Jean Dolbeault
http://www.ceremade.dauphine.fr/∼dolbeaul Ceremade, Université Paris-Dauphine
October 7, 2019 IIT Bombay
Outline
From ϕ-entropies to H1 hypocoercivity Bϕ-entropies and diffusions
Bϕ-hypocoercivity (H1 framework)
An L2 abstract result and mode-by-mode hypocoercivity BAbstract statement, toy model, global L2 hypocoercivity result BDiffusion limit, application to the torus and a more numerical point of view
BDecay rates in the Euclidean space without confinement Diffusion and kinetic transport with very weak confinement
The Vlasov-Poisson-Fokker-Planck system: linearization and hypocoercivity
From ϕ-entropies to H 1 hypocoercivity
BSome references of related works (Chafaï 2004),(Bolley, Gentil 2010) (Baudoin 2017)
(Monmarché),(Evans, 2017)
(Arnold, Erb, 2014),(Arnold, Stürzer), (Achleitner, Arnold, Stürzer, 2016),(Achleitner, Arnold, Carlen, 2017),(Arnold, Einav, Wöhrer, 2017)
BIn collaboration with X. Li
Definition of the ϕ-entropies
E[w] :=
Z
Rd
ϕ(w)dγ
ϕis a nonnegative convex continuous function onR+ such that ϕ(1) = 0 and 1/ϕ00is concave on (0,+∞):
ϕ00≥0, ϕ≥ϕ(1) = 0 and (1/ϕ00)00≤0 Classical examples
ϕp(w) :=p−11 wp−1−p(w−1)
p∈(1,2]
ϕ1(w) :=w logw−(w−1) The invariant measure
dγ=e−ψdx
whereψis apotentialsuch thate−ψ is in L1(Rd,dx) dγ is a probability measure
Diffusions
Ornstein-Uhlenbeck equationor backward Kolmogorov equation
∂w
∂t =Lw:= ∆w− ∇ψ· ∇w
− Z
Rd
(Lw1)w2dγ= Z
Rd
∇w1· ∇w2dγ ∀w1,w2∈H1(Rd,dγ) 1 =
Z
Rd
w0dγ= Z
Rd
w(t,·)dγ and limt→+∞w(t,·) = 1 d
dtE[w] =− Z
Rd
ϕ00(w)|∇xw|2dγ=:−I[w] (Fisher information) If for some Λ>0: entropy – entropy production inequality
I[w]≥ΛE[w] ∀w∈H1(Rd,dγ) E[w(t,·)]≤E[w0]e−Λt ∀t≥0 Fokker-Planck equation: u=wγ converges tou?=γ
∂u
∂t = ∆u+∇x·(u∇xψ)
Generalized Csiszár-Kullback-Pinsker inequality
(Pinsker),(Csiszár 1967),(Kullback 1967),(Cáceres, Carrillo, JD, 2002)
Proposition
Let p∈[1,2], w∈L1∩Lp(Rd,dγ)be a nonnegative function, and assume thatϕ∈C2(0,+∞)is a nonnegative strictly convex function such thatϕ(1) =ϕ0(1) = 0. If A:= infs∈(0,∞)s2−pϕ00(s)>0, then
E[w]≥2−2pAminn
1,kwkp−2Lp(Rd,dγ)
okw−1k2Lp(Rd,dγ)
Convexity, tensorization and sub-additivity
Z
Rdi
ϕ00(w)|∇w|2dγi =:Iγi[w]≥ΛiEγi[w] ∀w∈H1(Rdi,dγi)
Theorem
If dγ1 and dγ2 are two probability measures on Rd1×Rd2, then Iγ1⊗γ2[w] =
Z
Rd1×Rd2
ϕ00(w)|∇w|2dγ1dγ2
≥min{Λ1,Λ2}Eγ1⊗γ2[w] ∀w∈H1(Rd1×Rd2,dγ)
Iγ1⊗γ2[w] = Z
Rd2
Iγ1[w]dγ2+ Z
Rd1
Iγ2[w]dγ1
Eγ1⊗γ2[w]≤ Z
Rd2
Eγ1[w]dγ2+ Z
Rd1
Eγ2[w]dγ1 ∀w∈L1(dγ1⊗γ2)
Perturbation (Holley-Stroock type) results
Withw:=R
Rdw dγ, assume that Λ
Z
Rd
ϕ(w)dγ−ϕ(w)
≤ Z
Rd
ϕ00(w)|∇w|2dγ ∀w∈H1(dγ) and, for some constantsa, b∈R,
e−bdγ≤dµ≤e−adγ Lemma
Ifϕis a C2 function such thatϕ00>0andwe:=R
Rdw dµ /R
Rddµ, then
ea−bΛ Z
Rd
ϕ(w)−ϕ(w)e −ϕ0(w)(we −w)e dµ≤
Z
Rd
ϕ00(w)|∇w|2dµ
Entropy – entropy production inequalities, linear flows
On a smooth convex bounded domain Ω, consider
∂w
∂t =Lw:= ∆w− ∇ψ· ∇w, ∇w·ν = 0 on ∂Ω d
dt Z
Ω
wp−1
p−1 dγ=−4 p
Z
Ω
|∇z|2dγ and z =wp/2 d
dt Z
Ω
|∇z|2dγ≤ −2 Λ(p) Z
Ω
|∇z|2dγ where Λ(p)>0 is the best constant in the inequality
2 p(p−1)
Z
Ω
|∇X|2dγ+ Z
Ω
Hessψ:X⊗X dγ≥Λ(p) Z
Ω
|X|2dγ Proposition
Z
Ω
wp−1
p−1 dγ≤ 4 pΛ
Z
Ω
|∇wp/2|2dγ for any w s.t.
Z
Ω
w dγ= 1 (Bakry, Emery 1985)
An interpolation inequality
Corollary
Assume that q∈[1,2). WithΛ = Λ(2/q), we have kfk2L2(Rd,dγ)− kfk2Lq(Rd,dγ)
2−q ≤ 1
Λ Z
Rd
|∇f|2dγ ∀f ∈H1(Rd,dγ)
Improved entropy – entropy production inequalities
In the special caseψ(x) =|x|2/2, withz =wp/2, we obtain that 1
2 d dt
Z
Rd
|∇z|2dγ+ Z
Rd
|∇z|2dγ≤ −2 pκp
Z
Rd
|∇z|4 z2 dγ withκp= (p−1) (2−p)/p
Cauchy-Schwarz: R
Rd|∇z|2dγ2
≤R
Rd
|∇z|4 z2 dγR
Rdz2dγ d
dtI[w] + 2I[w]≤ −κp I[w]2 1 + (p−1)E[w]
Proposition
Assume that q∈(1,2) and dγ= (2π)−d/2e−|x|2/2dx. There exists a strictly convex function F such that F(0) = 0 and F0(0) = 1 and
F
kfk2L2(Rd,dγ)−1
≤ k∇fk2L2(Rd,dγ) if kfkLq(Rd,dγ)= 1
ϕ-hypocoercivity (H 1 framework)
Badapt the strategy ofϕ-entropies to kinetic equations
BVillani’s strategy: derive H1estimates (using a twisted Fisher information) and then use standard interpolation inequalities to establish entropy decay rates
The twisted Fisher informationis notthe derivative of theϕ-entropy
Thekinetic Fokker-Planck equation, orVlasov-Fokker-Planck equation:
∂f
∂t +v· ∇xf− ∇xψ· ∇vf = ∆vf+∇v·(v f) (1) withψ(x) =|x|2/2 andkfkL1(Rd×Rd)= 1 has a unique nonnegative stationary solution
f?(x,v) = (2π)−de−
1
2(|x|2+|v|2)
and the functiong=f/f? solves thekinetic Ornstein-Uhlenbeck equation
∂g
∂t +Tg=Lg
with transport operatorTand Ornstein-Uhlenbeck operatorLgiven by
Tg:=v· ∇xg−x· ∇vg and Lg:= ∆vg−v· ∇vg
Sharp rates for the kinetic Fokker-Planck equation
Letψ(x) =|x|2/2,dµ:=f?dx dv,E[g] :=
Z Z
Rd×Rd
ϕp(g)dµ Proposition
Let p∈[1,2]and consider a nonnegative solution g∈L1(Rd×Rd) of the kinetic Fokker-Planck equation. There is a constantC>0such that
E[g(t,·,·)]≤Ce−t ∀t ≥0 and the rate e−t is sharp as t→+∞
(Villani),(Arnold, Erb): a twistedFisher informationfunctional Jλ[h] = (1−λ)
Z
Rd
|∇vh|2dµ+(1−λ) Z
Rd
|∇xh|2dµ+λ Z
Rd
|∇xh+∇vh|2dµ (Arnold, Erb)relies onλ= 1/2 and dtdJ1/2[h(t,·)]≤ −J1/2[h(t,·)]
Improved rates (in the large entropy regime)
Rewrite the decay of theFisher informationfunctional as
−12 d dt
Z
Rd
X⊥·M0X dµ= Z
Rd
X⊥·M1X dµ+ Z
Rd
Y⊥·M2Y dµ whereX = (∇vh,∇xh), Y = (Hvv,Hxv,Mvv,Mxv)
M0=
1 λ λ ν
⊗IdRd, M1=
1−λ 1+λ−ν2
1+λ−ν
2 λ
⊗ IdRd
M2=
1 λ −κ2 −κ λ2 λ ν −κ λ2 −κ ν2
−κ2 −κ λ2 2κ 2κ λ
−κ λ2 −κ ν2 2κ λ 2κν
⊗IdRd×Rd
With constant coefficients λ?(λ, ν) = max
min
X
X⊥·M1X
X⊥·M0X : (λ, ν)∈R2s.t.M2≥0
For (λ, ν) = (1/2), λ?= 1/2 and the eigenvalues ofM2(12,1) are given as a function ofκ= 8 (2−p)/p∈[0,8] are all nonnegative
2 4 6 8
0.5 1.0 1.5
λ1(κ)
λ2(κ) λ3(κ)
λ4(κ)
κ
We know that
Y⊥·M2Y ≥λ1(p, λ)|Y|2
for someλ1(p, λ)>0 and|Y|2≥ kMvvk2so that, by Cauchy-Schwarz, Z
Rd
|∇vh|2dµ 2
≤ Z
Rd
h2dµ Z
Rd
kMvvk2dµ≤c0 Z
Rd
kMvvk2dµ Theorem
Let p∈(1,2)and h be a solution of the kinetic Ornstein-Uhlenbeck equation. Then there exists a functionλ:R+→[1/2,1)such that λ(0) = limt→+∞λ(t) = 1/2 and a function ρ >1/2 s.t.
d
dtJλ(t)[h(t,·)]≤ −2ρ(t)Jλ(t)[h(t,·)]
As a consequence, for any t≥0 we have the global estimate Jλ(t)[h(t,·)]≤J1/2[h0] exp
−2 Z t
0
ρ(s)ds
Let us definea:=etR
Rd|∇vh|2dµ,b:=etR
Rd∇vh· ∇xh dµ, c:=etR
Rd|∇xh|2dµandj:=a+b+c da
dt ≤a−2 (j−c), dc
dt ≤2 (j−a)−c and dj dt ≤0 with the constraintsa≥0,c≥0 andb2≤a c
0.0 0.5 1.0 1.5 2.0
J. Dolbeault Hypocoercivity
An abstract hypocoercivity result and mode-by-mode hypocoercivity
BAbstract statement, toy model and a global L2hypocoercivity result
BMode-by-mode hypocoercivity
BApplication to the torus and numerics BDecay rates in the whole space
Collaboration with E. Bouin, S. Mischler, C. Mouhot, C. Schmeiser
An abstract evolution equation
Let us consider the equation dF
dt +TF =LF (2)
In the framework of kinetic equations,TandLare respectively the transport and the collision operators
We assume thatTandL are respectively anti-Hermitian and Hermitian operators defined on the complex Hilbert space (H,h·,·i)
A:= 1 + (TΠ)∗TΠ−1 (TΠ)∗
∗denotes the adjoint with respect toh·,·i
Π is the orthogonal projection onto the null space ofL
The assumptions
λm, λM, andCM are positive constants such that, for anyF ∈H Bmicroscopic coercivity:
− hLF,Fi ≥λmk(1−Π)Fk2 (H1) Bmacroscopic coercivity:
kTΠFk2≥λMkΠFk2 (H2) Bparabolic macroscopic dynamics:
ΠTΠF = 0 (H3)
Bbounded auxiliary operators:
kAT(1−Π)Fk+kALFk ≤CMk(1−Π)Fk (H4) The estimate
1 2
d
dtkFk2=hLF,Fi ≤ −λmk(1−Π)Fk2 is not enough to conclude thatkF(t,·)k2 decays exponentially
Equivalence and entropy decay
For someδ >0 to be determined later, the L2 entropy / Lyapunov functional is defined by
H[F] := 12kFk2+δRehAF,Fi
as in(J.D.-Mouhot-Schmeiser) so thathATΠF,Fi ∼ kΠFk2 and
− d
dtH[F] = :D[F]
=− hLF,Fi+δhATΠF,Fi
−δRehTAF,Fi+δRehAT(1−Π)F,Fi −δRehALF,Fi
Bentropy decay rate: for anyδ >0 small enough andλ=λ(δ) λH[F]≤D[F]
Bnorm equivalenceofH[F] and kFk2 2− δ
4 kFk2≤H[F]≤2 +δ 4 kFk2
Exponential decay of the entropy
λ= 3 (1+λλM
M)minn
1, λm,(1+λλmλM
M)CM2
o
,δ= 12 min
1, λm,(1+λλmλM
M)CM2
h1(δ, λ) := (δCM)2−4
λm−δ−2 +δ
4 λ δ λM
1 +λM
−2 +δ 4 λ
Theorem
LetL andTbe closed linear operators (respectively Hermitian and anti-Hermitian) onH. Under (H1)–(H4), for any t ≥0
H[F(t,·)]≤H[F0]e−λ?t whereλ? is characterized by
λ?:= sup
λ >0 : ∃δ >0s.t. h1(δ, λ) = 0, λm−δ−14(2 +δ)λ >0
Sketch of the proof
SinceATΠ = 1 + (TΠ)∗TΠ−1
(TΠ)∗TΠ, from (H1) and (H2)
− hLF,Fi+δhATΠF,Fi ≥λmk(1−Π)Fk2+ δ λM
1 +λM kΠFk2 By (H4), we know that
|RehAT(1−Π)F,Fi+ RehALF,Fi| ≤CMkΠFk k(1−Π)Fk The equationG=AF is equivalent to (TΠ)∗F=G+ (TΠ)∗TΠG hTAF,Fi=hG,(TΠ)∗Fi=kGk2+kTΠGk2=kAFk2+kTAFk2 hG,(TΠ)∗Fi ≤ kTAFk k(1−Π)Fk ≤ 1
2µkTAFk2+µ
2 k(1−Π)Fk2 kAFk ≤ 1
2k(1−Π)Fk, kTAFk ≤ k(1−Π)Fk, |hTAF,Fi| ≤ k(1−Π)Fk2 WithX :=k(1−Π)Fkand Y :=kΠFk
D[F]−λH[F]≥(λm−δ)X2+ δ λM 1 +λM
Y2−δCMX Y−2 +δ
4 λ(X2+Y2)
Hypocoercivity
Corollary
For anyδ∈(0,2), ifλ(δ)is the largest positive root of h1(δ, λ) = 0for whichλm−δ−14(2 +δ)λ >0, then for any solution F of (2)
kF(t)k2≤ 2 +δ
2−δe−λ(δ)tkF(0)k2 ∀t≥0 From the norm equivalence ofH[F] andkFk2
2− δ
4 kFk2≤H[F]≤2 +δ 4 kFk2 We use 2−4δkF0k2≤H[F0] so thatλ?≥supδ∈(0,2)λ(δ)
Formal macroscopic (diffusion) limit
Scaled evolution equation εdF
dt +TF =1 εLF
on the Hilbert spaceH. Fε=F0+εF1+ε2F2+O(ε3) asε→0+ ε−1: LF0= 0,
ε0: TF0=LF1, ε1: dFdt0 +TF1=LF2
The first equation reads asF0= ΠF0
The second equation is simply solved byF1=−(TΠ)F0
After projection, the third equation is
d
dt(ΠF0)− ΠT(TΠ)F0= ΠLF2= 0
∂tu+ (TΠ)∗(TΠ)u= 0 is such that dtdkuk2=−2k(TΠ)uk2≤ −2λMkuk2
A toy problem
du
dt = (L−T)u, L=
0 0 0 −1
, T=
0 −k
k 0
, k2≥Λ>0 Non-monotone decay, a well known picture:
see for instance(Filbet, Mouhot, Pareschi, 2006) H-theorem: dtd|u|2=−2u22
macroscopic limit: dudt1 =−k2u1
generalized entropy: H(u) =|u|2− 1+kδk2u1u2
dH
dt = −
2− δk2 1 +k2
u22− δk2
1 +k2u21+ δk 1 +k2u1u2
≤ −(2−δ)u22− δΛ
1 + Λu21+δ 2u1u2
Plots for the toy problem
1 2 3 4 5 6
0.2 0.4 0.6 0.8 1 u12
1 2 3 4 5 6
0.1 0.2 0.3 0.4 0.5 0.6 0.7 u12+u22
1 2 3 4 5 6
0.2 0.4 0.6 0.8 u112
u12 u22
1 2 3 4 5 6
0.2 0.4 0.6 0.8 1 H
Mode-by-mode hypocoercivity
BFokker-Planck equation and scattering collision operators BA mode-by-mode hypocoercivity result
BEnlargement of the space by factorization
BApplication to the torus and some numerical results
(Bouin, J.D., Mischler, Mouhot, Schmeiser)
Fokker-Planck equation with general equilibria
We consider the Cauchy problem
∂tf +v· ∇xf =Lf, f(0,x,v) =f0(x,v) (3) for a distribution functionf(t,x,v), withpositionvariablex∈Rd or x∈Td the flat d-dimensional torus
Fokker-Planckcollision operator with a general equilibriumM Lf =∇v·h
M∇v M−1fi
Notation and assumptions: anadmissible local equilibrium M is positive, radially symmetric and
Z
Rd
M(v)dv= 1, dγ=γ(v)dv:= dv M(v) γis anexponential weightif
lim
|v|→∞
|v|k
γ(v)= lim
|v|→∞M(v)|v|k= 0 ∀k∈(d,∞)
Definitions
Θ = 1 d
Z
Rd
|v|2M(v)dv= Z
Rd
(v·e)2M(v)dv for an arbitrarye∈Sd−1
Z
Rd
v⊗v M(v)dv= Θ Id Then
θ= 1
d k∇vMk2L2(dγ)= 4 d
Z
Rd
|∇v
√ M
2dv<∞ IfM(v) = e−
1 2|v|2
(2π)d/2, then Θ = 1 andθ= 1 σ:= 1
2 pθ/Θ
Microscopic coercivity property(Poincaré inequality): for all u=M−1F∈H1(M dv)
Z
Rd
|∇u|2M dv≥λm
Z
Rd
u−
Z
Rd
u M dv 2
M dv
Scattering collision operators
Scatteringcollision operator Lf =
Z
Rd
σ(·,v0) f(v0)M(·)−f(·)M(v0) dv0
Main assumption on thescattering rateσ: for some positive, finiteσ 1≤σ(v,v0)≤σ ∀v,v0∈Rd
Example: linear BGK operator
Lf =Mρf−f, ρf(t,x) = Z
Rd
f(t,x,v)dv Local mass conservation
Z
Rd
Lf dv = 0 and we have
Z
Rd
|Lf|2dγ≤4σ2 Z
Rd
|Mρf−f|2dγ
The symmetry condition Z
Rd
σ(v,v0)−σ(v0,v)
M(v0)dv0 = 0 ∀v∈Rd implies thelocal mass conservationR
RdLf dv= 0
Micro-reversibility,i.e., the symmetry ofσ, is not required The null space ofLis spanned by the local equilibriumM Lonly acts on the velocity variable
Microscopic coercivity property: for someλm>0 1
2 Z Z
Rd×Rd
σ(v,v0)M(v)M(v0) (u(v)−u(v0))2dv0dv
≥λm
Z
Rd
(u−ρu M)2M dv holds according to Proposition 2.2 of(Degond, Goudon, Poupaud, 2000)for allu=M−1F ∈L2(M dv). Ifσ≡1, thenλm= 1
Fourier modes
In order to perform amode-by-mode hypocoercivityanalysis, we introduce the Fourier representation with respect tox,
f(t,x,v) = Z
Rd
ˆf(t, ξ,v)e−i x·ξdµ(ξ)
dµ(ξ) = (2π)−ddξ anddξ is the Lesbesgue measure ifx∈Rd dµ(ξ) = (2π)−d P
z∈Zdδ(ξ−z) is discrete forx∈Td
Parseval’s identity ifξ∈Zd and Plancherel’s formula ifx ∈Rd read kf(t,·,v)kL2(dx)=
ˆf(t,·,v)
L2(dµ(ξ)) The Cauchy problem is now decoupled in theξ-direction
∂tˆf +Tˆf =Lˆf, ˆf(0, ξ,v) = ˆf0(ξ,v) Tˆf =i(v·ξ) ˆf
For any fixedξ∈Rd, let us apply the abstract result with H= L2(dγ), kFk2=
Z
Rd
|F|2dγ , ΠF =M Z
Rd
F dv =MρF
andTˆf =i(v·ξ) ˆf,TΠF=i(v·ξ)ρFM, kTΠFk2=|ρF|2
Z
Rd
|v·ξ|2M(v)dv = Θ|ξ|2|ρF|2= Θ|ξ|2kΠFk2 (H2)Macroscopic coercivitykTΠFk2≥λMkΠFk2 : λM= Θ|ξ|2 (H3)R
Rdv M(v)dv= 0 The operatorAis given by
AF =−iξ·R
Rdv0F(v0)dv0 1 + Θ|ξ|2 M
A mode-by-mode hypocoercivity result
kAFk=kA(1−Π)Fk ≤ 1 1 + Θ|ξ|2
Z
Rd
|(1−Π)F|
√
M |v·ξ|√ M dv
≤ 1
1 + Θ|ξ|2k(1−Π)Fk Z
Rd
(v·ξ)2M dv 1/2
=
√ Θ|ξ|
1 + Θ|ξ|2k(1−Π)Fk Scattering operatorkLFk2≤4σ2k(1−Π)Fk2
Fokker-Planck (FP) operator kALFk ≤ 2
1 + Θ|ξ|2 Z
Rd
|(1−Π)F|
√
M |ξ·∇v√
M|dv≤
√θ|ξ|
1 + Θ|ξ|2k(1−Π)Fk In both cases withκ=√
θ(FP) orκ= 2σ√
Θ we obtain kALFk ≤ κ|ξ|
1 + Θ|ξ|2k(1−Π)Fk
TAF(v) =−(v·ξ)M 1 + Θ|ξ|2
Z
Rd
(v0·ξ) (1−Π)F(v0)dv0 is estimated by
kTAFk ≤ Θ|ξ|2
1 + Θ|ξ|2k(1−Π)Fk (H4) holds withCM= κ1+Θ|ξ|+Θ|ξ||ξ|22
Two elementary estimates Θ|ξ|2
1 + Θ|ξ|2 ≥ Θ max{1,Θ}
|ξ|2 1 +|ξ|2 λM
(1 +λM)CM2 = Θ (1 + Θ|ξ|2)
(κ+ Θ|ξ|)2 ≥ Θ κ2+ Θ
Mode-by-mode hypocoercivity with exponential weights
Theorem
Let us consider an admissible M and a collision operatorLsatisfying Assumption(H), and takeξ∈Rd. Ifˆf is a solution such that ˆf0(ξ,·)∈L2(dγ), then for any t≥0, we have
ˆf(t, ξ,·)
2 L2(dγ)
≤3e−µξt
ˆf0(ξ,·)
2 L2(dγ)
where
µξ := Λ|ξ|2
1 +|ξ|2 and Λ = Θ
3 max{1,Θ} min
1, λmΘ κ2+ Θ
withκ= 2σ√
Θfor scattering operators andκ=√
θfor (FP) operators
Exponential convergence to equilibrium in T
dThe unique global equilibrium in the casex∈Td is given by f∞(x,v) =ρ∞M(v) with ρ∞= 1
|Td| Z Z
Td×Rd
f0dx dv
Theorem
Assume thatγ has an exponential growth. We consider an admissible M , a collision operatorLsatisfying Assumption (H). There exists a positive constant C such that the solution f of (3) onTd×Rd with initial datum f0∈L2(dx dγ)satisfies
kf(t,·,·)−f∞kL2(dx dγ)≤C kf0−f∞kL2(dx dγ)e−14Λt ∀t≥0
Enlargement of the space by factorization
A simple case (factorization of order 1) of thefactorization method of(Gualdani, Mischler, Mouhot)
Theorem
LetB1,B2 be Banach spaces and letB2 be continuously imbedded in B1, i.e.,k · k1≤c1k · k2. Let BandA+B be the generators of the strongly continuous semigroups eBt and e(A+B)t onB1. If for all t≥0,
e(A+B)t
2→2≤c2e−λ2t, eBt
1→1≤c3e−λ1t, kAk1→2≤c4
wherek · ki→j denotes the operator norm for linear mappings fromBi
toBj. Then there exists a positive constant C =C(c1,c2,c3,c4)such that, for all t≥0,
e(A+B)t 1→1≤
( C(1 +|λ1−λ2|−1)e−min{λ1,λ2}t forλ16=λ2 C(1 +t)e−λ1t forλ1=λ2
Integrating the identity dsd e(A+B)seB(t−s)
=e(A+B)sAeB(t−s) with respect tos∈[0,t] gives
e(A+B)t =eBt+ Z t
0
e(A+B)sAeB(t−s)ds The proof is completed by the straightforward computation
e(A+B)t
1→1≤c3e−λ1t+c1 Z t
0
e(A+B)sAeB(t−s) 1→2ds
≤c3e−λ1t+c1c2c3c4e−λ1t Z t
0
e(λ1−λ2)sds
Weights with polynomial growth
Let us consider the measure
dγk:=γk(v)dv where γk(v) =πd/2 Γ((k−d)/2)
Γ(k/2) 1 +|v|2k/2 for an arbitraryk∈(d,+∞)
We chooseB1= L2(dγk) andB2= L2(dγ) Theorem
LetΛ = 3 max{1,Θ}Θ minn 1,κλ2m+ΘΘ
o
and k∈(d,∞]. For anyξ∈Rd if ˆf is a solution with initial datumˆf0(ξ,·)∈L2(dγk), then there exists a constant C =C(k,d, σ)such that
ˆf(t, ξ,·)
2 L2(dγk)
≤C e−µξt
ˆf0(ξ,·)
2 L2(dγk)
∀t≥0
Fokker-Planck: AF=NχRF andBF =−i(v·ξ)F+LF−AF N andRare two positive constants,χis a smooth cut-off function andχR:=χ(·/R)
For anyRandN large enough, according to Lemma 3.8 of(Mischler, Mouhot, 2016)
Z
Rd
(L−A)(F)F dγk ≤ −λ1 Z
Rd
F2dγk for someλ1>0 ifk>d, andλ2=µξ/2≤1/4
Scattering operator:
AF(v) = M(v) Z
Rd
σ(v,v0)F(v0)dv0 BF(v) = −
i(v·ξ) + Z
Rd
σ(v,v0)M(v0)dv0
F(v) Boundedness: kAFkL2(dγ)≤σ R
Rdγ−1k dv1/2
kFkL2(dγk)
λ1= 1 andλ2=µξ/2≤1/4
Exponential convergence to equilibrium in T
dThe unique global equilibrium in the casex∈Td is given by f∞(x,v) =ρ∞M(v) with ρ∞= 1
|Td| Z Z
Td×Rd
f0dx dv
Theorem
Assume that k∈(d,∞] andγ has an exponential growth if k=∞.
We consider an admissible M , a collision operatorLsatisfying Assumption(H), andΛgiven by (11)
There exists a positive constant Ck such that the solution f of (3) on Td×Rd with initial datum f0∈L2(dx dγk) satisfies
kf(t,·,·)−f∞kL2(dx dγk)≤Ck kf0−f∞kL2(dx dγk)e−14Λt ∀t ≥0
If we represent the flat torusTd by the box [0,2π)d with periodic boundary conditions, the Fourier variable satisfiesξ∈Zd. Forξ= 0, the microscopic coercivity implies
ˆf(t,0,·)−ˆf∞(0,·)
L2(dγ)≤
ˆf0(0,·)−ˆf∞(0,·)
L2(dγ)e−t Otherwiseµξ ≥Λ/2 for anyξ6= 0
Parseval’s identity applies, with measuredγ(v) andC∞=√ 3 The result with weightγk follows from the factorization result for someCk>0
Computation of the constants
BA more numerical point of view
Two simple examples: Ldenotes either theFokker-Planck operator L1f := ∆vf+∇v·(v f)
or thelinear BGK operator
L2f := Πf −f
Πf =ρfM is the projection operator on the normalized Gaussian function
M(v) = e−12|v|2 (2π)d/2 andρf :=R
Rdf dv is the spatial density
Where do we have space for improvements ?
WithX :=k(1−Π)Fkand Y :=kΠFk, we wrote D[F]−λH[F]
≥(λm− δ)X2+ δ λM
1 +λM
Y2−δCMX Y −λ
2 X2+Y2+δX Y
≥(λm−δ)X2+ δ λM
1 +λM
Y2− δCMX Y −2 +δ
4 λ(X2+Y2) We can directly study the positivity condition for the quadratic form
(λm− δ)X2+ δ λM
1 +λM
Y2−δCMX Y −λ
2 X2+Y2+δX Y λm = 1,λM =|ξ|2 andCM =|ξ|(1 +|ξ|)/(1 +|ξ|2)
2 4 6 8 10 0.1
0.2 0.3 0.4
Withλm= 1, λM =|ξ|2 andCM =|ξ|(1 +|ξ|)/(1 +|ξ|2), we optimizeλunder the condition that the quadratic form
(λm− δ)X2+ δ λM
1 +λM
Y2−δCMX Y −λ
2 X2+Y2+δX Y is positive, thus getting aλ(ξ)
By taking alsoδ=δ(ξ) whereξis seen as a parameter, we get a better estimate ofλ(ξ)
2 4 6 8 10 0.1
0.2 0.3 0.4 0.5
By takingδ=δ(ξ), for each value ofξ we build a different Lyapunov function, namely
Hξ[F] := 12kFk2+δ(ξ) RehAF,Fi where the operatorAis given by
AF =−iξ·R
Rdv0F(v0)dv0 1 +|ξ|2 M We can consider
AεF =−iξ·R
Rdv0F(v0)dv0 ε+|ξ|2 M and look for the optimal value ofε...
2 4 6 8 10 0.1
0.2 0.3 0.4 0.5
The dependence ofλin εis monotone, and the limit asε→0+ gives the optimal estimate ofλ. The operator
A0F =−iξ·R
Rdv0F(v0)dv0
|ξ|2 M
is not bounded anymore, but estimates still make sense and limξ→0δ(ξ) = 0 (see below)
2 4 6 8 10
0.05 0.10 0.15 0.20 0.25 0.30 0.35
2 4 6 8 10 0.1
0.2 0.3 0.4
Theorem (Hypocoercivity onTd with exponential weight) Assume thatL=L1 or L=L2. If f is a solution, then
kf(t,·,·)−f∞k2L2(dx dγ)≤C?kf0k2L2(dx dγ)e−λ?t ∀t≥0 with f∞(x,v) =M(v)RR
Td×Rdf0(x,v)dx dv C?≈1.75863 andλ?= 132(5−2√
3)≈0.236292.
(work in progress)
Some comments on recent works
A more algebraic approach based on the spectral analysis of symmetric and non-symmetric operators
On BGK models
(Achleitner, Arnold, Carlen) On Fokker-Planck models (Arnold, Erb)
(Arnold, Stürzer) (Arnold, Einav, Wöhrer)
Decay rates in the whole space
Algebraic decay rates in R
dOn the whole Euclidean space, we can define the entropy H[f] := 12kfk2L2(dx dγk)+δhAf,fidx dγk
Replacing themacroscopic coercivitycondition byNash’s inequality kuk2L2(dx)≤CNashkuk
4 d+2
L1(dx)k∇uk
2d d+2
L2(dx)
proves that
H[f]≤C
H[f0] +kf0k2L1(dx dv)
(1 +t)−d2 Theorem
Assume thatγk has an exponential growth (k =∞) or a polynomial growth of order k>d
There exists a constant C >0 such that, for any t≥0 kf(t,·,·)k2L2(dx dγk)≤C
kf0k2L2(dx dγk)+kf0k2L2(dγk; L1(dx))
(1 +t)−d2
A direct proof... Recall thatµξ= 1+|ξ|Λ|ξ|22 By the Plancherel formula
kf(t,·,·)k2L2(dx dγk)≤C Z
Rd
Z
Rd
e−µξt|ˆf0|2dξ
dγk
if|ξ|<1, thenµξ ≥Λ2 |ξ|2 Z
|ξ|≤1
e−µξt|ˆf0|2dξ≤C kf0(·,v)k2L1(dx)
Z
Rd
e−Λ2|ξ|2tdξ
≤C kf0(·,v)k2L1(dx)t−d2 if|ξ| ≥1, thenµξ ≥Λ/2 when |ξ| ≥1
Z
|ξ|>1
e−µξt|ˆf0|2dξ≤C e−Λ2t kf0(·,v)k2L2(dx)
Improved decay rate for zero average solutions
Theorem
Assume that f0∈L1loc(Rd×Rd)withRR
Rd×Rdf0(x,v)dx dv= 0 and C0:=kf0k2L2(dγk+2; L1(dx))+kf0k2L2(dγk; L1(|x|dx))+kf0k2L2(dx dγk)<∞ Then there exists a constant ck>0 such that
kf(t,·,·)k2L2(dx dγk)≤ckC0(1 +t)−(1+d2)