jagomart
digital resources
picture1_Geometry Pdf 168625 | Lisbonpaper


 149x       Filetype PDF       File size 0.21 MB       Source: www2.math.uconn.edu


File: Geometry Pdf 168625 | Lisbonpaper
riemannian geometry of di s1 s1 revisited mariagordina 1 1 abstract afurtherstudyofriemanniangeometrydi s s ispresented wedescribe hermitian and riemannian metrics on the complexication of the homogeneous space as well as ...

icon picture PDF Filetype PDF | Posted on 25 Jan 2023 | 2 years ago
Partial capture of text on file.
                                      RIEMANNIAN GEOMETRY OF Diff(S1)/S1 REVISITED
                                                                  MARIAGORDINA
                                                                                             1   1
                                        Abstract. AfurtherstudyofRiemanniangeometryDiff(S )/S ispresented.
                                        Wedescribe Hermitian and Riemannian metrics on the complexification of the
                                        homogeneous space, as well as the complexified symplectic form. It is based on
                                        the ideas from [12], where instead of using the K¨ahler structure symmetries to
                                        compute the Ricci curvature, the authors rely on classical finite-dimensional
                                        results of Nomizu et al on Riemannian geometry of homogeneous spaces.
                                                               Table of Contents
                                  1.  Introduction                                                                   1
                                  Acknowledgment.                                                                    2
                                  2.  Virasoro algebra                                                               2
                                  3.  Diff(S1)/S1 as a K¨ahler manifold                                               4
                                  References                                                                         9
                                                                 1. Introduction
                                  Let Diff(S1) be the Virasoro group of orientation-preserving diffeomorphisms of
                               the unit circle.  Then the quotient space Diff(S1)/S1 describes those diffeomor-
                               phisms that fix a point on the circle. The geometry of this infinite-dimensional
                               space has been of interest to physicists (e.g. [8], [7], [19]).
                                  We follow the approach taken in [8, 7, 19, 14] in that we describe the space
                               Diff(S1)/S1 as an infinite dimensional complex manifold with a K¨ahler metric.
                               Theorem3.3 describes properties of the Hermitian and Riemannian metrics, as well
                               as of the complexified symplectic form. Then we introduce the covariant derivative
                               ∇ which is consistent with the K¨ahler structure. We use the expression for the
                               derivative found in [12], where the classical finite-dimensional results of K.Nomizu
                               in [16] for homogeneous spaces were used in this infinite-dimensional setting. The
                               goal of the present article is to clarify certain parts of [12], in particular, Theorem
                               4.5. This theorem stated that the covariant derivative in question is Levi-Civita, but
                               the details were omitted. In the present paper we explicitly define the Riemannian
                               metric g for which ∇ is the Levi-Civita covariant derivative. This is proven in part
                               (3) of Theorem 3.5 of the present paper. To complete the exposition we present
                               the computation of the Riemannian curvature tensor and the Ricci curvature for
                               the covariant derivative ∇. This proof follows the one in [12].
                                  Date: April 10, 2007.
                                  Key words and phrases. Virasoro algebra, group of diffeomorphisms, Ricci curvature.
                                   The research of the author is partially supported by the NSF Grant DMS-0306468 and the
                               Humboldt Foundation Research Fellowship.
                                                                          1
                                 2                                      M. GORDINA
                                    Our interest to the geometry of this infinite-dimensional manifold comes from
                                 attempts to develop stochastic analysis on infinite-dimensional manifolds. Relevant
                                 references include works by H. Airault, V. Bogachev, P. Malliavin, A. Thalmaier
                                 ([2, 6, 3, 4, 5, 10]). A group Brownian motion in Diff(S1) has been constructed
                                 by P.Malliavin in [15]. From the finite-dimensional case we know that the lower
                                 bound of the Ricci curvature controls the growth of the Brownian motion, therefore
                                 a better understanding of the geometry of Diff(S1)/S1 might help in studying a
                                 Brownian motion on this homogeneous space. For further references to the works
                                 exploring the connections between stochastic analysis and Riemannian geometry
                                 in infinite dimensions, mostly in loop groups and their extensions such as current
                                 groups, path spaces and complex Wiener spaces see [9], [11], [17], [18].
                                 Acknowledgment. Theauthor thanks Ana Bela Cruzeiro and Jean-Claude Zam-
                                 brini for organizing a satellite conference of the International Congress of Math-
                                 ematicians on Stochastic Analysis in Mathematical Physics in September of 2006
                                 at the University of Lisbon, Portugal, where the results of this paper have been
                                 presented.
                                                                 2. Virasoro algebra
                                 Notation2.1. LetDiff(S1)bethegroupoforientation preserving C∞-diffeomorphisms
                                 of the unit circle, and diff(S1) its Lie algebra. The elements of diff(S1) will be iden-
                                 tified with the C∞ left-invariant vector fields f(t) d , with the Lie bracket given
                                 by                                                       dt
                                                            [f,g] = fg′ −f′g,f,g ∈ diff(S1).
                                 Definition 2.2. Suppose c,h are positive constants. Then the Virasoro algebra
                                 V    is the vector space R ⊕ diff(S1) with the Lie bracket given by
                                  c,h
                                 (2.1)                   [aκ+f,bκ+g]         =ω (f,g)κ+[f,g],
                                                                        V        c,h
                                                                         c,h
                                 where κ ∈ R is the central element, and ω is the bilinear symmetric form
                                                 ω (f,g)=Z 2π³(2h− c )f′(t)− c f(3)(t)´g(t)dt.
                                                   c,h                     12          12              2π
                                                                0
                                 Remark 2.3. If h = 0, c = 6, then ωc,h is the fundamental cocycle ω (see [3])
                                                            ω(f,g) = −Z 2π³f′+f(3)´g dt.
                                                                           0                4π
                                 Remark 2.4. A simple verification shows that V              with ω     satisfies the Jacobi
                                                                                        c,h        c,h
                                 identity, and therefore V     with this bracket is indeed a Lie algebra. In addition,
                                                            c,h
                                 by the integration by parts formula ω        satisfies
                                                                          c,h
                                 (2.2)                         ω    (f′,g) = −ω      (f,g′).
                                                                 c,h              c,h
                                 Moreover, ωc,h is anti-symmetric
                                 (2.3)                          ωc,h (f,g) = −ωc,h(g,f).
                                 Notation 2.5. Throughout this work we use k,m,n... ∈ N, and α,β,γ... ∈ Z.
                                                                    VIRASORO GROUP                                        3
                                    Below we introduce an inner product on the Lie algebra diff(S1) which has a
                                 natural basis
                                 (2.4)             fk = coskt,gm = sinmt, k = 0,1,2...,m = 1,2....
                                    The Lie bracket in this basis satisfies the following identities
                                                         1 µ                         m−n           ¶
                                             [f ,f ] =       (m−n)g       +(m+n)            g        , m 6= n,
                                               m n       2            m+n           |m−n| |m−n|
                                                         1 µ                         m−n           ¶
                                 (2.5)       [g  , g ] =     (n−m)g       +(m+n)            g        , m 6= n,
                                               m n       2            m+n           |m−n| |m−n|
                                                         1 ¡                               ¢
                                             [fm,gn] = 2 (n−m)fm+n+(m+n)f|m−n| .
                                 Notation 2.6. By diff (S1) we denote the space of functions having mean 0. This
                                                          0
                                                                     1    1          1
                                 space can be identified with diff(S )/S , where S is being viewed as constant vector
                                 fields corresponding to rotations of S1.
                                                                       1
                                    Then any element of f ∈ diff0(S ) can be written
                                                                         ∞
                                                                 f(t) = X(a f +b g ),
                                                                               k k    k k
                                                                        k=1
                                            ∞         ∞       2
                                 with {a }      , {b }     ∈ ℓ since f is smooth. We will also need the following
                                         k k=1     k k=1
                                                              1
                                 endomorphism J of diff0(S )
                                                                           ∞
                                 (2.6)                         J(f)(t) = X(b f −a g ).
                                                                                k k    k k
                                                                          k=1
                                 It satisfies J2 = −I.
                                 Notation 2.7. For any k ∈ Z we denote θ = 2hk+ c (k3−k).
                                                                                k          12
                                 Remark 2.8. Note that θ          =−θ , for any k ∈ Z.
                                                              −k       k
                                    The form ω      and the endomorphism J induce an inner product on diff (S1) by
                                                c,h                                                               0
                                                            hf,gi = ω    (f,Jg) = ω     (g,Jf).
                                                                      c,h            c,h
                                 The last identity follows from Equation (2.3).
                                 Proposition 2.9. hf,gi is an inner product on diff (S1).
                                                                                          0
                                 Proof. Let b = 0, then
                                              0
                                                  Z 2π ³        c    ′     c   (3)  ´          dt
                                   ωc,h(f,Jf) = 0        (2h−12)f (t)−12f         (t) (Jf)(t)2π =
                                        Z 2π Ã ∞                    !Ã ∞                     !            ∞
                                                X                        X                      dt     1 X        2   2
                                                    θ (b f −a g )            (b f −a g )            =        θ (a +b ).
                                                     k  k k    k k             m m     m m      2π     2      k   k   k
                                          0    k=1                      m=1                              k=1
                                 Then for any f ∈ diff (S1)
                                                        0
                                 4                                       M. GORDINA
                                                                         ∞
                                                                      1 X ¡           2        2¢
                                                            hf,fi =         θ   a (f) +b (f)      .
                                                                      2      k    k        k
                                                                        k=1
                                                                                                                            ¤
                                                                   (2n+m)θ
                                 Notation 2.10. Let λ           =          m for any n,m ∈ Z. Then it is easy to check
                                                           m,n       2θ
                                 that                                  m+n
                                 (2.7)                              λm,n = λn,m+m−n.
                                                                                       2
                                                                   1    1          ¨
                                                        3. Diff(S )/S as a Kahler manifold
                                    Denote g = diff(S1), m = diff (S1), h = f R, so that g = m ⊕ h. Then g is an
                                                                      0             0
                                 infinite-dimensional Lie algebra equipped with an inner product h·,·i. Note that
                                 for any n ∈ N
                                                        [f ,f ] = −ng ∈ m, [g ,g ] = nf ∈ m,
                                                          0   n         n           0  n       n
                                 and therefore [h,m] ⊂ m. In addition, h is a Lie subalgebra of g, but m is not a Lie
                                 subalgebra of g since [f ,g ] = mf .
                                                           m m           0
                                                     1                                            1                          1
                                    Let G = Diff(S ) with the associated Lie algebra diff(S ), the subgroup H = S
                                 with the Lie algebra h ⊂ g, then m is a tangent space naturally associated with
                                                       1    1
                                 the quotient Diff(S )/S . For any g ∈ g we denote by g                 (respectively g ) its
                                                                                                    m                   h
                                 m-(respectively h-)component, that is, g = g +g , g              ∈ m, g ∈ h. The fact
                                                                                      m    h   m          h
                                 [h,m] ⊂ mimpliesthatforanyh ∈ htheadjointrepresentationad(h) = [h,·] : g → g
                                 maps m into m. We will abuse notation by using ad(h) for the corresponding
                                 endomorphism of m.
                                    Recall that J : diff (S1) → diff (S1) is an endomorphism defined by (2.6), or
                                                          0             0
                                 equivalently, in the basis {f ,g }, m,n = 1,... by
                                                                m n
                                                                  Jf =−g , Jg =f .
                                                                     m        m      n     n
                                 This is an almost complex structure on diff (S1), and as was shown in [12] it is
                                                                                   0
                                 actually a complex structure for an appropriately chosen connection.
                                    Let g and m be the complexifications of g and m respectively. Now we would
                                          C        C
                                 like to introduce Hermitian metric, Riemannian metric and the complexified sym-
                                 plectic form ωC.
                                                                                                                  2
                                 Notation 3.1. For any f + ig,u + iv ∈ gC, where f,g,u,v ∈ g and i = −1 we
                                 denote
                                               h(f +ig,u+iv)=hf,ui+hg,vi+i(hg,ui−hf,vi);
                                               g(f +ig,u+iv)=hf,ui+hg,vi=Re(h(f +ig,u+iv));
                                              ωC(f +ig,u+iv)=hg,ui−hf,vi=Im(h(f +ig,u+iv)).
                                 We will call h a Hermitian metric, g a Riemannian metric, and ω                a symplectic
                                 form.                                                                        C
                                    The endomorphism J can be naturally extended to gC by J (f +ig) = Jf +iJg
                                 for any f,g ∈ g. We will abuse notation and use the same J for this extended
                                 endomorphism. It is easy to check that J is complex-linear.
The words contained in this file might help you see if this file matches what you are looking for:

...Riemannian geometry of di s revisited mariagordina abstract afurtherstudyofriemanniangeometrydi ispresented wedescribe hermitian and metrics on the complexication homogeneous space as well complexied symplectic form it is based ideas from where instead using k ahler structure symmetries to compute ricci curvature authors rely classical nite dimensional results nomizu et al spaces table contents introduction acknowledgment virasoro algebra a manifold references let be group orientation preserving dieomorphisms unit circle then quotient describes those dieomor phisms that x point this innite has been interest physicists e g we follow approach taken in describe an complex with metric theorem properties introduce covariant derivative which consistent use expression for found were used setting goal present article clarify certain parts particular stated question levi civita but details omitted paper explicitly dene proven part complete exposition computation tensor proof follows one date ap...

no reviews yet
Please Login to review.