an introduction to chromatic homotopy theory€¦ · an introduction to chromatic homotopy theory...
TRANSCRIPT
An Introduction to Chromatic Homotopy TheoryPart I : Spectra and Localization
Agnes Beaudry
May 14, 2019
Outline of the Course
(I) Spectra and Localization
(II) Complex Orientations and the Morava K -Theories
(III) The Chromatic Filtration
(IV) Morava E -Theory and the Stabilizer Group
Part I – Spectra and Localization
(1) Cohomology Theories
(2) Spectra
(3) Stable Homotopy Category
(4) Bousfield Localization
(5) What is Chromatic Homotopy Theory?
Cohomology with Coefficients in G
ĄHG˚p´q : CWop
` ÝÑ Ab
1. (Homotopy) If f » g , then ĄHG˚pf q “ ĄHG
˚pgq.
2. (Additivity)
ĄHG˚
˜
ž
iPI
Xi
¸
–ź
iPI
ĄHG˚pXi q
3. (Exactness) For A Ď X a subcomplex, the following sequence is exact:
ĄHG˚pX Aq Ñ ĄHG
˚pX q Ñ ĄHG
˚pAq
4. (Suspension) For each n, there is a natural isomorphism
ĄHGnpX q
–ÝÑ ĄHG
n`1pΣX q.
5. (Dimension) ĄHG˚pS0q “ G in ˚ “ 0.
In fact, it is representable
ĄHGnpX q – rX ,KpG , nqs, KpG , nq
»ÝÑ ΩKpG , n ` 1q.
Eilenberg-Steenrod Axioms
A reduced cohomology theory is a functor
rE˚ : CWop` ÝÑ Ab
which satisfies the following axioms
1. (Homotopy) If f » g then rE˚pf q “ rE˚pgq.
2. (Additivity)
rE˚
˜
ž
iPI
Xi
¸
–ź
iPI
rE˚pXi q
3. (Exactness) For A Ď X a subcomplex, the following sequence is exact:
rE˚pX Aq Ñ rE˚pX q Ñ rE˚pAq
4. (Suspension) For each n, there is a natural isomorphism
rEnpX q–ÝÑ rEn`1pΣX q.
5. (Dimension) rE˚pS0q “ G in ˚ “ 0.
E˚pX q “ rE˚pX`q.
The Brown Representability Theorem
Let E be a cohomology theory. There is a sequence of based spaces En, n ě 0with weak equivalences
ωn : En»ÝÑ ΩEn`1.
such thatrEnpX q – rX ,Ens.
The adjunctionrX ,ΩY s – rΣX ,Y s.
gives the suspension isomorphism
rEnpX q – rX ,Ens–
ωn˝´
// rX ,ΩEn`1s – rΣX ,En`1s – rEn`1pΣX q
The Category of Spectra Sp
Objects. An (Ω-)spectrum E is a sequence of based spaces En, n ě 0 withweak equivalences
ωn : En»ÝÑ ΩEn`1
Morphisms. A map of f : E Ñ F is a sequence of maps fn : En Ñ Fn such thatthe diagram commutes:
Enfn //
ωn
Fn
ωn
ΩEn`1
Ωfn`1 // ΩFn`1
We denote the category of spectra by Sp.
Spectrification
For a sequence E “ tEn : n ě 0u and inclusions ωn : En ãÑ ΩEn`1,
LEn “ limÝÑk
ΩkEn`k , Lωn “ limÝÑk
Ωkωn`k
is a spectrum. This is called spectrification.
Ordinary Cohomology with Coefficients in G
HG˚p´q is represented by
HGn “ KpG , nq ωn : KpG , nq»ÝÑ ΩKpG , n ` 1q.
Complex K -Theory
By Bott Periodicity, ΩU » Z ˆ BU and ΩpZ ˆ BUq » U. Complex K -theoryK˚p´q, is represented by
K “ tZˆ BU,U,Zˆ BU,U,Zˆ BU,U, . . .u.
Suspension Spectra
For a based space X , the suspension spectrum is the spectrification of
pΣ8X qn “ ΣnX ωn : ΣnX Ñ ΩΣn`1X
where ωn is the adjoint to the identity
σn : ΣΣnX“ÝÑ Σn`1X .
We often write X for Σ8X . The sphere spectrum is S0 “ Σ8S0.
Complex Cobordism
MU is the spectrification of the sequence
tMUp1q,ΣMUp1q,MUp2q,ΣMUp2q, . . .u
where MUpnq is the Thom space of the canonical complex vector bundle
γn Ñ GrnpC8q » BUpnq.
Note:i˚γn`1 – γn ‘ C
// γn`1
BUpnq
i // BUpn ` 1q
The map ω2n`1 : ΣMUpnq Ñ MUp2nq is adjoint to:
Σ2MUpnq – Thpi˚γn`1qσ2n`1 // Thpγn`1q “ MUpn ` 1q
The map ω2n : MUpnq Ñ ΩΣMUpnq is adjoint to the identity.
Homotopy Groups
If E is a spectrum, then the rth homotopy group of E is
πrE “ limÝÑn
πr`nEn.
A map f : E Ñ F is a weak equivalence if π˚f is an isomorphism.
Examples
§ The stable homotopy groups of spheres are are
πsr “ πrΣ8S0 – lim
ÝÑn
πr`nSn.
§ π˚HA – A concentrated in ˚ “ 0.
§ π˚K – Zrβ˘1s for β P π2K “ KpCP1q the Bott class, i.e.,
π2rK “ Ztβr u, π2r`1K “ 0.
§ π˚MU – Zrx1, x2, . . .s for xn P π2nMU related to rCPns.
π0MU “ Zt1u, π1MU “ 0, π2MU “ Ztx1u, π4MU “ Ztx21 , x2u.
Homotopy Category
Let C be a category and W a subcategory such that
‚ All isomorphisms of C are in W,
‚ If 2 out of 3 of tf , g , g ˝ f u are in W, then so is the third.
The homotopy category of C is a category HopC,Wq and a functor
ι : C Ñ HopC,Wq
such that, for F : C Ñ D which maps W to isomorphisms, there is
C
ι
F // D
HopC,WqD!FW
;;
such that F–ÝÑ FW ˝ ι.
Stable Homotopy Category
SH “ HopSp,WS0 q for WS0 the weak equivalences is called the stable homo-topy category.
§ rX ,Y s :“ SHpX ,Y q are abelian groups
§ Finite products and coproducts are equivalent in SH§ Closed symmetric monoidal with unit S0:
´^´ : SHˆ SHÑ SH F p´,´q : SHop ˆ SHÑ SH
F pX ^Y ,Zq – F pX ,F pY ,Zqq
§ Triangulated: Σ: SHÑ SH with
ΣX “ X ^S1 Σ´1X “ F pS1,X q.
§ Distinguished triangles
Xf // Y // Cf
// ΣX
where Cf “ Y Yf pX ^r0, 1sq is the cofiber of f .
§ Long exact sequences for r´,Z s and rZ ,´s. For example:
. . . // πnX // πnY // πnCf// πn´1X // . . .
Models for Spectra
There are other choices pC,Wq with SH “ HopC,Wq. In particular, there areclosed symmetric monoidal models for Sp.
Homology and Cohomology
If E is a spectrum, then E -homology is the functor rE˚ : SHÑ Ab
X ÞÑ rEnpX q :“ πnpE ^X q
and E -cohomology is the functor E˚ : SHop Ñ Ab
X ÞÑ rEnpX q :“ π´nF pX ,Eq “ rX ,ΣnE s.
We letEn :“ rEnpS
0q “ πnE “ rE´npS0q “: E´n
Stable Homotopy Groups as a Homology Theory
For E “ S0, this gives:ĂS0˚pX q “ π˚pX q
Weak equivalences are ĂS0˚-isomorphisms and SH is obtained from Sp by
inverting these.
Bousfield Localization
A map f : X Ñ Y is an E -equivalences if
rE˚pf q : rE˚pX q–ÝÑ rE˚pY q
Let WE Ď Sp be the subcategory of E -equivalences and
SHE :“ HopSp,WE q.
§ X is E -acyclic if rE˚pX q “ 0.
§ Y is E -local if rX ,Y s “ 0 when X is E -acyclic.
Let SpE be the full subcategory whose objects are E -local spectra. In fact:
SHE – HopSpE ,W X SpE q.
Exercise
A map f : X Ñ Y in SpE is a weak equivalence iff it is an E -equivalence.
Universal Property of Localization
An E -localization is an E -equivalence η : X Ñ LEX for LEX P SpE . Theseexist and are unique in SH:
LE : SHÑ SHE Ď SH, η : 1SH Ñ LE .
If f : X Ñ Y with Y P SpE , then
Xf //
η
Y
LEX
fE
==
and fE is unique in SH. There is a distinguished triangle
CEX Ñ X Ñ LEX
where CEX is the terminal E˚-acyclic spectrum with a map to X .
Exercise
There is a natural transformation LF_E Ñ LE and,
LE » LELE_F » LE_FLE .
Localization and Completion
For G an abelian group, let SG a a Moore spectrum
š
jPJ S0trju // š
iPI S0tgiu // SG // Σ
š
jPJ S0trju
for a presentationÀ
jPJ Ztrju ÑÀ
iPI Ztgiu Ñ G .
§ If G “ Zppq or Q, then LSGX » SG ^X and
π˚LSGX – π˚X b G .
The p-localization of X is:
Xppq :“ LSZppqX
The rationalization of X is:
XQ :“ LSQX
§ If the groups π˚X are finitely generated, then
π˚LSZpX – π˚X b Zp
where Zp is the p-adic integers. The p-completion of X is
Xp :“ LSZpX .
The Sphere
For the p-local sphere Sppq “ LSZppqS ,
πnSppq –
#
Zppq n “ 0
Torppπsnq n ą 0.
For the rarional sphere SQ “ LSQS ,
πnSQ –
#
Q n “ 0
0 n ą 0.
Connective Spectra
A spectrum is connective if πrX “ 0 for r ă 0. For such X
LHGX » LSGX .
Adams Operations
For k ě 0, there are unstable operations
ψk : K˚pX q Ñ K˚pX q.
They give stable operations
ψk : Kp Ñ Kp
for k P Zˆp The action of ψk on π˚Kp “ Zprβ˘1s is determined by
ψk pβq “ kβ.
K -Theory Localization
Let p be odd and KZp “ K ^SZp so that
π˚KZp – Zprv˘11 s.
There is a distinguished triangle
LKZpS0 // Kp
ψ`´1 // Kp // ΣLKZpS0
for ` a topological generator of Zˆp .
Image of J
Real Bott Periodicity gives π3`4kO – Z. The J homomorphism
J : πnO Ñ πsn
has image
n 3 7 11 15 19 23 27 31 35 39 43Z24 Z240 Z504 Z480 Z264 Z65520 Z24 Z16320 Z28728 Z13200 Z552
Lower Bounds on Im J Section 4.1 99
4.1 Lower Bounds for Im J
After starting this section with the definition of the J–homomorphism, we will
use a homomorphism K(X)→H∗(X;Q) known as the Chern character to show that
an/2 is a lower bound on the order of the image of J in dimension 4n − 1. Then
using real K–theory this bound will be improved to an , and we will take care of the
cases in which the domain of the J–homomorphism is Z2 .
The simplest definition of the J–homomorphism goes as follows. An element
[f ] ∈ πi(O(n)) is represented by a family of isometries fx ∈ O(n) , x ∈ Si , with fx
the identity when x is the basepoint of Si . Writing Sn+i as ∂(Di+1×Dn) = Si×Dn ∪
Di+1×Sn−1 and Sn as Dn/∂Dn , let Jf(x,y) = fx(y) for (x,y) ∈ Si×Dn and let
Jf(Di+1×Sn−1) = ∂Dn , the basepoint of Dn/∂Dn . Clearly f ≃ g implies Jf ≃ Jg ,
so we have a map J :πi(O(n))→πn+i(Sn) . We will tacitly exclude the trivial case
i = 0.
Proposition 4.1. J is a homomorphism.
Proof: We can view Jf as a map In+i→Sn = Dn/∂Dn which on Si×Dn ⊂ In+1 is
given by (x,v)!fx(v) and which sends the complement of Si×Dn to the basepoint
∂Dn . Taking a similar view of Jg , the sum Jf + Jg is obtained by juxtaposing these
two maps on either side of a hyperplane. We may assume fx is the identity for x in
the right half of Si and gx is the identity for x in the left half of Si . Then we obtain
a homotopy from Jf + Jg to J(f + g) by moving the two Si×Dn ’s together until
they coincide, as shown in the figure below. )⊓
→ →We know that πi(O(n)) and πn+i(S
n) are independent of n for n > i + 1, so
we would expect the J–homomorphism defined above
Exercise
Compute π˚LKZpS0 and compare with the p-component of impJq.
Next Time
Fix a prime p. There are spectra Kp0q,Kp1q,Kp2q, . . . ,Kp8q called theMorava K -theories:
LnX :“ LKp0q_..._KpnqX
In fact,LKp0q – LHQ LKp1q – LKZp LKp8q – LHZp
The chromatic convergence theorem (Hopkins–Ravenel) states:
S0ppq – lim
ÐÝn
LnS0
in SH. The nth chromatic layer is LnS0.
Chromatic homotopy theory studies the S0ppq
via this filtration.
π˚Sp2q (Illustration by Isaksen)
3
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 350
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
The E∞-page of the classical Adams spectral sequence
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60 62 64 66 68 70
0
2
4
6
8
10
12
14
16
18
20
22
24
26
28
30
32
34
36
h0 h1 h2 h3
c0
Ph1 Ph2
h23
d0 h30h4
h1h4
Pc0
P 2h1
h2h4
c1
P 2h2
g
Pd0
h4c0
h20i
P 2c0
P 3h1 P 3h2
d20
h24
n
h100 h5
h1h5
d1
q
P 3c0
p
P 4h1
h0h2h5
e20
P 4h2
t
h22h5
x
h20h3h5
h1h3h5
h5c0
h3d1
u
P 2h20i
f1
Ph1h5
g2
P 4c0
z
P 5h1
Ph2h5
d30
P 5h2
g2
h23h5
h5d0
w
B1
N
d0l
Ph5c0
e0r
Pu
h70Q′
B2
d0e20
P 5c0
P 6h1
h5c1
C
h3g2
gn
P 6h2
d1g
e0m
x′
d0u
h0h5i
e20g
P 4h20i
P 6c0
P 7h1
h1Q2
B21
d0w
P 7h2
B3
g3
d20l
D3
A′
d0e0r
h25
h5n
E1 + C0
Rh1X1
d20e2
0
h1H1
X2C′
h250 h6
h1h6
h3Q2
q1
P 7c0
k1
B23
Ph5j
gw
P 8h1
r1
B5 + D′2
d0e0m
Q3
X3
C11
d0x′ e0gr
d20u
P 8h2
d2
h22H1
h3A′
h30G21
h0h4x′
h22h6
p′
D′3
h3(E1 + C0)
h3h6
p1
h2Q3
h1h3H1
d1e1
h1W1
π˚Sp2q (Illustration by Isaksen)
π˚Sp2q (Illustration by Isaksen)
Telescope Conjecture (Ravenel)
The first n-rays are detected by LnS0.
Chromatic Splitting Conjecture (Hopkins)
The gluing data for the chromatic layers is simple.
Thank you!