diffusions
Jean Dolbeault
http://www.ceremade.dauphine.fr/∼dolbeaul Ceremade, Universit´e Paris-Dauphine
June 5, 2015
Conference onRecent trends in geometric analysis, Carry-Le-Rouet, June 1-5, 2015
Fast diffusion equations:
New points of view
1
improved inequalities and scalings
2
scalings and a concavity property
3
improved rates and best matching
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
Improved inequalities and scalings
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
The logarithmic Sobolev inequality
dµ=µdx,µ(x) = (2π)−d/2e−|x|2/2, onRd withd≥1 Gaussian logarithmic Sobolev inequality
Z
Rd|∇u|2dµ≥ 1 2
Z
Rd|u|2log|u|2dµ for any functionu∈H1(Rd,dµ)such thatR
Rd|u|2dµ= 1 ϕ(t) :=d
4
exp 2t
d
−1−2t d
∀t ∈R
[Stam, 1959],[Weissler, 1978],[Bakry, Ledoux (2006)],[Fathi et al.
(2014)],[Dolbeault, Toscani (2014)]
Proposition
Z
Rd|∇u|2dµ−1 2
Z
Rd|u|2 log|u|2dµ≥ϕ Z
Rd|u|2log|u|2dµ
∀u∈H1(Rd,dµ) s.t.
Z
Rd|u|2dµ= 1 and Z
Rd|x|2|u|2dµ=d
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
Consequences for the heat equation
Ornstein-Uhlenbeck equation (or backward Kolmogorov equation)
∂f
∂t = ∆f −x· ∇f
with initial datumf0∈L1+(Rd,(1 +|x|2)dµand define theentropy as E[f] :=
Z
Rd
f logf dµ , d
dtE[f] =−4 Z
Rd|∇√
f|2dµ≤ −2E[f] thus proving that E[f(t,·)]≤ E[f0]e−2t. Moreover,
d dt
Z
Rd
f|x|2dµ= 2 Z
Rd
f (d− |x|2)dµ Theorem
Assume that E[f0]is finite and R
Rdf0|x|2dµ=d R
Rdf0dµ. Then E[f(t,·)]≤ −d
2 logh 1−
1−e−2dE[f0] e−2ti
∀t ≥0
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
Gagliardo-Nirenberg inequalities and the FDE
k∇wkϑL2(Rd)kwk1L−q+1ϑ(Rd)≥CGNkwkL2q(Rd)
With the right choice of the constants, the functional J[w] := 14(q2−1)R
Rd|∇w|2dx+βR
Rd|w|q+1dx−KCαGN R
Rd|w|2qdx2qα is nonnegative and J[w]≥J[w∗] = 0
Theorem
[Dolbeault-Toscani]For some nonnegative, convex, increasing ϕ J[w]≥ϕ
β R
Rd|w∗|q+1dx−R
Rd|w|q+1dx for any w ∈Lq+1(Rd)such that R
Rd|∇w|2dx<∞and R
Rd|w|2q|x|2dx =R
Rdw∗2q|x|2dx
Consequence for decay rates of relative R´enyi entropies: fasterrates of convergence in intermediate asymptotics for ∂u∂t = ∆up
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
Scalings and a concavity property
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
The fast diffusion equation in original variables
Consider the nonlinear diffusion equation in Rd,d ≥1
∂u
∂t = ∆up
with initial datumu(x,t = 0) =u0(x)≥0such thatR
Rdu0dx= 1and R
Rd|x|2u0dx<+∞. The large time behavior of the solutions is governed by the source-type Barenblatt solutions
U⋆(t,x) := 1
κt1/µdB⋆ x κt1/µ
where
µ:= 2 +d(p−1), κ:=
2µp p−1
1/µ
and B⋆is the Barenblatt profile B⋆(x) :=
C⋆− |x|21/(p−1)
+ ifp>1 C⋆+|x|21/(p−1)
ifp<1
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
The entropy
Theentropy is defined by E :=
Z
Rd
updx and theFisher information by
I :=
Z
Rd
u|∇v|2dx with v = p p−1up−1 Ifu solves the fast diffusion equation, then
E′= (1−p) I To computeI′, we will use the fact that
∂v
∂t = (p−1)v∆v+|∇v|2 F := Eσ with σ= µ
d(1−p) = 1+ 2 1−p
1
d +p−1
= 2 d
1 1−p−1 has a linear growth asymptotically ast →+∞
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
The concavity property
Theorem
[Toscani-Savar´e]Assume that p≥1−1d if d >1 and p>0if d = 1.
Then F(t)is increasing,(1−p) F′′(t)≤0 and
t→lim+∞
1
t F(t) = (1−p)σ lim
t→+∞Eσ−1I = (1−p)σEσ⋆−1I⋆
[Dolbeault-Toscani]The inequality
Eσ−1I≥Eσ⋆−1I⋆
is equivalent to the Gagliardo-Nirenberg inequality k∇wkθL2(Rd)kwk1L−q+1θ(Rd)≥CGNkwkL2q(Rd)
if1−1d ≤p<1. Hint: up−1/2= kwkw
L2q(Rd),q=2p−11
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
The proof
Lemma
If u solves ∂u∂t = ∆up with 1d ≤p<1, then I′ = d
dt Z
Rd
u|∇v|2dx =−2 Z
Rd
up
kD2vk2+ (p−1) (∆v)2 dx
kD2vk2= 1
d(∆v)2+
D2v− 1 d∆vId
2
1
σ(1−p)E2−σ (Eσ)′′= (1−p) (σ−1) Z
Rd
u|∇v|2dx 2
−2 1
d +p−1 Z
Rd
updx Z
Rd
up(∆v)2dx
−2 Z
Rd
updx Z
Rd
up
D2v− 1 d ∆vId
2
dx
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
Best matching
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
Relative entropy and best matching
Consider the family of the Barenblatt profiles Bσ(x) :=σ−d2 C⋆+σ1|x|2p−11
∀x ∈Rd
The Barenblatt profile Bσ plays the role of alocal Gibbs stateifC⋆ is chosen so thatR
RdBσdx=R
Rdv dx The relative entropy is defined by
Fσ[v] := 1 p−1
Z
Rd
vp−Bσp−p Bσp−1(v−Bσ) dx To minimize Fσ[v]with respect to σis equivalent to fixσsuch that
σ Z
Rd|x|2B1dx= Z
Rd|x|2Bσdx= Z
Rd|x|2v dx
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
A Csisz´ar-Kullback(-Pinsker) inequality
Letp∈(d+2d ,1)and consider the relative entropy Fσ[u] := 1
p−1 Z
Rd
up−Bσp−p Bσp−1(u−Bσ) dx
Theorem
[J.D., Toscani] Assume that u is a nonnegative function inL1(Rd)such that upand x 7→ |x|2u are both integrable onRd. IfkukL1(Rd)=M and R
Rd|x|2u dx =R
Rd|x|2Bσdx, then Fσ[u]
σd2(1−p) ≥ p 8R
RdB1pdx
C⋆ku−BσkL1(Rd)+1 σ
Z
Rd|x|2|u−Bσ|dx 2
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
Temperature (fast diffusion case)
Thesecond moment functional(temperature) is defined by Θ(t) := 1
d Z
Rd|x|2u(t,x)dx and such that
Θ′= 2 E
0 t
τ(0)
τ(t) τ∞= lims→+∞τ(s) s Θµ/2(s) s (s τ(0))Θ
s (s+τ(t))Θ µ/2
µ/2 s (s+τ∞)Θµ/2 s
Fast diffusion case +
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
Temperature (porous medium case) and delay
LetU⋆s be thebest matching Barenblattfunction, in the sense of relative entropyF[u| U⋆s], among all Barenblatt functions(U⋆s)s>0. We define s as a function oft and consider thedelay given by
τ(t) :=
Θ(t) Θ⋆
µ2
−t
0
s τ(t) τ(0)
τ∞= lims→+∞τ(s) t
s (s+τ
(0)) Θµ/2 s Θµ/2(s)
s (s+τ (t))
Θµ/2 s (s+τ∞)Θµ/2 Porous medium case
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
A result on delays
Theorem
Assume that p≥1−1d and p6= 1. The best matching Barenblatt function of a solution u is (t,x)7→ U⋆(t+τ(t),x)and the function t 7→τ(t)is nondecreasing if p >1 and nonincreasing if1−d1 ≤p<1 With G := Θ1−η2,η=d(1−p) = 2−µ, theR´enyi entropy power functional H := Θ−η2 Eis such that
G′=µH with H := Θ−η2 E H′
1−p = Θ−1−η2 Θ I−dE2
= dE2
Θη2+1(q−1) with q := Θ I dE2 ≥1 dE2= 1
d
− Z
Rd
x· ∇(up)dx 2
= 1 d
Z
Rd
x·u∇v dx 2
≤ 1 d
Z
Rd
u|x|2dx Z
Rd
u|∇v|2dx= Θ I
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
An estimate of the delay
Theorem
If p >1−1d and p 6= 1, then the delay satisfies
t→lim+∞|τ(t)−τ(0)| ≥ |1−p|Θ(0)1−d2(1−p) 2 H⋆
H⋆−H(0)2
Θ(0) I(0)−dE(0)2
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
Fast diffusion equations on manifolds and sharp functional inequalities
1
The sphere
2
The line
3
Compact Riemannian manifolds
4
The cylinder: Caffarelli-Kohn-Nirenberg inequalities
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
The Cylinder: symmetry breaking in CKN inequalities
Interpolation inequalities on the sphere
Joint work with M.J. Esteban, M. Kowalczyk and M. Loss
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
A family of interpolation inequalities on the sphere
The following interpolation inequality holds on the sphere p−2
d Z
Sd|∇u|2d vg+ Z
Sd|u|2d vg ≥ Z
Sd|u|pd vg
2/p
∀u∈H1(Sd,dvg) for anyp∈(2,2∗]with2∗= d2d−2 ifd ≥3
for anyp∈(2,∞)ifd= 2
Here dvg is the uniform probability measure: vg(Sd) = 1
1 is the optimal constant, equality achieved by constants p= 2∗ corresponds to Sobolev’s inequality...
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
The Cylinder: symmetry breaking in CKN inequalities
Stereographic projection
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
Sobolev inequality
The stereographic projection of Sd ⊂Rd×R∋(ρ φ,z)ontoRd: to ρ2+z2= 1,z ∈[−1,1],ρ≥0,φ∈Sd−1 we associatex∈Rd such that r =|x|,φ= |xx|
z = r2−1
r2+ 1 = 1− 2
r2+ 1 , ρ= 2r r2+ 1
and transform any function uonSd into a functionv onRd using u(y) = ρrd−22
v(x) = r22+1d−22
v(x) = (1−z)−d−22v(x) p= 2∗,Sd =14d(d−2)|Sd|2/d: Euclidean Sobolev inequality
Z
Rd|∇v|2dx≥Sd
Z
Rd|v|d−2d2 dx d−2d
∀v∈ D1,2(Rd)
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
The Cylinder: symmetry breaking in CKN inequalities
Extended inequality
Z
Sd|∇u|2d vg ≥ d p−2
"
Z
Sd|u|pd vg
2/p
− Z
Sd|u|2d vg
#
∀u∈H1(Sd,dµ) is valid
for anyp∈(1,2)∪(2,∞)ifd= 1,2 for anyp∈(1,2)∪(2,2∗]ifd ≥3 Casep= 2: Logarithmic Sobolev inequality Z
Sd|∇u|2d vg≥ d 2
Z
Sd|u|2log
|u|2 R
Sd|u|2d vg
d vg ∀u∈H1(Sd,dµ) Casep= 1: Poincar´e inequality
Z
Sd|∇u|2d vg ≥d Z
Sd|u−u¯|2d vg with u¯:=
Z
Sd
u d vg ∀u∈H1(Sd,dµ)
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
Optimality: a perturbation argument
For anyp∈(1,2∗]ifd≥3, anyp>1 ifd = 1or2, it is remarkable that
Q[u] := (p−2)k∇uk2L2(Sd)
kuk2Lp(Sd)− kuk2L2(Sd)
≥ inf
u∈H1(Sd,dµ)Q[u] = 1 d is achieved in the limiting case
Q[1 +εv]∼ k∇vk2L2(Sd)
kvk2L2(Sd)
as ε→0 when v is an eigenfunction associated with the first nonzero eigenvalue of∆g, thus proving the optimality
p<2: a proof by semi-groups using Nelson’s hypercontractivity lemma. p>2: no simple proof based on spectral analysis is available:
[Beckner], an approach based on Lieb’s duality, the Funk-Hecke formula and some (non-trivial) computations
elliptic methods /Γ2formalism of Bakry-Emery / nonlinear flows
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
The Cylinder: symmetry breaking in CKN inequalities
Schwarz symmetrization and the ultraspherical setting
(ξ0, ξ1, . . . ξd)∈Sd,ξd =z, Pd
i=0|ξi|2= 1[Smets-Willem]
Lemma
Up to a rotation, any minimizer ofQdepends only onξd =z
• Letdσ(θ) := (sinZθ)dd−1dθ,Zd :=√ π Γ(
d 2 ) Γ(d+1 2 )
: ∀v ∈H1([0, π],dσ) p−2
d Z π
0 |v′(θ)|2dσ+ Z π
0 |v(θ)|2dσ≥ Z π
0 |v(θ)|pdσ 2p
• Change of variablesz = cosθ,v(θ) =f(z) p−2
d Z 1
−1|f′|2νdνd+ Z 1
−1|f|2dνd ≥ Z 1
−1|f|pdνd
2 p
where νd(z)dz =dνd(z) :=Zd−1νd2−1dz,ν(z) := 1−z2
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
The ultraspherical operator
With dνd=Zd−1νd2−1dz, ν(z) := 1−z2, consider the space L2((−1,1),dνd)with scalar product
hf1,f2i= Z 1
−1
f1f2dνd, kfkp= Z 1
−1
fp dνd
p1
The self-adjointultraspherical operator is
Lf := (1−z2)f′′−d z f′=νf′′+d 2 ν′f′ which satisfieshf1,Lf2i=−R1
−1f1′f2′ν dνd
Proposition
Let p∈[1,2)∪(2,2∗], d ≥1
−hf,Lfi= Z 1
−1|f′|2νdνd≥d kfk2p− kfk22
p−2 ∀f ∈H1([−1,1],dνd)
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
The Cylinder: symmetry breaking in CKN inequalities
Flows on the sphere
Heat flow and the Bakry-Emery method
Fast diffusion (porous media) flow and the choice of the exponents
Joint work with M.J. Esteban, M. Kowalczyk and M. Loss
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
Heat flow and the Bakry-Emery method
With g =fp,i.e. f =gαwithα= 1/p
(Ineq.) −hf,Lfi=−hgα,Lgαi=:I[g]≥d kgk21α− kg2αk1
p−2 =:F[g] Heat flow
∂g
∂t =Lg d
dt kgk1= 0, d
dt kg2αk1=−2 (p−2)hf,Lfi= 2 (p−2) Z 1
−1|f′|2ν dνd
which finally gives d
dtF[g(t,·)] =− d p−2
d
dtkg2αk1=−2dI[g(t,·)]
Ineq. ⇐⇒ d
dtF[g(t,·)]≤ −2dF[g(t,·)] ⇐= d
dtI[g(t,·)]≤ −2dI[g(t,·)]
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
The Cylinder: symmetry breaking in CKN inequalities
The equation forg =fp can be rewritten in terms off as
∂f
∂t =Lf + (p−1)|f′|2 f ν
−1 2
d dt
Z 1
−1|f′|2ν dνd =1 2
d
dthf,Lfi=hLf,Lfi+ (p−1)h|f′|2 f ν,Lfi d
dtI[g(t,·)] + 2dI[g(t,·)]= d dt
Z 1
−1|f′|2νdνd+ 2d Z 1
−1|f′|2νdνd
=−2 Z 1
−1
|f′′|2+ (p−1) d d+ 2
|f′|4
f2 −2 (p−1)d−1 d+ 2
|f′|2f′′
f
ν2dνd
is nonpositive if
|f′′|2+ (p−1) d d+ 2
|f′|4
f2 − 2 (p−1)d−1 d+ 2
|f′|2f′′
f is pointwise nonnegative, which is granted if
(p−1)d−1 d+ 2
2
≤(p−1) d
d+ 2 ⇐⇒ p≤ 2d2+ 1
(d−1)2 = 2#< 2d d−2 = 2∗
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
... up to the critical exponent: a proof in two slides
d dz,L
u= (Lu)′− Lu′ =−2z u′′−d u′ Z 1
−1
(Lu)2dνd = Z 1
−1|u′′|2ν2dνd+d Z 1
−1|u′|2ν dνd
Z 1
−1
(Lu)|u′|2
u νdνd = d d+ 2
Z 1
−1
|u′|4
u2 ν2dνd−2d−1 d+ 2
Z 1
−1
|u′|2u′′
u ν2dνd
On (−1,1), let us consider the porous medium (fast diffusion)flow ut=u2−2β
Lu+κ|u′|2 u ν
Ifκ=β(p−2) + 1, the Lp norm is conserved d
dt Z 1
−1
uβpdνd =βp(κ−β(p−2)−1) Z 1
−1
uβ(p−2)|u′|2ν dνd= 0
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
The Cylinder: symmetry breaking in CKN inequalities
f =uβ,kf′k2L2(Sd)+p−d2
kfk2L2(Sd)− kfk2Lp(Sd)
≥0 ?
A:=
Z 1
−1|u′′|2ν2dνd−2d−1
d+ 2(κ+β−1) Z 1
−1
u′′|u′|2 u ν2dνd
+
κ(β−1) + d
d+ 2(κ+β−1) Z 1
−1
|u′|4 u2 ν2dνd
Ais nonnegative for someβ if 8d2
(d+ 2)2(p−1) (2∗−p)≥0
Ais a sum of squares ifp∈(2,2∗)for an arbitrary choice ofβ in a certain interval (depending on pand
A= Z 1
−1
u′′−p+ 2 6−p
|u′|2 u
2
ν2dνd≥0 ifp= 2∗andβ = 4 6−p
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
The rigidity point of view
Which computation have we done ? ut=u2−2β
Lu+κ|uu′|2ν
− Lu−(β−1)|u′|2
u ν+ λ
p−2u= λ p−2uκ Multiply by Luand integrate
...
Z 1
−1Lu uκdνd=−κ Z 1
−1
uκ|u′|2 u dνd
Multiply by κ|uu′|2 and integrate ...= +κ
Z 1
−1
uκ|u′|2 u dνd
The two terms cancel and we are left only with the two-homogenous terms
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
The Cylinder: symmetry breaking in CKN inequalities
Improvements of the inequalities (subcritical range)
as long as the exponent is either in the range(1,2)or in the range (2,2∗), on can establishimproved inequalities
An improvement automatically gives an explicit stability result of the optimal functions in the (non-improved) inequality
By duality, this provides a stability result for Keller-Lieb-Tirring inequalities
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
What does “improvement” mean ?
An improvedinequality is
dΦ (e)≤i ∀u∈H1(Sd) s.t. kuk2L2(Sd)= 1 for some functionΦsuch thatΦ(0) = 0,Φ′(0) = 1,Φ′>0 and Φ(s)>s for anys. WithΨ(s) :=s−Φ−1(s)
i−de≥d(Ψ◦Φ)(e) ∀u∈H1(Sd) s.t. kuk2L2(Sd)= 1 Lemma (Generalized Csisz´ar-Kullback inequalities)
k∇uk2L2(Sd)− d p−2
hkuk2Lp(Sd)− kuk2L2(Sd)
i
≥dkuk2L2(Sd)(Ψ◦Φ)
Ckuk
2 (1−r) Ls(Sd)
kuk2L2 (Sd) kur−u¯rk2Lq(Sd)
∀u∈H1(Sd) s(p) := max{2,p} andp∈(1,2): q(p) := 2/p,r(p) :=p;p∈(2,4):
q=p/2,r = 2;p≥4: q=p/(p−2),r =p−2
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
The Cylinder: symmetry breaking in CKN inequalities
Linear flow: improved Bakry-Emery method
Cf. [Arnold, JD]
wt =Lw+κ|w′|2 w ν With 2♯:= (d−2d2+11)2
γ1:=
d−1 d+ 2
2
(p−1) (2#−p) if d>1, γ1:= p−1
3 if d= 1 Ifp∈[1,2)∪(2,2♯] andw is a solution, then
d
dt (i−de)≤ −γ1
Z 1
−1
|w′|4
w2 dνd ≤ −γ1 |e′|2 1−(p−2) e Recalling that e′=−i, we get a differential inequality
e′′+ de′ ≥γ1 |e′|2 1− (p−2) e After integration: dΦ(e(0))≤i(0)
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
Nonlinear flow: the H¨older estimate of J. Demange
wt =w2−2β
Lw+κ|w′|2 w
For allp∈[1,2∗],κ=β(p−2) + 1, dtd R1
−1wβpdνd = 0
−2β12 dtd R1
−1
|(wβ)′|2ν+p−d2 w2β−w2β
dνd ≥γR1
−1
|w′|4 w2 ν2dνd
Lemma
For all w ∈H1 (−1,1),dνd
, such thatR1
−1wβpdνd = 1 Z 1
−1
|w′|4
w2 ν2dνd ≥ 1 β2
R1
−1|(wβ)′|2νdνdR1
−1|w′|2ν dνd
R1
−1w2βdνd
δ
.... but there are conditions on β
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
The Cylinder: symmetry breaking in CKN inequalities
Admissible ( p, β ) for d = 5
0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
1 2 3 4 5 6
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
The line
A first example of a non-compact manifold
Joint work with M.J. Esteban, A. Laptev and M. Loss
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
The Cylinder: symmetry breaking in CKN inequalities
One-dimensional Gagliardo-Nirenberg-Sobolev inequalities
kfkLp(R)≤CGN(p)kf′kθL2(R)kfk1L−2(θR) if p∈(2,∞) kfkL2(R) ≤CGN(p)kf′kηL2(R)kfk1L−p(ηR) if p∈(1,2) withθ= p2−p2 andη=22+p−p
The threshold case corresponding to the limit as p→2 is the logarithmic Sobolev inequality
R
Ru2log
u2 kuk2L2 (R)
dx≤ 12kuk2L2(R) log
2 πe
ku′k2L2 (R)
kuk2L2 (R)
Ifp>2,u⋆(x) = (coshx)−p−22 solves
−(p−2)2u′′+ 4u−2p|u|p−2u= 0 Ifp∈(1,2)consider u∗(x) = (cosx)2−p2 ,x∈(−π/2, π/2)
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
Flow
Let us define on H1(R)the functional F[v] :=kv′k2L2(R)+ 4
(p−2)2kvk2L2(R)−Ckvk2Lp(R) s.t.F[u⋆] = 0 With z(x) := tanhx, consider theflow
vt= v1−p2
√1−z2
v′′+ 2p
p−2z v′+ p 2
|v′|2
v + 2
p−2v
Theorem (Dolbeault-Esteban-Laptev-Loss) Let p∈(2,∞). Then
d
dtF[v(t)]≤0 and lim
t→∞F[v(t)] = 0
d
dtF[v(t)] = 0 ⇐⇒ v0(x) =u⋆(x−x0) Similar results for p∈(1,2)
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
The Cylinder: symmetry breaking in CKN inequalities
The inequality ( p > 2) and the ultraspherical operator
The problem on the line is equivalent to the critical problem for the ultraspherical operator
Z
R
|v′|2dx+ 4 (p−2)2
Z
R
|v|2dx≥C Z
R
|v|pdx 2p With
z(x) = tanhx, v⋆= (1−z2)p−21 and v(x) =v⋆(x)f(z(x)) equality is achieved forf = 1and, if we letν(z) := 1−z2, then
Z 1
−1|f′|2νdνd+ 2p (p−2)2
Z 1
−1|f|2dνd ≥ 2p (p−2)2
Z 1
−1|f|pdνd
2 p
where dνp denotes the probability measuredνp(z) :=ζ1
pνp−22 dz d= p−2p2 ⇐⇒ p=d−2d2
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
Change of variables = stereographic projection + Emden-Fowler
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
The Cylinder: symmetry breaking in CKN inequalities
Compact Riemannian manifolds
no sign is required on the Ricci tensor and an improved integral criterion is established
the flow explores the energy landscape... and shows the non-optimality of the improved criterion
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
Riemannian manifolds with positive curvature
(M,g)is a smooth closed compact connected Riemannian manifold dimensiond, no boundary,∆g is the Laplace-Beltrami operator vol(M) = 1,R is the Ricci tensor,λ1=λ1(−∆g)
ρ:= inf
M inf
ξ∈Sd−1
R(ξ , ξ) Theorem (Licois-V´eron, Bakry-Ledoux)
Assume d ≥2 andρ >0. If λ≤(1−θ)λ1+θ dρ
d−1 where θ= (d−1)2(p−1) d(d+ 2) +p−1 >0 then for any p∈(2,2∗), the equation
−∆gv+ λ
p−2 v−vp−1
= 0 has a unique positive solution v ∈C2(M): v ≡1
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
The Cylinder: symmetry breaking in CKN inequalities
Riemannian manifolds: first improvement
Theorem (Dolbeault-Esteban-Loss) For any p∈(1,2)∪(2,2∗)
0< λ < λ⋆= inf
u∈H2(M)
Z
M
h(1−θ) (∆gu)2+ θd
d−1R(∇u,∇u)i d vg
R
M|∇u|2d vg
there is a unique positive solution in C2(M): u≡1 limp→1+θ(p) = 0 =⇒limp→1+λ⋆(p) =λ1 ifρis bounded λ⋆=λ1=dρ/(d−1) =d ifM=Sd sinceρ=d−1
(1−θ)λ1+θ dρ
d−1 ≤λ⋆≤λ1
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
Riemannian manifolds: second improvement
Hgu denotes Hessian ofuandθ= (d−1)2(p−1) d(d+ 2) +p−1 Qgu:=Hgu−g
d ∆gu−(d−1) (p−1) θ(d+ 3−p)
∇u⊗ ∇u
u −g
d
|∇u|2 u
Λ⋆:= inf
u∈H2(M)\{0}
(1−θ) Z
M
(∆gu)2d vg+ θd d−1
Z
M
hkQguk2+R(∇u,∇u)i Z
M|∇u|2d vg
Theorem (Dolbeault-Esteban-Loss)
Assume that Λ⋆>0. For any p∈(1,2)∪(2,2∗), the equation has a unique positive solution in C2(M)ifλ∈(0,Λ⋆): u≡1
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
The Cylinder: symmetry breaking in CKN inequalities
Optimal interpolation inequality
For any p∈(1,2)∪(2,2∗)orp= 2∗ ifd≥3 k∇vk2L2(M)≥ λ
p−2
hkvk2Lp(M)− kvk2L2(M)
i ∀v ∈H1(M) Theorem (Dolbeault-Esteban-Loss)
AssumeΛ⋆>0. The above inequality holds for someλ= Λ∈[Λ⋆, λ1] IfΛ⋆< λ1, then the optimal constantΛ is such that
Λ⋆<Λ≤λ1
If p = 1, thenΛ =λ1
Using u= 1 +ε ϕas a test function whereϕwe getλ≤λ1
A minimum of
v 7→ k∇vk2L2(M)−p−λ2 h
kvk2Lp(M)− kvk2L2(M)
i under the constraintkvkLp(M)= 1is negative ifλ > λ1
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
The flow
The key tools the flow ut =u2−2β
∆gu+κ|∇u|2 u
, κ= 1 +β(p−2) Ifv =uβ, then dtdkvkLp(M) = 0and the functional
F[u] :=
Z
M|∇(uβ)|2d vg+ λ p−2
"
Z
M
u2βd vg− Z
M
uβpd vg
2/p#
is monotone decaying
J. Demange, Improved Gagliardo-Nirenberg-Sobolev inequalities on manifolds with positive curvature, J. Funct. Anal., 254 (2008), pp. 593–611. Also see C. Villani, Optimal Transport, Old and New
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
The Cylinder: symmetry breaking in CKN inequalities
Elementary observations (1/2)
Letd ≥2,u∈C2(M), and consider the trace free Hessian Lgu:=Hgu−g
d ∆gu Lemma
Z
M
(∆gu)2d vg= d d−1
Z
M
kLguk2d vg+ d d−1
Z
M
R(∇u,∇u)d vg
Based on the Bochner-Lichnerovicz-Weitzenb¨ock formula 1
2∆|∇u|2=kHguk2+∇(∆gu)· ∇u+R(∇u,∇u)
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
Elementary observations (2/2)
Lemma Z
M
∆gu|∇u|2 u d vg
= d
d+ 2 Z
M
|∇u|4
u2 d vg− 2d d+ 2
Z
M
[Lgu] :
∇u⊗ ∇u u
d vg
Lemma
Z
M
(∆gu)2d vg≥λ1
Z
M
|∇u|2d vg ∀u∈H2(M) andλ1is the optimal constant in the above inequality
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
The Cylinder: symmetry breaking in CKN inequalities
The key estimates
G[u] :=R
M
hθ(∆gu)2+ (κ+β−1) ∆gu|∇u|u 2 +κ(β−1)|∇u|u24
id vg
Lemma 1 2β2
d
dtF[u] =−(1−θ) Z
M
(∆gu)2d vg− G[u] +λ Z
M
|∇u|2d vg
Qθgu:=Lgu−1θ dd+2−1(κ+β−1)h
∇u⊗∇u
u −gd |∇u|u 2i Lemma
G[u] = θd d−1
Z
MkQθguk2d vg+ Z
M
R(∇u,∇u)d vg
−µ Z
M
|∇u|4 u2 d vg
withµ:= 1 θ
d−1 d+ 2
2
(κ+β−1)2−κ(β−1)−(κ+β−1) d d+ 2
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
The end of the proof
Assume thatd ≥2. Ifθ= 1, thenµis nonpositive if β−(p)≤β≤β+(p) ∀p∈(1,2∗) where β±:= b±√2 ab2−a witha = 2−p+h
(d−1) (p−1) d+2
i2
andb = d+3d+2−p Notice that β−(p)< β+(p)ifp∈(1,2∗)andβ−(2∗) =β+(2∗)
θ= (d−1)2(p−1)
d(d+ 2) +p−1 and β= d+ 2 d+ 3−p Proposition
Let d ≥2, p∈(1,2)∪(2,2∗)(p6= 5or d 6= 2) 1
2β2 d
dtF[u]≤(λ−Λ⋆) Z
M
|∇u|2d vg
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
The Cylinder: symmetry breaking in CKN inequalities
The Moser-Trudinger-Onofri inequality on Riemannian manifolds
Joint work with G. Jankowiak and M.J. Esteban
Extension to compact Riemannian manifolds of dimension 2...
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
We shall also denote byRthe Ricci tensor, by Hgu the Hessian ofu and by
Lgu:=Hgu−g d ∆gu
the trace free Hessian. Let us denote by Mguthe trace free tensor Mgu:=∇u⊗ ∇u−g
d |∇u|2 We define
λ⋆:= inf
u∈H2(M)\{0}
Z
M
h
kLgu−12Mguk2+R(∇u,∇u)i
e−u/2d vg
Z
M|∇u|2e−u/2d vg
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
The Cylinder: symmetry breaking in CKN inequalities
Theorem
Assume that d = 2andλ⋆>0. If u is a smooth solution to
−1
2∆gu+λ=eu then u is a constant function if λ∈(0, λ⋆) The Moser-Trudinger-Onofri inequality onM
1
4k∇uk2L2(M)+λ Z
M
u d vg≥λlog Z
M
eud vg
∀u∈H1(M) for some constantλ >0. Let us denote byλ1the first positive eigenvalue of−∆g
Corollary
If d = 2, then the MTO inequality holds withλ= Λ := min{4π, λ⋆}. Moreover, if Λis strictly smaller thanλ1/2, then the optimal constant in the MTO inequality is strictly larger thanΛ
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
The flow
∂f
∂t = ∆g(e−f/2)−12|∇f|2e−f/2 Gλ[f] :=
Z
MkLgf −12Mgf k2e−f/2d vg+ Z
M
R(∇f,∇f)e−f/2d vg
−λ Z
M
|∇f|2e−f/2d vg Then for any λ≤λ⋆ we have
d
dtFλ[f(t,·)] = Z
M −12∆gf +λ
∆g(e−f/2)−12|∇f|2e−f/2 d vg
=− Gλ[f(t,·)]
Since Fλ is nonnegative andlimt→∞Fλ[f(t,·)] = 0, we obtain that Fλ[u]≥
Z ∞
0 Gλ[f(t,·)]dt
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
The Cylinder: symmetry breaking in CKN inequalities
Weighted Moser-Trudinger-Onofri inequalities on the two-dimensional Euclidean space
On the Euclidean spaceR2, given a general probability measureµ does the inequality
1 16π
Z
R2|∇u|2dx≥λ
log Z
Rd
eudµ
− Z
Rd
u dµ
hold for someλ >0? Let
Λ⋆:= inf
x∈R2
−∆ logµ 8π µ Theorem
Assume that µis a radially symmetric function. Then any radially symmetric solution to the EL equation is a constant if λ <Λ⋆and the inequality holds withλ= Λ⋆ if equality is achieved among radial functions
J. Dolbeault Sharp functional inequalities and nonlinear diffusions
Caffarelli-Kohn-Nirenberg inequalities
Work in progress with M.J. Esteban and M. Loss
J. Dolbeault Sharp functional inequalities and nonlinear diffusions