  • ϕ(10)

= ϕ(9)

  • w4
  • (α) = (α6, α1223345, −α1234, α4, α0123, α1, α2),
  • ϕ(11)

= ϕ(10)

  • w2
  • (α) = (α6, α235, α1234, −α123, α0123, −α234, α2),

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 28 / 41

slide-122
SLIDE 122

α0 α1 α2 α3 α4 α5 α6

  • ϕ(10)

= ϕ(9)

  • w4
  • (α) = (α6, α1223345, −α1234, α4, α0123, α1, α2),
  • ϕ(11)

= ϕ(10)

  • w2
  • (α) = (α6, α235, α1234, −α123, α0123, −α234, α2),
  • ϕ(12)

= ϕ(11)

  • w5
  • (α) = (α6, α235, α1, −α123, α0123, α234, −α34),

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 28 / 41

slide-123
SLIDE 123

α0 α1 α2 α3 α4 α5 α6

  • ϕ(10)

= ϕ(9)

  • w4
  • (α) = (α6, α1223345, −α1234, α4, α0123, α1, α2),
  • ϕ(11)

= ϕ(10)

  • w2
  • (α) = (α6, α235, α1234, −α123, α0123, −α234, α2),
  • ϕ(12)

= ϕ(11)

  • w5
  • (α) = (α6, α235, α1, −α123, α0123, α234, −α34),
  • ϕ(13)

= ϕ(12)

  • w5
  • (α) = (α6, α235, α1, −α123, α0123, α2, α34),

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 28 / 41

slide-124
SLIDE 124

α0 α1 α2 α3 α4 α5 α6

  • ϕ(10)

= ϕ(9)

  • w4
  • (α) = (α6, α1223345, −α1234, α4, α0123, α1, α2),
  • ϕ(11)

= ϕ(10)

  • w2
  • (α) = (α6, α235, α1234, −α123, α0123, −α234, α2),
  • ϕ(12)

= ϕ(11)

  • w5
  • (α) = (α6, α235, α1, −α123, α0123, α234, −α34),
  • ϕ(13)

= ϕ(12)

  • w5
  • (α) = (α6, α235, α1, −α123, α0123, α2, α34),
  • ϕ(14)

= ϕ(13)

  • w3
  • (α) = (α6, α235, −α23, α123, α0, α2, α34),

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 28 / 41

slide-125
SLIDE 125

α0 α1 α2 α3 α4 α5 α6

  • ϕ(10)

= ϕ(9)

  • w4
  • (α) = (α6, α1223345, −α1234, α4, α0123, α1, α2),
  • ϕ(11)

= ϕ(10)

  • w2
  • (α) = (α6, α235, α1234, −α123, α0123, −α234, α2),
  • ϕ(12)

= ϕ(11)

  • w5
  • (α) = (α6, α235, α1, −α123, α0123, α234, −α34),
  • ϕ(13)

= ϕ(12)

  • w5
  • (α) = (α6, α235, α1, −α123, α0123, α2, α34),
  • ϕ(14)

= ϕ(13)

  • w3
  • (α) = (α6, α235, −α23, α123, α0, α2, α34),
  • ϕ(15)

= ϕ(14)

  • w2
  • (α) = (α6, α5, α23, α1, α0, −α3, α34),

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 28 / 41

slide-126
SLIDE 126

α0 α1 α2 α3 α4 α5 α6

  • ϕ(10)

= ϕ(9)

  • w4
  • (α) = (α6, α1223345, −α1234, α4, α0123, α1, α2),
  • ϕ(11)

= ϕ(10)

  • w2
  • (α) = (α6, α235, α1234, −α123, α0123, −α234, α2),
  • ϕ(12)

= ϕ(11)

  • w5
  • (α) = (α6, α235, α1, −α123, α0123, α234, −α34),
  • ϕ(13)

= ϕ(12)

  • w5
  • (α) = (α6, α235, α1, −α123, α0123, α2, α34),
  • ϕ(14)

= ϕ(13)

  • w3
  • (α) = (α6, α235, −α23, α123, α0, α2, α34),
  • ϕ(15)

= ϕ(14)

  • w2
  • (α) = (α6, α5, α23, α1, α0, −α3, α34),
  • ϕ(16)

= ϕ(15)

  • w5
  • (α) = (α6, α5, α2, α1, α0, α3, α4),

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 28 / 41

slide-127
SLIDE 127

α0 α1 α2 α3 α4 α5 α6

  • ϕ(10)

= ϕ(9)

  • w4
  • (α) = (α6, α1223345, −α1234, α4, α0123, α1, α2),
  • ϕ(11)

= ϕ(10)

  • w2
  • (α) = (α6, α235, α1234, −α123, α0123, −α234, α2),
  • ϕ(12)

= ϕ(11)

  • w5
  • (α) = (α6, α235, α1, −α123, α0123, α234, −α34),
  • ϕ(13)

= ϕ(12)

  • w5
  • (α) = (α6, α235, α1, −α123, α0123, α2, α34),
  • ϕ(14)

= ϕ(13)

  • w3
  • (α) = (α6, α235, −α23, α123, α0, α2, α34),
  • ϕ(15)

= ϕ(14)

  • w2
  • (α) = (α6, α5, α23, α1, α0, −α3, α34),
  • ϕ(16)

= ϕ(15)

  • w5
  • (α) = (α6, α5, α2, α1, α0, α3, α4),
  • ϕ(17)

= ϕ(15)

  • σr2
  • (α) = (α0, α1, α2, α3, α4, α5, α6).

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 28 / 41

slide-128
SLIDE 128

α0 α1 α2 α3 α4 α5 α6

  • ϕ(10)

= ϕ(9)

  • w4
  • (α) = (α6, α1223345, −α1234, α4, α0123, α1, α2),
  • ϕ(11)

= ϕ(10)

  • w2
  • (α) = (α6, α235, α1234, −α123, α0123, −α234, α2),
  • ϕ(12)

= ϕ(11)

  • w5
  • (α) = (α6, α235, α1, −α123, α0123, α234, −α34),
  • ϕ(13)

= ϕ(12)

  • w5
  • (α) = (α6, α235, α1, −α123, α0123, α2, α34),
  • ϕ(14)

= ϕ(13)

  • w3
  • (α) = (α6, α235, −α23, α123, α0, α2, α34),
  • ϕ(15)

= ϕ(14)

  • w2
  • (α) = (α6, α5, α23, α1, α0, −α3, α34),
  • ϕ(16)

= ϕ(15)

  • w5
  • (α) = (α6, α5, α2, α1, α0, α3, α4),
  • ϕ(17)

= ϕ(15)

  • σr2
  • (α) = (α0, α1, α2, α3, α4, α5, α6).

Thus, ϕ∗ = σr ◦ w5 ◦ w2 ◦ w3 ◦ w6 ◦ w5 ◦ w2 ◦ w4 ◦ w3 ◦ w1 ◦ w2 ◦ w5 ◦ w0 ◦ w1 ◦ w2 ◦ w6 ◦ w5

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 28 / 41

slide-129
SLIDE 129

Our Question

Recall our question:

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 29 / 41

slide-130
SLIDE 130

Our Question

Recall our question: Using our understanding of W (E (1)

6

), compare the following

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 29 / 41

slide-131
SLIDE 131

Our Question

Recall our question: Using our understanding of W (E (1)

6

), compare the following Equation obtained by B. Grammaticos, A. Ramani, and Y. Ohta by the application of the singularity confinement criterion to a deautonomization of a QRT map.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 29 / 41

slide-132
SLIDE 132

Our Question

Recall our question: Using our understanding of W (E (1)

6

), compare the following Equation obtained by B. Grammaticos, A. Ramani, and Y. Ohta by the application of the singularity confinement criterion to a deautonomization of a QRT map. Equation obtained by T. Takenawa and A. D. as a reduction of an elementary Schlesinger transformation of a Fuchsian system.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 29 / 41

slide-133
SLIDE 133

Our Question

Recall our question: Using our understanding of W (E (1)

6

), compare the following Equation obtained by B. Grammaticos, A. Ramani, and Y. Ohta by the application of the singularity confinement criterion to a deautonomization of a QRT map. Equation obtained by T. Takenawa and A. D. as a reduction of an elementary Schlesinger transformation of a Fuchsian system. First let us review these equations.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 29 / 41

slide-134
SLIDE 134

Difference Painlev´ e Equation of Type d-P(A(1)∗

2

): Deautonomization

The following example of a d-P(A(1)∗

2

) equation was first obtained by B. Grammaticos,

  • A. Ramani, and Y. Ohta back around 1996 by applying the singularity confinement criterion to

deautonomization of an integrable discrete autonomous mapping; due to the simplicity structure

  • f the equation we will refer to it as a model example.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 30 / 41

slide-135
SLIDE 135

Difference Painlev´ e Equation of Type d-P(A(1)∗

2

): Deautonomization

The following example of a d-P(A(1)∗

2

) equation was first obtained by B. Grammaticos,

  • A. Ramani, and Y. Ohta back around 1996 by applying the singularity confinement criterion to

deautonomization of an integrable discrete autonomous mapping; due to the simplicity structure

  • f the equation we will refer to it as a model example.

The Model Example of d-P(A(1)∗

2

) We consider a birational map ϕ : P1 × P1 P1 × P1 with parameters b1, . . . , b8: ϕ : b1 b2 b3 b4 b5 b6 b7 b8; f , g

¯ b1 ¯ b2 ¯ b3 ¯ b4 ¯ b5 ¯ b6 ¯ b7 ¯ b8; ¯ f , ¯ g

  • ,

δ = b1 + b2 + b3 + b4 + b5 + b6 + b7 + b8 ¯ b1 = b1, ¯ b3 = b3, ¯ b5 = b5 + δ, ¯ b7 = b7 − δ ¯ b2 = b2, ¯ b4 = b4, ¯ b6 = b6 + δ, ¯ b8 = b8 − δ, and ¯ f and ¯ g are given by the equation          (f + g)(¯ f + g) = (g + b1)(g + b2)(g + b3)(g + b4) (g − b5)(g − b6) (¯ f + g)(¯ f + ¯ g) = (¯ f − ¯ b1)(¯ f − ¯ b2)(¯ f − ¯ b3)(¯ f − ¯ b4) (¯ f + ¯ b7)(¯ f + ¯ b8) .

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 30 / 41

slide-136
SLIDE 136

Difference Painlev´ e Equation of Type d-P(A(1)∗

2

): Deautonomization

The singularity structure of this example is the same as in our model:

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 31 / 41

slide-137
SLIDE 137

Difference Painlev´ e Equation of Type d-P(A(1)∗

2

): Deautonomization

The singularity structure of this example is the same as in our model:

p7(−b7, ∞) p8(−b8, ∞) p5(∞, b5) p6(∞, b6) p1(b1, −b1) p2(b2, −b2) p3(b3, −b3) p4(b4, −b4) g = ∞ Hg g = 0 Hg f = 0 f = ∞ Hf Hf Hf + Hg f + g = 0 Hg − E7 − E8 Hf − E5 − E6 Hf + Hg − E1 − E2 − E3 − E4 E1 E2 E3 E4 E5 E6 E7 E8

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 31 / 41

slide-138
SLIDE 138

Difference Painlev´ e Equation of Type d-P(A(1)∗

2

): Deautonomization

The singularity structure of this example is the same as in our model:

p7(−b7, ∞) p8(−b8, ∞) p5(∞, b5) p6(∞, b6) p1(b1, −b1) p2(b2, −b2) p3(b3, −b3) p4(b4, −b4) g = ∞ Hg g = 0 Hg f = 0 f = ∞ Hf Hf Hf + Hg f + g = 0 Hg − E7 − E8 Hf − E5 − E6 Hf + Hg − E1 − E2 − E3 − E4 E1 E2 E3 E4 E5 E6 E7 E8

Now let us compute the action of this mapping on Pic(X)

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 31 / 41

slide-139
SLIDE 139

Difference Painlev´ e Equation of Type d-P(A(1)∗

2

): Deautonomization

The action of ϕ∗ on Pic(X) Finally, we compute the action of ϕ∗ on Pic(X) to be Hf → 6Hf + 3Hg − 2E1 − 2E2 − 2E3 − 2E4 − E5 − E6 − 3E7 − 3E8, Hg → 3Hf + Hg − E1 − E2 − E3 − E4 − E7 − E8, E1 → 2Hf + Hg − E2 − E3 − E4 − E7 − E8, E2 → 2Hf + Hg − E1 − E3 − E4 − E7 − E8, E3 → 2Hf + Hg − E1 − E2 − E4 − E7 − E8, E4 → 2Hf + Hg − E1 − E2 − E3 − E7 − E8, E5 → 3Hf + Hg − E1 − E2 − E3 − E4 − E6 − E7 − E8, E6 → 3Hf + Hg − E1 − E2 − E3 − E4 − E5 − E7 − E8, E7 → Hf − E8, E8 → Hf − E7,

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 32 / 41

slide-140
SLIDE 140

Difference Painlev´ e Equation of Type d-P(A(1)∗

2

): Deautonomization

The action of ϕ∗ on Pic(X) Finally, we compute the action of ϕ∗ on Pic(X) to be Hf → 6Hf + 3Hg − 2E1 − 2E2 − 2E3 − 2E4 − E5 − E6 − 3E7 − 3E8, Hg → 3Hf + Hg − E1 − E2 − E3 − E4 − E7 − E8, E1 → 2Hf + Hg − E2 − E3 − E4 − E7 − E8, E2 → 2Hf + Hg − E1 − E3 − E4 − E7 − E8, E3 → 2Hf + Hg − E1 − E2 − E4 − E7 − E8, E4 → 2Hf + Hg − E1 − E2 − E3 − E7 − E8, E5 → 3Hf + Hg − E1 − E2 − E3 − E4 − E6 − E7 − E8, E6 → 3Hf + Hg − E1 − E2 − E3 − E4 − E5 − E7 − E8, E7 → Hf − E8, E8 → Hf − E7, and so the induced action ϕ∗ on the sub-lattice R⊥ is given by the following translation: (α0, α1, α2, α3, α4, α5, α6) → (α0, α1, α2, α3, α4, α5, α6) + (0, 0, 0, 1, 0, −1, 0)δ, as well as the permutation σr = (D0D1D2) of the irreducible components of −KX .

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 32 / 41

slide-141
SLIDE 141

Difference Painlev´ e Equation of Type d-P(A(1)∗

2

): Deautonomization

The action of ϕ∗ on Pic(X) Finally, we compute the action of ϕ∗ on Pic(X) to be Hf → 6Hf + 3Hg − 2E1 − 2E2 − 2E3 − 2E4 − E5 − E6 − 3E7 − 3E8, Hg → 3Hf + Hg − E1 − E2 − E3 − E4 − E7 − E8, E1 → 2Hf + Hg − E2 − E3 − E4 − E7 − E8, E2 → 2Hf + Hg − E1 − E3 − E4 − E7 − E8, E3 → 2Hf + Hg − E1 − E2 − E4 − E7 − E8, E4 → 2Hf + Hg − E1 − E2 − E3 − E7 − E8, E5 → 3Hf + Hg − E1 − E2 − E3 − E4 − E6 − E7 − E8, E6 → 3Hf + Hg − E1 − E2 − E3 − E4 − E5 − E7 − E8, E7 → Hf − E8, E8 → Hf − E7, and so the induced action ϕ∗ on the sub-lattice R⊥ is given by the following translation: (α0, α1, α2, α3, α4, α5, α6) → (α0, α1, α2, α3, α4, α5, α6) + (0, 0, 0, 1, 0, −1, 0)δ, as well as the permutation σr = (D0D1D2) of the irreducible components of −KX . Hence ϕ∗ = σr ◦ w5 ◦ w2 ◦ w3 ◦ w6 ◦ w5 ◦ w2 ◦ w4 ◦ w3 ◦ w1 ◦ w2 ◦ w5 ◦ w0 ◦ w1 ◦ w2 ◦ w6 ◦ w5.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 32 / 41

slide-142
SLIDE 142

Sakai’s Classification Scheme for Discrete Painlev´ e Equations.

Recall Sakai’s classification scheme:

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 33 / 41

slide-143
SLIDE 143

Sakai’s Classification Scheme for Discrete Painlev´ e Equations.

Recall Sakai’s classification scheme:

  • E(1)

8 e

  • E(1)

8 q

  • E(1)

7 q

  • E(1)

6 q

  • D(1)

5 q

  • A(1)

4 q

  • (A2 + A1)(1)q
  • (A1 + A1)(1)q

|α|2 = 14

  • A(1)

1 q |α|2 = 4

  • A(1)

q

  • A(1)

1 q |α|2=8

  • E(1)

8 δ

  • E(1)

7 δ

  • E(1)

6 δ

  • D(1)

4 c,δ

  • A(1)

3 c,δ

  • 2A(1)

1 c,δ

  • A(1)

1 c,δ

  • A(1)

c

  • 2A(1)

1 c,δ

  • A(1)

1 c,δ

  • A(1)

c q-PVI q-PV q-PIV, q-PIII q-PI PVI, d-PV PV, d-PIV d-PIII PIII

  • alt. d-PII

PIII PIII PIV, d-PII PII, alt.d-PI PI

Symmetry-type classification scheme for Painlev´ e equations

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 33 / 41

slide-144
SLIDE 144

Sakai’s Classification Scheme for Discrete Painlev´ e Equations.

Recall Sakai’s classification scheme:

  • E(1)

8 e

  • E(1)

8 q

  • E(1)

7 q

  • E(1)

6 q

  • D(1)

5 q

  • A(1)

4 q

  • (A2 + A1)(1)q
  • (A1 + A1)(1)q

|α|2 = 14

  • A(1)

1 q |α|2 = 4

  • A(1)

q

  • A(1)

1 q |α|2=8

  • E(1)

8 δ

  • E(1)

7 δ

  • E(1)

6 δ

  • D(1)

4 c,δ

  • A(1)

3 c,δ

  • 2A(1)

1 c,δ

  • A(1)

1 c,δ

  • A(1)

c

  • 2A(1)

1 c,δ

  • A(1)

1 c,δ

  • A(1)

c q-PVI q-PV q-PIV, q-PIII q-PI PVI, d-PV PV, d-PIV d-PIII PIII

  • alt. d-PII

PIII PIII PIV, d-PII PII, alt.d-PI PI

The differential part of the classification scheme

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 33 / 41

slide-145
SLIDE 145

Sakai’s Classification Scheme for Discrete Painlev´ e Equations.

Recall Sakai’s classification scheme:

  • E(1)

8 e

  • E(1)

8 q

  • E(1)

7 q

  • E(1)

6 q

  • D(1)

5 q

  • A(1)

4 q

  • (A2 + A1)(1)q
  • (A1 + A1)(1)q

|α|2 = 14

  • A(1)

1 q |α|2 = 4

  • A(1)

q

  • A(1)

1 q |α|2=8

  • E(1)

8 δ

  • E(1)

7 δ

  • E(1)

6 δ

  • D(1)

4 c,δ

  • A(1)

3 c,δ

  • 2A(1)

1 c,δ

  • A(1)

1 c,δ

  • A(1)

c

  • 2A(1)

1 c,δ

  • A(1)

1 c,δ

  • A(1)

c q-PVI q-PV q-PIV, q-PIII q-PI PVI, d-PV PV, d-PIV d-PIII PIII

  • alt. d-PII

PIII PIII PIV, d-PII PII, alt.d-PI PI

The purely discrete part of the classification scheme: why Painlev´ e?

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 33 / 41

slide-146
SLIDE 146

Sakai’s Classification Scheme for Discrete Painlev´ e Equations.

Recall Sakai’s classification scheme:

  • E(1)

8 e

  • E(1)

8 q

  • E(1)

7 q

  • E(1)

6 q

  • D(1)

5 q

  • A(1)

4 q

  • (A2 + A1)(1)q
  • (A1 + A1)(1)q

|α|2 = 14

  • A(1)

1 q |α|2 = 4

  • A(1)

q

  • A(1)

1 q |α|2=8

  • E(1)

8 δ

  • E(1)

7 δ

  • E(1)

6 δ

  • D(1)

4 c,δ

  • A(1)

3 c,δ

  • 2A(1)

1 c,δ

  • A(1)

1 c,δ

  • A(1)

c

  • 2A(1)

1 c,δ

  • A(1)

1 c,δ

  • A(1)

c q-PVI q-PV q-PIV, q-PIII q-PI PVI, d-PV PV, d-PIV d-PIII PIII

  • alt. d-PII

PIII PIII PIV, d-PII PII, alt.d-PI PI

Isomonodromic approach: difference Painlev´ e equations as reductions from Schlesinger transformations of Fuchsian systems

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 33 / 41

slide-147
SLIDE 147

Difference Painlev´ e Equation of Type d-P(A(1)∗

2

): Schlesinger Transformations

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 34 / 41

slide-148
SLIDE 148

Difference Painlev´ e Equation of Type d-P(A(1)∗

2

): Schlesinger Transformations

So pure difference Painlev´ e equations in Sakai’s scheme are:

  • E (1)

8

δ →

  • E (1)

7

δ →

  • E (1)

6

δ → · · · (symmetry) or

  • A(1)∗∗

δ →

  • A(1)∗

1

δ →

  • A(1)∗

2

δ → · · · (surface)

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 34 / 41

slide-149
SLIDE 149

Difference Painlev´ e Equation of Type d-P(A(1)∗

2

): Schlesinger Transformations

So pure difference Painlev´ e equations in Sakai’s scheme are:

  • E (1)

8

δ →

  • E (1)

7

δ →

  • E (1)

6

δ → · · · (symmetry) or

  • A(1)∗∗

δ →

  • A(1)∗

1

δ →

  • A(1)∗

2

δ → · · · (surface)

P.Boalch has identified the Fuchsian systems whose Schlesinger transformations have the required symmetry type (spectral type 131313 for d-P( A∗

2)).

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 34 / 41

slide-150
SLIDE 150

Difference Painlev´ e Equation of Type d-P(A(1)∗

2

): Schlesinger Transformations

So pure difference Painlev´ e equations in Sakai’s scheme are:

  • E (1)

8

δ →

  • E (1)

7

δ →

  • E (1)

6

δ → · · · (symmetry) or

  • A(1)∗∗

δ →

  • A(1)∗

1

δ →

  • A(1)∗

2

δ → · · · (surface)

P.Boalch has identified the Fuchsian systems whose Schlesinger transformations have the required symmetry type (spectral type 131313 for d-P( A∗

2)).

However, each discrete Painlev´ e equation is characterized not only by the symmetry or the surface type, but also by the actual translation direction in Pic(X) and to identify that explicit computations are needed.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 34 / 41

slide-151
SLIDE 151

Difference Painlev´ e Equation of Type d-P(A(1)∗

2

): Schlesinger Transformations

So pure difference Painlev´ e equations in Sakai’s scheme are:

  • E (1)

8

δ →

  • E (1)

7

δ →

  • E (1)

6

δ → · · · (symmetry) or

  • A(1)∗∗

δ →

  • A(1)∗

1

δ →

  • A(1)∗

2

δ → · · · (surface)

P.Boalch has identified the Fuchsian systems whose Schlesinger transformations have the required symmetry type (spectral type 131313 for d-P( A∗

2)).

However, each discrete Painlev´ e equation is characterized not only by the symmetry or the surface type, but also by the actual translation direction in Pic(X) and to identify that explicit computations are needed. Take n = 2 finite poles z0 = 0, z1 = 1, matrix size m = 3, and rank(Ai) = 2: A(z) = A0 z + A1 z − 1 , Ai = BiC†

i =

bi,1 bi,2

  • c1†

i

c2†

i

  • .

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 34 / 41

slide-152
SLIDE 152

Difference Painlev´ e Equation of Type d-P(A(1)∗

2

): Schlesinger Transformations

So pure difference Painlev´ e equations in Sakai’s scheme are:

  • E (1)

8

δ →

  • E (1)

7

δ →

  • E (1)

6

δ → · · · (symmetry) or

  • A(1)∗∗

δ →

  • A(1)∗

1

δ →

  • A(1)∗

2

δ → · · · (surface)

P.Boalch has identified the Fuchsian systems whose Schlesinger transformations have the required symmetry type (spectral type 131313 for d-P( A∗

2)).

However, each discrete Painlev´ e equation is characterized not only by the symmetry or the surface type, but also by the actual translation direction in Pic(X) and to identify that explicit computations are needed. Take n = 2 finite poles z0 = 0, z1 = 1, matrix size m = 3, and rank(Ai) = 2: A(z) = A0 z + A1 z − 1 , Ai = BiC†

i =

bi,1 bi,2

  • c1†

i

c2†

i

  • .

The corresponding Riemann scheme and the Fuchs relation are        z = 0 z = 1 z = ∞ θ1 θ1

1

κ1 θ2 θ2

1

κ2 κ3        , θ1

0 + θ2 0 + θ1 1 + θ2 1 + 3

  • j=1

κj = 0.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 34 / 41

slide-153
SLIDE 153

Difference Painlev´ e Equation of Type d-P(A(1)∗

2

): Schlesinger Transformations

So pure difference Painlev´ e equations in Sakai’s scheme are:

  • E (1)

8

δ →

  • E (1)

7

δ →

  • E (1)

6

δ → · · · (symmetry) or

  • A(1)∗∗

δ →

  • A(1)∗

1

δ →

  • A(1)∗

2

δ → · · · (surface)

P.Boalch has identified the Fuchsian systems whose Schlesinger transformations have the required symmetry type (spectral type 131313 for d-P( A∗

2)).

However, each discrete Painlev´ e equation is characterized not only by the symmetry or the surface type, but also by the actual translation direction in Pic(X) and to identify that explicit computations are needed. Take n = 2 finite poles z0 = 0, z1 = 1, matrix size m = 3, and rank(Ai) = 2: A(z) = A0 z + A1 z − 1 , Ai = BiC†

i =

bi,1 bi,2

  • c1†

i

c2†

i

  • .

The corresponding Riemann scheme and the Fuchs relation are        z = 0 z = 1 z = ∞ θ1 θ1

1

κ1 θ2 θ2

1

κ2 κ3        , θ1

0 + θ2 0 + θ1 1 + θ2 1 + 3

  • j=1

κj = 0. No continuous deformations but non-trivial Schlesinger transformations.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 34 / 41

slide-154
SLIDE 154

Using various gauge transformations we can normalize the b-vectors, and then use the condition C†

i Bi = ˆi to parameterize the c†-vectors:

B0 =   1 1   , C†

0 =

θ1 α θ2 β

  • , B1 =

  1 1 1 1   , C†

1 =

−γ − θ1

1

γ θ1

1

θ2

1 − δ

δ

  • .

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 35 / 41

slide-155
SLIDE 155

Using various gauge transformations we can normalize the b-vectors, and then use the condition C†

i Bi = ˆi to parameterize the c†-vectors:

B0 =   1 1   , C†

0 =

θ1 α θ2 β

  • , B1 =

  1 1 1 1   , C†

1 =

−γ − θ1

1

γ θ1

1

θ2

1 − δ

δ

  • .

Requiring that the eigenvalues of A∞ = −A0 − A1 are κ1, κ2, and κ3: tr(A∞) = κ1 + κ2 + κ3 (the Fuchs relation) |A∞|11 + |A∞|22 + |A∞|33 = κ2κ3 + κ3κ1 + κ1κ2 det(A∞) = κ1κ2κ3 imposes two linear constraints on four parameters α, β, γ, and δ.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 35 / 41

slide-156
SLIDE 156

Using various gauge transformations we can normalize the b-vectors, and then use the condition C†

i Bi = ˆi to parameterize the c†-vectors:

B0 =   1 1   , C†

0 =

θ1 α θ2 β

  • , B1 =

  1 1 1 1   , C†

1 =

−γ − θ1

1

γ θ1

1

θ2

1 − δ

δ

  • .

Requiring that the eigenvalues of A∞ = −A0 − A1 are κ1, κ2, and κ3: tr(A∞) = κ1 + κ2 + κ3 (the Fuchs relation) |A∞|11 + |A∞|22 + |A∞|33 = κ2κ3 + κ3κ1 + κ1κ2 det(A∞) = κ1κ2κ3 imposes two linear constraints on four parameters α, β, γ, and δ. We can write them as a linear system on α and β: (γ + δ + θ1

1 − θ2 1)α − (γ + δ)β = κ2κ3 + κ3κ1 + κ1κ2 + (θ2 0 − θ1 0)δ

− (θ2

0 + θ1 1)(θ1 0 + θ2 1) − θ2 0θ1 1),

−(θ2

0(γ + δ + θ1 1 − θ2 1) + θ2 1γ + θ1 1δ)α + (θ1 0(γ + δ) + θ2 1γ + θ1 1δ)β = κ1κ2κ3

+ θ1

1((θ1 0 − θ2 0)δ + θ2 0(θ1 0 + θ2 1)).

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 35 / 41

slide-157
SLIDE 157

Using various gauge transformations we can normalize the b-vectors, and then use the condition C†

i Bi = ˆi to parameterize the c†-vectors:

B0 =   1 1   , C†

0 =

θ1 α θ2 β

  • , B1 =

  1 1 1 1   , C†

1 =

−γ − θ1

1

γ θ1

1

θ2

1 − δ

δ

  • .

Requiring that the eigenvalues of A∞ = −A0 − A1 are κ1, κ2, and κ3: tr(A∞) = κ1 + κ2 + κ3 (the Fuchs relation) |A∞|11 + |A∞|22 + |A∞|33 = κ2κ3 + κ3κ1 + κ1κ2 det(A∞) = κ1κ2κ3 imposes two linear constraints on four parameters α, β, γ, and δ. We can write them as a linear system on α and β: (γ + δ + θ1

1 − θ2 1)α − (γ + δ)β = κ2κ3 + κ3κ1 + κ1κ2 + (θ2 0 − θ1 0)δ

− (θ2

0 + θ1 1)(θ1 0 + θ2 1) − θ2 0θ1 1),

−(θ2

0(γ + δ + θ1 1 − θ2 1) + θ2 1γ + θ1 1δ)α + (θ1 0(γ + δ) + θ2 1γ + θ1 1δ)β = κ1κ2κ3

+ θ1

1((θ1 0 − θ2 0)δ + θ2 0(θ1 0 + θ2 1)).

Notice that the coefficients of the matrix of the above linear system are written in terms of the expressions γ + δ, γ + δ + θ1

1 − θ2 1, and θ2 1γ + θ1 1δ.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 35 / 41

slide-158
SLIDE 158

Choose parameterization variables x and y to simplify the structure of the substitution rule (matrix entries and the determinant): x = (γ + δ)(θ1

0 − θ2 0)

θ1

1 − θ2 1

, y = θ2

1γ + θ1 1δ

γ + δ + θ1

1 − θ2 1

.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 36 / 41

slide-159
SLIDE 159

Choose parameterization variables x and y to simplify the structure of the substitution rule (matrix entries and the determinant): x = (γ + δ)(θ1

0 − θ2 0)

θ1

1 − θ2 1

, y = θ2

1γ + θ1 1δ

γ + δ + θ1

1 − θ2 1

. This gives: α(x, y) =

  • yr1 + x(θ2

0r1+r2)

x+θ1

0−θ2

  • (x + y)(θ1

1 − θ2 1) ,

β(x, y) =

  • (y + θ2

0)r1 + r2

  • (x + y)(θ1

1 − θ2 1) ,

where r1 and r2 are the right-hand-sides of our linear system on α and β r1 = r1(x, y) = κ1κ2 + κ2κ3 + κ3κ1 − (y − θ2

1)(x − θ2 0) − θ1 0(y + θ2 0)

− θ1

1(θ1 0 + θ2 0 + θ2 1),

r2 = r2(x, y) = κ1κ2κ3 + θ1

1((y − θ2 1)(x − θ2 0) + θ1 0(y + θ2 0)).

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 36 / 41

slide-160
SLIDE 160

Choose parameterization variables x and y to simplify the structure of the substitution rule (matrix entries and the determinant): x = (γ + δ)(θ1

0 − θ2 0)

θ1

1 − θ2 1

, y = θ2

1γ + θ1 1δ

γ + δ + θ1

1 − θ2 1

. This gives: α(x, y) =

  • yr1 + x(θ2

0r1+r2)

x+θ1

0−θ2

  • (x + y)(θ1

1 − θ2 1) ,

β(x, y) =

  • (y + θ2

0)r1 + r2

  • (x + y)(θ1

1 − θ2 1) ,

where r1 and r2 are the right-hand-sides of our linear system on α and β r1 = r1(x, y) = κ1κ2 + κ2κ3 + κ3κ1 − (y − θ2

1)(x − θ2 0) − θ1 0(y + θ2 0)

− θ1

1(θ1 0 + θ2 0 + θ2 1),

r2 = r2(x, y) = κ1κ2κ3 + θ1

1((y − θ2 1)(x − θ2 0) + θ1 0(y + θ2 0)).

Schlesinger evolution equations give us the map ψ : (x, y) → (¯ x, ¯ y):          ¯ x = (α − β)(αx(θ1

1 − θ2 1) + (1 + θ2 0)(x(y − θ2 1) + y(θ1 0 − θ2 0)))

(α − β)(x(y − θ2

1) + (θ1 0 − θ2 0)y) − α(θ1 1 + 1)(θ1 0 − θ2 0)

¯ y = (α − β)(y(x + θ1

0 − θ2 0) − θ2 1x)

α(θ1

0 − θ2 0)

.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 36 / 41

slide-161
SLIDE 161

Choose parameterization variables x and y to simplify the structure of the substitution rule (matrix entries and the determinant): x = (γ + δ)(θ1

0 − θ2 0)

θ1

1 − θ2 1

, y = θ2

1γ + θ1 1δ

γ + δ + θ1

1 − θ2 1

. This gives: α(x, y) =

  • yr1 + x(θ2

0r1+r2)

x+θ1

0−θ2

  • (x + y)(θ1

1 − θ2 1) ,

β(x, y) =

  • (y + θ2

0)r1 + r2

  • (x + y)(θ1

1 − θ2 1) ,

where r1 and r2 are the right-hand-sides of our linear system on α and β r1 = r1(x, y) = κ1κ2 + κ2κ3 + κ3κ1 − (y − θ2

1)(x − θ2 0) − θ1 0(y + θ2 0)

− θ1

1(θ1 0 + θ2 0 + θ2 1),

r2 = r2(x, y) = κ1κ2κ3 + θ1

1((y − θ2 1)(x − θ2 0) + θ1 0(y + θ2 0)).

Schlesinger evolution equations give us the map ψ : (x, y) → (¯ x, ¯ y):          ¯ x = (α − β)(αx(θ1

1 − θ2 1) + (1 + θ2 0)(x(y − θ2 1) + y(θ1 0 − θ2 0)))

(α − β)(x(y − θ2

1) + (θ1 0 − θ2 0)y) − α(θ1 1 + 1)(θ1 0 − θ2 0)

¯ y = (α − β)(y(x + θ1

0 − θ2 0) − θ2 1x)

α(θ1

0 − θ2 0)

. Very complicated! (Finding a simple form for this equation was one of the main motivations behind this project)

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 36 / 41

slide-162
SLIDE 162

Difference Painlev´ e Equation of Type d-P(A(1)∗

2

): Schlesinger Transformations

The Okamoto surface for the map ψ : (x, y) → (¯ x, ¯ y) is given by the blow-up diagram:

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 37 / 41

slide-163
SLIDE 163

Difference Painlev´ e Equation of Type d-P(A(1)∗

2

): Schlesinger Transformations

The Okamoto surface for the map ψ : (x, y) → (¯ x, ¯ y) is given by the blow-up diagram:

π7 π8 π5 π6 π1 π2 π3 π4 y = ∞ Hy y = 0 Hy x = 0 x = ∞ Hx Hx Hx + Hy x + y = 0 Hy − F7 − F8 Hx − F5 − F6 Hx + Hy − F1 − F2 − F3 − F4 Hy − F4 Hx − F4 F4 F1 F2 F3 F5 F6 F7 F8

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 37 / 41

slide-164
SLIDE 164

Difference Painlev´ e Equation of Type d-P(A(1)∗

2

): Schlesinger Transformations

The Okamoto surface for the map ψ : (x, y) → (¯ x, ¯ y) is given by the blow-up diagram:

π7 π8 π5 π6 π1 π2 π3 π4 y = ∞ Hy y = 0 Hy x = 0 x = ∞ Hx Hx Hx + Hy x + y = 0 Hy − F7 − F8 Hx − F5 − F6 Hx + Hy − F1 − F2 − F3 − F4 Hy − F4 Hx − F4 F4 F1 F2 F3 F5 F6 F7 F8

So we see that the configuration structure is the same, but the coordinates of the blowup points are now expressed in terms of the characteristic indices: pi(θ2

0+κi, −θ2 0−κi),

p4(0, 0), p5

  • ∞, θ1

1

  • ,

p6

  • ∞, θ2

1

  • ,

p7

  • θ2

0 − θ1 0, ∞

  • ,

p8

  • θ2

0 + 1, ∞

  • .

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 37 / 41

slide-165
SLIDE 165

Difference Painlev´ e Equation of Type d-P(A(1)∗

2

): Schlesinger Transformations

The action on the Picard lattice is, however, quite different:

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 38 / 41

slide-166
SLIDE 166

Difference Painlev´ e Equation of Type d-P(A(1)∗

2

): Schlesinger Transformations

The action on the Picard lattice is, however, quite different: The action of ψ∗ on Pic(X) Hf → 2Hf + 3Hg − E1 − E2 − E3 − E4 − 2E5 − 2E8, Hg → 3Hf + 5Hg − 2E1 − 2E2 − 2E3 − 2E4 − 3E5 − E6 − 2E8, E1 → Hf + 2Hg − E2 − E3 − E4 − E5 − E8, E2 → Hf + 2Hg − E1 − E3 − E4 − E5 − E8, E3 → Hf + 2Hg − E1 − E2 − E4 − E5 − E8, E4 → Hf + 2Hg − E1 − E2 − E3 − E5 − E8, E5 → E7, E6 → 2Hf + 2Hg − E1 − E2 − E3 − E4 − 2E5 − E8, E7 → 2Hf + 3Hg − E1 − E2 − E3 − E4 − 2E5 − E6 − 2E8, E8 → Hg − E5, and so the induced action ϕ∗ on the sub-lattice R⊥ is given by the following translation: (α0, α1, α2, α3, α4, α5, α6) → (α0, α1, α2, α3, α4, α5, α6) + (0, 0, 0, −1, 1, 1, −1)δ,

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 38 / 41

slide-167
SLIDE 167

Comparison between different forms of d-P( A∗

2) To compare between these two examples, we can do the following:

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 39 / 41

slide-168
SLIDE 168

Comparison between different forms of d-P( A∗

2) To compare between these two examples, we can do the following: Compare the parameters and the dynamic on the level of parameters: bi = θ2

0 + κi, b4 = 0, b5 = θ1 1, b6 = θ2 1, b7 = θ1 0 − θ2 0, b8 = −θ2 0 − 1.

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 39 / 41

slide-169
SLIDE 169

Comparison between different forms of d-P( A∗

2) To compare between these two examples, we can do the following: Compare the parameters and the dynamic on the level of parameters: bi = θ2

0 + κi, b4 = 0, b5 = θ1 1, b6 = θ2 1, b7 = θ1 0 − θ2 0, b8 = −θ2 0 − 1.

This can also be written as follows, with δ = χ(−KX ) = b1 + · · · + b8(= −1): ϕ : b1 b2 b3 b4 b5 b6 b7 b8

  • b1

b2 b3 b4 b5 + δ b6 + δ b7 − δ b8 − δ

  • deautonomization

ψ : b1 b2 b3 b4 b5 b6 b7 b8

  • b1

b2 b3 b4 b5 − δ b6 b7 + δ b8

  • Schlesinger Transformations

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 39 / 41

slide-170
SLIDE 170

Comparison between different forms of d-P( A∗

2) To compare between these two examples, we can do the following: Compare the parameters and the dynamic on the level of parameters: bi = θ2

0 + κi, b4 = 0, b5 = θ1 1, b6 = θ2 1, b7 = θ1 0 − θ2 0, b8 = −θ2 0 − 1.

This can also be written as follows, with δ = χ(−KX ) = b1 + · · · + b8(= −1): ϕ : b1 b2 b3 b4 b5 b6 b7 b8

  • b1

b2 b3 b4 b5 + δ b6 + δ b7 − δ b8 − δ

  • deautonomization

ψ : b1 b2 b3 b4 b5 b6 b7 b8

  • b1

b2 b3 b4 b5 − δ b6 b7 + δ b8

  • Schlesinger Transformations

Riemann scheme

  • which gave d-P(A(1)∗

2

) = Σ0(1, 3) ◦ 1 0

2 1

  • Σ0(1, 3) ◦

1 0

1 1

  • .
  • :

       z = 0 z = 1 z = ∞ θ1 θ1

1

κ1 θ2 θ2

1

κ2 κ3       

0 1 1 1

→        z = 0 z = 1 z = ∞ θ1

0 − 1

θ1

1 + 1

κ1 θ2 θ2

1

κ2 κ3        ,        z = 0 z = 1 z = ∞ θ1 θ1

1

κ1 θ2 θ2

1

κ2 κ3       

d-P(A(1)∗

2

)

− →        z = 0 z = 1 z = ∞ θ1 θ1

1 − 1

κ1 + 1 θ2

0 − 1

θ2

1 − 1

κ2 + 1 κ3 + 1        .

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 39 / 41

slide-171
SLIDE 171

Comparison between different forms of d-P( A∗

2) Translation directions: ϕ∗ : (α0, α1, α2, α3, α4, α5, α6) → (α0, α1, α2, α3, α4, α5, α6) + (0, 0, 0, 1, 0, −1, 0)δ ψ∗ : (α0, α1, α2, α3, α4, α5, α6) → (α0, α1, α2, α3, α4, α5, α6) + (0, 0, 0, −1, 1, 1, −1)δ

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 40 / 41

slide-172
SLIDE 172

Comparison between different forms of d-P( A∗

2) Translation directions: ϕ∗ : (α0, α1, α2, α3, α4, α5, α6) → (α0, α1, α2, α3, α4, α5, α6) + (0, 0, 0, 1, 0, −1, 0)δ ψ∗ : (α0, α1, α2, α3, α4, α5, α6) → (α0, α1, α2, α3, α4, α5, α6) + (0, 0, 0, −1, 1, 1, −1)δ The best approach, however, is through the decomposition. In the same way as we did for ϕ∗, we can compute and compare the decomposition for ψ∗; ϕ∗ = σr ◦ w5 ◦ w2 ◦ w3 ◦ w6 ◦ w5 ◦ w2 ◦ w4 ◦ w3 ◦ w1 ◦ w2 ◦ w5 ◦ w0 ◦ w1 ◦ w2 ◦ w6 ◦ w5 ψ∗ = σr ◦ w1 ◦ w2 ◦ w3 ◦ w6 ◦ w5 ◦ w2 ◦ w4 ◦ w3 ◦ w1 ◦ w2 ◦ w5 ◦ w0 ◦ w1 ◦ w2 ◦ w6 ◦ w3

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 40 / 41

slide-173
SLIDE 173

Comparison between different forms of d-P( A∗

2) Translation directions: ϕ∗ : (α0, α1, α2, α3, α4, α5, α6) → (α0, α1, α2, α3, α4, α5, α6) + (0, 0, 0, 1, 0, −1, 0)δ ψ∗ : (α0, α1, α2, α3, α4, α5, α6) → (α0, α1, α2, α3, α4, α5, α6) + (0, 0, 0, −1, 1, 1, −1)δ The best approach, however, is through the decomposition. In the same way as we did for ϕ∗, we can compute and compare the decomposition for ψ∗; ϕ∗ = σr ◦ w5 ◦ w2 ◦ w3 ◦ w6 ◦ w5 ◦ w2 ◦ w4 ◦ w3 ◦ w1 ◦ w2 ◦ w5 ◦ w0 ◦ w1 ◦ w2 ◦ w6 ◦ w5 ψ∗ = σr ◦ w1 ◦ w2 ◦ w3 ◦ w6 ◦ w5 ◦ w2 ◦ w4 ◦ w3 ◦ w1 ◦ w2 ◦ w5 ◦ w0 ◦ w1 ◦ w2 ◦ w6 ◦ w3 This gives us the equivalence! ψ∗ = σr ◦ w1 ◦ w5 ◦ σr2 ◦ ϕ∗ ◦ w5 ◦ w3 = (w3 ◦ w5) ◦ ϕ∗ ◦ (w3 ◦ w5)−1

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 40 / 41

slide-174
SLIDE 174

Comparison between different forms of d-P( A∗

2) Translation directions: ϕ∗ : (α0, α1, α2, α3, α4, α5, α6) → (α0, α1, α2, α3, α4, α5, α6) + (0, 0, 0, 1, 0, −1, 0)δ ψ∗ : (α0, α1, α2, α3, α4, α5, α6) → (α0, α1, α2, α3, α4, α5, α6) + (0, 0, 0, −1, 1, 1, −1)δ The best approach, however, is through the decomposition. In the same way as we did for ϕ∗, we can compute and compare the decomposition for ψ∗; ϕ∗ = σr ◦ w5 ◦ w2 ◦ w3 ◦ w6 ◦ w5 ◦ w2 ◦ w4 ◦ w3 ◦ w1 ◦ w2 ◦ w5 ◦ w0 ◦ w1 ◦ w2 ◦ w6 ◦ w5 ψ∗ = σr ◦ w1 ◦ w2 ◦ w3 ◦ w6 ◦ w5 ◦ w2 ◦ w4 ◦ w3 ◦ w1 ◦ w2 ◦ w5 ◦ w0 ◦ w1 ◦ w2 ◦ w6 ◦ w3 This gives us the equivalence! ψ∗ = σr ◦ w1 ◦ w5 ◦ σr2 ◦ ϕ∗ ◦ w5 ◦ w3 = (w3 ◦ w5) ◦ ϕ∗ ◦ (w3 ◦ w5)−1 The mapping w5 ◦ w3 gives us the change of variables between the two equations, f = x(y − θ1

1) + y(θ1 0 + κ1) + (θ2 0 + κ1)(θ1 0 + θ2 0 + θ1 1 + 2κ1)

y + θ2

0 + κ1

g = x(y − θ2

0 − θ1 1 − κ1) + y(θ1 0 − θ2 0) + (θ2 0 + κ1)(θ1 0 + θ2 0 + 2κ1)

x − θ2

0 − κ1

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 40 / 41

slide-175
SLIDE 175

References

References Hidetaka Sakai, Rational surfaces associated with affine root systems and geometry of the Painlev´ e equations, Comm. Math. Phys. 220 (2001), no. 1, 165–229. Hidetaka Sakai, Problem: discrete Painlev´ e equations and their Lax forms, Algebraic, analytic and geometric aspects of complex differential equations and their deformations. Painlev´ e hierarchies, RIMS Kˆ

  • kyˆ

uroku Bessatsu, B2, Res. Inst. Math. Sci. (RIMS), Kyoto, 2007, pp. 195–208. Philip Boalch, Quivers and difference Painlev´ e equations, Groups and symmetries, CRM

  • Proc. Lecture Notes, vol. 47, Amer. Math. Soc., Providence, RI, 2009, pp. 25–51.

Anton Dzhamay, Hidetaka Sakai, Tomoyuki Takenawa, Discrete Hamiltonian Structure of Schlesinger Transformations, arXiv:1302.2972v2 [math-ph] Anton Dzhamay and Tomoyuki Takenawa, Geometric Analysis of Reductions from Schlesinger Transformations to Difference Painlev´ e Equations, arXiv:1408.3778 [math-ph] Masatoshi Noumi, Painlev´ e equations through symmetry., Translations of Mathematical Monographs, 223. American Mathematical Society, (2004) 156 pp. Kenji Kajiwara, Masatoshi Noumi, Yasuhiko Yamada, Geometric Aspects of Painlev´ e Equations, J. Phys. A: Math. Theor. 50 (2017) 073001 (164pp), arXiv:1509.08186 [nlin.SI]

Anton Dzhamay (UNC) Geometry of Discrete Painlev´ e Equations August 25, 2017 41 / 41