In this paragraph, X denotes a locally finite-dimensional manifold of class Cr, where r∈Nk; E and F denote two vector bundles of finite rank of class Cr and with basis X.
All manifolds and vector bundles considered are assumed to be locally finite-dimensional.
14.1. Differential Operators
14.1.1. Let k be an integer such that 0≤k≤r, and let Pk(E) be the bundle of jets of sections of order k of E (12.6.1). Put
Dk(E,F)=L(Pk(E);F).
This is a vector bundle with basis X and of class Cr−k. A section of this bundle over an open set U of X is identified with a mapping
D:Pk(E)∣U⟶F∣U
which commutes with the projection onto U and is linear on each fibre; such a section is called a differential operator on U, of type E→F, and of order ≤k. Let h∈Nk∪{0}, with 0≤h≤r−k. We denote by DUk,h(E,F) the set of differential operators on U, of type E→F and of order ≤k, which are of class Ch (as sections of Dk(E,F), or as morphisms of Pk(E)∣U into F∣U, this amounts to the same thing). If M=Dk(E,F), we have DUk,h(E,F)=FMh(U); this is a module over Ch(U), cf. 7.4.1.
If 0≤k′≤k, the morphism rk,k′:Pk(E)⟶Pk′(E) (cf. 12.6.6) defines an injection of Dk′(E,F) into Dk(E,F); every differential operator of order ≤k′ is thus identified with a differential operator of order ≤k. A differential operator of order ≤k which is not of order ≤k−1 is sometimes said to be of order k. If h∈Nk∪{0} and h≤r−k, and if U is an open set of X, we have:
0⊂DU0,h(E,F)⊂DU1,h(E,F)⊂⋯⊂DUk,h(E,F).
If r=∞ or ω, we denote by DUω,h(E,F) the union of the DUk,h(E,F) for k≥0; an element of this union is called a differential operator on U of bounded order (of type E→F, of class Ch) or sometimes simply a differential operator on U.
14.1.2 (« Operators of order zero »). We have P0(E)=E (12.5.1) and D0(E,F)=L(E;F). If U is open in X, and if h∈Nk∪{0} and h≤r, an element of DU0,h(E,F) is a morphism of class Ch of E∣U into F∣U.
14.1.3 (“Weakening of structure”). Suppose that K=R, and let r′∈NR with r′≤r. Let X′,E′ and F′ be the manifold and the vector bundles of class Cr′ obtained by weakening of structure from X,E, and F respectively. Let k be an integer such that 0≤k≤r′. Then Pk(E′) is the bundle of class Cr′−k deduced from Pk(E) by weakening of structure, and there is an analogous result for Dk(E′,F′). In particular, if U is open in X, and if 0≤h≤r−k′, one has DUk′,h(E′,F′)=DUk,h(E,F).
14.1.4. Let D be a differential operator on X, of type E→F, of order ≤k (where 0≤k≤r). Let U be an open set in X, and let s be a section, of class Cm, of E over U, with m∈NK∪{0} and k≤m≤r. Then jk(s) is a section of class Cm−k of Pk(E) over U (12.6.1); its image under D is denoted by DU(s), or D(s). It is a section of F∣U. Suppose that D is of class Ch, where h=m−k. Then DU(s) is of class Ch, and the mappings
DU:SEm(U)→SFh(U)
thus obtained enjoy the following properties:
(1) DU is K-linear. (2) For every s∈SEm(U) and every open set V of U, one has DV(s∣V)=DU(s)∣V. (3) For every s∈SEm(U) and every x∈U such that jxk(s)=0, one has DU(s)(x)=0. (3') For every coordinate system ξ=(ξ1,…,ξn) in U and every frame s=(s1,…,sd) of E over U (7.4.4), there exist sections ni,α(1≤i≤d,α∈Nn,∣α∣≤k) of F over U, of class Ch and such that, for every family (fi)1≤i≤d of elements of Cm(U), one has
(4) For every function f∈Cm(U) and every mapping θ of SUm(E) into SUh(F), let ad(f)θ denote the mapping
s↦f.θ(s)−θ(f.s)
of SUm(E) into SUh(F). Then, for every function f∈Cm(U), there exists a differential operator L on U, of type E→F, of order ≤k−1 and of class Ch such that
(ad(f)DU)(s)=LU(s)for every s∈SUm(E).
(5) One has
ad(f0)…ad(fk)DU=0for arbitrary f0,…,fk in Cm(U).
If one sets I={0,…,k} and fH=∏i∈Hfi for every subset H of I, the preceding relation is equivalent to:
H⊂I∑(−1)Card(H)fH.DU(fI−H.s)=0
for arbitrary f0,…,fk in Cm(U) and s in SEm(U).
14.1.5 ("Characterization of differential operators"). Let k,m and h=m−k be as in 14.1.4. For every open set U of X, let DU be a mapping of SEm(U) into SFh(U). Suppose that conditions 1, 2, 3 (resp. 1, 2, 3') of 14.1.4 are satisfied for every open subset U of X. Then there exists an element D of DXk,h(E,F), and only one, such that the DU are the corresponding mappings.(1) In what follows, D is identified with the family of the DU.
When m=∞ or ω, we have h=m and one may replace conditions (3) and (3') by either of conditions (4) and (5).
14.1.6 (“Scalar differential operators”). Suppose that E and F are equal to the trivial vector bundle KX. A differential operator of type E→F is then called a scalar differential operator or simply a differential operator; if it is of order ≤k, it is a section of the vector bundle L(Pk(X);KX)=Pk(X)∗, dual of the vector bundle Pk(X); using the isomorphism
i−1:Pk(X)∗→T(k)(X)(cf. 13.2.5),
one sees that a scalar differential operator of order ≤k is identified with a section of the vector bundle T(k)(X), i.e. with a field of point distributions of order≤k.
Take in particular for X an open subset of Kn, where n is an integer ≥0. The fields of point distributions
Δα:x↦Δxα(cf. 13.2.6)
are scalar differential operators of class Cω. For every h∈NK, the Δα for (∣α∣≤k) form a basis of the Ch(X)-module DXk,h(KX,KX).
14.1.7 (“Complex differential operators on a real manifold”). Suppose K=R and that the vector bundles E and F are endowed with complex structures (8.8.1); let k be an integer such that 0≤k≤r. The vector bundle Pk(E) is then endowed with a complex structure (12.6.3); the vector bundle Lc(Pk(E);F) (7.8.5) of complex linear mappings from Pk(E) into F is a vector subbundle of L(Pk(E);F)=Dk(E,F); we denote it by Dck(E,F). A section of Dck(E,F) is called a complex differential operator, of type E→F and of order ≤k. If h∈NK and h≤r−k, and if U is an open subset of X, we denote likewise by DUk,h(E,F)c the space of sections of class Ch of Dck(E,F) over U; it is a C-vector space. The other definitions and results of this paragraph extend analogously to complex differential operators; we leave their formulation to the reader.
14.1.8 (« Composition »). Let G be a vector bundle of class Cr and with base X. Let k′ and k′′ be positive integers with sum k≤r, and let D′ (resp. D′′) be a differential operator on X, of type E→F (resp. of type F→G), of order ≤k′ (resp. ≤k′′). Suppose D′:Pk′(E)→F is of class Ch′, with h′∈NK∪{0} and k′′≤h′≤r−k′; the mapping
D∗′=Pk′′(D′):Pk′′(Pk′(E))→Pk′′(F)
is of class Ch′−k′′ (12.6.3). Let β be the canonical homomorphism from Pk(E) into Pk′′(Pk′(E)), cf. 12.6.5. By composing the mappings
Pk(E)βPk′′(Pk′(E))D∗′Pk′′(F)D′′G,
one obtains a differential operator of type E→G and of order ≤k. This operator is called the composite of D′′ and D′; it is denoted by D′′∘D′.
1 It is in fact sufficient that condition (3') be satisfied for a family (ξλ) of systems of coordinates and a family of frames (sλ) whose domains cover X.
If U is an open subset of X, and if s ∈ SEk(U), one has
(D′′∘D′)(s)=D′′(D′(s))
(where the two members are sections of G over U); this justifies the terminology and notation adopted.
Suppose now that D'' is of class Ch′′, with h′′∈NK∪{0} and h′′≤r−k′′. Then D′′∘D′ is of class Ch, with h=inf(h′−k′′,h′′).
Suppose that one has E=F=G, and that k′≤h′′, in which case D′∘D′′ is defined. Then D′∘D′′−D′′∘D′ is a differential operator of type E→E and of order ≤k−1; it is denoted by [D′,D′′].
14.1.9 (« Associativity »). With the notation and hypotheses of 14.1.8, let H be a vector bundle of class Cr and with base X, and let D′′′ be a differential operator on X, of type G→H and of order ≤k′′′, with k′+k′′+k′′′≤r. Suppose that D′ is of class Ch′, with h′∈NK∪{0} and k′′+k′′′≤h′≤r−k′ and that D′′ is of class Ch′′, with h′′∈NK∪{0} and k′′′≤h′′≤r−k′′. Then the composites
D′′′∘(D′′∘D′)and(D′′′∘D′′)∘D′
are defined and are equal.
14.1.10. Examples
a) Let U be an open subset of X, let D∈DUk,h(E,F) with h∈NK∪{0} and h≤r−k, and let f∈Ch+k(U) (resp. f∈Ch(U)). Multiplication by f in E (resp. F) is a differential operator on U of order ≤0, and of type E→E (resp. F→F), cf. 14.1.2. The composites D∘f and f∘D are defined and belong to DUk,h(E;F); we put
ad(f)D=f∘D−D∘f∈DUk−1,h(E,F),
cf. 14.1.4, (4). Thus DUk,h(E,F) is endowed with a Ch+k(U)-module structure on the right (resp. a Ch(U)-module structure on the left). The left module structure coincides with that of 14.1.1.
b) Suppose r=∞ or r=ω. The differential operators of type E→E and of class Cr form an associative K-algebra with unit element.
c) Take for X an open subset of Kn. The scalar differential operators Δα (14.1.6) satisfy the formulas
Δα∘Δβ=((α,β))Δα+β.
They commute with one another. Denote by Di the operator Δεi (« i-th partial derivative »). If K is of characteristic 0, we have
Δα=α!1D1α1…Dnαn.
If K is of characteristic p=0, we have Dip=0 for every i.
14.1.11 (« Multidifferential operators »). Let E1,…,En be vector bundles of class Cr and with base X, and let k,k1,…,kn be integers belonging to [0,r]. Put
L(k1,…,kn)=L(Pk1(E1)⊗⋯⊗Pkn(En);F).
If ki≤k for all i, the projections rki,ki:Pk(Ei)→Pki(Ei) make it possible to identify the bundle L(k1,…,kn) with a sub-bundle of the bundle L(k)=L(k,…,k). There exists a smallest sub-bundle M(k) of L(k) containing all the sub-bundles L(k1,…,kn), for k1+⋯+kn=k. A section H of M(k) is called an n-differential (or multidifferential) operator of type (E1,…,En)→F and of order ≤k. If si (1≤i≤n) is a section of class Cm (m∈NK,m≥k) of Ei over an open set U of X, one defines HU(s1,…,sn)=H(s1,…,sn) as the image under H of the section jk(s1)⊗⋯⊗jk(sn) of the bundle
Pk(E1)⊗⋯⊗Pk(En).
It is a section of F∣U. If H is of class Ch, this section is of class Cp with p=inf(m−k,h).
When E1,…,En and F are equal to KX, one says that H is a scalar n-differential (or multidifferential) operator.
Example
Let J be a finite set, and X an open set of KJ. Let H be a scalar n-differential operator of order ≤k on X. There exists a unique family of scalar functions
for every family (f1,…,fn) of functions of class Ck on X. In order that H be of class Ch, it is necessary and sufficient that the functions c(α(1),…,α(n)) be of class Ch.
14.2. Symbols
14.2.1. Let k be a positive integer ≤r. Consider the exact sequence
0→Pk(T(X);E)i′Pk(E)j′Pk−1(E)→0
of No. 12.6.8, where j=rk,k−1. Applying the vector bundle functor M↦L(M;F) to this sequence, we obtain the exact sequence
The homomorphism j′=L(j;IdF) is the canonical inclusion of Dk−1(E,F) in Dk(E,F); the homomorphism σk is, by definition, equal to L(i;IdF). If D is a differential operator of type E→F and of order ≤k, its image under σk is a section σk(D) of Sk(E,F) which is called the k-symbol (or simply the symbol) of D; we have σk(D)=0 if and only if D is of order ≤k−1. If D is of class Ch, the same is true of its symbol.
14.2.2 (« Identifications of the symbol bundle »). The bundle Sk(E,F)=L(Pk(T(X);E);F) can be written in various ways. First, if x∈X, an element of Pk(Tx(X);Ex) is identified (13.1.2) with an element of L(TSk(Tx(X));Ex). We thus obtain identifications of vector bundles
Pk(T(X);E)=L(TSk(T(X));E)=(TSk(T(X)))∗⊗E
and, applying the vector bundle functor M↦L(M;F)=M∗⊗F, we obtain:
Sk(E,F)=TSk(T(X))⊗E∗⊗F=TSk(T(X))⊗L(E;F).
The preceding expression can be transformed further. If M and N are two vector bundles, let Symk(M;N) denote the subbundle of L(M,…,M;N) formed by symmetric k-linear mappings. We have a canonical isomorphism
Symk(M;N)→Symk(M;KX)⊗N;
on the other hand, the duality between ⊗kM and ⊗kM∗ identifies Symk(M;KX) with TSk(M∗). We thus obtain an identification
Symk(M;N)=TSk(M∗)⊗N.
Applying this formula to M=T(X)∗ and N=L(E;F), we have:
Sk(E,F)=TSk(T(X))⊗L(E;F)=Symk(T(X)∗;L(E;F)).
Suppose K is of characteristic 0 or >k. If u is a symmetric k-linear mapping, let u~ denote the polynomial mapping
x↦k!1u(x,…,x)
corresponding to it. The mapping u↦u~ defines an isomorphism
Sk(E,F)=Symk(T(X)∗,L(E;F))→Pk(T(X)∗;L(E;F)).
Thus, every element σ of Sk(E,F)x, with x∈X, may be identified with a homogeneous polynomial mapping σ~ of degree k from Tx(X)∗ into L(Ex;Fx).
14.2.3 (“Calculation of the symbol of a differential operator”). Let x∈X and let D be a differential operator of type E→F, of order ≤k, defined in an open neighbourhood of x and of class Cr−k. We shall make explicit the value σk(D)(x) of the symbol of D at x. Consider it as an element of Symk(Tx(X)∗;L(Ex;Fx)) (cf. 14.2.2). Let ω1,…,ωk be covectors at x and choose functions f1,…,fk of class Cr in an open neighbourhood U of x such that dxfi=ωi for i=1,…,k. Put
D′=(−1)kad(fk)…ad(f1)D(cf. 14.1.10,a).
The operator D′ is of order ≤0; it is a section of L(E;F)∣U. The value of D′ at x is an element λ(ω1,…,ωk) of L(Ex;Fx) which depends only on (ω1,…,ωk). The resulting mapping λ is symmetric. It is equal to the symbol σk(D)(x) of D at x.
Suppose that K=R or C, and let us make explicit σk(D)(x) considered as a homogeneous polynomial mapping of degree k from Tx(X)∗ into L(Ex;Fx) (cf. 14.2.2). Let ω∈Tx(X)∗ and v∈Ex; choose a function f of class Cr in an open neighbourhood U of x such that dxf=ω, and a section s of class Cr of E in U such that s(x)=v. There exists a family (φ0,φ1,…,φk) of sections of F in U, and only one, such that
e−tfD(etfs)=j=0∑ktjφjfor all t∈K.
Then φk(x) does not depend on the choice of f and s such that dxf=ω,v(x)=s; the mapping v↦φk(x) is an element λ(ω) of L(Ex;Fx), and λ is a homogeneous polynomial mapping of degree k from Tx(X)∗ into L(Ex;Fx). This mapping is equal to the symbol σk(D)(x) of D at x.
14.2.4 ("The scalar case"). Take E=F=KX, in which case one identifies Dk(E,F) with the bundle T(k)(X) of point distributions of order ≤k, cf. 14.1.6. One has
Sk(E,F)=TSk(T(X))
and the symbol
σk:T(k)(X)→TSk(T(X))
is none other than the composite
T(k)(X)→T(k)(X)/T(k−1)(X)ikTSk(T(X)),
where ik is the isomorphism defined in 13.3.2.
Take for example X to be an open subset of Kn, and let {e1,…,en} be the canonical basis of Kn (identified with the tangent spaces Tx(X)). If α∈Nn is such that ∣α∣≤k, the symbol of the operator Δα is given by the formulas (cf. 13.3.3):
cf. 14.2.2, one deduces from this, by tensor product, a pairing (*)
Sk′′(F,G)×Sk′(E,F)→Sk(E,G).
One then has the formula
σk(D′′∘D′)=σk′′(D′′)⋅σk′(D′),
where the product occurring in the right-hand term is defined by the pairing (*) above.
Now suppose K of characteristic zero, and let x∈X. Denote by σ (resp. σ′,σ′′) the k-symbol (resp. the k′-symbol, the k′′-symbol) of D′′∘D′ (resp. D′,D′′) at x. For every ω∈Tx(X)∗, one then has (with the notations at the end of 14.2.2):
In this No., it is assumed that X is purely of dimension n.
14.3. Transposition
14.3.1. Let Ω=det(T(X)∗), cf. 7.9.9; it is a vector bundle of rank 1 at each point, with base X, and of class Cr−1. One has
Ω=∧nT(X)∗=Altn(T(X);KX).
Let M be a vector bundle with base X. Put
M~=L(M;Ω)=M∗⊗Ω.
[^1]
If M is of class Cn, with h∈NK∪{0} and h≤r−1, the same is true of M~. The canonical mapping of M into M~=L(L(M;Ω);Ω) is an isomorphism; it is used to identify M with M~.
14.3.2 ("Definition of the transpose"). Let k be a positive integer ≤r−1, and let D be a differential operator on X, of type E→F, of order ≤k, and of class Cn, with h∈NK∪{0} and k≤h≤r−k. Then there exists a differential operator tD on X, of type Fˉ→Eˉ and of order ≤k, having the following property:
Let ξ=(ξ1,…,ξn) be a system of coordinates in an open set U of X and let s=(si)1≤i≤e (resp. t=(tj)1≤j≤f) be a frame of E (resp. of F) on U. Let (si∗) (resp. (tj∗)) be the frame of E∗ (resp. of F∗) such that ⟨si,sj∗⟩=δij (resp. ⟨ti,tj∗⟩=δij), and let (s~i) (resp. (t~j)) be the frame of Eˉ (resp. Fˉ) on U obtained by forming the tensor product of (si∗) (resp. of (tj∗)) with the frame ω=dξ1∧⋯∧dξn of Ω. Let cαij (1≤i≤e, 1≤j≤f, ∣α∣≤k) be the functions of class Cn on U such that one has
D(i∑fisi)=i,j,α∑cαijΔξα(fi)⋅tj
for arbitrary functions fi in Ch+k(U), cf. 14.1.4, (3'). One then has
tD(j∑gjt~j)=i,j,α∑(−1)∣α∣Δξα(cαijgj)s~i
for arbitrary functions gj in Ch+k(U).
80 DIFFERENTIAL AND ANALYTIC MANIFOLDS § 14
The preceding property (which must be verified whatever ξ, s and t may be) determines tD uniquely.[^2] The operator tD is called the transpose of D; it is of class Ch−k. If h≥2k, the transpose of tD is defined and equal to D (taking into account the identifications of E and F with E and F respectively).
The mapping D↦tD is K-linear. If k=0, i.e. if D is a morphism from E into F, tD is the morphism L(D;IdΩ) from F into E.
The symbol of tD is deduced from the symbol of D by means of the isomorphism
TSk(T(X))⊗L(E;F)⟶TSk(T(X))⊗L(F;E)
which is the tensor product of the automorphism (−1)k of TSk(T(X)) and the isomorphism u↦L(u;IdΩ) of L(E;F) onto L(F;E).
14.3.3 (“Transpose of a Composite”). With the notation and hypotheses of 14.1.8, suppose that D′ is of class Ch′ and D′′ of class Ch′′, with h′,h′′ in NK∪{0}, h′≥k′+2k′′ and h′′≥k′′+2k′.
Then the transposes of D′, D′′ and D′′∘D′ are defined, as is the composite of tD′ and tD′′, and one has
t(D′′∘D′)=tD′∘tD′′.
In particular, with the hypotheses and notation of 14.3.2, and f a function of class Ch′ on X with 2k≤h′, one has t(D∘f)=f∘tD.
Example
Let X be an open subset of Kn, let ξ1,…,ξn denote the coordinate functions on X, and identify Ω with KX by means of the frame ω=dξ1∧⋯∧dξn. If E=F=KX, one has
E=F=L(KX;Ω)=L(KX;KX)=KX,
and the operation D↦tD transforms scalar differential operators into scalar differential operators. One has, for example,
t(Δα)=(−1)∣α∣Δαfor every α∈Nn.
If r is infinite, transposition is an anti-automorphism of the algebra DKX∞(KX,KX).
14.3.4. Let k be a positive integer such that K is of characteristic 0 or >k. Let Ω1 be the vector bundle ⋀n−1T(X)∗=L(T(X);KX). The notation and hypotheses being those of 14.3.2, let moreover D′ be a differential operator on X, from F→E, of order ≤k; a Green operator for (D,D′) is defined to be any multidifferential operator G (cf. 14.1.11) of type (E,F)→Ω1, of order ≤k−1[^3] and of class Ch (with h≥1), such that one has the identity
(1)
⟨D(u),v⟩−⟨u,D′(v)⟩=d(G(u,v))
for every open set U of X and every elements u∈SEk(U), v∈SFk(U). Let us specify that, in this formula, d denotes exterior differentiation (8.3.5) and ⟨D(u),v⟩ (resp.
⟨u,D′(v)⟩) denotes the section of Ω on U obtained from the pair (D(u),v) (resp. from the pair (u,D′(v))) by the canonical pairing of F×F~ (resp. of E×E~) into Ω. The existence of G implies D′=tD; one also says that G is a Green operator for D.
14.3.5. Let D be a differential operator on X, of type E→F, of order ≤k, of class Ch with h∈NK and k≤h≤r−k. There exists a Green operator of class Ch−k+1 for D in each of the following cases:
a) K=R, r=∞ and X is paracompact. b) K is of characteristic 0 or >k, X is isomorphic to an open set of a space Kn, and the bundles E and F are isomorphic to trivial bundles. c) The operator D is of order ≤1 (cf. 14.3.7).
14.3.6. The hypotheses and notations of 14.3.3 are retained. If G′ (resp. G′′) is a Green operator for D′ (resp. D′′), there exists one and only one Green operator H for D′′∘D′ such that one has
H(u,v)=G′′(D′(u),v)+G′(u,tD′′(v))
for all open sets U of X and elements u∈SEk(U) and v∈SG~k(U).
14.3.7. One has S1(E,F)=T(X)⊗E∗⊗F (cf. 14.2.2). On the other hand, the right interior product (ξ,ω)↦i(ξ)ω (cf. A, III, p. 158) defines an isomorphism of T(X)⊗Ω onto Ω1; by tensor product with the dual Ω∗ of Ω, one thus obtains an isomorphism of T(X) onto Ω∗⊗Ω1, whence identifications of vector bundles
This being set, let D be a differential operator on X, of type E→F, of order ≤1, of class Ch with h∈NK∪{0} and h≤r−1. There exists a unique Green operator for D; it is of order 0; it is a morphism of class Ch from E⊗F~ into Ω1, and it is deduced from the symbol of D by the preceding identification.
14.3.8. Suppose that one has K=R, that X is separated and oriented (10.2.4) and let A be a closed piece (11.1.2) of X. Equip A with the orientation induced by that of X, and ∂A with the corresponding orientation (11.2.1). Let k be an integer such that 2k≤r, D a differential operator on X, of type E→F, of order ≤k and of class Ck. Let G be a Green operator for D. Let U be an open set of X, u∈SEk(U), v∈SF~k(U) and suppose that the supports of u and v meet A in a compact set. One then has
∫A⟨D(u),v⟩−∫A⟨u,tD(v)⟩=∫∂AG(u,v)
(Green's formula). In particular, if ∂A=∅, one has
∫A⟨D(u),v⟩=∫A⟨u,tD(v)⟩.
14.4. Examples
Suppose X purely of finite dimension n. Suppose it is also endowed with a structure of manifold of class C^{r+1} compatible with the given structure of class C^r.
14.4.1 (« Infinitesimal transformations »). Let τ be a vector functor in finite dimension for isomorphisms, of class C^r; suppose that, for every finite-dimensional vector space V, τ(V) is finite-dimensional. We denote by Eτ the vector bundle τ(T(X)); it is of class C^r.
Let ξ be a vector field of class C^r on X. There exists a differential operator D of type Eτ→Eτ and of order ≤1 such that
DU(s)=θξ.s
for every open set U of X and every s∈SEτ(U), cf. 8.4.3. Such a differential operator D is unique; we denote it by (θξ)τ or simply θξ. Its symbol is the section ξ⊗IdEτ of the vector bundle
S1(Eτ,Eτ)=T(X)⊗L(Eτ;Eτ).
If V is a finite-dimensional K-vector space, put
τ~(V)=L(τ(V);Altn(V;K))=τ(V)∗⊗∧nV∗,
and if u:V1→V2 is an isomorphism, define τ~(u):τ~(V1)→τ~(V2) by transport of structure. We thus obtain a vector functor in finite dimension τ~. The bundle Eτ~=τ~(T(X)) identifies in the obvious way with the bundle (Eτ)∗; the transpose of (θξ)τ is −(θξ)τ~ and formula (1) of no. 14.3.4 takes the form
⟨θξ.u,v⟩+⟨u,θξ.v⟩=d(i(ξ)(u.v)).
14.4.2 (« Exterior differential »). For every integer p≥0 put
Ωp=Altp(T(X);KX).
There exists a differential operator D of type Ωp→Ωp+1 and of order ≤1 such that
DU(ω)=dω
for every open set U of X and every ω∈SΩp(U). Such a differential operator D is unique; we denote it by d (or dp if we wish to specify the integer p).
Its symbol is the element of S1(Ωp,Ωp+1) obtained as follows: first of all we have
S1(Ωp,Ωp+1)=T(X)⊗L(Ωp;Ωp+1)
=L(Ω1;L(Ωp;Ωp+1))=L(Ω1,Ωp;Ωp+1),
where L(Ω1,Ωp;Ωp+1) denotes the bundle of bilinear mappings of Ω1×Ωp into Ωp+1. Now the exterior product (α,β)↦α∧β defines a canonical section of this latter bundle; this section is the symbol of dp.
To determine the transpose of dp, we first remark that the exterior product defines a pairing Ωp×Ωn−p→Ωn=Ω which permits identifying (Ωp)∗ with Ωn−p. We identify (Ωp+1)∗ with Ωn−p−1 in the same way. With these conventions, the transpose of dp is (−1)p+1dn−p−1 and the corresponding Green operator, of type
(Ωp,Ωn−p−1)⟶Ω1=Ωn−1, is simply the exterior product (on each fibre).
14.4.3 (« Laplacian »).
Let K=R, r=∞, X=Rn, E=F=KX; let ξi(1≤i≤n) denote the coordinate functions on X; put Di=∂/∂ξi, and L=∑i=1nDi2. The operator L is a scalar differential operator of order ≤2. If we identify Ω with KX by means of the frame ω=dξ1∧⋯∧dξn, we have E~=F~=KX (cf. 14.3.3, example) and L~=L.
For 1≤i≤n, put ωi=(−1)i−1dξ1∧⋯∧dξi−1∧dξi+1∧⋯∧dξn. The ωi form a frame of Ω1=Ωn−1. If f∈C1(X), let grad(f) denote the section ∑i=1nDi(f)ωi of Ω1. There exists a Green operator G for L such that
G(u,v)=v.grad(u)−u.grad(v)
for every open set U of X and u,v in C1(U).
In particular, let A be a compact piece of Rn. Orient Rn, endow A with the orientation induced by that of Rn, and ∂A with the corresponding orientation (11.2.1). We have
∫A(Lu).vω−∫Au.(Lv)ω=∫∂A(v.grad(u)−u.grad(v))
for u,v in C1(X).
[^1]: Care should be taken not to confuse M~ with the covering of M defined in 10.2.4 and denoted in the same way. [^2]: In fact, it is sufficient to verify the property in question for triples (ξλ,sλ,tλ) such that the domains of the ξλ cover X. [^3]: When k=0, it is agreed that this means that G=0.