Isogenies, power operations, and homotopy theory Charles Rezk - - PowerPoint PPT Presentation

isogenies power operations and homotopy theory
SMART_READER_LITE
LIVE PREVIEW

Isogenies, power operations, and homotopy theory Charles Rezk - - PowerPoint PPT Presentation

Isogenies, power operations, and homotopy theory Charles Rezk University of Illinois at Urbana-Champaign Seoul, August 18, 2014 Overview Plan. Context: power operations in cohomology theories. Recent advances: Morava E -theories. Formal


slide-1
SLIDE 1

Isogenies, power operations, and homotopy theory

Charles Rezk

University of Illinois at Urbana-Champaign

Seoul, August 18, 2014

slide-2
SLIDE 2

Overview

Plan. Context: “power operations” in cohomology theories. Recent advances: Morava E-theories. Formal groups and isogenies. Applications and vistas.

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 2 / 29

slide-3
SLIDE 3

K-theory

Motivating example. K-theory (Grothendieck; Atiyah-Hirzebruch; 1950s) K(X) = K 0(X) := isomorphism classes of vector bundles /X

  • / ∼

V ∼ V1 + V2 if 0 → V1 → V → V2 → 0. Functors on vector bundles give operations on K(X), e.g.,: V , W → V ⊗ W , V → ΛnV , V → SymnV . K(X) is a Λ-ring (Grothendieck) Functions λn : K(X) → K(X) satisfying axioms λn(x + y) = · · · , λn(xy) = · · · , λmλn(x) = · · · . . . = explicit polynomials in λi(x), λj(y).

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 3 / 29

slide-4
SLIDE 4

K-theory

Motivating example. K-theory (Grothendieck; Atiyah-Hirzebruch; 1950s) K(X) = K 0(X) := isomorphism classes of vector bundles /X

  • / ∼

V ∼ V1 + V2 if 0 → V1 → V → V2 → 0. Functors on vector bundles give operations on K(X), e.g.,: V , W → V ⊗ W , V → ΛnV , V → SymnV . K(X) is a Λ-ring (Grothendieck) Functions λn : K(X) → K(X) satisfying axioms λn(x + y) = · · · , λn(xy) = · · · , λmλn(x) = · · · . . . = explicit polynomials in λi(x), λj(y).

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 3 / 29

slide-5
SLIDE 5

K-theory

Motivating example. K-theory (Grothendieck; Atiyah-Hirzebruch; 1950s) K(X) = K 0(X) := isomorphism classes of vector bundles /X

  • / ∼

V ∼ V1 + V2 if 0 → V1 → V → V2 → 0. Functors on vector bundles give operations on K(X), e.g.,: V , W → V ⊗ W , V → ΛnV , V → SymnV . K(X) is a Λ-ring (Grothendieck) Functions λn : K(X) → K(X) satisfying axioms λn(x + y) = · · · , λn(xy) = · · · , λmλn(x) = · · · . . . = explicit polynomials in λi(x), λj(y).

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 3 / 29

slide-6
SLIDE 6

Equivariant K-theory

Compact Lie G X. Equivariant K-theory KG(X) = K(X G) := {G equivariant vb /X}/ ∼ . (Atiyah, 1966) tensor power is an operation V → V ⊗n : KG(X) → KG×Σn(X) ≈ KG(X) ⊗ RΣn. KG(point) = K(pointG) = RG = representation ring of G. Σn = symmetric group. Λ-rings ⇐ ⇒ representation theory of symmetric groups

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 4 / 29

slide-7
SLIDE 7

Equivariant K-theory

Compact Lie G X. Equivariant K-theory KG(X) = K(X G) := {G equivariant vb /X}/ ∼ . (Atiyah, 1966) tensor power is an operation V → V ⊗n : KG(X) → KG×Σn(X) ≈ KG(X) ⊗ RΣn. KG(point) = K(pointG) = RG = representation ring of G. Σn = symmetric group. Λ-rings ⇐ ⇒ representation theory of symmetric groups

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 4 / 29

slide-8
SLIDE 8

Equivariant K-theory

Compact Lie G X. Equivariant K-theory KG(X) = K(X G) := {G equivariant vb /X}/ ∼ . (Atiyah, 1966) tensor power is an operation V → V ⊗n : KG(X) → KG×Σn(X) ≈ KG(X) ⊗ RΣn. KG(point) = K(pointG) = RG = representation ring of G. Σn = symmetric group. Λ-rings ⇐ ⇒ representation theory of symmetric groups

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 4 / 29

slide-9
SLIDE 9

Equivariant K-theory

Compact Lie G X. Equivariant K-theory KG(X) = K(X G) := {G equivariant vb /X}/ ∼ . (Atiyah, 1966) tensor power is an operation V → V ⊗n : KG(X) → KG×Σn(X) ≈ KG(X) ⊗ RΣn. KG(point) = K(pointG) = RG = representation ring of G. Σn = symmetric group. Λ-rings ⇐ ⇒ representation theory of symmetric groups

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 4 / 29

slide-10
SLIDE 10

Properties of Λ-rings, 1

Λ-ring structure is complicated to describe, but is easy for “nice” rings. (Wilkerson, 1982) R torsion free comm. ring:

  • Λ-ring

structures on R

{ψp : R → R}p prime lifts of Frobenius, ψpψq = ψqψp.

  • Adams operations ψn, n ≥ 1; ψmψn = ψmn, ring homomorphisms

Adams congruence ψp(x) ≡ xp mod p, p prime any Λ-ring has ψp,θp : R → R satisfying ψp is a ring homomorphism, ψp(x) = xp + p θp(x) (say θp is a witness to the pth Adams congruence)

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 5 / 29

slide-11
SLIDE 11

Properties of Λ-rings, 1

Λ-ring structure is complicated to describe, but is easy for “nice” rings. (Wilkerson, 1982) R torsion free comm. ring:

  • Λ-ring

structures on R

{ψp : R → R}p prime lifts of Frobenius, ψpψq = ψqψp.

  • Adams operations ψn, n ≥ 1; ψmψn = ψmn, ring homomorphisms

Adams congruence ψp(x) ≡ xp mod p, p prime any Λ-ring has ψp,θp : R → R satisfying ψp is a ring homomorphism, ψp(x) = xp + p θp(x) (say θp is a witness to the pth Adams congruence)

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 5 / 29

slide-12
SLIDE 12

Properties of Λ-rings, 1

Λ-ring structure is complicated to describe, but is easy for “nice” rings. (Wilkerson, 1982) R torsion free comm. ring:

  • Λ-ring

structures on R

{ψp : R → R}p prime lifts of Frobenius, ψpψq = ψqψp.

  • Adams operations ψn, n ≥ 1; ψmψn = ψmn, ring homomorphisms

Adams congruence ψp(x) ≡ xp mod p, p prime any Λ-ring has ψp,θp : R → R satisfying ψp is a ring homomorphism, ψp(x) = xp + p θp(x) (say θp is a witness to the pth Adams congruence)

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 5 / 29

slide-13
SLIDE 13

Properties of Λ-rings, 1

Λ-ring structure is complicated to describe, but is easy for “nice” rings. (Wilkerson, 1982) R torsion free comm. ring:

  • Λ-ring

structures on R

{ψp : R → R}p prime lifts of Frobenius, ψpψq = ψqψp.

  • Adams operations ψn, n ≥ 1; ψmψn = ψmn, ring homomorphisms

Adams congruence ψp(x) ≡ xp mod p, p prime any Λ-ring has ψp,θp : R → R satisfying ψp is a ring homomorphism, ψp(x) = xp + p θp(x) (say θp is a witness to the pth Adams congruence)

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 5 / 29

slide-14
SLIDE 14

Properties of Λ-rings, 2

Line bundle L → X = ⇒ ψn(L) = L⊗n in K(X) Multiplicative group scheme Gm = Spec

  • Z[T, T −1]
  • ≈ Spec
  • K(ptU(1))
  • Adams operation =

⇒ isogeny of Gm:

  • K(ptU(1))

ψn

− → K(ptU(1))

⇒ Gm

[n]

− → Gm Isogeny: finite flat homomorphism of group schemes Remarks.

  • Gm = Spf K(BU(1)), multiplicative formal group

These properties useful in classical applications (e.g., Adams work on vector fields on spheres, image of J, . . . )

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 6 / 29

slide-15
SLIDE 15

Properties of Λ-rings, 2

Line bundle L → X = ⇒ ψn(L) = L⊗n in K(X) Multiplicative group scheme Gm = Spec

  • Z[T, T −1]
  • ≈ Spec
  • K(ptU(1))
  • Adams operation =

⇒ isogeny of Gm:

  • K(ptU(1))

ψn

− → K(ptU(1))

⇒ Gm

[n]

− → Gm Isogeny: finite flat homomorphism of group schemes Remarks.

  • Gm = Spf K(BU(1)), multiplicative formal group

These properties useful in classical applications (e.g., Adams work on vector fields on spheres, image of J, . . . )

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 6 / 29

slide-16
SLIDE 16

Properties of Λ-rings, 2

Line bundle L → X = ⇒ ψn(L) = L⊗n in K(X) Multiplicative group scheme Gm = Spec

  • Z[T, T −1]
  • ≈ Spec
  • K(ptU(1))
  • Adams operation =

⇒ isogeny of Gm:

  • K(ptU(1))

ψn

− → K(ptU(1))

⇒ Gm

[n]

− → Gm Isogeny: finite flat homomorphism of group schemes Remarks.

  • Gm = Spf K(BU(1)), multiplicative formal group

These properties useful in classical applications (e.g., Adams work on vector fields on spheres, image of J, . . . )

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 6 / 29

slide-17
SLIDE 17

Properties of Λ-rings, 2

Line bundle L → X = ⇒ ψn(L) = L⊗n in K(X) Multiplicative group scheme Gm = Spec

  • Z[T, T −1]
  • ≈ Spec
  • K(ptU(1))
  • Adams operation =

⇒ isogeny of Gm:

  • K(ptU(1))

ψn

− → K(ptU(1))

⇒ Gm

[n]

− → Gm Isogeny: finite flat homomorphism of group schemes Remarks.

  • Gm = Spf K(BU(1)), multiplicative formal group

These properties useful in classical applications (e.g., Adams work on vector fields on spheres, image of J, . . . )

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 6 / 29

slide-18
SLIDE 18

Properties of Λ-rings, 2

Line bundle L → X = ⇒ ψn(L) = L⊗n in K(X) Multiplicative group scheme Gm = Spec

  • Z[T, T −1]
  • ≈ Spec
  • K(ptU(1))
  • Adams operation =

⇒ isogeny of Gm:

  • K(ptU(1))

ψn

− → K(ptU(1))

⇒ Gm

[n]

− → Gm Isogeny: finite flat homomorphism of group schemes Remarks.

  • Gm = Spf K(BU(1)), multiplicative formal group

These properties useful in classical applications (e.g., Adams work on vector fields on spheres, image of J, . . . )

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 6 / 29

slide-19
SLIDE 19

Other examples of power operations

h∗(−) = generalized cohomology theory, commutative ring valued Would like to have h∗(X) Pn − → h∗

Σn(X) = h∗(X × BΣn)

refines of nth power x → xn Do these exist? Yes if h∗(−) represented by a structured commutative ring spectrum (= commutative S-algebra = E∞-ring spectrum = . . . ) Examples. (Steenrod, 1953) reduced power operations in H∗(−, Fp) (Sqi for p = 2, Pi for p odd) (Voevodsky, 2001) motivic reduced power operations (Quillen, 1971) power operations in bordism theories based on M → M×n Σn used to prove π∗MU classifies formal group laws

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 7 / 29

slide-20
SLIDE 20

Other examples of power operations

h∗(−) = generalized cohomology theory, commutative ring valued Would like to have h∗(X) Pn − → h∗

Σn(X) = h∗(X × BΣn)

refines of nth power x → xn Do these exist? Yes if h∗(−) represented by a structured commutative ring spectrum (= commutative S-algebra = E∞-ring spectrum = . . . ) Examples. (Steenrod, 1953) reduced power operations in H∗(−, Fp) (Sqi for p = 2, Pi for p odd) (Voevodsky, 2001) motivic reduced power operations (Quillen, 1971) power operations in bordism theories based on M → M×n Σn used to prove π∗MU classifies formal group laws

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 7 / 29

slide-21
SLIDE 21

Other examples of power operations

h∗(−) = generalized cohomology theory, commutative ring valued Would like to have h∗(X) Pn − → h∗

Σn(X) = h∗(X × BΣn)

refines of nth power x → xn Do these exist? Yes if h∗(−) represented by a structured commutative ring spectrum (= commutative S-algebra = E∞-ring spectrum = . . . ) Examples. (Steenrod, 1953) reduced power operations in H∗(−, Fp) (Sqi for p = 2, Pi for p odd) (Voevodsky, 2001) motivic reduced power operations (Quillen, 1971) power operations in bordism theories based on M → M×n Σn used to prove π∗MU classifies formal group laws

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 7 / 29

slide-22
SLIDE 22

Other examples of power operations

h∗(−) = generalized cohomology theory, commutative ring valued Would like to have h∗(X) Pn − → h∗

Σn(X) = h∗(X × BΣn)

refines of nth power x → xn Do these exist? Yes if h∗(−) represented by a structured commutative ring spectrum (= commutative S-algebra = E∞-ring spectrum = . . . ) Examples. (Steenrod, 1953) reduced power operations in H∗(−, Fp) (Sqi for p = 2, Pi for p odd) (Voevodsky, 2001) motivic reduced power operations (Quillen, 1971) power operations in bordism theories based on M → M×n Σn used to prove π∗MU classifies formal group laws

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 7 / 29

slide-23
SLIDE 23

Other examples of power operations

h∗(−) = generalized cohomology theory, commutative ring valued Would like to have h∗(X) Pn − → h∗

Σn(X) = h∗(X × BΣn)

refines of nth power x → xn Do these exist? Yes if h∗(−) represented by a structured commutative ring spectrum (= commutative S-algebra = E∞-ring spectrum = . . . ) Examples. (Steenrod, 1953) reduced power operations in H∗(−, Fp) (Sqi for p = 2, Pi for p odd) (Voevodsky, 2001) motivic reduced power operations (Quillen, 1971) power operations in bordism theories based on M → M×n Σn used to prove π∗MU classifies formal group laws

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 7 / 29

slide-24
SLIDE 24

Other examples of power operations

h∗(−) = generalized cohomology theory, commutative ring valued Would like to have h∗(X) Pn − → h∗

Σn(X) = h∗(X × BΣn)

refines of nth power x → xn Do these exist? Yes if h∗(−) represented by a structured commutative ring spectrum (= commutative S-algebra = E∞-ring spectrum = . . . ) Examples. (Steenrod, 1953) reduced power operations in H∗(−, Fp) (Sqi for p = 2, Pi for p odd) (Voevodsky, 2001) motivic reduced power operations (Quillen, 1971) power operations in bordism theories based on M → M×n Σn used to prove π∗MU classifies formal group laws

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 7 / 29

slide-25
SLIDE 25

Other examples of power operations

h∗(−) = generalized cohomology theory, commutative ring valued Would like to have h∗(X) Pn − → h∗

Σn(X) = h∗(X × BΣn)

refines of nth power x → xn Do these exist? Yes if h∗(−) represented by a structured commutative ring spectrum (= commutative S-algebra = E∞-ring spectrum = . . . ) Examples. (Steenrod, 1953) reduced power operations in H∗(−, Fp) (Sqi for p = 2, Pi for p odd) (Voevodsky, 2001) motivic reduced power operations (Quillen, 1971) power operations in bordism theories based on M → M×n Σn used to prove π∗MU classifies formal group laws

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 7 / 29

slide-26
SLIDE 26

Elliptic cohomology

What is elliptic cohomology? Theory K-theory elliptic cohomology Group scheme Gm elliptic curve Cycles vector bundles ??? ??? = 2-dim conformal field theories? (Segal, . . . ) Examples: (Goerss-Hopkins-Miller) tmf = “topological modular forms” associated to universal elliptic curve over MEll structured comm ring spectrum = ⇒ power operations! (Lurie) Equivariant elliptic cohomology theories Open question: Which equivariant elliptic cohomology theories admit power operations?

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 8 / 29

slide-27
SLIDE 27

Elliptic cohomology

What is elliptic cohomology? Theory K-theory elliptic cohomology Group scheme Gm elliptic curve Cycles vector bundles ??? ??? = 2-dim conformal field theories? (Segal, . . . ) Examples: (Goerss-Hopkins-Miller) tmf = “topological modular forms” associated to universal elliptic curve over MEll structured comm ring spectrum = ⇒ power operations! (Lurie) Equivariant elliptic cohomology theories Open question: Which equivariant elliptic cohomology theories admit power operations?

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 8 / 29

slide-28
SLIDE 28

Elliptic cohomology

What is elliptic cohomology? Theory K-theory elliptic cohomology Group scheme Gm elliptic curve Cycles vector bundles ??? ??? = 2-dim conformal field theories? (Segal, . . . ) Examples: (Goerss-Hopkins-Miller) tmf = “topological modular forms” associated to universal elliptic curve over MEll structured comm ring spectrum = ⇒ power operations! (Lurie) Equivariant elliptic cohomology theories Open question: Which equivariant elliptic cohomology theories admit power operations?

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 8 / 29

slide-29
SLIDE 29

Elliptic cohomology

What is elliptic cohomology? Theory K-theory elliptic cohomology Group scheme Gm elliptic curve Cycles vector bundles ??? ??? = 2-dim conformal field theories? (Segal, . . . ) Examples: (Goerss-Hopkins-Miller) tmf = “topological modular forms” associated to universal elliptic curve over MEll structured comm ring spectrum = ⇒ power operations! (Lurie) Equivariant elliptic cohomology theories Open question: Which equivariant elliptic cohomology theories admit power operations?

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 8 / 29

slide-30
SLIDE 30

Elliptic cohomology

What is elliptic cohomology? Theory K-theory elliptic cohomology Group scheme Gm elliptic curve Cycles vector bundles ??? ??? = 2-dim conformal field theories? (Segal, . . . ) Examples: (Goerss-Hopkins-Miller) tmf = “topological modular forms” associated to universal elliptic curve over MEll structured comm ring spectrum = ⇒ power operations! (Lurie) Equivariant elliptic cohomology theories Open question: Which equivariant elliptic cohomology theories admit power operations?

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 8 / 29

slide-31
SLIDE 31

Elliptic cohomology

What is elliptic cohomology? Theory K-theory elliptic cohomology Group scheme Gm elliptic curve Cycles vector bundles ??? ??? = 2-dim conformal field theories? (Segal, . . . ) Examples: (Goerss-Hopkins-Miller) tmf = “topological modular forms” associated to universal elliptic curve over MEll structured comm ring spectrum = ⇒ power operations! (Lurie) Equivariant elliptic cohomology theories Open question: Which equivariant elliptic cohomology theories admit power operations?

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 8 / 29

slide-32
SLIDE 32

Elliptic cohomology

What is elliptic cohomology? Theory K-theory elliptic cohomology Group scheme Gm elliptic curve Cycles vector bundles ??? ??? = 2-dim conformal field theories? (Segal, . . . ) Examples: (Goerss-Hopkins-Miller) tmf = “topological modular forms” associated to universal elliptic curve over MEll structured comm ring spectrum = ⇒ power operations! (Lurie) Equivariant elliptic cohomology theories Open question: Which equivariant elliptic cohomology theories admit power operations?

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 8 / 29

slide-33
SLIDE 33

Elliptic cohomology

What is elliptic cohomology? Theory K-theory elliptic cohomology Group scheme Gm elliptic curve Cycles vector bundles ??? ??? = 2-dim conformal field theories? (Segal, . . . ) Examples: (Goerss-Hopkins-Miller) tmf = “topological modular forms” associated to universal elliptic curve over MEll structured comm ring spectrum = ⇒ power operations! (Lurie) Equivariant elliptic cohomology theories Open question: Which equivariant elliptic cohomology theories admit power operations?

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 8 / 29

slide-34
SLIDE 34

Elliptic cohomology

What is elliptic cohomology? Theory K-theory elliptic cohomology Group scheme Gm elliptic curve Cycles vector bundles ??? ??? = 2-dim conformal field theories? (Segal, . . . ) Examples: (Goerss-Hopkins-Miller) tmf = “topological modular forms” associated to universal elliptic curve over MEll structured comm ring spectrum = ⇒ power operations! (Lurie) Equivariant elliptic cohomology theories Open question: Which equivariant elliptic cohomology theories admit power operations?

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 8 / 29

slide-35
SLIDE 35

Elliptic cohomology

What is elliptic cohomology? Theory K-theory elliptic cohomology Group scheme Gm elliptic curve Cycles vector bundles ??? ??? = 2-dim conformal field theories? (Segal, . . . ) Examples: (Goerss-Hopkins-Miller) tmf = “topological modular forms” associated to universal elliptic curve over MEll structured comm ring spectrum = ⇒ power operations! (Lurie) Equivariant elliptic cohomology theories Open question: Which equivariant elliptic cohomology theories admit power operations?

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 8 / 29

slide-36
SLIDE 36

A nice example: Elliptic cohomology at the Tate curve

Tate curve T[ [q] ] = “C×/qZ”, defined over Spec Z[ [q] ]. Equivariant elliptic cohomology at Tate curve EllTate(X G) :=

approx K

  • Lghost(X G)U(1)
  • “ghost loops” = contstant loops; RHS is K of “twisted sectors” (see e.g.,

Ruan 2000, Lupercio-Uribe 2002) (Ganter, 2007, 2013) Power operations for EllTate EllTate(X G) is an elliptic Λ-ring: two families of operations λn : EllTate → EllTate, µm : EllTate → EllTate ⊗Z[

[q] ] Z[

[q1/m] ] {λn} are Λ-ring structure, {µm} are Λ-ring homomorphisms

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 9 / 29

slide-37
SLIDE 37

A nice example: Elliptic cohomology at the Tate curve

Tate curve T[ [q] ] = “C×/qZ”, defined over Spec Z[ [q] ]. Equivariant elliptic cohomology at Tate curve EllTate(X G) :=

approx K

  • Lghost(X G)U(1)
  • “ghost loops” = contstant loops; RHS is K of “twisted sectors” (see e.g.,

Ruan 2000, Lupercio-Uribe 2002) (Ganter, 2007, 2013) Power operations for EllTate EllTate(X G) is an elliptic Λ-ring: two families of operations λn : EllTate → EllTate, µm : EllTate → EllTate ⊗Z[

[q] ] Z[

[q1/m] ] {λn} are Λ-ring structure, {µm} are Λ-ring homomorphisms

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 9 / 29

slide-38
SLIDE 38

A nice example: Elliptic cohomology at the Tate curve

Tate curve T[ [q] ] = “C×/qZ”, defined over Spec Z[ [q] ]. Equivariant elliptic cohomology at Tate curve EllTate(X G) :=

approx K

  • Lghost(X G)U(1)
  • “ghost loops” = contstant loops; RHS is K of “twisted sectors” (see e.g.,

Ruan 2000, Lupercio-Uribe 2002) (Ganter, 2007, 2013) Power operations for EllTate EllTate(X G) is an elliptic Λ-ring: two families of operations λn : EllTate → EllTate, µm : EllTate → EllTate ⊗Z[

[q] ] Z[

[q1/m] ] {λn} are Λ-ring structure, {µm} are Λ-ring homomorphisms

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 9 / 29

slide-39
SLIDE 39

A nice example: Elliptic cohomology at the Tate curve

Tate curve T[ [q] ] = “C×/qZ”, defined over Spec Z[ [q] ]. Equivariant elliptic cohomology at Tate curve EllTate(X G) :=

approx K

  • Lghost(X G)U(1)
  • “ghost loops” = contstant loops; RHS is K of “twisted sectors” (see e.g.,

Ruan 2000, Lupercio-Uribe 2002) (Ganter, 2007, 2013) Power operations for EllTate EllTate(X G) is an elliptic Λ-ring: two families of operations λn : EllTate → EllTate, µm : EllTate → EllTate ⊗Z[

[q] ] Z[

[q1/m] ] {λn} are Λ-ring structure, {µm} are Λ-ring homomorphisms

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 9 / 29

slide-40
SLIDE 40

Morava E-theory: introduction

Morava E-theories are “designer cohomology theories” — manufactured using homotopy theory, not coming from “nature” some arise as completions of “natural” theories, e.g. K ∧

p ,

Ell∧

s.-s. point

have rich theory of power operations (Ando, Hopkins, Strickland, R.) Goal: describe what we know about this theory (a lot) Recall: Power operations for K-theory are “controlled” by isogenies of Gm Slogan Power operations for Morava E-theories are “controlled” by “deformations” of Frobenius isogenies of 1-dimensional formal groups

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 10 / 29

slide-41
SLIDE 41

Morava E-theory: introduction

Morava E-theories are “designer cohomology theories” — manufactured using homotopy theory, not coming from “nature” some arise as completions of “natural” theories, e.g. K ∧

p ,

Ell∧

s.-s. point

have rich theory of power operations (Ando, Hopkins, Strickland, R.) Goal: describe what we know about this theory (a lot) Recall: Power operations for K-theory are “controlled” by isogenies of Gm Slogan Power operations for Morava E-theories are “controlled” by “deformations” of Frobenius isogenies of 1-dimensional formal groups

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 10 / 29

slide-42
SLIDE 42

Morava E-theory: introduction

Morava E-theories are “designer cohomology theories” — manufactured using homotopy theory, not coming from “nature” some arise as completions of “natural” theories, e.g. K ∧

p ,

Ell∧

s.-s. point

have rich theory of power operations (Ando, Hopkins, Strickland, R.) Goal: describe what we know about this theory (a lot) Recall: Power operations for K-theory are “controlled” by isogenies of Gm Slogan Power operations for Morava E-theories are “controlled” by “deformations” of Frobenius isogenies of 1-dimensional formal groups

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 10 / 29

slide-43
SLIDE 43

Morava E-theory: introduction

Morava E-theories are “designer cohomology theories” — manufactured using homotopy theory, not coming from “nature” some arise as completions of “natural” theories, e.g. K ∧

p ,

Ell∧

s.-s. point

have rich theory of power operations (Ando, Hopkins, Strickland, R.) Goal: describe what we know about this theory (a lot) Recall: Power operations for K-theory are “controlled” by isogenies of Gm Slogan Power operations for Morava E-theories are “controlled” by “deformations” of Frobenius isogenies of 1-dimensional formal groups

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 10 / 29

slide-44
SLIDE 44

Morava E-theory: introduction

Morava E-theories are “designer cohomology theories” — manufactured using homotopy theory, not coming from “nature” some arise as completions of “natural” theories, e.g. K ∧

p ,

Ell∧

s.-s. point

have rich theory of power operations (Ando, Hopkins, Strickland, R.) Goal: describe what we know about this theory (a lot) Recall: Power operations for K-theory are “controlled” by isogenies of Gm Slogan Power operations for Morava E-theories are “controlled” by “deformations” of Frobenius isogenies of 1-dimensional formal groups

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 10 / 29

slide-45
SLIDE 45

Morava E-theory: introduction

Morava E-theories are “designer cohomology theories” — manufactured using homotopy theory, not coming from “nature” some arise as completions of “natural” theories, e.g. K ∧

p ,

Ell∧

s.-s. point

have rich theory of power operations (Ando, Hopkins, Strickland, R.) Goal: describe what we know about this theory (a lot) Recall: Power operations for K-theory are “controlled” by isogenies of Gm Slogan Power operations for Morava E-theories are “controlled” by “deformations” of Frobenius isogenies of 1-dimensional formal groups

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 10 / 29

slide-46
SLIDE 46

Morava E-theory: summary

Let G0/Fp = one dimensional commutative formal group of height n ∈ {1, 2, . . . }. (Morava, 1978; Goerss-Hopkins-Miller 1993–2004) There exists a cohomology theory EG0 (Morava E-theory) which is represented by a structured commutative ring spectrum is complex orientable; formal group Spf(E 0CP∞) = universal deformation of G0 (in sense of Lubin-Tate) E 0

G0(pt) = Zp[

[a1, . . . , an−1] ] E ∗

G0(pt) = E 0 G0(pt)[u, u−1],

u ∈ E 2

G0(pt)

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 11 / 29

slide-47
SLIDE 47

Morava E-theory: summary

Let G0/Fp = one dimensional commutative formal group of height n ∈ {1, 2, . . . }. (Morava, 1978; Goerss-Hopkins-Miller 1993–2004) There exists a cohomology theory EG0 (Morava E-theory) which is represented by a structured commutative ring spectrum is complex orientable; formal group Spf(E 0CP∞) = universal deformation of G0 (in sense of Lubin-Tate) E 0

G0(pt) = Zp[

[a1, . . . , an−1] ] E ∗

G0(pt) = E 0 G0(pt)[u, u−1],

u ∈ E 2

G0(pt)

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 11 / 29

slide-48
SLIDE 48

Morava E-theory: summary

Let G0/Fp = one dimensional commutative formal group of height n ∈ {1, 2, . . . }. (Morava, 1978; Goerss-Hopkins-Miller 1993–2004) There exists a cohomology theory EG0 (Morava E-theory) which is represented by a structured commutative ring spectrum is complex orientable; formal group Spf(E 0CP∞) = universal deformation of G0 (in sense of Lubin-Tate) E 0

G0(pt) = Zp[

[a1, . . . , an−1] ] E ∗

G0(pt) = E 0 G0(pt)[u, u−1],

u ∈ E 2

G0(pt)

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 11 / 29

slide-49
SLIDE 49

Morava E-theory: summary

Let G0/Fp = one dimensional commutative formal group of height n ∈ {1, 2, . . . }. (Morava, 1978; Goerss-Hopkins-Miller 1993–2004) There exists a cohomology theory EG0 (Morava E-theory) which is represented by a structured commutative ring spectrum is complex orientable; formal group Spf(E 0CP∞) = universal deformation of G0 (in sense of Lubin-Tate) E 0

G0(pt) = Zp[

[a1, . . . , an−1] ] E ∗

G0(pt) = E 0 G0(pt)[u, u−1],

u ∈ E 2

G0(pt)

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 11 / 29

slide-50
SLIDE 50

Morava E-theory: summary

Let G0/Fp = one dimensional commutative formal group of height n ∈ {1, 2, . . . }. (Morava, 1978; Goerss-Hopkins-Miller 1993–2004) There exists a cohomology theory EG0 (Morava E-theory) which is represented by a structured commutative ring spectrum is complex orientable; formal group Spf(E 0CP∞) = universal deformation of G0 (in sense of Lubin-Tate) E 0

G0(pt) = Zp[

[a1, . . . , an−1] ] E ∗

G0(pt) = E 0 G0(pt)[u, u−1],

u ∈ E 2

G0(pt)

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 11 / 29

slide-51
SLIDE 51

Formal groups and complex oriented theories

Formal group is object locally described by a formal group law. Formal group law (commutative, 1-dimensional) S(x, y) ∈ R[ [x, y] ] satisfying axioms for abelian group: S(x, 0) = x = S(0, x), S(x, y) = S(y, x), S(S(x, y), z) = S(x, S(y, z)). For future reference, we note the p-series of G0: [p](x) = S(x, S(x, . . . S(x, x)))

  • x appears p times

Complex oriented cohomology theory Ring-valued cohomology theory E such that E ∗(CP∞) = E ∗[ [x] ], and x restricts to fundamental class of CP1 = S2. Examples: H∗(−, Z), K-theory, Ell, Morava E-theories,. . .

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 12 / 29

slide-52
SLIDE 52

Formal groups and complex oriented theories

Formal group is object locally described by a formal group law. Formal group law (commutative, 1-dimensional) S(x, y) ∈ R[ [x, y] ] satisfying axioms for abelian group: S(x, 0) = x = S(0, x), S(x, y) = S(y, x), S(S(x, y), z) = S(x, S(y, z)). For future reference, we note the p-series of G0: [p](x) = S(x, S(x, . . . S(x, x)))

  • x appears p times

Complex oriented cohomology theory Ring-valued cohomology theory E such that E ∗(CP∞) = E ∗[ [x] ], and x restricts to fundamental class of CP1 = S2. Examples: H∗(−, Z), K-theory, Ell, Morava E-theories,. . .

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 12 / 29

slide-53
SLIDE 53

Formal groups and complex oriented theories

Formal group is object locally described by a formal group law. Formal group law (commutative, 1-dimensional) S(x, y) ∈ R[ [x, y] ] satisfying axioms for abelian group: S(x, 0) = x = S(0, x), S(x, y) = S(y, x), S(S(x, y), z) = S(x, S(y, z)). For future reference, we note the p-series of G0: [p](x) = S(x, S(x, . . . S(x, x)))

  • x appears p times

Complex oriented cohomology theory Ring-valued cohomology theory E such that E ∗(CP∞) = E ∗[ [x] ], and x restricts to fundamental class of CP1 = S2. Examples: H∗(−, Z), K-theory, Ell, Morava E-theories,. . .

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 12 / 29

slide-54
SLIDE 54

Deformations of formal groups

G0/Fp formal group of height n (i.e., [p]G0(x) = c xpn + O(xpn+1), c = 0) R = complete local ring, Fp ⊂ R/m Groupoid Def0

G0(R) of deformations of G0/Fp to R

Deformation (G, α): G is a formal group over R, iso α: G0

− → GR/m of formal groups over Fp Isomorphism (G, α) → (G ′, α′) of deformations: iso f : G → G ′ compatible with id of G0 Classified up to canonical iso by Lubin and Tate: (Lubin-Tate, 1966) ∃ universal deformation (Guniv, αuniv) over A ≈ Zp[ [a1, . . . , an−1] ] Guniv is the formal group of Morava E-theory EG0

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 13 / 29

slide-55
SLIDE 55

Deformations of formal groups

G0/Fp formal group of height n (i.e., [p]G0(x) = c xpn + O(xpn+1), c = 0) R = complete local ring, Fp ⊂ R/m Groupoid Def0

G0(R) of deformations of G0/Fp to R

Deformation (G, α): G is a formal group over R, iso α: G0

− → GR/m of formal groups over Fp Isomorphism (G, α) → (G ′, α′) of deformations: iso f : G → G ′ compatible with id of G0 Classified up to canonical iso by Lubin and Tate: (Lubin-Tate, 1966) ∃ universal deformation (Guniv, αuniv) over A ≈ Zp[ [a1, . . . , an−1] ] Guniv is the formal group of Morava E-theory EG0

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 13 / 29

slide-56
SLIDE 56

Deformations of formal groups

G0/Fp formal group of height n (i.e., [p]G0(x) = c xpn + O(xpn+1), c = 0) R = complete local ring, Fp ⊂ R/m Groupoid Def0

G0(R) of deformations of G0/Fp to R

Deformation (G, α): G is a formal group over R, iso α: G0

− → GR/m of formal groups over Fp Isomorphism (G, α) → (G ′, α′) of deformations: iso f : G → G ′ compatible with id of G0 Classified up to canonical iso by Lubin and Tate: (Lubin-Tate, 1966) ∃ universal deformation (Guniv, αuniv) over A ≈ Zp[ [a1, . . . , an−1] ] Guniv is the formal group of Morava E-theory EG0

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 13 / 29

slide-57
SLIDE 57

Deformations of formal groups

G0/Fp formal group of height n (i.e., [p]G0(x) = c xpn + O(xpn+1), c = 0) R = complete local ring, Fp ⊂ R/m Groupoid Def0

G0(R) of deformations of G0/Fp to R

Deformation (G, α): G is a formal group over R, iso α: G0

− → GR/m of formal groups over Fp Isomorphism (G, α) → (G ′, α′) of deformations: iso f : G → G ′ compatible with id of G0 Classified up to canonical iso by Lubin and Tate: (Lubin-Tate, 1966) ∃ universal deformation (Guniv, αuniv) over A ≈ Zp[ [a1, . . . , an−1] ] Guniv is the formal group of Morava E-theory EG0

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 13 / 29

slide-58
SLIDE 58

Deformations of formal groups

G0/Fp formal group of height n (i.e., [p]G0(x) = c xpn + O(xpn+1), c = 0) R = complete local ring, Fp ⊂ R/m Groupoid Def0

G0(R) of deformations of G0/Fp to R

Deformation (G, α): G is a formal group over R, iso α: G0

− → GR/m of formal groups over Fp Isomorphism (G, α) → (G ′, α′) of deformations: iso f : G → G ′ compatible with id of G0 Classified up to canonical iso by Lubin and Tate: (Lubin-Tate, 1966) ∃ universal deformation (Guniv, αuniv) over A ≈ Zp[ [a1, . . . , an−1] ] Guniv is the formal group of Morava E-theory EG0

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 13 / 29

slide-59
SLIDE 59

Isogenies

Isogeny of formal groups over R Homomorphism f : G → G ′ given locally over R by f (x) = cxn+ higher degree terms, c ∈ R×. (n = deg f ) G0/Fp has a distinguished family of Frobenius isogenies Frobr : G0 → G0, r ≥ 0, given locally by Frobr(x) = xpr . Category DefG0(R) of deformations of Frobenius Objects: deformations (G, α) to R (= objects of Def0

G0(R))

Morphisms (G, α) → (G ′, α′): isogenies f : G → G ′ compatible with Frobr : G0 → G0, some r ≥ 0

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 14 / 29

slide-60
SLIDE 60

Isogenies

Isogeny of formal groups over R Homomorphism f : G → G ′ given locally over R by f (x) = cxn+ higher degree terms, c ∈ R×. (n = deg f ) G0/Fp has a distinguished family of Frobenius isogenies Frobr : G0 → G0, r ≥ 0, given locally by Frobr(x) = xpr . Category DefG0(R) of deformations of Frobenius Objects: deformations (G, α) to R (= objects of Def0

G0(R))

Morphisms (G, α) → (G ′, α′): isogenies f : G → G ′ compatible with Frobr : G0 → G0, some r ≥ 0

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 14 / 29

slide-61
SLIDE 61

Isogenies

Isogeny of formal groups over R Homomorphism f : G → G ′ given locally over R by f (x) = cxn+ higher degree terms, c ∈ R×. (n = deg f ) G0/Fp has a distinguished family of Frobenius isogenies Frobr : G0 → G0, r ≥ 0, given locally by Frobr(x) = xpr . Category DefG0(R) of deformations of Frobenius Objects: deformations (G, α) to R (= objects of Def0

G0(R))

Morphisms (G, α) → (G ′, α′): isogenies f : G → G ′ compatible with Frobr : G0 → G0, some r ≥ 0

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 14 / 29

slide-62
SLIDE 62

Isogenies

Isogeny of formal groups over R Homomorphism f : G → G ′ given locally over R by f (x) = cxn+ higher degree terms, c ∈ R×. (n = deg f ) G0/Fp has a distinguished family of Frobenius isogenies Frobr : G0 → G0, r ≥ 0, given locally by Frobr(x) = xpr . Category DefG0(R) of deformations of Frobenius Objects: deformations (G, α) to R (= objects of Def0

G0(R))

Morphisms (G, α) → (G ′, α′): isogenies f : G → G ′ compatible with Frobr : G0 → G0, some r ≥ 0

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 14 / 29

slide-63
SLIDE 63

The “pile” Def = DefG0

We have assignments complete local ring R = ⇒ category Def(R) local homomorphism R → R′ = ⇒ functor Def(R) → Def(R′) If Def(R) were a groupoid, we would call it a (pre-)stack Def is the “pile” of deformations of powers of Frob Sheaves on Def A sheaf of modules on Def is a collection of functors AR : Def(R) →

  • R-modules
  • with compatibility wrt base change along local homomorphisms R → R′

Likewise, a sheaf of commutative rings on Def is . . . Notation: Mod(Def), Com(Def). Mod(Def) = Mod(Γ) for a certain ring Γ

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 15 / 29

slide-64
SLIDE 64

The “pile” Def = DefG0

We have assignments complete local ring R = ⇒ category Def(R) local homomorphism R → R′ = ⇒ functor Def(R) → Def(R′) If Def(R) were a groupoid, we would call it a (pre-)stack Def is the “pile” of deformations of powers of Frob Sheaves on Def A sheaf of modules on Def is a collection of functors AR : Def(R) →

  • R-modules
  • with compatibility wrt base change along local homomorphisms R → R′

Likewise, a sheaf of commutative rings on Def is . . . Notation: Mod(Def), Com(Def). Mod(Def) = Mod(Γ) for a certain ring Γ

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 15 / 29

slide-65
SLIDE 65

The “pile” Def = DefG0

We have assignments complete local ring R = ⇒ category Def(R) local homomorphism R → R′ = ⇒ functor Def(R) → Def(R′) If Def(R) were a groupoid, we would call it a (pre-)stack Def is the “pile” of deformations of powers of Frob Sheaves on Def A sheaf of modules on Def is a collection of functors AR : Def(R) →

  • R-modules
  • with compatibility wrt base change along local homomorphisms R → R′

Likewise, a sheaf of commutative rings on Def is . . . Notation: Mod(Def), Com(Def). Mod(Def) = Mod(Γ) for a certain ring Γ

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 15 / 29

slide-66
SLIDE 66

The “pile” Def = DefG0

We have assignments complete local ring R = ⇒ category Def(R) local homomorphism R → R′ = ⇒ functor Def(R) → Def(R′) If Def(R) were a groupoid, we would call it a (pre-)stack Def is the “pile” of deformations of powers of Frob Sheaves on Def A sheaf of modules on Def is a collection of functors AR : Def(R) →

  • R-modules
  • with compatibility wrt base change along local homomorphisms R → R′

Likewise, a sheaf of commutative rings on Def is . . . Notation: Mod(Def), Com(Def). Mod(Def) = Mod(Γ) for a certain ring Γ

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 15 / 29

slide-67
SLIDE 67

The “pile” Def = DefG0

We have assignments complete local ring R = ⇒ category Def(R) local homomorphism R → R′ = ⇒ functor Def(R) → Def(R′) If Def(R) were a groupoid, we would call it a (pre-)stack Def is the “pile” of deformations of powers of Frob Sheaves on Def A sheaf of modules on Def is a collection of functors AR : Def(R) →

  • R-modules
  • with compatibility wrt base change along local homomorphisms R → R′

Likewise, a sheaf of commutative rings on Def is . . . Notation: Mod(Def), Com(Def). Mod(Def) = Mod(Γ) for a certain ring Γ

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 15 / 29

slide-68
SLIDE 68

The “pile” Def = DefG0

We have assignments complete local ring R = ⇒ category Def(R) local homomorphism R → R′ = ⇒ functor Def(R) → Def(R′) If Def(R) were a groupoid, we would call it a (pre-)stack Def is the “pile” of deformations of powers of Frob Sheaves on Def A sheaf of modules on Def is a collection of functors AR : Def(R) →

  • R-modules
  • with compatibility wrt base change along local homomorphisms R → R′

Likewise, a sheaf of commutative rings on Def is . . . Notation: Mod(Def), Com(Def). Mod(Def) = Mod(Γ) for a certain ring Γ

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 15 / 29

slide-69
SLIDE 69

The “pile” Def = DefG0

We have assignments complete local ring R = ⇒ category Def(R) local homomorphism R → R′ = ⇒ functor Def(R) → Def(R′) If Def(R) were a groupoid, we would call it a (pre-)stack Def is the “pile” of deformations of powers of Frob Sheaves on Def A sheaf of modules on Def is a collection of functors AR : Def(R) →

  • R-modules
  • with compatibility wrt base change along local homomorphisms R → R′

Likewise, a sheaf of commutative rings on Def is . . . Notation: Mod(Def), Com(Def). Mod(Def) = Mod(Γ) for a certain ring Γ

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 15 / 29

slide-70
SLIDE 70

Morava E-theory takes values in sheaves on Def

(Ando-Hopkins-Strickland 2004; see R. 2009) Power operations make Morava E-cohomology EG0 a functor E ∗(−): Spaces → Com∗(Def) Key step (Strickland 1997, 1998): E 0BΣpr /I classifies subgroups of rank pr of deformations Broader context: We have E ∗(X) = π∗(E X+) where A = E X+ is (i) a structured commutative E-algebra spectrum, (ii) K(n)-local (⇔ π∗A complete wrt (a1, . . . , an−1) in a suitable sense) The real theorem is (ibid) π∗ lifts to a functor π∗ : hCom(E)K(n) → Com∗(Def)

  • n homotopy category of K(n)-local commutative E-algebra spectra

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 16 / 29

slide-71
SLIDE 71

Morava E-theory takes values in sheaves on Def

(Ando-Hopkins-Strickland 2004; see R. 2009) Power operations make Morava E-cohomology EG0 a functor E ∗(−): Spaces → Com∗(Def) Key step (Strickland 1997, 1998): E 0BΣpr /I classifies subgroups of rank pr of deformations Broader context: We have E ∗(X) = π∗(E X+) where A = E X+ is (i) a structured commutative E-algebra spectrum, (ii) K(n)-local (⇔ π∗A complete wrt (a1, . . . , an−1) in a suitable sense) The real theorem is (ibid) π∗ lifts to a functor π∗ : hCom(E)K(n) → Com∗(Def)

  • n homotopy category of K(n)-local commutative E-algebra spectra

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 16 / 29

slide-72
SLIDE 72

Morava E-theory takes values in sheaves on Def

(Ando-Hopkins-Strickland 2004; see R. 2009) Power operations make Morava E-cohomology EG0 a functor E ∗(−): Spaces → Com∗(Def) Key step (Strickland 1997, 1998): E 0BΣpr /I classifies subgroups of rank pr of deformations Broader context: We have E ∗(X) = π∗(E X+) where A = E X+ is (i) a structured commutative E-algebra spectrum, (ii) K(n)-local (⇔ π∗A complete wrt (a1, . . . , an−1) in a suitable sense) The real theorem is (ibid) π∗ lifts to a functor π∗ : hCom(E)K(n) → Com∗(Def)

  • n homotopy category of K(n)-local commutative E-algebra spectra

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 16 / 29

slide-73
SLIDE 73

Morava E-theory takes values in sheaves on Def

(Ando-Hopkins-Strickland 2004; see R. 2009) Power operations make Morava E-cohomology EG0 a functor E ∗(−): Spaces → Com∗(Def) Key step (Strickland 1997, 1998): E 0BΣpr /I classifies subgroups of rank pr of deformations Broader context: We have E ∗(X) = π∗(E X+) where A = E X+ is (i) a structured commutative E-algebra spectrum, (ii) K(n)-local (⇔ π∗A complete wrt (a1, . . . , an−1) in a suitable sense) The real theorem is (ibid) π∗ lifts to a functor π∗ : hCom(E)K(n) → Com∗(Def)

  • n homotopy category of K(n)-local commutative E-algebra spectra

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 16 / 29

slide-74
SLIDE 74

Examples

Mod(Def) = modules for a certain ring Γ Height 1 G0 = multiplicative formal group; EG0 = K ∧

p

Γ = Zp[ψp]

  • gen. by Adams operation ψp

Height 2 (R., arXiv:0812.1320) G0/F2 = completion of s.-s. elliptic curve y2 + y = x3 over F2 Γ = Z2[ [a] ]Q0, Q1, Q2

  

Q0a = a2 Q0 − 2a Q1 + 6 Q2 Q1a = 3 Q0 + a Q2 Q2a = −a Q0 + 3 Q1 Q1Q0 = 2 Q2Q1 − 2 Q0Q2 Q2Q0 = Q0Q1 + a Q0Q2 − 2 Q1Q2

    (Y. Zhu, 2014) gives similar description at height 2, p = 3 There is a uniform description of Γ/p at height 2, all primes p

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 17 / 29

slide-75
SLIDE 75

Examples

Mod(Def) = modules for a certain ring Γ Height 1 G0 = multiplicative formal group; EG0 = K ∧

p

Γ = Zp[ψp]

  • gen. by Adams operation ψp

Height 2 (R., arXiv:0812.1320) G0/F2 = completion of s.-s. elliptic curve y2 + y = x3 over F2 Γ = Z2[ [a] ]Q0, Q1, Q2

  

Q0a = a2 Q0 − 2a Q1 + 6 Q2 Q1a = 3 Q0 + a Q2 Q2a = −a Q0 + 3 Q1 Q1Q0 = 2 Q2Q1 − 2 Q0Q2 Q2Q0 = Q0Q1 + a Q0Q2 − 2 Q1Q2

    (Y. Zhu, 2014) gives similar description at height 2, p = 3 There is a uniform description of Γ/p at height 2, all primes p

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 17 / 29

slide-76
SLIDE 76

Examples

Mod(Def) = modules for a certain ring Γ Height 1 G0 = multiplicative formal group; EG0 = K ∧

p

Γ = Zp[ψp]

  • gen. by Adams operation ψp

Height 2 (R., arXiv:0812.1320) G0/F2 = completion of s.-s. elliptic curve y2 + y = x3 over F2 Γ = Z2[ [a] ]Q0, Q1, Q2

  

Q0a = a2 Q0 − 2a Q1 + 6 Q2 Q1a = 3 Q0 + a Q2 Q2a = −a Q0 + 3 Q1 Q1Q0 = 2 Q2Q1 − 2 Q0Q2 Q2Q0 = Q0Q1 + a Q0Q2 − 2 Q1Q2

    (Y. Zhu, 2014) gives similar description at height 2, p = 3 There is a uniform description of Γ/p at height 2, all primes p

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 17 / 29

slide-77
SLIDE 77

Examples

Mod(Def) = modules for a certain ring Γ Height 1 G0 = multiplicative formal group; EG0 = K ∧

p

Γ = Zp[ψp]

  • gen. by Adams operation ψp

Height 2 (R., arXiv:0812.1320) G0/F2 = completion of s.-s. elliptic curve y2 + y = x3 over F2 Γ = Z2[ [a] ]Q0, Q1, Q2

  

Q0a = a2 Q0 − 2a Q1 + 6 Q2 Q1a = 3 Q0 + a Q2 Q2a = −a Q0 + 3 Q1 Q1Q0 = 2 Q2Q1 − 2 Q0Q2 Q2Q0 = Q0Q1 + a Q0Q2 − 2 Q1Q2

    (Y. Zhu, 2014) gives similar description at height 2, p = 3 There is a uniform description of Γ/p at height 2, all primes p

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 17 / 29

slide-78
SLIDE 78

Examples

Mod(Def) = modules for a certain ring Γ Height 1 G0 = multiplicative formal group; EG0 = K ∧

p

Γ = Zp[ψp]

  • gen. by Adams operation ψp

Height 2 (R., arXiv:0812.1320) G0/F2 = completion of s.-s. elliptic curve y2 + y = x3 over F2 Γ = Z2[ [a] ]Q0, Q1, Q2

  

Q0a = a2 Q0 − 2a Q1 + 6 Q2 Q1a = 3 Q0 + a Q2 Q2a = −a Q0 + 3 Q1 Q1Q0 = 2 Q2Q1 − 2 Q0Q2 Q2Q0 = Q0Q1 + a Q0Q2 − 2 Q1Q2

    (Y. Zhu, 2014) gives similar description at height 2, p = 3 There is a uniform description of Γ/p at height 2, all primes p

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 17 / 29

slide-79
SLIDE 79

Examples

Mod(Def) = modules for a certain ring Γ Height 1 G0 = multiplicative formal group; EG0 = K ∧

p

Γ = Zp[ψp]

  • gen. by Adams operation ψp

Height 2 (R., arXiv:0812.1320) G0/F2 = completion of s.-s. elliptic curve y2 + y = x3 over F2 Γ = Z2[ [a] ]Q0, Q1, Q2

  

Q0a = a2 Q0 − 2a Q1 + 6 Q2 Q1a = 3 Q0 + a Q2 Q2a = −a Q0 + 3 Q1 Q1Q0 = 2 Q2Q1 − 2 Q0Q2 Q2Q0 = Q0Q1 + a Q0Q2 − 2 Q1Q2

    (Y. Zhu, 2014) gives similar description at height 2, p = 3 There is a uniform description of Γ/p at height 2, all primes p

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 17 / 29

slide-80
SLIDE 80

Properties of Γ

n = height of G0/Fp (Ando 1995) Center(Γ) = Zp[ ˜ T1, . . . , ˜ Tn], (Hecke algebra) (R. arXiv:1204.4831) Γ is quadratic, i.e., Γ ≈ Tensor alg.(C1) / (ideal gen. by C2) where C1 and C2 ⊆ C1 ⊗E0 C1 are E0 = Zp[ [a1, . . . , an−1] ] bimodules (ibid) Γ is Koszul: have Γ-bimodule resolution 0 ← Γ ← Γ ⊗E0 C0 ⊗E0 Γ ← · · · ← Γ ⊗E0 Cn ⊗E0 Γ ← 0, each Ck is E0-bimod, free and f.g. as right E0-mod; C0 = E0 = ⇒ gl. dim(Γ) = 2n

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 18 / 29

slide-81
SLIDE 81

Properties of Γ

n = height of G0/Fp (Ando 1995) Center(Γ) = Zp[ ˜ T1, . . . , ˜ Tn], (Hecke algebra) (R. arXiv:1204.4831) Γ is quadratic, i.e., Γ ≈ Tensor alg.(C1) / (ideal gen. by C2) where C1 and C2 ⊆ C1 ⊗E0 C1 are E0 = Zp[ [a1, . . . , an−1] ] bimodules (ibid) Γ is Koszul: have Γ-bimodule resolution 0 ← Γ ← Γ ⊗E0 C0 ⊗E0 Γ ← · · · ← Γ ⊗E0 Cn ⊗E0 Γ ← 0, each Ck is E0-bimod, free and f.g. as right E0-mod; C0 = E0 = ⇒ gl. dim(Γ) = 2n

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 18 / 29

slide-82
SLIDE 82

Properties of Γ

n = height of G0/Fp (Ando 1995) Center(Γ) = Zp[ ˜ T1, . . . , ˜ Tn], (Hecke algebra) (R. arXiv:1204.4831) Γ is quadratic, i.e., Γ ≈ Tensor alg.(C1) / (ideal gen. by C2) where C1 and C2 ⊆ C1 ⊗E0 C1 are E0 = Zp[ [a1, . . . , an−1] ] bimodules (ibid) Γ is Koszul: have Γ-bimodule resolution 0 ← Γ ← Γ ⊗E0 C0 ⊗E0 Γ ← · · · ← Γ ⊗E0 Cn ⊗E0 Γ ← 0, each Ck is E0-bimod, free and f.g. as right E0-mod; C0 = E0 = ⇒ gl. dim(Γ) = 2n

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 18 / 29

slide-83
SLIDE 83

Remark about “Koszul” property

(R. ibid) Γ is Koszul This was conjectured by Ando-Hopkins-Strickland It is purely a theorem about formal algebraic geometry Only general proof is a purely “topological” proof, using ingredients:

(1) Γ = “primitives” of the Hopf algebra

m≥0 E0(BΣm) (Strickland)

(2) bar complex of Γ in degree k is “primitives” in

  • m1,...,mk E0B(Σm1 ≀ · · · ≀ Σmk)

(3) vanishing results for Bredon homology of partition complexes with

  • coeff. in appropriate Mackey functors (Arone-Dwyer-Lesh 2013)

Proof inspired by role of partition complexes as “derivatives of identity functor” in Goodwillie’s functor calculus (R. 2012) purely alg. geom. proof in height 2 case, using results on moduli of subgroups of elliptic curves

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 19 / 29

slide-84
SLIDE 84

Remark about “Koszul” property

(R. ibid) Γ is Koszul This was conjectured by Ando-Hopkins-Strickland It is purely a theorem about formal algebraic geometry Only general proof is a purely “topological” proof, using ingredients:

(1) Γ = “primitives” of the Hopf algebra

m≥0 E0(BΣm) (Strickland)

(2) bar complex of Γ in degree k is “primitives” in

  • m1,...,mk E0B(Σm1 ≀ · · · ≀ Σmk)

(3) vanishing results for Bredon homology of partition complexes with

  • coeff. in appropriate Mackey functors (Arone-Dwyer-Lesh 2013)

Proof inspired by role of partition complexes as “derivatives of identity functor” in Goodwillie’s functor calculus (R. 2012) purely alg. geom. proof in height 2 case, using results on moduli of subgroups of elliptic curves

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 19 / 29

slide-85
SLIDE 85

Remark about “Koszul” property

(R. ibid) Γ is Koszul This was conjectured by Ando-Hopkins-Strickland It is purely a theorem about formal algebraic geometry Only general proof is a purely “topological” proof, using ingredients:

(1) Γ = “primitives” of the Hopf algebra

m≥0 E0(BΣm) (Strickland)

(2) bar complex of Γ in degree k is “primitives” in

  • m1,...,mk E0B(Σm1 ≀ · · · ≀ Σmk)

(3) vanishing results for Bredon homology of partition complexes with

  • coeff. in appropriate Mackey functors (Arone-Dwyer-Lesh 2013)

Proof inspired by role of partition complexes as “derivatives of identity functor” in Goodwillie’s functor calculus (R. 2012) purely alg. geom. proof in height 2 case, using results on moduli of subgroups of elliptic curves

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 19 / 29

slide-86
SLIDE 86

Remark about “Koszul” property

(R. ibid) Γ is Koszul This was conjectured by Ando-Hopkins-Strickland It is purely a theorem about formal algebraic geometry Only general proof is a purely “topological” proof, using ingredients:

(1) Γ = “primitives” of the Hopf algebra

m≥0 E0(BΣm) (Strickland)

(2) bar complex of Γ in degree k is “primitives” in

  • m1,...,mk E0B(Σm1 ≀ · · · ≀ Σmk)

(3) vanishing results for Bredon homology of partition complexes with

  • coeff. in appropriate Mackey functors (Arone-Dwyer-Lesh 2013)

Proof inspired by role of partition complexes as “derivatives of identity functor” in Goodwillie’s functor calculus (R. 2012) purely alg. geom. proof in height 2 case, using results on moduli of subgroups of elliptic curves

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 19 / 29

slide-87
SLIDE 87

Remark about “Koszul” property

(R. ibid) Γ is Koszul This was conjectured by Ando-Hopkins-Strickland It is purely a theorem about formal algebraic geometry Only general proof is a purely “topological” proof, using ingredients:

(1) Γ = “primitives” of the Hopf algebra

m≥0 E0(BΣm) (Strickland)

(2) bar complex of Γ in degree k is “primitives” in

  • m1,...,mk E0B(Σm1 ≀ · · · ≀ Σmk)

(3) vanishing results for Bredon homology of partition complexes with

  • coeff. in appropriate Mackey functors (Arone-Dwyer-Lesh 2013)

Proof inspired by role of partition complexes as “derivatives of identity functor” in Goodwillie’s functor calculus (R. 2012) purely alg. geom. proof in height 2 case, using results on moduli of subgroups of elliptic curves

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 19 / 29

slide-88
SLIDE 88

Remark about “Koszul” property

(R. ibid) Γ is Koszul This was conjectured by Ando-Hopkins-Strickland It is purely a theorem about formal algebraic geometry Only general proof is a purely “topological” proof, using ingredients:

(1) Γ = “primitives” of the Hopf algebra

m≥0 E0(BΣm) (Strickland)

(2) bar complex of Γ in degree k is “primitives” in

  • m1,...,mk E0B(Σm1 ≀ · · · ≀ Σmk)

(3) vanishing results for Bredon homology of partition complexes with

  • coeff. in appropriate Mackey functors (Arone-Dwyer-Lesh 2013)

Proof inspired by role of partition complexes as “derivatives of identity functor” in Goodwillie’s functor calculus (R. 2012) purely alg. geom. proof in height 2 case, using results on moduli of subgroups of elliptic curves

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 19 / 29

slide-89
SLIDE 89

Remark about “Koszul” property

(R. ibid) Γ is Koszul This was conjectured by Ando-Hopkins-Strickland It is purely a theorem about formal algebraic geometry Only general proof is a purely “topological” proof, using ingredients:

(1) Γ = “primitives” of the Hopf algebra

m≥0 E0(BΣm) (Strickland)

(2) bar complex of Γ in degree k is “primitives” in

  • m1,...,mk E0B(Σm1 ≀ · · · ≀ Σmk)

(3) vanishing results for Bredon homology of partition complexes with

  • coeff. in appropriate Mackey functors (Arone-Dwyer-Lesh 2013)

Proof inspired by role of partition complexes as “derivatives of identity functor” in Goodwillie’s functor calculus (R. 2012) purely alg. geom. proof in height 2 case, using results on moduli of subgroups of elliptic curves

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 19 / 29

slide-90
SLIDE 90

Congruence criterion

Homotopy groups of K(n)-local E-algebras have some more structure: π∗ : hCom(E)K(n) →

  • T-algebras
  • “T-algebras” = a complicated algebraic catgeory (like Λ-rings)

(R. 2009) R p-torsion free commutative E∗-algebra:

  • T-algebra

structures on R

A ∈ Com∗(Def) with A(Guniv) = R satisfying “Frobenius congruence”

  • Frobenius congruence: Qx ≡ xp mod pR for a certain Q ∈ Γ

There is a (non-additive) witness to the Frobenius congruence: θ: R → R satisfying Qx = xp + p θ(x) where R is a T-algebra

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 20 / 29

slide-91
SLIDE 91

Congruence criterion

Homotopy groups of K(n)-local E-algebras have some more structure: π∗ : hCom(E)K(n) →

  • T-algebras
  • “T-algebras” = a complicated algebraic catgeory (like Λ-rings)

(R. 2009) R p-torsion free commutative E∗-algebra:

  • T-algebra

structures on R

A ∈ Com∗(Def) with A(Guniv) = R satisfying “Frobenius congruence”

  • Frobenius congruence: Qx ≡ xp mod pR for a certain Q ∈ Γ

There is a (non-additive) witness to the Frobenius congruence: θ: R → R satisfying Qx = xp + p θ(x) where R is a T-algebra

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 20 / 29

slide-92
SLIDE 92

Congruence criterion

Homotopy groups of K(n)-local E-algebras have some more structure: π∗ : hCom(E)K(n) →

  • T-algebras
  • “T-algebras” = a complicated algebraic catgeory (like Λ-rings)

(R. 2009) R p-torsion free commutative E∗-algebra:

  • T-algebra

structures on R

A ∈ Com∗(Def) with A(Guniv) = R satisfying “Frobenius congruence”

  • Frobenius congruence: Qx ≡ xp mod pR for a certain Q ∈ Γ

There is a (non-additive) witness to the Frobenius congruence: θ: R → R satisfying Qx = xp + p θ(x) where R is a T-algebra

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 20 / 29

slide-93
SLIDE 93

Application 1: nilpotence

Easy consequence of existence of “witness” θ such that Q(x) = xp + p θ(x), Q(x + y) = Q(x) + Q(y): If A ∈ Com(E)K(n), then x ∈ π∗A, prx = 0 = ⇒ x(p+1)r = 0. Idea: deduce relation θ(px) = pp−1x − Q(x) = (pp−1 − 1)xp − p θ(x). If px = 0, then 0 = x θ(px) = −xp+1. Mathew-Noel-Naumann observe this, and use it (with Nilpotence Theorem

  • f Devinatz-Hopkins-Smith) to give an easy proof of a conjecture of May:

(Mathhew-Noel-Naumann 2014) If R = structured commutative ring spectrum, then the kernel of the Hurewicz map π∗R → H∗(R, Z) consists of nilpotent elements

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 21 / 29

slide-94
SLIDE 94

Application 1: nilpotence

Easy consequence of existence of “witness” θ such that Q(x) = xp + p θ(x), Q(x + y) = Q(x) + Q(y): If A ∈ Com(E)K(n), then x ∈ π∗A, prx = 0 = ⇒ x(p+1)r = 0. Idea: deduce relation θ(px) = pp−1x − Q(x) = (pp−1 − 1)xp − p θ(x). If px = 0, then 0 = x θ(px) = −xp+1. Mathew-Noel-Naumann observe this, and use it (with Nilpotence Theorem

  • f Devinatz-Hopkins-Smith) to give an easy proof of a conjecture of May:

(Mathhew-Noel-Naumann 2014) If R = structured commutative ring spectrum, then the kernel of the Hurewicz map π∗R → H∗(R, Z) consists of nilpotent elements

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 21 / 29

slide-95
SLIDE 95

Application 1: nilpotence

Easy consequence of existence of “witness” θ such that Q(x) = xp + p θ(x), Q(x + y) = Q(x) + Q(y): If A ∈ Com(E)K(n), then x ∈ π∗A, prx = 0 = ⇒ x(p+1)r = 0. Idea: deduce relation θ(px) = pp−1x − Q(x) = (pp−1 − 1)xp − p θ(x). If px = 0, then 0 = x θ(px) = −xp+1. Mathew-Noel-Naumann observe this, and use it (with Nilpotence Theorem

  • f Devinatz-Hopkins-Smith) to give an easy proof of a conjecture of May:

(Mathhew-Noel-Naumann 2014) If R = structured commutative ring spectrum, then the kernel of the Hurewicz map π∗R → H∗(R, Z) consists of nilpotent elements

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 21 / 29

slide-96
SLIDE 96

Application 2: units and orientations

A = structured commutative ring = ⇒ units spectrum gl1A (gl1A)0(X) = (A0(X))×

  • Question. Does there exist structured commutative ring map MG → A,

where MG = spectrum representing bordism (G ∈ {U, SU, O, SO, Spin, . . . })? Answer (May-Quinn-Ray-Tornehave 1977). Yes iff the composite g → o J − → gl1S → gl1A is null-homotopic as map of spectra, where g = infinite delooping of G (Ando-Hopkins-R.; see Hopkins 2002) There is a map of structured commutative ring spectra MString → tmf which realizes the “Witten genus”; String = six-connected cover of Spin

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 22 / 29

slide-97
SLIDE 97

Application 2: units and orientations

A = structured commutative ring = ⇒ units spectrum gl1A (gl1A)0(X) = (A0(X))×

  • Question. Does there exist structured commutative ring map MG → A,

where MG = spectrum representing bordism (G ∈ {U, SU, O, SO, Spin, . . . })? Answer (May-Quinn-Ray-Tornehave 1977). Yes iff the composite g → o J − → gl1S → gl1A is null-homotopic as map of spectra, where g = infinite delooping of G (Ando-Hopkins-R.; see Hopkins 2002) There is a map of structured commutative ring spectra MString → tmf which realizes the “Witten genus”; String = six-connected cover of Spin

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 22 / 29

slide-98
SLIDE 98

Application 2: units and orientations

A = structured commutative ring = ⇒ units spectrum gl1A (gl1A)0(X) = (A0(X))×

  • Question. Does there exist structured commutative ring map MG → A,

where MG = spectrum representing bordism (G ∈ {U, SU, O, SO, Spin, . . . })? Answer (May-Quinn-Ray-Tornehave 1977). Yes iff the composite g → o J − → gl1S → gl1A is null-homotopic as map of spectra, where g = infinite delooping of G (Ando-Hopkins-R.; see Hopkins 2002) There is a map of structured commutative ring spectra MString → tmf which realizes the “Witten genus”; String = six-connected cover of Spin

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 22 / 29

slide-99
SLIDE 99

Application 2: units and orientations

A = structured commutative ring = ⇒ units spectrum gl1A (gl1A)0(X) = (A0(X))×

  • Question. Does there exist structured commutative ring map MG → A,

where MG = spectrum representing bordism (G ∈ {U, SU, O, SO, Spin, . . . })? Answer (May-Quinn-Ray-Tornehave 1977). Yes iff the composite g → o J − → gl1S → gl1A is null-homotopic as map of spectra, where g = infinite delooping of G (Ando-Hopkins-R.; see Hopkins 2002) There is a map of structured commutative ring spectra MString → tmf which realizes the “Witten genus”; String = six-connected cover of Spin

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 22 / 29

slide-100
SLIDE 100

Logarithmic operations

Logarithmic operation: spectrum map ℓ: gl1A → A (tom Dieck 1989) A = K ∧

p : exists ℓ: gl1K ∧ p → K ∧ p , giving ℓ: K ∧ p (X)× → K ∧ p (X) by

ℓ(x) = log(x) − 1

p log(ψp(x))

log = Taylor exp. at 1 = 1

p log

  • xp/ψp(x)
  • ψp(x) ≡ xp mod p

=

m≥1(−1)m pm−1 m

(θp(x)/x)m ψp(x) = xp + p θp(x) (R. 2006) E = EG0, height G0 = n; exists ℓ: gl1E → E giving E 0(X)× → E 0(X) by ℓ(x) =

n

  • k=0

(−1)k p(k

2)−k log ˜

Tk(x) where ˜ Tk ∈ Zp[ ˜ T1, . . . , ˜ Tn] = Center(Γ)

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 23 / 29

slide-101
SLIDE 101

Logarithmic operations

Logarithmic operation: spectrum map ℓ: gl1A → A (tom Dieck 1989) A = K ∧

p : exists ℓ: gl1K ∧ p → K ∧ p , giving ℓ: K ∧ p (X)× → K ∧ p (X) by

ℓ(x) = log(x) − 1

p log(ψp(x))

log = Taylor exp. at 1 = 1

p log

  • xp/ψp(x)
  • ψp(x) ≡ xp mod p

=

m≥1(−1)m pm−1 m

(θp(x)/x)m ψp(x) = xp + p θp(x) (R. 2006) E = EG0, height G0 = n; exists ℓ: gl1E → E giving E 0(X)× → E 0(X) by ℓ(x) =

n

  • k=0

(−1)k p(k

2)−k log ˜

Tk(x) where ˜ Tk ∈ Zp[ ˜ T1, . . . , ˜ Tn] = Center(Γ)

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 23 / 29

slide-102
SLIDE 102

Logarithmic operations

Logarithmic operation: spectrum map ℓ: gl1A → A (tom Dieck 1989) A = K ∧

p : exists ℓ: gl1K ∧ p → K ∧ p , giving ℓ: K ∧ p (X)× → K ∧ p (X) by

ℓ(x) = log(x) − 1

p log(ψp(x))

log = Taylor exp. at 1 = 1

p log

  • xp/ψp(x)
  • ψp(x) ≡ xp mod p

=

m≥1(−1)m pm−1 m

(θp(x)/x)m ψp(x) = xp + p θp(x) (R. 2006) E = EG0, height G0 = n; exists ℓ: gl1E → E giving E 0(X)× → E 0(X) by ℓ(x) =

n

  • k=0

(−1)k p(k

2)−k log ˜

Tk(x) where ˜ Tk ∈ Zp[ ˜ T1, . . . , ˜ Tn] = Center(Γ)

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 23 / 29

slide-103
SLIDE 103

Logarithmic operations

Logarithmic operation: spectrum map ℓ: gl1A → A (tom Dieck 1989) A = K ∧

p : exists ℓ: gl1K ∧ p → K ∧ p , giving ℓ: K ∧ p (X)× → K ∧ p (X) by

ℓ(x) = log(x) − 1

p log(ψp(x))

log = Taylor exp. at 1 = 1

p log

  • xp/ψp(x)
  • ψp(x) ≡ xp mod p

=

m≥1(−1)m pm−1 m

(θp(x)/x)m ψp(x) = xp + p θp(x) (R. 2006) E = EG0, height G0 = n; exists ℓ: gl1E → E giving E 0(X)× → E 0(X) by ℓ(x) =

n

  • k=0

(−1)k p(k

2)−k log ˜

Tk(x) where ˜ Tk ∈ Zp[ ˜ T1, . . . , ˜ Tn] = Center(Γ)

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 23 / 29

slide-104
SLIDE 104

Logarithmic operations

Logarithmic operation: spectrum map ℓ: gl1A → A (tom Dieck 1989) A = K ∧

p : exists ℓ: gl1K ∧ p → K ∧ p , giving ℓ: K ∧ p (X)× → K ∧ p (X) by

ℓ(x) = log(x) − 1

p log(ψp(x))

log = Taylor exp. at 1 = 1

p log

  • xp/ψp(x)
  • ψp(x) ≡ xp mod p

=

m≥1(−1)m pm−1 m

(θp(x)/x)m ψp(x) = xp + p θp(x) (R. 2006) E = EG0, height G0 = n; exists ℓ: gl1E → E giving E 0(X)× → E 0(X) by ℓ(x) =

n

  • k=0

(−1)k p(k

2)−k log ˜

Tk(x) where ˜ Tk ∈ Zp[ ˜ T1, . . . , ˜ Tn] = Center(Γ)

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 23 / 29

slide-105
SLIDE 105

Application to String-orientation of tmf

(Ando-Hopkins-R.) Exists MString → tmf realizing Witten genus Must construct null-homotopy of α: string → gl1tmf Proof idea: Above techniques give “locally defined” logarithms ℓn : gl1tmf∧

p → tmfK(n), n = 1, 2, all primes p

Work one prime at a time; have “fracture squares” gl1tmf∧

p ℓ2

  • ℓ1
  • tmfK(2)

ι1

  • tmfK(1)

γ

(tmfK(1))K(2)

Map(string, tmfK(2)) ≈ ∗, so reduce to string → HoFib(γ) Explicit formulas for ℓ1, ℓ2 identify γ = ι2 ◦ (id − U), where U : tmfK(1) → tmfK(1) is topological lift of “Atkin operator” on p-adic modular forms

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 24 / 29

slide-106
SLIDE 106

Application to String-orientation of tmf

(Ando-Hopkins-R.) Exists MString → tmf realizing Witten genus Must construct null-homotopy of α: string → gl1tmf Proof idea: Above techniques give “locally defined” logarithms ℓn : gl1tmf∧

p → tmfK(n), n = 1, 2, all primes p

Work one prime at a time; have “fracture squares” gl1tmf∧

p ℓ2

  • ℓ1
  • tmfK(2)

ι1

  • tmfK(1)

γ

(tmfK(1))K(2)

Map(string, tmfK(2)) ≈ ∗, so reduce to string → HoFib(γ) Explicit formulas for ℓ1, ℓ2 identify γ = ι2 ◦ (id − U), where U : tmfK(1) → tmfK(1) is topological lift of “Atkin operator” on p-adic modular forms

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 24 / 29

slide-107
SLIDE 107

Application to String-orientation of tmf

(Ando-Hopkins-R.) Exists MString → tmf realizing Witten genus Must construct null-homotopy of α: string → gl1tmf Proof idea: Above techniques give “locally defined” logarithms ℓn : gl1tmf∧

p → tmfK(n), n = 1, 2, all primes p

Work one prime at a time; have “fracture squares” gl1tmf∧

p ℓ2

  • ℓ1
  • tmfK(2)

ι1

  • tmfK(1)

γ

(tmfK(1))K(2)

Map(string, tmfK(2)) ≈ ∗, so reduce to string → HoFib(γ) Explicit formulas for ℓ1, ℓ2 identify γ = ι2 ◦ (id − U), where U : tmfK(1) → tmfK(1) is topological lift of “Atkin operator” on p-adic modular forms

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 24 / 29

slide-108
SLIDE 108

Application to String-orientation of tmf

(Ando-Hopkins-R.) Exists MString → tmf realizing Witten genus Must construct null-homotopy of α: string → gl1tmf Proof idea: Above techniques give “locally defined” logarithms ℓn : gl1tmf∧

p → tmfK(n), n = 1, 2, all primes p

Work one prime at a time; have “fracture squares” gl1tmf∧

p ℓ2

  • ℓ1
  • tmfK(2)

ι1

  • tmfK(1)

γ

(tmfK(1))K(2)

Map(string, tmfK(2)) ≈ ∗, so reduce to string → HoFib(γ) Explicit formulas for ℓ1, ℓ2 identify γ = ι2 ◦ (id − U), where U : tmfK(1) → tmfK(1) is topological lift of “Atkin operator” on p-adic modular forms

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 24 / 29

slide-109
SLIDE 109

Application to String-orientation of tmf

(Ando-Hopkins-R.) Exists MString → tmf realizing Witten genus Must construct null-homotopy of α: string → gl1tmf Proof idea: Above techniques give “locally defined” logarithms ℓn : gl1tmf∧

p → tmfK(n), n = 1, 2, all primes p

Work one prime at a time; have “fracture squares” gl1tmf∧

p ℓ2

  • ℓ1
  • tmfK(2)

ι1

  • tmfK(1)

γ

(tmfK(1))K(2)

Map(string, tmfK(2)) ≈ ∗, so reduce to string → HoFib(γ) Explicit formulas for ℓ1, ℓ2 identify γ = ι2 ◦ (id − U), where U : tmfK(1) → tmfK(1) is topological lift of “Atkin operator” on p-adic modular forms

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 24 / 29

slide-110
SLIDE 110

Application 3: derived indecomposables

Commutative ring k; augmented comm. k-algebra π: R → k Indecomposables Qk(R) := I/I 2, I = Ker

  • π: R → k
  • (“cotangent space at π”)

Tk(R) := Homk(Qk(R), k) (“tangent space at π”) Commutative ring spectrum k; augmented comm. k-algebra π: R → k (Basterra 1999, Basterra-Mandell 2005) Derived version TQk(R) := “I/I 2” = hocolim Ωn

nuΣn nuI,

nu = non-unital k-algebras TTk(R) := Homk(TQk(R), k) Also called reduced topological Andr´ e-Quillen homology/cohomology

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 25 / 29

slide-111
SLIDE 111

Application 3: derived indecomposables

Commutative ring k; augmented comm. k-algebra π: R → k Indecomposables Qk(R) := I/I 2, I = Ker

  • π: R → k
  • (“cotangent space at π”)

Tk(R) := Homk(Qk(R), k) (“tangent space at π”) Commutative ring spectrum k; augmented comm. k-algebra π: R → k (Basterra 1999, Basterra-Mandell 2005) Derived version TQk(R) := “I/I 2” = hocolim Ωn

nuΣn nuI,

nu = non-unital k-algebras TTk(R) := Homk(TQk(R), k) Also called reduced topological Andr´ e-Quillen homology/cohomology

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 25 / 29

slide-112
SLIDE 112

Rational homotopy and variants

HA = Eilenberg-MacLane spectrum, representing H∗(−, A) (Sullivan 1977) X = simply connected f. type space; HQX+ = rational cochains spectrum π∗TTHQ(HQX+) ≈ π∗X ⊗ Q (Mandell 2006) X = simply connected f. type space; HF

X+ p

= mod p cochains spectrum π∗TTHFp(HF

X+ p )∧ p ≈ 0

(X ∧

p can be recovered from HF X+ p , but not this way)

Q: Are there structured commutative rings R that behave like HQ? Yes: K(n)-local R, such as Morava E-theories

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 26 / 29

slide-113
SLIDE 113

Rational homotopy and variants

HA = Eilenberg-MacLane spectrum, representing H∗(−, A) (Sullivan 1977) X = simply connected f. type space; HQX+ = rational cochains spectrum π∗TTHQ(HQX+) ≈ π∗X ⊗ Q (Mandell 2006) X = simply connected f. type space; HF

X+ p

= mod p cochains spectrum π∗TTHFp(HF

X+ p )∧ p ≈ 0

(X ∧

p can be recovered from HF X+ p , but not this way)

Q: Are there structured commutative rings R that behave like HQ? Yes: K(n)-local R, such as Morava E-theories

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 26 / 29

slide-114
SLIDE 114

Rational homotopy and variants

HA = Eilenberg-MacLane spectrum, representing H∗(−, A) (Sullivan 1977) X = simply connected f. type space; HQX+ = rational cochains spectrum π∗TTHQ(HQX+) ≈ π∗X ⊗ Q (Mandell 2006) X = simply connected f. type space; HF

X+ p

= mod p cochains spectrum π∗TTHFp(HF

X+ p )∧ p ≈ 0

(X ∧

p can be recovered from HF X+ p , but not this way)

Q: Are there structured commutative rings R that behave like HQ? Yes: K(n)-local R, such as Morava E-theories

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 26 / 29

slide-115
SLIDE 115

Rational homotopy and variants

HA = Eilenberg-MacLane spectrum, representing H∗(−, A) (Sullivan 1977) X = simply connected f. type space; HQX+ = rational cochains spectrum π∗TTHQ(HQX+) ≈ π∗X ⊗ Q (Mandell 2006) X = simply connected f. type space; HF

X+ p

= mod p cochains spectrum π∗TTHFp(HF

X+ p )∧ p ≈ 0

(X ∧

p can be recovered from HF X+ p , but not this way)

Q: Are there structured commutative rings R that behave like HQ? Yes: K(n)-local R, such as Morava E-theories

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 26 / 29

slide-116
SLIDE 116

Indecomposables and Bousfield-Kuhn functor

(Behrens-R., in progress) E = Morava E-theory at height n; X = S2d−1 odd dimensional sphere π∗TTE(E X+) ≈ E ∗ΦnX Bousfield-Kuhn functor Φn : Spaces∗ → SpectraK(n) Φn carries part of the “vn-local homotopy groups of X” Spectral sequence computing derived tangent space E = Morava E-theory, π∗R smooth over π∗E, E 2

s,t = Exts Γ(ω−1/2 ⊗ Qπ∗E(π∗R), ω(t−1)/2 ⊗ nul) =

⇒ π∗TTE(R) ωt/2 ≈ ˜ E 0(St), nul = E0 with trivial Γ-action; E 2

s,t = 0 if s > n

Combine E 2

s,t = Exts Γ(ωd−1, ω(t−1)/2 ⊗ nul) =

⇒ E ∗ΦnS2d−1 Recovers known calc at n = 1; collapses to E ∗Φ2S2d−1 = Ext2 at n = 2

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 27 / 29

slide-117
SLIDE 117

Indecomposables and Bousfield-Kuhn functor

(Behrens-R., in progress) E = Morava E-theory at height n; X = S2d−1 odd dimensional sphere π∗TTE(E X+) ≈ E ∗ΦnX Bousfield-Kuhn functor Φn : Spaces∗ → SpectraK(n) Φn carries part of the “vn-local homotopy groups of X” Spectral sequence computing derived tangent space E = Morava E-theory, π∗R smooth over π∗E, E 2

s,t = Exts Γ(ω−1/2 ⊗ Qπ∗E(π∗R), ω(t−1)/2 ⊗ nul) =

⇒ π∗TTE(R) ωt/2 ≈ ˜ E 0(St), nul = E0 with trivial Γ-action; E 2

s,t = 0 if s > n

Combine E 2

s,t = Exts Γ(ωd−1, ω(t−1)/2 ⊗ nul) =

⇒ E ∗ΦnS2d−1 Recovers known calc at n = 1; collapses to E ∗Φ2S2d−1 = Ext2 at n = 2

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 27 / 29

slide-118
SLIDE 118

Indecomposables and Bousfield-Kuhn functor

(Behrens-R., in progress) E = Morava E-theory at height n; X = S2d−1 odd dimensional sphere π∗TTE(E X+) ≈ E ∗ΦnX Bousfield-Kuhn functor Φn : Spaces∗ → SpectraK(n) Φn carries part of the “vn-local homotopy groups of X” Spectral sequence computing derived tangent space E = Morava E-theory, π∗R smooth over π∗E, E 2

s,t = Exts Γ(ω−1/2 ⊗ Qπ∗E(π∗R), ω(t−1)/2 ⊗ nul) =

⇒ π∗TTE(R) ωt/2 ≈ ˜ E 0(St), nul = E0 with trivial Γ-action; E 2

s,t = 0 if s > n

Combine E 2

s,t = Exts Γ(ωd−1, ω(t−1)/2 ⊗ nul) =

⇒ E ∗ΦnS2d−1 Recovers known calc at n = 1; collapses to E ∗Φ2S2d−1 = Ext2 at n = 2

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 27 / 29

slide-119
SLIDE 119

Vista: power operations in equivariant elliptic cohomology

Q: Does equivariant elliptic cohomology admit power operations? Analogue of Def: Isog = “pile” of all elliptic curves and isogenies between them = ⇒ Mod(Isog), Com(Isog) Mod(Isog) has analog of Koszul property Mod(Isogp) has homological dimension 2 rel to Qcoh(MEll) Known power operations for EllTate and Ell∧

s.-s. are consistent with

this picture Conjecturally, equivariant elliptic cohomologies which are etale over MEll should be “classified” by the etale site of Isog

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 28 / 29

slide-120
SLIDE 120

Vista: power operations in equivariant elliptic cohomology

Q: Does equivariant elliptic cohomology admit power operations? Analogue of Def: Isog = “pile” of all elliptic curves and isogenies between them = ⇒ Mod(Isog), Com(Isog) Mod(Isog) has analog of Koszul property Mod(Isogp) has homological dimension 2 rel to Qcoh(MEll) Known power operations for EllTate and Ell∧

s.-s. are consistent with

this picture Conjecturally, equivariant elliptic cohomologies which are etale over MEll should be “classified” by the etale site of Isog

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 28 / 29

slide-121
SLIDE 121

Vista: power operations in equivariant elliptic cohomology

Q: Does equivariant elliptic cohomology admit power operations? Analogue of Def: Isog = “pile” of all elliptic curves and isogenies between them = ⇒ Mod(Isog), Com(Isog) Mod(Isog) has analog of Koszul property Mod(Isogp) has homological dimension 2 rel to Qcoh(MEll) Known power operations for EllTate and Ell∧

s.-s. are consistent with

this picture Conjecturally, equivariant elliptic cohomologies which are etale over MEll should be “classified” by the etale site of Isog

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 28 / 29

slide-122
SLIDE 122

Vista: power operations in equivariant elliptic cohomology

Q: Does equivariant elliptic cohomology admit power operations? Analogue of Def: Isog = “pile” of all elliptic curves and isogenies between them = ⇒ Mod(Isog), Com(Isog) Mod(Isog) has analog of Koszul property Mod(Isogp) has homological dimension 2 rel to Qcoh(MEll) Known power operations for EllTate and Ell∧

s.-s. are consistent with

this picture Conjecturally, equivariant elliptic cohomologies which are etale over MEll should be “classified” by the etale site of Isog

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 28 / 29

slide-123
SLIDE 123

Vista: power operations in equivariant elliptic cohomology

Q: Does equivariant elliptic cohomology admit power operations? Analogue of Def: Isog = “pile” of all elliptic curves and isogenies between them = ⇒ Mod(Isog), Com(Isog) Mod(Isog) has analog of Koszul property Mod(Isogp) has homological dimension 2 rel to Qcoh(MEll) Known power operations for EllTate and Ell∧

s.-s. are consistent with

this picture Conjecturally, equivariant elliptic cohomologies which are etale over MEll should be “classified” by the etale site of Isog

Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 28 / 29

slide-124
SLIDE 124

Some references

http://www.math.uiuc.edu/~rezk/rezk-icm-2014-slides.pdf

Ando, “Isogenies of formal group laws and power operations in the cohomology theories En”, Duke J., 1995 Ando, Hopkins, Strickland, “The sigma orientation is an H∞ map”, Amer. J. 2004; arXiv:math/0204053 Arone, Dwyer, Lesh, “Bredon homology of partition complexes”; http://arxiv.org/abs/1306.0056 Atiyah, “Power operations in K-theory”, Q. J. Math Oxford, 1966 Basterra, “Andr´ e-Quillen cohomology of commutative S-algebras”, J. Pure Appl. Alg., 1999 Basterra, Mandell, “Homology and cohomology of E∞-ring spectra”, Math. Z., 2005; arXiv:math/0407209 Ganter, “Stringy power operations in Tate K-theory”, Homology, Homotopy, Appl., 2013; arXiv:math/0701565 Ganter, “Power operations in orbifold Tate K-theory”; arXiv:1301.2754 Goerss, Hopkins, “Moduli spaces of commutative ring spectra”, LMS Lecture Notes 315, 2004 Hopkins, “Algebraic topology and modular forms”, Proc. ICM, 2002 Lubin, Tate, “Formal moduli for one-parameter formal Lie groups”, Bull. Soc. Math. France, 1966 Lupercio, Uribe, “Loop groupoids, gerbes, and twisted sectors on orbifolds”, Contemp. Math. 310, 2002; arXiv:math/0110207 Mathew, Noel, Naumann, “On a nilpotence conjecture of J. P. May”; http://arxiv.org/abs/1403.2023 May, Quinn, Ray, Tornehave, “E∞-ring spaces and E∞-ring spectra”, Springer LNM 577, 1977 Morava, “Completions of complex cobordism”, Springer LNM 658, 1978 Rezk, “The units of a ring spectrum and a logarithmic cohomology operation”, J. AMS, 2006; arXiv:math/0407022 Rezk, “Power operations for Morava E-theory of height 2 at the prime 2”; arXiv:0812.1320. Rezk, “The congruence criterion for power operations in Morava E-theory”, Homology, Homotopy Appl., 2009; arXiv:0902.2499 Rezk, “Rings of power operations for Morava E-theories are Koszul”; http://arxiv.org/abs/1204.4831 Ruan, “Stringy geometry and topology of orbifolds”, Contemp. Math. 310, 2002; arXiv:math/0011149 Strickland, “Finite subgroups of formal groups”, J. Pure Appl. Alg., 1997 Strickland, “Morava E-theory of symmetric groups”, Topology, 1998; arXiv:math/9801125 Sullivan, “Infinitesimal computations in topology”, IHES, 1977 tom Dieck, “The Artin-Hasse logarithm for λ-rings”, Springer LNM 1370, 1989 Wilkerson, “Λ-rings, binomial domains, and vector bundles over CP∞”, Comm. Alg., 1982 Zhu, “The power operation structure on Morava E-theory of height 2 and prime 3”, Alg. & Geom. Top., 2014 Charles Rezk (UIUC) Power operations Seoul, August 18, 2014 29 / 29