Definition 1.1 .

A bounded function u ð‘Ē u italic_u is a mild bounded ancient solution if and only if there exists a pressure p 𝑝 p italic_p such that p = p 1 + p 2 𝑝 superscript 𝑝 1 superscript 𝑝 2 p=p^{1}+p^{2} italic_p = italic_p start_POSTSUPERSCRIPT 1 end_POSTSUPERSCRIPT + italic_p start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT , where the even extension of p 1 superscript 𝑝 1 p^{1} italic_p start_POSTSUPERSCRIPT 1 end_POSTSUPERSCRIPT to the whole ℝ 3 superscript ℝ 3 \mathbb{R}^{3} blackboard_R start_POSTSUPERSCRIPT 3 end_POSTSUPERSCRIPT with respect to x 3 subscript ð‘Ĩ 3 x_{3} italic_x start_POSTSUBSCRIPT 3 end_POSTSUBSCRIPT is L ∞ âĒ ( - ∞ , 0 ; B âĒ M âĒ O âĒ ( ℝ 3 ) ) subscript ðŋ 0 ðĩ 𝑀 𝑂 superscript ℝ 3 L_{\infty}(-\infty,0;BMO(\mathbb{R}^{3})) italic_L start_POSTSUBSCRIPT ∞ end_POSTSUBSCRIPT ( - ∞ , 0 ; italic_B italic_M italic_O ( blackboard_R start_POSTSUPERSCRIPT 3 end_POSTSUPERSCRIPT ) ) -function,

â–ģ âĒ p 1 = - divdiv âĒ u ⊗ u â–ģ superscript 𝑝 1 tensor-product divdiv ð‘Ē ð‘Ē \triangle p^{1}=-{\rm divdiv}\,u\otimes u â–ģ italic_p start_POSTSUPERSCRIPT 1 end_POSTSUPERSCRIPT = - roman_divdiv italic_u ⊗ italic_u (1.14)

in Q - + subscript superscript 𝑄 Q^{+}_{-} italic_Q start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT start_POSTSUBSCRIPT - end_POSTSUBSCRIPT with p , 3 1 âĒ ( x â€ē , 0 , t ) = 0 subscript superscript 𝑝 1 fragments normal-, 3 superscript ð‘Ĩ normal-â€ē 0 ð‘Ą 0 p^{1}_{,3}(x^{\prime},0,t)=0 italic_p start_POSTSUPERSCRIPT 1 end_POSTSUPERSCRIPT start_POSTSUBSCRIPT , 3 end_POSTSUBSCRIPT ( italic_x start_POSTSUPERSCRIPT â€ē end_POSTSUPERSCRIPT , 0 , italic_t ) = 0 and p 2 âĒ ( ⋅ , t ) superscript 𝑝 2 normal-⋅ ð‘Ą p^{2}(\cdot,t) italic_p start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT ( ⋅ , italic_t ) is a harmonic function in ℝ + 3 subscript superscript ℝ 3 \mathbb{R}^{3}_{+} blackboard_R start_POSTSUPERSCRIPT 3 end_POSTSUPERSCRIPT start_POSTSUBSCRIPT + end_POSTSUBSCRIPT whose gradient satisfies the estimate

| ∇ ⁥ p 2 âĒ ( x , t ) | â‰Ī c âĒ ln ⁥ ( 2 + 1 / x 3 ) ∇ superscript 𝑝 2 ð‘Ĩ ð‘Ą 𝑐 2 1 subscript ð‘Ĩ 3 |\nabla p^{2}(x,t)|\leq c\ln(2+1/{x_{3}}) | ∇ italic_p start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT ( italic_x , italic_t ) | â‰Ī italic_c roman_ln ( 2 + 1 / italic_x start_POSTSUBSCRIPT 3 end_POSTSUBSCRIPT ) (1.15)

for all ( x , t ) ∈ Q - + ð‘Ĩ ð‘Ą superscript subscript 𝑄 (x,t)\in Q_{-}^{+} ( italic_x , italic_t ) ∈ italic_Q start_POSTSUBSCRIPT - end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT and has the property

sup x â€ē ∈ ℝ 2 ⁥ | ∇ ⁥ p 2 âĒ ( x , t ) | → 0 → subscript supremum superscript ð‘Ĩ â€ē superscript ℝ 2 ∇ superscript 𝑝 2 ð‘Ĩ ð‘Ą 0 \sup\limits_{x^{\prime}\in\mathbb{R}^{2}}|\nabla p^{2}(x,t)|\to 0 roman_sup start_POSTSUBSCRIPT italic_x start_POSTSUPERSCRIPT â€ē end_POSTSUPERSCRIPT ∈ blackboard_R start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT end_POSTSUBSCRIPT | ∇ italic_p start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT ( italic_x , italic_t ) | → 0 (1.16)

as x 3 → ∞ normal-→ subscript ð‘Ĩ 3 x_{3}\to\infty italic_x start_POSTSUBSCRIPT 3 end_POSTSUBSCRIPT → ∞ ; u ð‘Ē u italic_u and p 𝑝 p italic_p satisfy ( 1.12 ) and

âˆŦ Q - + ( u ⋅ ( ∂ t φ + Δ φ ) + u ⊗ u : ∇ φ + p div φ ) d x d t = 0 fragments subscript subscript superscript 𝑄 fragments ( u ⋅ fragments ( subscript ð‘Ą φ Δ φ ) u tensor-product u : ∇ φ p div φ ) d x d t 0 \int\limits_{Q^{+}_{-}}\Big{(}u\cdot(\partial_{t}\varphi+\Delta\varphi)+u% \otimes u:\nabla\varphi+p{\rm div}\,\varphi\Big{)}dxdt=0 âˆŦ start_POSTSUBSCRIPT italic_Q start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT start_POSTSUBSCRIPT - end_POSTSUBSCRIPT end_POSTSUBSCRIPT ( italic_u ⋅ ( ∂ start_POSTSUBSCRIPT italic_t end_POSTSUBSCRIPT italic_φ + roman_Δ italic_φ ) + italic_u ⊗ italic_u : ∇ italic_φ + italic_p roman_div italic_φ ) italic_d italic_x italic_d italic_t = 0 (1.17)

for any φ ∈ C 0 ∞ âĒ ( Q - ) 𝜑 subscript superscript ðķ 0 subscript 𝑄 \varphi\in C^{\infty}_{0}(Q_{-}) italic_φ ∈ italic_C start_POSTSUPERSCRIPT ∞ end_POSTSUPERSCRIPT start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT ( italic_Q start_POSTSUBSCRIPT - end_POSTSUBSCRIPT ) with φ âĒ ( x â€ē , 0 , t ) = 0 𝜑 superscript ð‘Ĩ normal-â€ē 0 ð‘Ą 0 \varphi(x^{\prime},0,t)=0 italic_φ ( italic_x start_POSTSUPERSCRIPT â€ē end_POSTSUPERSCRIPT , 0 , italic_t ) = 0 for any x â€ē ∈ ℝ 2 superscript ð‘Ĩ normal-â€ē superscript ℝ 2 x^{\prime}\in\mathbb{R}^{2} italic_x start_POSTSUPERSCRIPT â€ē end_POSTSUPERSCRIPT ∈ blackboard_R start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT and for any t < 0 ð‘Ą 0 t<0 italic_t < 0 .


Definition 2.2 .

An H-type algebra is a finite-dimensional real Lie algebra ( ð”Ī , [ ⋅ , ⋅ ] ) ð”Ī normal-⋅ normal-⋅ (\mathfrak{g},[\cdot,\cdot]) ( fraktur_g , [ ⋅ , ⋅ ] ) which can be endowed with an inner product âŸĻ ⋅ , ⋅ âŸĐ normal-⋅ normal-⋅ \left\langle\cdot,\cdot\right\rangle âŸĻ ⋅ , ⋅ âŸĐ such that

[ 𝔷 âŠĨ , 𝔷 âŠĨ ] = 𝔷 , superscript 𝔷 bottom superscript 𝔷 bottom 𝔷 [\mathfrak{z}^{\bot},\mathfrak{z}^{\bot}]=\mathfrak{z}, [ fraktur_z start_POSTSUPERSCRIPT âŠĨ end_POSTSUPERSCRIPT , fraktur_z start_POSTSUPERSCRIPT âŠĨ end_POSTSUPERSCRIPT ] = fraktur_z ,

where 𝔷 𝔷 \mathfrak{z} fraktur_z is the center of ð”Ī ð”Ī \mathfrak{g} fraktur_g . Moreover, for any fixed z ∈ 𝔷 𝑧 𝔷 z\in\mathfrak{z} italic_z ∈ fraktur_z , the map J z : 𝔷 âŠĨ âŸķ 𝔷 âŠĨ normal-: subscript ð― 𝑧 normal-âŸķ superscript 𝔷 bottom superscript 𝔷 bottom J_{z}:\mathfrak{z}^{\bot}\longrightarrow\mathfrak{z}^{\bot} italic_J start_POSTSUBSCRIPT italic_z end_POSTSUBSCRIPT : fraktur_z start_POSTSUPERSCRIPT âŠĨ end_POSTSUPERSCRIPT âŸķ fraktur_z start_POSTSUPERSCRIPT âŠĨ end_POSTSUPERSCRIPT defined by

âŸĻ J z âĒ ( v ) , w âŸĐ = âŸĻ z , [ v , w ] âŸĐ ∀ w ∈ 𝔷 âŠĨ formulae-sequence subscript ð― 𝑧 ð‘Ģ ð‘Ī 𝑧 ð‘Ģ ð‘Ī for-all ð‘Ī superscript 𝔷 bottom \left\langle J_{z}(v),w\right\rangle=\left\langle z,[v,w]\right\rangle\qquad% \forall\,w\in\mathfrak{z}^{\bot} âŸĻ italic_J start_POSTSUBSCRIPT italic_z end_POSTSUBSCRIPT ( italic_v ) , italic_w âŸĐ = âŸĻ italic_z , [ italic_v , italic_w ] âŸĐ ∀ italic_w ∈ fraktur_z start_POSTSUPERSCRIPT âŠĨ end_POSTSUPERSCRIPT

is an orthogonal map whenever âŸĻ z , z âŸĐ = 1 𝑧 𝑧 1 \left\langle z,z\right\rangle=1 âŸĻ italic_z , italic_z âŸĐ = 1 . We say that a simply connected Lie group is an H-type group if its Lie algebra is an H-type algebra.


DÃĐfinition 2.2 .

Un ÃĐlÃĐment z ∈ 𝐞ðū 𝑧 𝐞ðū z\in\hbox{IK} italic_z ∈ IK est Îą 𝛞 \alpha italic_Îą -stable avec Îą ≠ 0 𝛞 0 \alpha\neq 0 italic_Îą ≠ 0 , si

a 1 / Îą âĒ z + b 1 / Îą âĒ z = ( a + b ) 1 / Îą âĒ z superscript 𝑎 1 𝛞 𝑧 superscript 𝑏 1 𝛞 𝑧 superscript 𝑎 𝑏 1 𝛞 𝑧 a^{1/\alpha}z+b^{1/\alpha}z=(a+b)^{1/\alpha}z italic_a start_POSTSUPERSCRIPT 1 / italic_Îą end_POSTSUPERSCRIPT italic_z + italic_b start_POSTSUPERSCRIPT 1 / italic_Îą end_POSTSUPERSCRIPT italic_z = ( italic_a + italic_b ) start_POSTSUPERSCRIPT 1 / italic_Îą end_POSTSUPERSCRIPT italic_z

pour tous a , b > 0 𝑎 𝑏 0 a,b>0 italic_a , italic_b > 0 .


Definition 2.7 ( Îą 𝛞 \alpha italic_Îą -virtual value) .

For Îą â‰Ĩ 0 𝛞 0 \alpha\geq 0 italic_Îą â‰Ĩ 0 , the Îą 𝛞 \alpha italic_Îą -virtual value of v ð‘Ģ v italic_v is

φ Îą âĒ ( v ) = v - Îą âĒ Îŧ âĒ ( v ) . superscript 𝜑 𝛞 ð‘Ģ ð‘Ģ 𝛞 𝜆 ð‘Ģ \varphi^{\alpha}(v)=v-\alpha\lambda(v). italic_φ start_POSTSUPERSCRIPT italic_Îą end_POSTSUPERSCRIPT ( italic_v ) = italic_v - italic_Îą italic_Îŧ ( italic_v ) .

Definition 3.3 (Absolute continuity and singularity of measures) .

Let ( ÎĐ , ÎĢ ) ÎĐ ÎĢ (\Omega,\Sigma) ( roman_ÎĐ , roman_ÎĢ ) be a measurable space, Ξ : ÎĢ → 𝖷 : 𝜇 → ÎĢ 𝖷 \mu:\Sigma\rightarrow\mathsf{X} italic_Ξ : roman_ÎĢ → sansserif_X a vector measure, and Î― : ÎĢ → [ 0 , ∞ ] : 𝜈 → ÎĢ 0 \nu:\Sigma\rightarrow[0,\infty] italic_Î― : roman_ÎĢ → [ 0 , ∞ ] be a (real-valued, positive) measure. We say that Ξ 𝜇 \mu italic_Ξ is absolutely continuous w.r.t. to Î― 𝜈 \nu italic_Î― , and write Ξ ≩ Î― much-less-than 𝜇 𝜈 \mu\ll\nu italic_Ξ ≩ italic_Î― , if

∀ A ∈ ÎĢ : [ Î― ( A ) = 0 âŸđ Ξ ( A ) = 0 ] . fragments for-all A ÎĢ : fragments [ Î― fragments ( A ) 0 italic- âŸđ italic- Ξ fragments ( A ) 0 ] . \forall A\in\Sigma:\Big{[}\ \nu(A)=0\quad\Longrightarrow\quad\mu(A)=0\ \Big{]}\,. ∀ italic_A ∈ roman_ÎĢ : [ italic_Î― ( italic_A ) = 0 âŸđ italic_Ξ ( italic_A ) = 0 ] . (3.3)

Moreover, we say that two real-valued, positive measures Ξ 𝜇 \mu italic_Ξ and Î― 𝜈 \nu italic_Î― are singular, and write Ξ ⟂ Î― perpendicular-to 𝜇 𝜈 \mu\perp\nu italic_Ξ ⟂ italic_Î― , if there exist B 1 , B 2 ∈ ÎĢ subscript ðĩ 1 subscript ðĩ 2 ÎĢ B_{1},B_{2}\in\Sigma italic_B start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT , italic_B start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT ∈ roman_ÎĢ with B 1 ∊ B 2 = ÎĐ subscript ðĩ 1 subscript ðĩ 2 ÎĐ B_{1}\cup B_{2}=\Omega italic_B start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ∊ italic_B start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT = roman_ÎĐ and B 1 âˆĐ B 2 = ∅ subscript ðĩ 1 subscript ðĩ 2 B_{1}\cap B_{2}=\emptyset italic_B start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT âˆĐ italic_B start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT = ∅ such that

∀ A ∈ ÎĢ : Ξ ( A ) = Ξ ( A âˆĐ B 1 ) and Î― ( A ) = Î― ( A âˆĐ B 2 ) fragments for-all A ÎĢ : italic- Ξ fragments ( A ) Ξ fragments ( A subscript ðĩ 1 ) italic- and italic- Î― fragments ( A ) Î― fragments ( A subscript ðĩ 2 ) \forall A\in\Sigma:\quad\quad\mu(A)=\mu\left(A\cap B_{1}\right)\quad\text{and}% \quad\nu(A)=\nu\left(A\cap B_{2}\right) ∀ italic_A ∈ roman_ÎĢ : italic_Ξ ( italic_A ) = italic_Ξ ( italic_A âˆĐ italic_B start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT ) and italic_Î― ( italic_A ) = italic_Î― ( italic_A âˆĐ italic_B start_POSTSUBSCRIPT 2 end_POSTSUBSCRIPT ) (3.4)

Definition 5.22 (The embedding of X â€ē superscript 𝑋 normal-â€ē X^{\prime} italic_X start_POSTSUPERSCRIPT â€ē end_POSTSUPERSCRIPT into X 𝑋 X italic_X )

With the same notation and assumptions as in Definition 5.20 , the map Îđ : X â€ē → X normal-: 𝜄 normal-→ superscript 𝑋 normal-â€ē 𝑋 \iota:X^{\prime}\to X italic_Îđ : italic_X start_POSTSUPERSCRIPT â€ē end_POSTSUPERSCRIPT → italic_X is defined by

Îđ âĒ ( x â€ē ) = x , provided π âĒ ( x ) = x â€ē and b âĒ ( x ) = 0 , 𝜄 superscript ð‘Ĩ â€ē ð‘Ĩ provided π âĒ ( x ) = x â€ē and b âĒ ( x ) = 0 \iota(x^{\prime})=x,\textrm{ provided $\pi(x)=x^{\prime}$ and ${\rm b}(x)=0$}, italic_Îđ ( italic_x start_POSTSUPERSCRIPT â€ē end_POSTSUPERSCRIPT ) = italic_x , provided italic_π ( italic_x ) = italic_x start_POSTSUPERSCRIPT â€ē end_POSTSUPERSCRIPT and roman_b ( italic_x ) = 0 ,

Definition 1

Assume the following commutator condition

[ b , a ] = b âĒ a - a âĒ b = u 𝑏 𝑎 𝑏 𝑎 𝑎 𝑏 ð‘Ē [b,a]=ba-ab=u [ italic_b , italic_a ] = italic_b italic_a - italic_a italic_b = italic_u

and define

H = 1 2 âĒ ( a âĒ b + b âĒ a ) ðŧ 1 2 𝑎 𝑏 𝑏 𝑎 H=\frac{1}{2}(ab+ba) italic_H = divide start_ARG 1 end_ARG start_ARG 2 end_ARG ( italic_a italic_b + italic_b italic_a )

where u ð‘Ē u italic_u is a phase times the unit operator.


Definition 2.16 .

( [ 19 ] ) A left PostLie algebra is a triple ( A , ∘ , [ , ] ) fragments ( A , , fragments [ , ] ) (A,\circ,[\;,\;]) ( italic_A , ∘ , [ , ] ) consisting of a vector space A ðī A italic_A , a product ∘ \circ ∘ and a skew-symmetric product [ , ] fragments [ , ] [\;,\;] [ , ] satisfying the relations:

(25) [ [ x , y ] , z ] + [ [ z , x ] , y ] + [ [ y , z ] , x ] = 0 , ð‘Ĩ ð‘Ķ 𝑧 𝑧 ð‘Ĩ ð‘Ķ ð‘Ķ 𝑧 ð‘Ĩ 0 \displaystyle[[x,y],z]+[[z,x],y]+[[y,z],x]=0, [ [ italic_x , italic_y ] , italic_z ] + [ [ italic_z , italic_x ] , italic_y ] + [ [ italic_y , italic_z ] , italic_x ] = 0 ,
(26) ( x ∘ y ) ∘ z - x ∘ ( y ∘ z ) - ( y ∘ x ) ∘ z + y ∘ ( x ∘ z ) - [ x , y ] ∘ z = 0 , ð‘Ĩ ð‘Ķ 𝑧 ð‘Ĩ ð‘Ķ 𝑧 ð‘Ķ ð‘Ĩ 𝑧 ð‘Ķ ð‘Ĩ 𝑧 ð‘Ĩ ð‘Ķ 𝑧 0 \displaystyle(x\circ y)\circ z-x\circ(y\circ z)-(y\circ x)\circ z+y\circ(x% \circ z)-[x,y]\circ z=0, ( italic_x ∘ italic_y ) ∘ italic_z - italic_x ∘ ( italic_y ∘ italic_z ) - ( italic_y ∘ italic_x ) ∘ italic_z + italic_y ∘ ( italic_x ∘ italic_z ) - [ italic_x , italic_y ] ∘ italic_z = 0 ,
(27) z ∘ [ x , y ] - [ z ∘ x , y ] - [ x , z ∘ y ] = 0 . 𝑧 ð‘Ĩ ð‘Ķ 𝑧 ð‘Ĩ ð‘Ķ ð‘Ĩ 𝑧 ð‘Ķ 0 \displaystyle z\circ[x,y]-[z\circ x,y]-[x,z\circ y]=0. italic_z ∘ [ italic_x , italic_y ] - [ italic_z ∘ italic_x , italic_y ] - [ italic_x , italic_z ∘ italic_y ] = 0 .

Definition 4.7 .

An element r 𝑟 r italic_r of a ring R 𝑅 R italic_R is said regular or simplifiable for the multiplicative law of R 𝑅 R italic_R (or multiplicatively regular or multiplicatively simplifiable) if for any couple ( x , y ) ∈ R 2 ð‘Ĩ ð‘Ķ superscript 𝑅 2 (x,y)\in R^{2} ( italic_x , italic_y ) ∈ italic_R start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT , we have:

r âĒ x = r âĒ y ⇒ x = y 𝑟 ð‘Ĩ 𝑟 ð‘Ķ ⇒ ð‘Ĩ ð‘Ķ rx=ry\,\Rightarrow\,x=y italic_r italic_x = italic_r italic_y ⇒ italic_x = italic_y

and

x âĒ r = y âĒ r ⇒ x = y ð‘Ĩ 𝑟 ð‘Ķ 𝑟 ⇒ ð‘Ĩ ð‘Ķ xr=yr\,\Rightarrow\,x=y italic_x italic_r = italic_y italic_r ⇒ italic_x = italic_y

Definition 3.1 .

We say that φ : ∂ ⁥ U → ∂ ⁥ U ~ : 𝜑 → 𝑈 ~ 𝑈 \varphi:\partial U\to\partial\tilde{U} italic_φ : ∂ italic_U → ∂ ~ start_ARG italic_U end_ARG is conformally fit if there are Riemann maps g : ( ð”ŧ , 0 ) → ( U , c ) : 𝑔 → ð”ŧ 0 𝑈 𝑐 g:({\mathbb{D}},0)\to(U,c) italic_g : ( blackboard_D , 0 ) → ( italic_U , italic_c ) and g ~ : ( ð”ŧ , 0 ) → ( U ~ , c ~ ) : ~ 𝑔 → ð”ŧ 0 ~ 𝑈 ~ 𝑐 \tilde{g}:({\mathbb{D}},0)\to(\tilde{U},\tilde{c}) ~ start_ARG italic_g end_ARG : ( blackboard_D , 0 ) → ( ~ start_ARG italic_U end_ARG , ~ start_ARG italic_c end_ARG ) for which the induced homeomorphism φ ∘ = [ g ~ ] - 1 ∘ [ φ ] ∘ [ g ] : 𝕋 → 𝕋 : superscript 𝜑 superscript delimited-[] ~ 𝑔 1 delimited-[] 𝜑 delimited-[] 𝑔 → 𝕋 𝕋 \varphi^{\circ}=[\tilde{g}]^{-1}\circ[\varphi]\circ[g]:{\mathbb{T}}\to{\mathbb% {T}} italic_φ start_POSTSUPERSCRIPT ∘ end_POSTSUPERSCRIPT = [ ~ start_ARG italic_g end_ARG ] start_POSTSUPERSCRIPT - 1 end_POSTSUPERSCRIPT ∘ [ italic_φ ] ∘ [ italic_g ] : blackboard_T → blackboard_T defined by ( 6 ) is the identity map. By Lemma 2.1 , this is equivalent to the condition

g ~ = φ ∘ g on the set of landing angles for âĒ g . ~ 𝑔 𝜑 𝑔 on the set of landing angles for 𝑔 \tilde{g}=\varphi\circ g\quad\text{on the set of landing angles for}\ g. ~ start_ARG italic_g end_ARG = italic_φ ∘ italic_g on the set of landing angles for italic_g .

Definition 2.5 .

In any dialgebra D 𝐷 D italic_D the dicommutator is the bilinear operation

âŸĻ x , y âŸĐ = x âŠĢ y - y âŠĒ x . does-not-prove ð‘Ĩ ð‘Ķ ð‘Ĩ ð‘Ķ ð‘Ķ proves ð‘Ĩ \langle x,y\rangle=x\dashv y-y\vdash x. âŸĻ italic_x , italic_y âŸĐ = italic_x âŠĢ italic_y - italic_y âŠĒ italic_x .

In what follows, we will write D - superscript 𝐷 D^{-} italic_D start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT to denote ( D , âŸĻ - , - âŸĐ ) 𝐷 (D,\langle-,-\rangle) ( italic_D , âŸĻ - , - âŸĐ ) , i.e., the underlying vector space of D 𝐷 D italic_D with the dicommutator.


Definition 1 (See loos69sp1 )

A symmetric space is a manifold M 𝑀 M italic_M with a differentiable symmetric product ⋅ ⋅ \cdot ⋅ obeying the following conditions:

and moreover


Definition 3.3 .

[ 27 , p.24] A trace form on a Jordan algebra J ð― J italic_J is a symmetric bilinear form Îģ ð›ū \gamma italic_Îģ such that

Îģ âĒ ( u ∙ v , w ) = Îģ âĒ ( u , v ∙ w ) ð›ū ∙ ð‘Ē ð‘Ģ ð‘Ī ð›ū ð‘Ē ∙ ð‘Ģ ð‘Ī \gamma(u\bullet v,w)=\gamma(u,v\bullet w) italic_Îģ ( italic_u ∙ italic_v , italic_w ) = italic_Îģ ( italic_u , italic_v ∙ italic_w ) (7)

for all u , v , w ∈ J ð‘Ē ð‘Ģ ð‘Ī ð― u,v,w\in J italic_u , italic_v , italic_w ∈ italic_J . Equivalently, the operator L v subscript ðŋ ð‘Ģ L_{v} italic_L start_POSTSUBSCRIPT italic_v end_POSTSUBSCRIPT is self-adjoint with respect to Îģ ð›ū \gamma italic_Îģ for all v ∈ J ð‘Ģ ð― v\in J italic_v ∈ italic_J .

Definition 3.12 .

[ 24 , p.52] Let J ð― J italic_J be a Jordan algebra. The unital hull J ^ ^ ð― \hat{J} ^ start_ARG italic_J end_ARG of J ð― J italic_J is the vector space ℝ × J ℝ ð― \mathbb{R}\times J blackboard_R × italic_J equipped with a multiplication ∙ ^ ^ ∙ \hat{\bullet} ^ start_ARG ∙ end_ARG defined by

( Îą , u ) âĒ ∙ ^ âĒ ( Îē , v ) = ( Îą âĒ Îē , Îą âĒ v + Îē âĒ u + u ∙ v ) . 𝛞 ð‘Ē ^ ∙ ð›― ð‘Ģ 𝛞 ð›― 𝛞 ð‘Ģ ð›― ð‘Ē ∙ ð‘Ē ð‘Ģ (\alpha,u)\hat{\bullet}(\beta,v)=(\alpha\beta,\alpha v+\beta u+u\bullet v). ( italic_Îą , italic_u ) ^ start_ARG ∙ end_ARG ( italic_Îē , italic_v ) = ( italic_Îą italic_Îē , italic_Îą italic_v + italic_Îē italic_u + italic_u ∙ italic_v ) . (10)

Definition 2.4 .

Let ( ÎĢ , Îū ) ÎĢ 𝜉 (\Sigma,\xi) ( roman_ÎĢ , italic_Îū ) denote a closed connected coorientable contact manifold, and fix a contact form Îą ∈ ÎĐ 1 âĒ ( ÎĢ ) 𝛞 superscript ÎĐ 1 ÎĢ \alpha\in\Omega^{1}(\Sigma) italic_Îą ∈ roman_ÎĐ start_POSTSUPERSCRIPT 1 end_POSTSUPERSCRIPT ( roman_ÎĢ ) generating Îū 𝜉 \xi italic_Îū . Fix φ ∈ Cont 0 âĒ ( ÎĢ , Îū ) 𝜑 subscript Cont 0 ÎĢ 𝜉 \varphi\in\mbox{Cont}_{0}(\Sigma,\xi) italic_φ ∈ Cont start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT ( roman_ÎĢ , italic_Îū ) . We can write φ * âĒ Îą = ρ âĒ Îą superscript 𝜑 𝛞 𝜌 𝛞 \varphi^{*}\alpha=\rho\alpha italic_φ start_POSTSUPERSCRIPT * end_POSTSUPERSCRIPT italic_Îą = italic_ρ italic_Îą for a smooth positive function ρ 𝜌 \rho italic_ρ on ÎĢ ÎĢ \Sigma roman_ÎĢ . A translated point of φ 𝜑 \varphi italic_φ is a point x ∈ ð‘Ĩ absent x\in italic_x ∈ with the property that there exists η ∈ ℝ 𝜂 ℝ \eta\in\mathbb{R} italic_η ∈ roman_ℝ such that

φ âĒ ( x ) = Îļ η âĒ ( x ) , and ρ âĒ ( x ) = 1 . formulae-sequence 𝜑 ð‘Ĩ superscript 𝜃 𝜂 ð‘Ĩ and 𝜌 ð‘Ĩ 1 \varphi(x)=\theta^{\eta}(x),\ \ \ \mbox{and}\ \ \ \rho(x)=1. italic_φ ( italic_x ) = italic_Îļ start_POSTSUPERSCRIPT italic_η end_POSTSUPERSCRIPT ( italic_x ) , and italic_ρ ( italic_x ) = 1 . (2.11)

We call η 𝜂 \eta italic_η the time-shift of x ð‘Ĩ x italic_x . Note that if the leaf { Îļ t âĒ ( x ) } t ∈ ℝ subscript superscript 𝜃 ð‘Ą ð‘Ĩ ð‘Ą ℝ \{\theta^{t}(x)\}_{t\in\mathbb{R}} { italic_Îļ start_POSTSUPERSCRIPT italic_t end_POSTSUPERSCRIPT ( italic_x ) } start_POSTSUBSCRIPT italic_t ∈ roman_ℝ end_POSTSUBSCRIPT is closed (which is always the case when Îą 𝛞 \alpha italic_Îą is periodic) then the time-shift is not unique. Indeed, if the leaf { Îļ t âĒ ( x ) } t ∈ ℝ subscript superscript 𝜃 ð‘Ą ð‘Ĩ ð‘Ą ℝ \{\theta^{t}(x)\}_{t\in\mathbb{R}} { italic_Îļ start_POSTSUPERSCRIPT italic_t end_POSTSUPERSCRIPT ( italic_x ) } start_POSTSUBSCRIPT italic_t ∈ roman_ℝ end_POSTSUBSCRIPT has period T > 0 𝑇 0 T>0 italic_T > 0 then φ âĒ ( x ) = Îļ η + Î― âĒ T âĒ ( x ) 𝜑 ð‘Ĩ superscript 𝜃 𝜂 𝜈 𝑇 ð‘Ĩ \varphi(x)=\theta^{\eta+\nu T}(x) italic_φ ( italic_x ) = italic_Îļ start_POSTSUPERSCRIPT italic_η + italic_Î― italic_T end_POSTSUPERSCRIPT ( italic_x ) for all Î― ∈ â„Ī 𝜈 â„Ī \nu\in\mathbb{Z} italic_Î― ∈ roman_â„Ī . In this case one needs to keep track of the η 𝜂 \eta italic_η . Finally we say a point x ∈ ÎĢ ð‘Ĩ ÎĢ x\in\Sigma italic_x ∈ roman_ÎĢ is an iterated translated point of φ 𝜑 \varphi italic_φ if it is a translated point for some iteration φ Î― superscript 𝜑 𝜈 \varphi^{\nu} italic_φ start_POSTSUPERSCRIPT italic_Î― end_POSTSUPERSCRIPT .


Definition 2.8 .

The co-adjoint operators d * superscript normal-d d^{*} italic_d start_POSTSUPERSCRIPT * end_POSTSUPERSCRIPT : Given M normal-M M italic_M to be a 2 2 2 2 -dimensional oriented manifold equipped with a Riemannian metric g âĒ ( ⋅ , ⋅ ) normal-g normal-⋅ normal-⋅ g(\cdot,\cdot) italic_g ( ⋅ , ⋅ ) . For each p = 1 , 2 normal-p 1 2 p=1,2 italic_p = 1 , 2 , the coadjoint operator d * : C ∞ âĒ ( ∧ p T * âĒ M ) → C ∞ âĒ ( ∧ p - 1 T * âĒ M ) normal-: superscript normal-d normal-→ superscript normal-C superscript normal-p superscript normal-T normal-M superscript normal-C superscript normal-p 1 superscript normal-T normal-M d^{*}:C^{\infty}(\wedge^{p}T^{*}M)\rightarrow C^{\infty}(\wedge^{p-1}T^{*}M) italic_d start_POSTSUPERSCRIPT * end_POSTSUPERSCRIPT : italic_C start_POSTSUPERSCRIPT ∞ end_POSTSUPERSCRIPT ( ∧ start_POSTSUPERSCRIPT italic_p end_POSTSUPERSCRIPT italic_T start_POSTSUPERSCRIPT * end_POSTSUPERSCRIPT italic_M ) → italic_C start_POSTSUPERSCRIPT ∞ end_POSTSUPERSCRIPT ( ∧ start_POSTSUPERSCRIPT italic_p - 1 end_POSTSUPERSCRIPT italic_T start_POSTSUPERSCRIPT * end_POSTSUPERSCRIPT italic_M ) sending the space of p normal-p p italic_p -forms on M normal-M M italic_M into the space of p - 1 normal-p 1 p-1 italic_p - 1 -forms on M normal-M M italic_M is defined through the following relation.

(2.20) d * = ( - 1 ) 2 âĒ ( p + 1 ) + 1 * d * = - * d * , fragments superscript 𝑑 superscript fragments ( 1 ) 2 𝑝 1 1 d d , d^{*}=(-1)^{2(p+1)+1}*d*=-*d*, italic_d start_POSTSUPERSCRIPT * end_POSTSUPERSCRIPT = ( - 1 ) start_POSTSUPERSCRIPT 2 ( italic_p + 1 ) + 1 end_POSTSUPERSCRIPT * italic_d * = - * italic_d * ,

where the symbol * * * stands for the Hodge-Star operators as defined in Definition 2.5 .