# The Heat Asymptotics on Filtered Manifolds

- 65 Downloads

## Abstract

The short-time heat kernel expansion of elliptic operators provides a link between local and global features of classical geometries. For many geometric structures related to (non-)involutive distributions, the natural differential operators tend to be Rockland, hence hypoelliptic. In this paper, we establish a universal heat kernel expansion for formally self-adjoint non-negative Rockland differential operators on general closed filtered manifolds. The main ingredient is the analysis of parametrices in a recently constructed calculus adapted to these geometric structures. The heat expansion implies that the new calculus, a more general version of the Heisenberg calculus, also has a non-commutative residue. Many of the well-known implications of the heat expansion such as, the structure of the complex powers, the heat trace asymptotics, the continuation of the zeta function, as well as Weyl’s law for the eigenvalue asymptotics, can be adapted to this calculus. Other consequences include a McKean–Singer type formula for the index of Rockland differential operators. We illustrate some of these results by providing a more explicit description of Weyl’s law for Rumin–Seshadri operators associated with curved BGG sequences over 5-manifolds equipped with a rank-two distribution of Cartan type.

## Keywords

Filtered manifold Hypoelliptic operator Heat kernel expansion Zeta function Non-commutative residue Generic rank-two distribution in dimension## Mathematics Subject Classification

58J35 (58A30, 58J20, 58J40, 58J42)## 1 Introduction

Many geometric structures related to (non-)involutive distributions can be described in terms of an underlying filtered manifold. These include contact structures, Engel manifolds, and all regular parabolic geometries. Filtered analogues of classical (elliptic) operators are usually hypoelliptic [21]. It is an old dream to link local and global aspects of filtered geometry with the spectrum and index of hypoelliptic operators by studying the short-time asymptotics of heat kernels, similarly to the elliptic case. Active research in this direction started in the 1970s, see [22, 25, 46], and was quite popular in 1980s, see [3, 26]. The main accomplishment had been the development of the Heisenberg calculus [3, 43, 52] and its application to operators in contact and CR geometries. On graded nilpotent Lie groups, the representation theory allowed a harmonic analysis perspective of hypoellipticity [16, 34, 35].

The progress on these problems has been revived by the advent of new tangent groupoid techniques to study analysis on these manifolds [15, 54, 55, 56, 57]. When the classical (elliptic) pseudodifferential calculus was described in [23] in a coordinate-free way using the tangent groupoid, it facilitated the construction of a pseudodifferential calculus on general filtered manifolds [56, 57]. This calculus has helped resolve two main hitherto unsurmountable challenges in analysis. It first allowed an explicit handle on the parametrix of a hypoelliptic operator similar to the one familiar from the elliptic case or nilpotent Lie groups as in [16]. The universality of this calculus and the existence of the parametrix [21] provide a clean way to obtain a general short-time heat kernel expansion for a large class of differential operators. Remarkably, this heat expansion has the same structure as the one for elliptic operators, bringing back the old expectation that the analysis should relate local geometric properties to global invariants.

The differential operators we consider satisfy a pointwise Rockland condition, a condition on their non-commutative principal symbols that guarantees that these operators are hypoelliptic. We also present here several consequences of the heat kernel asymptotics, which are well known for elliptic operators, but are not known for Rockland operators in this generality. These include the explicit structure of complex powers of the operators, Weyl’s asymptotic formula for the growth of eigenvalues, the McKean–Singer index formula, the description of a non-commutative residue for Heisenberg calculus, and the construction of a K-homology class associated to Rockland operators.

Let us look at the manifolds, the operators, and their heat kernel expansions in more detail.

### 1.1 Filtered Manifolds

Filtered manifolds provide a very general setup to study geometry [39, 40]. These geometries include foliations, contact manifolds, Engel structures on 4-manifolds, graded nilpotent Lie groups, and all regular parabolic geometries. The equiregular Carnot–Carathéodory spaces considered in [28] are also filtered manifolds.

A filtered manifold *M* has a naturally associated non-commutative tangent bundle, that is, a simply connected nilpotent Lie group \(\mathcal {T}_xM\) attached to each point *x* in *M*. The harmonic analysis of this non-commutative tangent space \(\mathcal {T}_xM\) is related to the analysis of differential operators on *M*. An effective way to exhibit this relationship uses the so-called Heisenberg tangent groupoid [15, 57]. The pseudodifferential calculus on filtered manifolds [56] was constructed using the Heisenberg tangent groupoid. It was further studied in [21] where we show that a pointwise Rockland condition implies hypoellipticity. We remark that the idea of Heisenberg calculus is based on the work of Debourd–Skandalis [23], who studied the standard pseudodifferential calculus using Connes’ tangent groupoid construction.

*M*using the Heisenberg tangent groupoid, \(\mathbb {T}M\). This is a groupoid that provides a deformation of the pair groupoid \(M\times M\) into the non-commutative tangent bundle \({\mathcal {T}}M\), which is obtained by giving

There is an equivalent local description of the kernels in suitably chosen coordinates. In this paper, we will mostly work with the local version; however, the global description allows a neater formulation of the concept of holomorphic families described below. In the local description, the kernels are described in a tubular neighborhood of the diagonal and the germs of diffeomorphism of these tubular neighborhood to the normal bundle is determined by the Euler-like vector field provided by the filtration [30]. Here, we will stick to the original Heisenberg tangent groupoid description for global invariance of the calculus.

### 1.2 Rockland Differential Operators

Let us briefly recall the kind of operators studied in this article. These are hypoelliptic operators on filtered manifolds which are elliptic in the Heisenberg calculus described above.

Let *M* be a closed filtered manifold and suppose *E* is a complex vector bundle over *M*. Let *A* be a differential operator acting on sections of *E*. In coordinates adapted to the filtration on *M*, we can assign a (co)symbol to *A* by freezing the coefficients at a point \(x\in M\), thereby obtaining a left invariant differential operator \(\sigma _x(A)\) on the osculating group \(\mathcal {T}_xM\). We require that each \(\sigma _x(A)\) satisfies the Rockland condition and furthermore *A* is symmetric and non-negative on \(L^2(E)\). Examples of such operators include the sub-Laplacian for many classes of sub-Riemannian geometries, the Rumin–Seshadri operators associated with (ungraded) BGG operators on parabolic geometries, and the square of the Connes–Moscovici transverse signature operator [18]. On trivially filtered manifolds, these operators are elliptic in the usual sense, and on contact or Heisenberg manifolds they are elliptic in the Heisenberg calculus of [3, 43].

### 1.3 Heat Kernel Asymptotics

In this paper, we study the kernel of \(e^{-tA}\) for \(t>0\) and the complex powers \(A^{-z}\) where \(z\in \mathbb {C}\) for a differential operator *A* described in the paragraph above. Rather than using the standard approach by Seeley [50], we follow Beals–Greiner–Stanton [4] who use a Volterra–Heisenberg calculus on \(M\times \mathbb {R}\) to describe the heat kernel on CR manifolds, see also [43, Chapter 5]. In our general setting, however, no new calculus has to be developed. Instead, we may regard \(M\times \mathbb {R}\) as a filtered manifold such that the heat operator \(A+\frac{\partial }{\partial t}\) becomes Rockland on \(M\times \mathbb {R}\). The general Rockland theorem established in [21, Theorem 3.13] thus yields a parametrix for the heat operator \(A+\frac{\partial }{\partial t}\) in the pseudodifferential operator calculus developed by van Erp and Yuncken in [56]. The asymptotic expansion of its Schwartz kernel along the diagonal immediately yields the heat kernel asymptotics for *A*.

Let us emphasize that this approach to the heat kernel asymptotics seems to only work for Rockland differential operators. For more general pseudodifferential operators *P* on the filtered manifold *M*, the corresponding heat operator \(\frac{\partial }{\partial t}+P\) on \(M\times \mathbb {R}\) is no longer pseudolocal in general and hence not in the Heisenberg pseudodifferential calculus. This implies that we can only investigate ungraded Rockland sequences in the sense of [21] and, in particular, the results can be applied only to ungraded BGG sequences. For example, the BGG sequences for Cartan geometries of generic rank-2 distributions in dimension five are always ungraded, and we will study them in greater detail here.

We mention here two prominent features of the heat asymptotics of Rockland differential operators.

As is well known in the classical (trivially filtered) case, certain terms in the heat kernel asymptotics of elliptic pseudodifferential operators vanish for differential operators. We show that this continues to hold for Rockland differential operators on filtered manifolds. Their expansion takes a form similar to the expansion of classical elliptic differential operators. In particular, there are no log terms, and every other polynomial term vanishes. To see this, we introduce the class of projective operators, see Definition 2, which have the desired asymptotic expansion and contain the parametrices of Rockland differential operators.

Another key feature of the heat asymptotics of a non-negative elliptic operator is the positivity of its leading term. This term has geometric significance: for instance, the leading term in the heat trace expansion of a Laplace operator encodes the volume of the corresponding Riemannian metric. Let us emphasize that this positivity of the leading term remains true for non-negative Rockland differential operators. In particular, this permits to derive a Weyl law for the eigenvalue asymptotics using a Tauberian theorem.

### 1.4 Applications

- 1.
The structure of the complex powers \(A^{-z}\) follows via Mellin transform from the heat asymptotics in the usual way. Analogous to the celebrated result of Seeley [50], the complex powers are pseudodifferential operators in the calculus of van Erp and Yuncken [56]. We interpret the notion of a holomorphic family of pseudodifferential operators [43], in terms of an essential homogeneity criterion on the Heisenberg tangent groupoid. As can be expected, \(A^{-z}\) is a holomorphic family for every non-negative Rockland differential operator

*A*. - 2.
Taking the trace of a holomorphic family, whenever it is defined, produces a holomorphic function. We will show that the various spectral zeta functions have meromorphic continuation with simple poles. In particular, the algebra of integer-order pseudodifferential operators on a filtered manifold admits a non-commutative residue.

- 3.
Weyl’s law for the eigenvalue asymptotics also follows from the positivity of the constant term in the heat expansion, or from the position and residue at the first pole of the spectral zeta function as usual. We will work out Weyl’s law more explicitly for Cartan geometries in dimension five using suitable geometric choices. Surprisingly, the constant in Weyl’s law is universal, depending only on the irreducible representation defining the BGG sequence.

- 4.
Rockland operators are hypoelliptic and hence on a closed manifold they are Fredholm. In particular, we note two consequences towards the study of their index, namely, the description of the K-homology class associated to them, and a generalization of the McKean–Singer formula.

### 1.5 Structure of the Paper

The remaining part of the paper is organized as follows. In Sect. 2 we formulate our main results: Theorem 1 on the heat kernel asymptotics and Theorem 2 on the structure of complex powers. Furthermore, we derive several immediate consequences: Corollary 1 on the heat trace expansion, Corollary 2 on the zeta function, Corollary 3 on Weyl’s law, Corollary 4 on the McKean–Singer formula, as well as Corollary 5 on the K-homology class. In Sect. 3, we briefly recall some background for the calculus of pseudodifferential operators on filtered manifolds and introduce the class of projective pseudodifferential operators of integral Heisenberg order, see Definition 2. In Sects. 4 and 5, we present proofs of Theorems 1 and 2 , respectively. In Sect. 6 we discuss holomorphic families of Heisenberg pseudodifferential operators. In Sect. 7, we construct a non-commutative residue, see Corollary 6. In Sect. 8 we will work out Weyl’s law more explicitly for Rumin–Seshadri operators, see Corollary 7, and specialize further to BGG operators on 5-manifolds equipped with a rank-two distribution of Cartan type, see Corollary 8.

We would like to thank an anonymous referee for helpful remarks and another anonymous referee for thorough reading and many useful comments.

## 2 Statement of the Main Results

*filtered manifold*[21, 39, 40, 41] is a smooth manifold

*M*together with a filtration of the tangent bundle

*TM*by smooth subbundles,

*Levi bracket.*This turns the associated graded vector bundle \(\mathfrak tM:=\bigoplus _p\mathfrak t^pM\) into a bundle of graded nilpotent Lie algebras called the

*bundle of osculating algebras.*The Lie algebra structure on the fiber \(\mathfrak t_xM=\bigoplus _p\mathfrak t^p_xM\) depends smoothly on the base point \(x\in M\), but is not assumed to be locally trivial. In particular, the Lie algebras \(\mathfrak t_xM\) might be non-isomorphic for different \(x\in M\).

Using negative degrees, we are following a convention prevalent in parabolic geometry, see [11, 40] for instance. The other convention, where everything is concentrated in positive degrees, is the one that has been adopted in [56, 57].

*homogeneous dimension*of

*M*by

*Heisenberg filtration*on differential operators. More explicitly, if

*E*and

*F*are two vector bundles over

*M*, then a differential operator \(\Gamma ^\infty (E)\rightarrow \Gamma ^\infty (F)\) is said to have

*Heisenberg order*at most

*r*if, locally, it can be written as a finite sum of operators of the form \(\Phi \nabla _{X_k}\cdots \nabla _{X_1}\) where \(\Phi \in \Gamma ^\infty (\hom (E,F))\), \(\nabla \) is a linear connection on

*E*, and \(X_j\in \Gamma ^\infty (T^{p_j}M)\) such that \(-(p_1+\cdots +p_k)\le r\). Denoting the space of differential operators of Heisenberg order at most

*r*by \(\mathcal {D}\mathcal {O}^r(E,F)\), we obtain a filtration

*G*is another vector bundle, then \(BA\in \mathcal {D}\mathcal {O}^{s+r}(E,G)\) and \(A^*\in \mathcal {D}\mathcal {O}^r(F,E)\). As usual, the formal adjoint is with respect to standard \(L^2\) inner products of the form

*M*and

*h*is a fiberwise Hermitian inner product on

*E*. The formal adjoint can be characterized by the equation \(\langle \!\langle A^*\phi ,\psi \rangle \!\rangle _E=\langle \!\langle \phi ,A\psi \rangle \!\rangle _F\) for all \(\psi \in \Gamma ^\infty _c(E)\) and \(\phi \in \Gamma ^\infty _c(F)\).

*Heisenberg principal symbol*, \(\sigma ^r_x(A)\in \mathcal {U}_{-r}(\mathfrak t_xM)\otimes \hom (E_x,F_x)\) at every point \(x\in M\). Here

For \(x\in M\) we let \(\mathcal {T}_xM\) denote the *osculating group* at *x*, that is, the simply connected nilpotent Lie group with Lie algebra \(\mathfrak t_xM\). The individual osculating groups can be put together to form a bundle of nilpotent Lie groups \({\mathcal {T}}M\) over *M* such that the fiberwise exponential map, \(\exp :\mathfrak tM\rightarrow {\mathcal {T}}M\), becomes a diffeomorphism of locally trivial bundles over *M*. The scaling automorphisms \(\dot{\delta }_{\lambda ,x}\in {{\,\mathrm{Aut}\,}}(\mathfrak t_xM)\) integrate to Lie group automorphisms \(\delta _{\lambda ,x}\in {{\,\mathrm{Aut}\,}}(\mathcal {T}_xM)\) which combine to form bundle diffeomorphisms \(\delta _\lambda :{\mathcal {T}}M\rightarrow {\mathcal {T}}M\) such that \(\delta _\lambda \circ \exp =\exp \circ \dot{\delta }_\lambda \). The Heisenberg principal symbol \(\sigma ^r_x(A)\) of an operator \(A\in \mathcal {D}\mathcal {O}^r(E,F)\) can be regarded as a left invariant differential operator on \(\mathcal {T}_xM\) which is homogeneous of degree *r*. More precisely, regarding \(\sigma ^r_x(A):C^\infty (\mathcal {T}_xM,E_x)\rightarrow C^\infty (\mathcal {T}_xM,F_x)\), we have \(\sigma ^r_x(A)\circ l_g^*=l_g^*\circ \sigma ^r_x(A)\) and \(\sigma ^r_x(A)\circ \delta _{\lambda ,x}^*=\lambda ^r\delta _{\lambda ,x}^*\circ \sigma ^r_x(A)\) for all \(x\in M\), \(\lambda \ne 0\) and \(g\in \mathcal {T}_xM\). Here \(l_g^*\) denotes pull back along the left translation, \(l_g:\mathcal {T}_xM\rightarrow \mathcal {T}_xM\), \(l_g(h):=gh\), and \(\delta _{\lambda ,x}^*\) denotes pull back along \(\delta _{\lambda ,x}:\mathcal {T}_xM\rightarrow \mathcal {T}_xM\).

A differential operator \(A\in \mathcal {D}\mathcal {O}^r(E,F)\) is said to satisfy the *Rockland condition* [44] at \(x\in M\) if, for every non-trivial irreducible unitary representation of the osculating group, \(\pi :\mathcal {T}_xM\rightarrow U(\mathcal {H})\), on a Hilbert space \(\mathcal {H}\), the linear operator \(\pi (\sigma ^r_x(A)):\mathcal {H}_\infty \otimes E_x\rightarrow \mathcal {H}_\infty \otimes F_x\) is injective. Here \(\mathcal {H}_\infty \) denotes the subspace of smooth vectors in \(\mathcal {H}\). An operator is called a *Rockland operator* if it satisfies the Rockland condition at every point \(x\in M\). We refer to [21, Sect. 2.3] for more details and references.

According to [21, Theorem 3.13] every Rockland operator \(A\in \mathcal {D}\mathcal {O}^r(E,F)\) admits a properly supported left parametrix \(B\in \Psi ^{-r}_\text {prop}(F,E)\) such that \(BA-{{\,\mathrm{id}\,}}\) is a smoothing operator. Here \(\Psi ^{-r}\) denotes the class of *pseudodifferential operators of Heisenberg order*\(-r\) which has recently been introduced by van Erp and Yuncken, see [56] and Sect. 3. These are operators whose Schwartz kernels have a wave front set which is contained in the conormal of the diagonal. In particular, their kernels are smooth away from the diagonal. Moreover, their kernels admit an asymptotic expansion along the diagonal with respect to a tubular neighborhood which is adapted to the filtration. In particular, the left parametrix *B* induces a continuous operator \(\Gamma ^\infty (F)\rightarrow \Gamma ^\infty (E)\) which extends continuously to a pseudolocal operator on distributional sections, \(\Gamma ^{-\infty }(F)\rightarrow \Gamma ^{-\infty }(E)\). Consequently, Rockland operators are *hypoelliptic*, that is, if \(\psi \) is a distributional section of *E* such that \(A\psi \) is smooth on an open subset *U* of *M*, then \(\psi \) was smooth on *U*, cf. [21, Corollary 2.10].

*M*to be closed. We consider a Rockland differential operator \(A\in \mathcal {D}\mathcal {O}^r(E)\) of Heisenberg order \(r\ge 1\) which is formally self-adjoint with respect to an \(L^2\) inner product of the form (4), that is, for all \(\psi _1,\psi _2\in \Gamma ^\infty (E)\) we have

### Lemma 1

With the above hypotheses, *A* is essentially self-adjoint with compact resolvent on \(L^2(E)\).

### Proof

Indeed, *A* is symmetric with dense domain \(\Gamma ^\infty (E)\) and thus closeable. By regularity, the domain of its adjoint coincides with the Heisenberg Sobolev space \(H^{r}(E)\), see [21, Corollary 3.24]. To see that this coincides with the domain of the closure, let \(\Lambda \in \Psi ^r(E)\) and \(\Lambda '\in \Psi ^{-r}(E)\) such that \(R=\Lambda '\Lambda -{{\,\mathrm{id}\,}}\) is a smoothing operator.^{1} Given \(\phi \in H^r(E)\), choose \(\phi _j\in \Gamma ^\infty (E)\) such that \(\phi _j\rightarrow \Lambda \phi \) in \(L^2(E)\). Since \(\Lambda '\) and \(A\Lambda '\) are both bounded on \(L^2(E)\), see [21, Proposition 3.9(a)], we also have \(\Lambda '\phi _j\rightarrow \Lambda '\Lambda \phi \) and \(A\Lambda '\phi _j\rightarrow A\Lambda '\Lambda \phi \) in \(L^2(E)\). Putting \(\psi _j:=\Lambda '\phi _j-R\phi \in \Gamma ^\infty (E)\), we obtain \(\psi _j\rightarrow \phi \) and \(A\psi _j\rightarrow A\phi \) in \(L^2(E)\), hence \(\phi \) is in the domain of the closure of *A*. The resolvent of *A* is compact, since \(A-z\) has a parametrix in \(\Psi ^{-r}(E)\) for every \(z\in \mathbb {C}\), and these operators are compact on \(L^2(E)\), see [21, Proposition 3.9(b)]. \(\square \)

*A*is non-negative, that is,

*strongly differentiable semi-group*\(e^{-tA}\) for \(t\ge 0\). More precisely, for each \(\psi \in L^2(E)\) the vector \(e^{-tA}\psi \) is contained in the domain of

*A*, and we have

*A*is non-negative, each \(e^{-tA}\) is a contraction on \(L^2(E)\). According to the Schwartz kernel theorem, it thus has a distributional kernel, \(k_t\in \Gamma ^{-\infty }(E\boxtimes E')\). More explicitly, we have

*M*and \(\langle -,-\rangle \) denotes the canonical pairing between \(\mathcal {D}(E)\) and \(\Gamma ^{-\infty }(E)=\mathcal {D}'(E)\). This pairing will also be denoted by \(\langle \phi ,\xi \rangle =\int _M\phi \,\xi =\int _{x\in M}\phi (x)\xi (x)\), where \(\phi \in \mathcal {D}(E)\) and \(\xi \in \Gamma ^{-\infty }(E)\). In particular, the right-hand side in (7) denotes the canonical pairing of \(\phi \boxtimes \psi \in \Gamma ^\infty (E'\boxtimes E)=\mathcal {D}(E\boxtimes E')\) with \(k_t\in \Gamma ^{-\infty }(E\boxtimes E')=\mathcal {D}'(E\boxtimes E')\).

One main aim of this paper is to establish the following heat kernel asymptotics, generalizing a result of Beals–Greiner–Stanton for CR manifolds [4, Theorems 5.6 and 4.5], see also [43, Theorem 5.1.26 and Proposition 5.1.15].

### Theorem 1

*E*be a vector bundle over a closed filtered manifold

*M*. Suppose \(A\in \mathcal {D}\mathcal {O}^r(E)\) is a Rockland differential operator of even

^{2}Heisenberg order \(r>0\) which is formally self-adjoint and non-negative with respect to an \(L^2\) inner product of the form (4), that is, \(\langle \!\langle A\psi _1,\psi _2\rangle \!\rangle =\langle \!\langle \psi _1,A\psi _2\rangle \!\rangle \) and \(\langle \!\langle A\psi ,\psi \rangle \!\rangle \ge 0\), for all \(\psi ,\psi _1,\psi _2\in \Gamma ^\infty (E)\). Then \(e^{-tA}\) is a smoothing operator for each \(t>0\), and the corresponding heat kernels \(k_t\in \Gamma ^\infty (E\boxtimes E')\) depend smoothly on \(t>0\). Moreover, as \(t\searrow 0\), we have an asymptotic expansion

*N*we have \(k_t(x,x)=\sum _{j=0}^{N-1}t^{(j-n)/r}q_j(x)+O(t^{(N-n)/r})\), uniformly in

*x*as \(t\searrow 0\). Moreover, \(q_j(x)=0\) for all odd

*j*, and \(q_0(x)>0\) in \({{\,\mathrm{end}\,}}(E_x)\otimes |\Lambda _{M,x}|\) for each \(x\in M\).

The terms \(q_j\in \Gamma ^\infty \bigl ({{\,\mathrm{end}\,}}(E)\otimes |\Lambda _M|\bigr )\) in Theorem 1 are (in principle) locally computable, they can be read off any parametrix for the heat operator \(A+\frac{\partial }{\partial t}\), see Remark 2 for more details. The leading term \(q_0(x)\) can also be obtained by evaluating the heat kernel of the Heisenberg principle symbol \(\sigma _x^r(A)\) at the point \((o_x,1)\in \mathcal {T}_xM\times (0,\infty )\), see (65). Here \(o_x\in \mathcal {T}_xM\) denotes the neutral element of the osculating group.

The spectral theorem also permits to construct complex powers \(A^z\) for every \(z\in \mathbb {C}\). These are unbounded operators on \(L^2(E)\) satisfying \(A^{z_1+z_2}=A^{z_1}A^{z_2}\) for all \(z_1,z_2\in \mathbb {C}\). The powers are defined such that \(A^z\) vanishes on \(\ker (A)\) and commutes with the orthogonal projection onto \(\ker (A)\). In particular, \(A^0\) is the orthogonal projection onto the orthogonal complement of \(\ker (A)\), and \(A^{-1}\) is the pseudoinverse of *A*. If \(z\in \mathbb {N}\), then \(A^z\) coincides with the ordinary power, i.e., the *z*-fold product of *A* with itself.

Another main goal of this paper is the following result about the structure of complex powers generalizing [43, Theorems 5.3.1 and 5.3.4].

### Theorem 2

Before turning to the proof of Theorems 1 and 2, we will now formulate several immediate corollaries. The next one generalizes [4, Theorem 5.6], see also [43, Proposition 6.1.1].

### Corollary 1

*N*we have \({{\,\mathrm{tr}\,}}(e^{-tA})=\sum _{j=0}^{N-1}t^{(j-n)/r}a_j+O(t^{(N-n)/r})\) as \(t\searrow 0\). Moreover, \(a_j=\int _M{{\,\mathrm{tr}\,}}_E(q_j)\) where \(q_j\in \Gamma ^\infty \bigl ({{\,\mathrm{end}\,}}(E)\otimes |\Lambda _M|\bigr )\) is as in Theorem 1, \(a_j=0\) for all odd

*j*, and \(a_0>0\).

### Proof

### Corollary 2

### Proof

*z*since the kernel \(k_{A^{-z}}(x,x)\), considered as a family in \(\Gamma ^\infty ({{\,\mathrm{end}\,}}(E)\otimes |\Lambda _M|)\), is holomorphic for \(\mathfrak {R}(z)>n/r\). Since \(k_{A^{-z}}(x,x)\) can be extended meromorphically to the entire complex plane, the same holds true for \(\zeta (z)\). The pole structure, residues, and special values follow immediately from the corresponding statements in Theorem 2. \(\square \)

The next corollary generalizes [43, Proposition 6.1.2].

### Corollary 3

*A*is essentially self-adjoint on \(L^2(E)\) with compact resolvent. There exists a complete orthonormal system of smooth eigenvectors \(\psi _j\in \Gamma ^\infty (E)\) with non-negative eigenvalues \(\lambda _j\ge 0\), that is, \(A\psi _j=\lambda _j\psi _j\) for all \(j\in \mathbb {N}\). Moreover,

### Proof

We have already shown that *A* is essentially self-adjoint with compact resolvent, see Lemma 1. It is well known that the spectrum of these operators is discrete [37, Theorem 6.29 in Chapter III§6.8] and real [37, Chapter V§3.5]. Since *A* is non-negative, each eigenvalue has to be non-negative. By hypoellipticity, its eigenfunctions are smooth, see [21, Corollary 2.10]. Since \(e^{-tA}\) and \(A^{-z}\) are trace class for \(t>0\) and \(\mathfrak {R}(z)>n/r\), respectively, the expressions (9) follow immediately, see [37, Chapter X§1.4]. Using the Tauberian theorem of Karamata, see [33, Theorem 108] or [51, Problem 14.2], Weyl’s law for the asymptotics of eigenvalues follows from the heat trace asymptotics in Corollary 1. \(\square \)

*E*and

*F*are two vector bundles over a closed filtered manifold

*M*and let \(D\in \mathcal {D}\mathcal {O}^k(E,F)\) be a differential operator of Heisenberg order at most \(k\ge 1\) such that

*D*and \(D^*\) are both Rockland. Then

*D*induces a Fredholm operator between Heisenberg Sobolev spaces, \(D:H_s(E)\rightarrow H_{s-k}(F)\), for every real

*s*, see [21, Corollary 3.28]. Moreover, its index does not depend on

*s*and can be expressed as

*k*which satisfy the Rockland condition. Clearly, they are formally self-adjoint and non-negative. According to Theorem 1, their heat kernels admit asymptotic expansions as \(t\searrow 0\),

### Corollary 4

*E*and

*F*be two vector bundles over a closed filtered manifold

*M*. Moreover, let \(D\in \mathcal {D}\mathcal {O}^k(E,F)\) be a differential operator of Heisenberg order at most \(k\ge 1\) such that

*D*and \(D^*\) both satisfy the Rockland condition. Then

*n*is odd.

### Proof

*t*and Corollary 1 yields

*n*is odd, then \(a_n^{D^*D}=0=a_n^{DD^*}\) according to Corollary 1, whence \({{\,\mathrm{ind}\,}}(D)=0\). \(\square \)

According to Atiyah [2] elliptic differential operators represent K-homology classes of the underlying manifold, see also [36, 6, Sect. 17], or [45, Sect. 5]. We have the following generalization for Rockland differential operators.

### Corollary 5

*E*and

*F*be two vector bundles over a closed filtered manifold

*M*. Moreover, let \(D\in \mathcal {D}\mathcal {O}^k(E,F)\) be a differential operator of Heisenberg order at most \(k\ge 1\) such that

*D*and \(D^*\) both satisfy the Rockland condition. Then the following holds true:

- (a)
\(P:=D({{\,\mathrm{id}\,}}_E+D^*D)^{-1/2}=({{\,\mathrm{id}\,}}_F+DD^*)^{-1/2}D\) is bounded from \(L^2(E)\) to \(L^2(F)\).

- (b)
\(P^*P-{{\,\mathrm{id}\,}}_E\) is compact on \(L^2(E)\).

- (c)
\(PP^*-{{\,\mathrm{id}\,}}_F\) is compact on \(L^2(F)\).

- (d)
[

*f*,*P*] is compact from \(L^2(E)\) to \(L^2(F)\), for each \(f\in C^\infty (M,\mathbb {C})\).

*C*(

*M*) by multiplication constitutes a graded Fredholm module, representing a K-homology class in \(K_0(M)=KK(C(M),\mathbb {C})\).

### Proof

According to Theorem 2 we have \(({{\,\mathrm{id}\,}}_E+D^*D)^{-1/2}\in \Psi ^{-k}(E)\), and thus \(P\in \Psi ^0(E,F)\). Hence, *P* represents a bounded operator \(L^2(E)\rightarrow L^2(F)\), see [21, Proposition 3.9(a)]. This shows (a).

Clearly, \(P^*P-{{\,\mathrm{id}\,}}_E=-({{\,\mathrm{id}\,}}_E+D^*D)^{-1}\in \Psi ^{-2k}(E)\). Hence, \(P^*P-{{\,\mathrm{id}\,}}_E\) represents a compact operator on \(L^2(E)\), see [21, Proposition 3.9(b)]. This shows (b), and (c) can be proved analogously.

*f*,

*P*] is compact, see [21, Proposition 3.9(b)]. \(\square \)

## 3 Pseudodifferential Operators on Filtered Manifolds

In this section, we will briefly recall van Erp and Yuncken’s pseudodifferential operator calculus on filtered manifolds, see [56] and [21]. Moreover, we will introduce a subclass of operators of integral order, characterized by an additional symmetry, and containing all differential operators. This class will be used to show that the heat kernel expansion for a differential operator has no log terms, and every other polynomial term vanishes.

*M*be a filtered manifold. For any two complex vector bundles

*E*and

*F*over

*M*, and every complex number

*s*, there is a class of operators called

*pseudodifferential operators of Heisenberg order s*and denoted by \(\Psi ^s(E,F)\), mapping sections of

*E*to sections of

*F*. Every \(A\in \Psi ^s(E,F)\) has a distributional Schwartz kernel \(k\in \Gamma ^{-\infty }(F\boxtimes E')\) with wave front set contained in the

*conormal*of the diagonal. In particular,

*k*is smooth away from the diagonal and

*A*induces a continuous operator \(\Gamma ^\infty _c(E)\rightarrow \Gamma ^\infty (F)\) which extends continuously to a pseudolocal operator on distributional sections, \(\Gamma ^{-\infty }_c(E)\rightarrow \Gamma ^{-\infty }(F)\). Here \(E':=E^*\otimes |\Lambda _M|\) where \(E^*\) denotes the dual bundle and \(|\Lambda _M|\) is the line bundle of 1-densities on

*M*. Moreover, \(F\boxtimes E'=p_1^*F\otimes p_2^*E'\) where \(p_i:M\times M\rightarrow M\) denote the canonical projections, \(i=1,2\). As usual,

*asymptotic expansion along the diagonal*in carefully chosen coordinates.

*M*. These are constructed using two geometrical choices: (1) a splitting of the filtration on

*TM*, i.e., a vector bundle isomorphism \(S:\mathfrak tM\rightarrow TM\) mapping \(\mathfrak t^pM\) into \(T^pM\) such that the composition with the canonical projection \(T^pM\rightarrow T^pM/T^{p+1}M=\mathfrak t^pM\) is the identity on \(\mathfrak t^pM\); and (2) a linear connection \(\nabla \) on

*TM*preserving the decomposition \(TM=\bigoplus _pS(\mathfrak t^pM)\). Denoting the corresponding exponential map by \(\exp ^\nabla :TM\rightarrow M\), we obtain a commutative diagram:Here \(\exp :\mathfrak tM\rightarrow {\mathcal {T}}M\) denotes the fiberwise exponential map, all downwards heading vertical arrows indicate canonical bundle projections, and the upwards pointing vertical arrows denote the corresponding zero (neutral) sections. In particular, \(\Delta (x)=(x,x)\) denotes the diagonal mapping, and \({{\,\mathrm{pr}\,}}_1(x,y)=x\). Using \(-S\) to identify \(\mathfrak tM\) with

*TM*, mediates between two common, yet conflicting, conventions we have adopted: The Lie algebra of a Lie group is defined using

*left*invariant vector fields, while the Lie algebroid of a smooth Lie groupoid is defined using

*right*invariant vector fields.

*U*of the zero section in \({\mathcal {T}}M\), it gives rise to a diffeomorphism \(\varphi :U\rightarrow V\) onto an open neighborhood

*V*of the diagonal in \(M\times M\) such that the rectangle at the bottom of diagram (16) commutes:Possibly shrinking

*U*, there exists a vector bundle isomorphism \(\phi \) over \(\varphi \) which restricts to the tautological identification over the diagonal/zero section,

*exponential coordinates adapted to the filtration.*Only the germ of \((\varphi ,\phi )\) along the zero section is relevant for formulating the asymptotic expansion. We will occasionally suppress the restriction to the neighborhoods

*U*or

*V*in our notation.

Put \(\mathcal {K}^\infty ({\mathcal {T}}M;E,F):=\Gamma ^\infty \bigl (\hom (\pi ^*E,\pi ^*F)\otimes \Omega _\pi \bigr )\), and let \(\mathcal {K}({\mathcal {T}}M;E,F)\) denote the space of distributional sections of \(\hom (\pi ^*E,\pi ^*F)\otimes \Omega _\pi \) with wave front set contained in the conormal of the zero section *o*(*M*). In particular, elements in \(\mathcal {K}({\mathcal {T}}M;E,F)\) are assumed to be smooth away from the zero section. Equivalently, these can be characterized as families \(a_x\in \Gamma ^{-\infty }(|\Lambda _{\mathcal {T}_xM}|)\otimes \hom (E_x,F_x)\) which are smooth away from the origin (regular) and depend smoothly on \(x\in M\).

*s*be a complex number. An element \(a\in \mathcal {K}({\mathcal {T}}M;E,F)\) is called

*essentially homogeneous of order s*if \((\delta _\lambda )_*a=\lambda ^sa\) mod \(\mathcal {K}^\infty ({\mathcal {T}}M;E,F)\), for all \(\lambda >0\). The space of principal cosymbols of order

*s*will be denoted by

*N*there exists an integer \(j_N\) such that \(\phi ^*(k|_V)-\sum _{j=0}^{j_N}k_j\) is of class \(C^N\). Strictly speaking, the right-hand side of (18) involves distributions representing the classes \(k_j\in \Sigma ^{s-j}(E,F)\), restricted to

*U*. Clearly, the condition expressed in (18) does not depend on the choice of these representatives. We will continue to suppress this in our notation in similar formulas below.

*Heisenberg principal symbol*of \(A\in \Psi ^s(E,F)\). It provides a short exact sequence

*r*in the sense of Sect. 2, and in this case the Heisenberg principal symbol discussed in Sect. 2 is related to the one introduced in the preceding paragraph by the canonical inclusion

*b*for the principal symbol exists if and only if \(\sigma ^s_x(A)\) satisfies the Rockland condition at each point \(x\in M\). The operator class \(\Psi ^s\) gives rise to a Heisenberg Sobolev scale with the expected mapping properties, see [21, Proposition 3.21] for details.

*zoom action*by putting \(\delta _\lambda ^{\mathbb {T}M}(g,0):=(\delta ^\text {op}_\lambda (g),0)\) and \(\delta _\lambda ^{\mathbb {T}M}(x,y,t):=(x,y,t/\lambda )\) for \(t\in \mathbb {R}^\times \) and all \(\lambda \ne 0\), see [56, Definition 17].

*E*and

*F*over

*M*, we consider the vector bundle

A conormal kernel \(k\in \mathcal {K}(M\times M;E,F)\) corresponds to an operator in \(\Psi ^s(E,F)\) if and only if it admits an extension across the tangent groupoid which is essentially homogeneous of order *s*. More precisely, iff there exists \(\mathbb {K}\in \mathcal {K}(\mathbb {T}M;E,F)\) such that \({{\,\mathrm{ev}\,}}_1(\mathbb {K})=k\) and \((\delta ^{\mathbb {T}M}_\lambda )_*\mathbb {K}=\lambda ^s\mathbb {K}\) mod \(\mathcal {K}^\infty (\mathbb {T}M;E,F)\) for all \(\lambda >0\), see [56, Definition 19] or [21, Definition 3.2]. In this case, we have \(\sigma ^s(A)=\nu ^*({{\,\mathrm{ev}\,}}_0(\mathbb {K}))\).

*t*in (21) is induced by multiplication with the function \(t:\mathbb {T}M\rightarrow \mathbb {R}\) given by the composition of the source (or target) map \(\mathbb {T}M\rightarrow M\times \mathbb {R}\) with the canonical projection onto \(\mathbb {R}\). Note that

*N*there exists an integer \(j_N\) such that \(\Phi ^*\bigl (\mathbb {K}|_{\mathbb {V}}\bigr )-\sum _{j=0}^{j_N}\mathbb {K}_jt^j\) is of class \(C^N\). The coefficients are related to the terms in the asymptotic expansion (18) of \(k={{\,\mathrm{ev}\,}}_1(\mathbb {K})\) via

*r*.

### Definition 1

*essentially projectively homogeneous of order r*if \((\delta _\lambda )_*a=\lambda ^ra\) mod \(\mathcal {K}^\infty ({\mathcal {T}}M;E,F)\) holds for all \(\lambda \ne 0\). Correspondingly, we put

### Lemma 2

*a*is essentially projectively homogeneous of order

*r*, i.e., represents an element in \(\Sigma ^r_2(E,F)\), if and only if there exist \(a_\infty \in \mathcal {K}^\infty ({\mathcal {T}}M;E,F)\), \(q\in \mathcal {K}({\mathcal {T}}M;E,F)\) and \(p\in \mathcal {P}^r({\mathcal {T}}M;E,F)\) such that \(a=a_\infty +q\) and

*q*is only unique mod \(\mathcal {P}^r({\mathcal {T}}M;E,F)\), but

*p*is without ambiguity.

### Definition 2

An operator \(A\in \Psi ^r(E,F)\) of integral order \(r\in \mathbb {Z}\) is called *projective* if its Schwartz kernel *k* admits an extension across the tangent groupoid, \(\mathbb {K}\in \mathcal {K}(\mathbb {T}M;E,F)\), which is essentially projectively homogeneous of order *r*, that is, \({{\,\mathrm{ev}\,}}_1(\mathbb {K})=k\), and for all \(\lambda \ne 0\) we have \((\delta ^{\mathbb {T}M}_\lambda )_*\mathbb {K}=\lambda ^r\mathbb {K}\) mod \(\mathcal {K}^\infty (\mathbb {T}M;E,F)\). We will denote these operators by \(\Psi ^r_2(E,F)\).

### Proposition 1

- (a)
If \(A\in \Psi ^r_2(E,F)\) and \(B\in \Psi ^l_2(F,G)\) then \(BA\in \Psi ^{l+r}_2(E,G)\), provided at least one of

*A*and*B*is properly supported. - (b)
If \(A\in \Psi ^r_2(E,F)\), then \(A^t\in \Psi ^r_2(F',E')\).

- (c)If \(A\in \Psi ^r_2(E,F)\), then \(\sigma ^r(A)\in \Sigma ^r_2(E,F)\). Moreover,is a natural short exact sequence.$$\begin{aligned} 0\rightarrow \Psi ^{r-1}_2(E,F)\rightarrow \Psi ^r_2(E,F)\xrightarrow {\sigma ^r}\Sigma ^r_2(E,F)\rightarrow 0 \end{aligned}$$
- (d)
\(\bigcap _{r\in \mathbb {Z}}\Psi ^r_2(E,F)=\mathcal {O}^{-\infty }(E,F)\), the smoothing operators.

- (e)
For \(r\in \mathbb {N}_0\) we have \(\mathcal {D}\mathcal {O}^r(E,F)=\mathcal {D}\mathcal {O}(E,F)\cap \Psi ^r_2(E,F)\).

- (f)
If an operator in \(\Psi ^r_2(E,F)\) admits a left parametrix in \(\Psi ^{-r}(F,E)\), then it also admits a left parametrix \(\Psi ^{-r}_2(F,E)\). An analogous statement holds true for right parametrices.

- (g)
Suppose \(A\in \Psi ^r(E,F)\) and let \(\phi ^*(k|_V)\sim \sum _{j=0}^\infty k_j\) with \(k_j\in \Sigma ^{r-j}(E,F)\) denote the asymptotic expansion of its kernel along the diagonal with respect to adapted exponential coordinates. Then \(A\in \Psi ^r_2(E,F)\) if and only if \(k_j\in \Sigma ^{r-j}_2(E,F)\) for all \(j\in \mathbb {N}_0\).

### Proof

Parts (a) and (b) can be proved exactly as in [56], see also [21, Proposition 3.4]. Indeed, all maps in the commutative diagram (20) are multiplicative.

Part (d) follows immediately from \(\bigcap _{r\in \mathbb {Z}}\Psi ^r(E,F)=\mathcal {O}^{-\infty }(E,F)\), see [56, Corollary 53] or [21, Proposition 3.4(c)], for we have the obvious inclusions \(\mathcal {O}^{-\infty }(E,F)\subseteq \Psi ^r_2(E,F)\subseteq \Psi ^r(E,F)\).

The proof of [21, Proposition 3.4(f)] actually shows (e), see also [56, Sect. 10.3].

To see (f), consider \(A\in \Psi ^r_2(E,F)\) and \(B\in \Psi ^{-r}(F,E)\) such that \(BA-{{\,\mathrm{id}\,}}\) is a smoothing operator. Choose \(\mathbb {K}\in \Sigma ^r_2(\mathbb {T}M;E,F)\) such that \({{\,\mathrm{ev}\,}}_1(\mathbb {K})\) represents *A* mod smoothing operators. Moreover, choose \(\mathbb {L}\in \Sigma ^{-r}(\mathbb {T}M;F,E)\) such that \({{\,\mathrm{ev}\,}}_1(\mathbb {L})\) represents *B* mod smoothing operators. Since \({{\,\mathrm{ev}\,}}_1\) induces a multiplicative isomorphism as indicated in (21), we conclude \(\mathbb {L}\mathbb {K}=1\) in \(\Sigma ^0(\mathbb {T}M;E,E)\). Consider \(\tilde{\mathbb {L}}:=\frac{1}{2}\bigl (\mathbb {L}+(-1)^r(\delta ^{\mathbb {T}M}_{-1})_*\mathbb {L}\bigr )\in \Sigma ^{-r}_2(\mathbb {T}M;\mathbb {F},\mathbb {E})\). In view of \(1=(\delta ^{\mathbb {T}M}_{-1})_*1=(\delta ^{\mathbb {T}M}_{-1})_*(\mathbb {L}\mathbb {K})=\bigl ((\delta ^{\mathbb {T}M}_{-1})_*\mathbb {L}\bigr )\bigl ((\delta ^{\mathbb {T}M}_{-1})_*\mathbb {K}\bigr )=\bigl ((\delta ^{\mathbb {T}M}_{-1})_*\mathbb {L}\bigr )\bigl ((-1)^r\mathbb {K}\bigr )=\bigl ((-1)^r(\delta ^{\mathbb {T}M}_{-1})_*\mathbb {L}\bigr )\mathbb {K}\), we conclude \(\tilde{\mathbb {L}}\mathbb {K}=1\) in \(\Sigma ^0_2(\mathbb {T}M;E,E)\). Hence, \({{\,\mathrm{ev}\,}}_1(\tilde{\mathbb {L}})\) gives rise to a left parametrix \({\tilde{B}}\in \Psi ^{-r}_2(F,E)\) such that \({\tilde{B}}A-{{\,\mathrm{id}\,}}\) is a smoothing operator.

*j*. Using (28), we see that this holds iff \(\mathbb {K}_j\in \Sigma ^{r-j}_2(\mathcal {T}^\text {op}M;E,F)\), for all

*j*. In view of (24), this is in turn equivalent to \(k_j\in \Sigma ^{r-j}_2(E,F)\), for all

*j*. \(\square \)

## 4 Heat Kernel Asymptotics

The aim of this section is to prove Theorem 1. To this end let *E* be a vector bundle over a closed filtered manifold *M*, and suppose \(A\in \mathcal {D}\mathcal {O}^r(E)\) is a differential Rockland operator of even Heisenberg order \(r>0\) which is formally self-adjoint and non-negative with respect to the \(L^2\) inner product induced by a volume density \(\mathrm{d}x\) on *M* and a fiberwise Hermitian metric *h* on *E*, see (4).

*A*and \(\frac{\partial }{\partial t}\) as operators over \(M\times \mathbb {R}\) acting on sections of the pull back bundle \({\tilde{E}}:=\pi _1^*E\). The fiberwise Hermitian metric

*h*on

*E*induces a fiberwise Hermitian metric \({\tilde{h}}:=\pi _1^*h\) on \({\tilde{E}}\). Furthermore, the volume density \(\mathrm{d}x\) on

*M*and the standard volume density \(\mathrm{d}t\) on \(\mathbb {R}\) provide a volume density \(\mathrm{d}x\mathrm{d}t\) on \(M\times \mathbb {R}\). We will use the \(L^2\) inner product on sections of \({\tilde{E}}\) induced by \(\mathrm{d}x\mathrm{d}t\) and \({\tilde{h}}\), see (4).

This filtration on \(M\times \mathbb {R}\) is motivated by the fact that the heat operator \(A+\frac{\partial }{\partial t}\) becomes a Rockland operator. More precisely, we have:

### Lemma 3

The differential operator \(A+\frac{\partial }{\partial t}\) is a Rockland operator of Heisenberg order *r*. The same is true for \((A+\frac{\partial }{\partial t})^*=A-\frac{\partial }{\partial t}\).

### Proof

Clearly, \(\frac{\partial }{\partial t}\in \Gamma ^\infty (T^{-r}(M\times \mathbb {R}))\), hence \(\frac{\partial }{\partial t}\) is a differential operator of Heisenberg order at most *r* on \(M\times \mathbb {R}\). Since \(\pi _1^*T^pM\subseteq T^p(M\times \mathbb {R})\), the operator *A* has Heisenberg order at most *r* when considered on \(M\times \mathbb {R}\). Consequently, \(A+\frac{\partial }{\partial t}\) has Heisenberg order at most *r*.

*A*. For the Heisenberg principal symbol of \(A+\tfrac{\partial }{\partial t}\) we clearly have

*T*of homogeneous degree \(-r\).

*A*is assumed to be formally self-adjoint, we have \(a^*=a\) and \(b^*=-b\), see (5), hence

*a*or

*b*acts injectively on \(\mathcal {H}_\infty \), see (35). If \(\bar{\pi }\) is non-trivial, then

*a*acts injectively for

*A*is assumed to satisfy the Rockland condition. If \(\bar{\pi }\) is trivial, then

*b*acts injectively, for it has to be a non-trivial scalar.

^{3}\(\square \)

*M*as described in Sect. 3 and summarized in the commutative diagram (16). The exponential coordinates on \(M\times \mathbb {R}\) will be defined such that the following diagram commutes:Here \(\tilde{\pi }:\mathcal {T}(M\times \mathbb {R})\rightarrow M\times \mathbb {R}\) denotes the bundle projection, \({\tilde{o}}\) is the zero section, and \(\tilde{\Delta }\) denotes the diagonal mapping. Moreover,

*TM*and the trivial connection on \(T\mathbb {R}\).

*N*there exists an integer \(j_N\) such that \(\tilde{\phi }^*(k_{{\tilde{Q}}}|_{{\tilde{V}}})-\sum _{j=0}^{j_N}{\tilde{q}}_j\) is of class \(C^N\) on \({\tilde{U}}\). According to Lemma 2, we may fix representatives \({\tilde{q}}_j\in \mathcal {K}\bigl (\mathcal {T}(M\times \mathbb {R});{\tilde{E}}\bigr )\) satisfying

*Q*is continuous, when \(C_+(\mathbb {R},L^2(E))=\varinjlim C_{[t_0,\infty )}(\mathbb {R},L^2(E))\) is equipped with the inductive limit topology over \(t_0\in \mathbb {R}\), and \(C_{[t_0,\infty )}(\mathbb {R},L^2(E))\) denotes the subspace of continuous maps which are supported on \([t_0,\infty )\) carrying the topology of uniform convergence on compact subsets. Composing

*Q*with the continuous inclusions \(\Gamma _c^\infty ({\tilde{E}})\subseteq C_+(\mathbb {R},L^2(E))\) and \(C_+(\mathbb {R},L^2(E))\subseteq \Gamma ^{-\infty }({\tilde{E}})\), we may thus regard it as a continuous operator, \(Q:\Gamma ^\infty _c({\tilde{E}})\rightarrow \Gamma ^{-\infty }({\tilde{E}})\). By the Schwartz kernel theorem,

*Q*has a distributional kernel, \(k_Q\in \Gamma ^{-\infty }({\tilde{E}}\boxtimes {\tilde{E}}')\), that is,

*Q*maps \(\Gamma ^\infty _c({\tilde{E}})\) continuously into \(\Gamma ^\infty ({\tilde{E}})\). Hence, \(QR_2\) is a smoothing operator, for it maps \(\Gamma ^{-\infty }_c({\tilde{E}})\) continuously into \(\Gamma ^\infty ({\tilde{E}})\). Using the equality on the right-hand side of (44), we conclude that \({\tilde{Q}}-Q\) is a smoothing operator.

*Q*differs from \({\tilde{Q}}\) by a smoothing operator, \(k_Q\) is smooth away from the diagonal and (39) gives an asymptotic expansion

*N*there exists an integer \(j_N\) such that

*x*,

*s*) in compact subsets of \(M\times \mathbb {R}\). Here \(|-|\) denotes a fiberwise homogeneous norm on \(\mathcal {T}(M\times \mathbb {R})\). Comparing (7), (41), and (42), we find

*s*, see (37) and (38). Combining this with (46) and (40), we conclude that

*x*. Using (49), (38), (33), and (2), where

*n*has to be replaced by the homogeneous dimension \(n+r\) of \(M\times \mathbb {R}\), we obtain

*x*.

Using \(\lambda =-1\) in (49) we see that \({\tilde{q}}_{j,(x,s)}(o_x,t)=0\) for all odd *j* and \(t\ne 0\), see (33) and (2). Hence, \(q_j(x)=0\) for all odd *j*, see (52).

### Remark 2

The asymptotic term \(q_j\in \Gamma ^\infty \bigl ({{\,\mathrm{end}\,}}(E)\otimes |\Lambda _M|\bigr )\) in Theorem 1 can be read off the homogeneous term \({\tilde{q}}_j\in \Sigma ^{-r-j}({\tilde{E}})\) in the asymptotic expansion (39) of any parametrix \({\tilde{Q}}\) for the heat operator \(A+\frac{\partial }{\partial t}\) on \(M\times \mathbb {R}\). Indeed, the representative \({\tilde{q}}_j\in \mathcal {K}(\mathcal {T}(M\times \mathbb {R});{\tilde{E}})\) used in (52) is uniquely characterized by (51) and (48). Note that these representatives are also translation invariant, corresponding precisely to the homogeneous terms on which the Volterra–Heisenberg calculus is based on, see [4, Sect. 3] and [43, Sect. 5.1].

To complete the proof of Theorem 1 it remains to show \(q_0(x)>0\). This will be accomplished using the subsequent lemma concerning the heat kernel on the osculating groups, cf. [43, Lemma 6.1.4].

### Lemma 4

### Proof

*dg*on \(\mathcal {T}_xM\). Writing \(k_t^{\sigma ^r_x(A)}=k_t'dg\) and \({\tilde{q}}_{0,x}=q'\mathrm{d}g\mathrm{d}t\), homogeneity implies

*B*is some left invariant homogeneous differential operator on \(\mathcal {T}_xM\). This shows that \(k_t'\) is indeed in the Schwartz space \(\mathcal {S}(\mathcal {T}_xM)\otimes {{\,\mathrm{end}\,}}(E_x)\), whence (53).

*t*. Testing (57) with \(\chi \in \Gamma ^\infty _c(\mathbb {R},E_x)\), considered as function on \(\mathcal {T}_xM\times \mathbb {R}\) which is constant in \(\mathcal {T}_xM\), we obtain

*A*is formally self-adjoint, we also have \(\bigl (k^{\sigma ^r_x(A)}_t\bigr )^*=k^{\sigma ^r_x(A)}_t\), that is,

## 5 Complex Powers

In this section, we will present a proof of Theorem 2 following the approach in [43, Sect. 5.3]. Throughout this section *E* denotes a vector bundle over a closed filtered manifold *M*, and \(A\in \mathcal {D}\mathcal {O}^r(E)\) is a Rockland differential operator of even Heisenberg order \(r>0\) which is formally self-adjoint and non-negative as in Theorem 1.

*P*denote the orthogonal projection onto \(\ker (A)\). To express the complex powers using the Mellin formula,

### Lemma 5

*P*onto \(\ker (A)\). Then there exists \(\varepsilon >0\) such that

*x*and

*y*.

### Proof

*A*is self-adjoint and non-negative, its spectrum is contained in \([0,\infty )\). Moreover, since

*A*has compact resolvent, zero is isolated in its spectrum. Hence, by functional calculus, there exists \(\varepsilon >0\) such that

*z*. If \(x\ne y\), then \(k_t(x,y)\) vanishes to infinite order at \(t=0\), see (47), hence the first integral in (69) converges in the \(C^\infty \)-topology and defines a smooth function on \(\{x\ne y\}\times \mathbb {C}\) which depends holomorphically on

*z*.

*s*, use (47), and move to exponential coordinates, see (37) and (38),

*N*, we choose an integer \(j_N\) such that \(\tilde{\phi }^*(k_Q|_{{\tilde{V}}})-\sum _{j=0}^{j_N}{\tilde{q}}_j\) is of class \(C^N\), see (46), and rewrite (71) in the form

*z*.

*z*and defines an entire function. In view of Theorem 1, the second integral on the right-hand side of (77) converges for \(\mathfrak {R}(z)>(n-N)/r\) and defines a holomorphic function on \(\{\mathfrak {R}(z)>(n-N)/r\}\). Note that these considerations are all uniform in

*x*. Combining this with (76), we see that \(k_{A^{-z}}(x,x)\) can be extended meromorphically to all complex

*z*with poles and special values as specified in Theorem 2. Here we also use the classical fact that \(1/\Gamma (z)\) is an entire function with zero set \(-\mathbb {N}_0\) and \((1/\Gamma )'(-l)=(-1)^ll!\) for all \(l\in \mathbb {N}_0\).

To complete the proof of Theorem 2, it remains to show that the powers \(A^{-z}\) form a holomorphic family of Heisenberg pseudodifferential operators in a sense analogous to [43, Sect. 4]. This will be established in Sect. 6.

## 6 Holomorphic Families of Heisenberg Pseudodifferential Operators

In this section, we will extend the concept of a holomorphic family of pseudodifferential operators [43, Chapter 4] to the Heisenberg calculus on general filtered manifolds. We will show that the complex powers discussed above do indeed form a holomorphic family as stated in Theorem 2. Holomorphic families will also be used to construct a non-commutative residue in Sect. 7.

### Remark 3

*N*, and suppose

*S*is a closed submanifold of

*N*. We let \(\mathcal {K}\subseteq \Gamma ^{-\infty }(\xi )\) denote the vector space of all distributional sections of \(\xi \) whose wave front set is contained in the conormal of

*S*. In particular, these distributions are smooth on \(N{\setminus } S\), hence we have a canonical map

*S*in

*N*, that is, \(T^\perp S:=TN|_S/TS\). Suppose \(\varphi :T^\perp S\rightarrow W\subseteq N\) is a tubular neighborhood, i.e., a diffeomorphism onto an open neighborhood

*W*of

*S*in

*N*, which restricts to the identity along

*S*. Moreover, let \(\phi \) be a vector bundle isomorphism \(\pi ^*(\xi |_S)\cong \varphi ^*(\xi |_W)\). If \(a\in \mathcal {K}\), then \(\phi ^*a\) is a distributional sections of \(\pi ^*(\xi |_S)\) whose wave front set is conormal to the zero section \(S\subseteq T^\perp S\). In particular, \(\phi ^*a\) is \(\pi \)-fibered. Hence, we obtain a map

Suppose *E* and *F* are two vector bundles over a filtered manifold *M*. Moreover, let \(\Omega \) be a domain in the complex plane and suppose \(s:\Omega \rightarrow \mathbb {C}\) is a holomorphic function. We intend to make precise when a family of operators \(A_z\in \Psi ^{s(z)}(E,F)\), parametrized by \(z\in \Omega \), is considered to be a holomorphic family.

Recall from Sect. 3 that \(\mathcal {K}(\mathbb {T}M;E,F)\) denotes the space of all distributional sections of the vector bundle (19) whose wave front set is contained in the conormal of the units, \(M\times \mathbb {R}\subseteq \mathbb {T}M\). We equip \(\mathcal {K}(\mathbb {T}M;E,F)\) with the topology described in Remark 3, and let \(\mathcal {K}_\Omega (\mathbb {T}M;E,F)\) denote the space of holomorphic curves from \(\Omega \) into \(\mathcal {K}(\mathbb {T}M;E,F)\). Moreover, we let \(\mathcal {K}_\Omega ^\infty (\mathbb {T}M;E,F)\) denote the space of holomorphic curves from \(\Omega \) into \(\mathcal {K}^\infty (\mathbb {T}M;E,F)\) where the latter space carries the \(C^\infty \)-topology.

### Definition 3

Let \(\Omega \) be a domain in the complex plane and suppose \(s:\Omega \rightarrow \mathbb {C}\) is a holomorphic function. A family of operators \(A_z\in \Psi ^{s(z)}(E,F)\), parametrized by \(z\in \Omega \), is called a *holomorphic family of Heisenberg pseudodifferential operators* if there exists a family \(\mathbb {K}_z\in \mathcal {K}_\Omega (\mathbb {T}M;E,F)\) such that \({{\,\mathrm{ev}\,}}_1(\mathbb {K}_z)\) is the Schwartz kernel of \(A_z\) for all \(z\in \Omega \), and \(\mathbb {K}_z\) is essentially homogeneous of order *s*(*z*) in the sense that \((\delta ^{\mathbb {T}M}_\lambda )_*\mathbb {K}_z-\lambda ^{s(z)}\mathbb {K}_z\) is a family in \(\mathcal {K}^\infty _\Omega (\mathbb {T}M;E,F)\), for all \(\lambda >0\).

### Lemma 6

Suppose \(A_z\in \Psi ^{s(z)}(E,F)\) and \(B_z\in \Psi ^{t(z)}(F,G)\) are two holomorphic families of Heisenberg pseudodifferential operators, where \(z\in \Omega \) and \(s,t:\Omega \rightarrow \mathbb {C}\). Then \(B_zA_z\in \Psi ^{(t+s)(z)}(E,G)\) is a holomorphic family of Heisenberg pseudodifferential operators, provided at least one of the two families is properly supported (locally uniformly in *z*). Moreover, the transpose \(A_z^t\in \Psi ^{s(z)}(F',E')\) is a holomorphic family of Heisenberg pseudodifferential operators.

### Proof

Recall from Sect. 3 that \(\mathcal {K}({\mathcal {T}}M;E,F)\) denotes the space of all distributional sections of the vector bundle \(\hom (\pi ^*E,\pi ^*F)\otimes \Omega _\pi \) over \({\mathcal {T}}M\) whose wave front set is contained in the conormal of the units, \(M\subseteq {\mathcal {T}}M\). We equip \(\mathcal {K}({\mathcal {T}}M;E,F)\) with the topology described in Remark 3, and let \(\mathcal {K}_\Omega ({\mathcal {T}}M;E,F)\) denote the space of holomorphic curves from \(\Omega \) into \(\mathcal {K}({\mathcal {T}}M;E,F)\). Moreover, we let \(\mathcal {K}_\Omega ^\infty ({\mathcal {T}}M;E,F)\) denote the space of holomorphic curves from \(\Omega \) into \(\mathcal {K}^\infty ({\mathcal {T}}M;E,F)\) where the latter space carries the \(C^\infty \)-topology.

### Definition 4

Let \(\Omega \) be a domain in the complex plane, and suppose \(s:\Omega \rightarrow \mathbb {C}\) is a holomorphic function. A family \(k_z\in \mathcal {K}_\Omega ({\mathcal {T}}M;E,F)\) is called *essentially homogeneous* of order *s*(*z*) if \((\delta _\lambda )_*k_z-\lambda ^{s(z)}k_z\in \mathcal {K}_\Omega ^\infty ({\mathcal {T}}M;E,F)\), for all \(\lambda >0\). This generalizes [43, Definition 4.4.1] where the term *almost homogeneous* is used.

### Lemma 7

- (a)
\(A_z\) is a holomorphic family of Heisenberg pseudodifferential operators.

- (b)Away from the diagonal, the Schwartz kernels \(k_z\) are smooth and depend holomorphically on \(z\in \Omega \). Moreover, with respect to some (and then every) exponential coordinates adapted to the filtration as in Sect. 3, see (16), we have an asymptotic expansion of the formwhere \(k_{j,z}\in \mathcal {K}_\Omega ({\mathcal {T}}M;E,F)\) is an essentially homogeneous family of order \(s(z)-j\), cf. Definition 4. More precisely, for every \(z_0\in \Omega \) and every integer$$\begin{aligned} \phi ^*(k_z|_V)\sim \sum _{j=0}^\infty k_{j,z} \end{aligned}$$(80)
*N*there exists a neighborhood*W*of \(z_0\) in \(\Omega \) such that, for sufficiently large \(J\in \mathbb {N}\), the expression \(\phi ^*(k_z|_V)-\sum _{j=0}^{J-1}k_{j,z}\) restricts to a holomorphic curve from*W*into the space of \(C^N\)-sections of \(\hom (\pi ^*E,\pi ^*F)\otimes \Omega _\pi \) over*U*. - (c)Away from the diagonal, the Schwartz kernels \(k_z\) are smooth and depend holomorphically on \(z\in \Omega \). Moreover, with respect to some (and then every) exponential coordinates adapted to the filtration as in Sect. 3, see (16), and for some (and then any) properly supported bump function \(\chi \in C^\infty _\text {prop}({\mathcal {T}}M)\) with \({{\,\mathrm{supp}\,}}(\chi )\subseteq U\) and \(\chi \equiv 1\) in a neighborhood of the zero section, the fiber wise Fourier transform,is smooth on \(\Omega \times \mathfrak t^*M\), depends holomorphically on$$\begin{aligned} \mathcal {F}\bigl (\chi \cdot \phi ^*(k_z|_V)\bigr )(\xi ) =\int _{X\in \mathfrak t_xM}e^{-2\pi \mathbf i\langle \xi ,X\rangle }\bigl (\chi \cdot \phi ^*(k_z|_V)\bigr )(\exp (X)),\quad \xi \in \mathfrak t_x^*M, \end{aligned}$$
*z*, and admits a uniform asymptotic expansion of the formwhere \({\hat{k}}_{j,z}\in \Gamma ^\infty (\hom (p^*E,p^*F)|_{\mathfrak t^*M{\setminus } M})\) are homogeneous of order \(s(z)-j\), i.e., \((\dot{\delta }'_\lambda )^*{\hat{k}}_{j,z}=\lambda ^{s(z)-j}{\hat{k}}_{j,z}\) for all \(\lambda >0\). Here \(\dot{\delta }_\lambda '\) denotes the grading automorphism on \(\mathfrak t^*M\) dual to \(\dot{\delta }_\lambda \) on \(\mathfrak tM\), and \(p:\mathfrak t^*M\rightarrow M\) denotes the vector bundle projection. More precisely, for all \(a,N\in \mathbb {N}\), and for every homogeneous differential operator$$\begin{aligned} \mathcal {F}\bigl (\chi \cdot \phi ^*(k_z|_V)\bigr )\sim \sum _{j=0}^\infty {\hat{k}}_{j,z}, \end{aligned}$$(81)*D*acting on \(\Gamma ^\infty (\hom (p^*E,p^*F))\), i.e., \((\dot{\delta }'_\lambda )_*D=\lambda ^aD\), for every compact \(K\subseteq \Omega \), and for some (and then every) fiberwise homogeneous norm on \(\mathfrak t^*M\), and for some (and then every) fiberwise Hermitian metric on \(\hom (E,F)\), there exists a constant \(C\ge 0\) such thatholds for all \(z\in K\) and all \(\xi \in \mathfrak t^*M\) with \(|\xi |\ge 1\).$$\begin{aligned} \left| D\left( \mathcal {F}\bigl (\chi \cdot \phi ^*(k_z|_V)\bigr )-\sum _{j=0}^{N-1}{\hat{k}}_{j,z}\right) (\xi )\right| \le C|\xi |^{s(z)-N-a} \end{aligned}$$(82)

### Proof

*W*be an open subset with compact closure in \(\Omega \), and suppose \(N\in \mathbb {N}\). Then, for sufficiently large

*J*, the family \(\mathbb {L}_{J,z}\) restricts to a holomorphic curve from

*W*into the space of \(C^N\)-sections over \(\mathbb {U}\). Indeed, this follows from a parametrized version of [56, Theorem 52]. We conclude that \(\Phi ^*\bigl (\mathbb {K}_z|_{\mathbb {V}}\bigr )-\sum _{j=0}^{J-1}\mathbb {K}_{j,z}t^j\) restricts to a holomorphic curve from

*W*into the space of \(C^N\)-sections over \(\mathbb {U}\). Evaluating at \(t=1\), we obtain the asymptotic expansion (80) with \(k_{j,z}=\nu ^*(\mathbb {K}_{j,z})\).

To see (b)\(\Rightarrow \)(a) we observe that \(t^jk_{j,z}\in \mathcal {K}_\Omega (\mathbb {U};E,F)\) is essentially homogeneous of order *s*(*z*). Parametrizing the proof of [56, Theorem 59], we obtain a holomorphic family \(\mathbb {L}_z\in \mathcal {K}_\Omega (\mathbb {U};E,F)\) which is essentially homogeneous of order *s*(*z*) and such that \({{\,\mathrm{ev}\,}}_1(\mathbb {L}_z)\sim \sum _{j=0}^\infty k_{j,z}\) in the same sense as (80). Clearly, this implies that \(\phi ^*(k_z|_V)-{{\,\mathrm{ev}\,}}_1(\mathbb {L}_z)\) is a holomorphic curve into the space of smooth sections over \(\mathbb {U}\). Adding a term in \(\mathcal {K}_\Omega ^\infty (\mathbb {U};E,F)\) to \(\mathbb {L}_z\), we may, moreover, assume \(\phi ^*(k_z|_V)={{\,\mathrm{ev}\,}}_1(\mathbb {L}_z)\). Using a bump function one readily constructs \(\mathbb {K}_z\in \mathcal {K}_\Omega (\mathbb {T}M;E,F)\), essentially homogeneous of order *s*(*z*), such that \(\Phi ^*(\mathbb {K}_z)\) coincides with \(\mathbb {L}_z\) on neighborhood of \(M\times \mathbb {R}\). Hence, \(k_z\) coincides with \({{\,\mathrm{ev}\,}}_1(\mathbb {K}_z)\) in a neighborhood of the diagonal. Clearly, this implies (a).

*D*, and every integer \(b\in \mathbb {N}\), there exists a constant \(C\ge 0\) such that

*D*, and every sufficiently large \(J\in \mathbb {N}\), there exists a constant \(C\ge 0\) such that

*D*, and every sufficiently large \(J\in \mathbb {N}\), there exists a constant \(C\ge 0\) such that

*D*is homogeneous of degree \(a\in \mathbb {N}\), i.e., \((\dot{\delta }'_\lambda )_*D=\lambda ^aD\), then \(|D{\hat{k}}_{j,z}(\xi )|\le C|\xi |^{s(z)-j-a}\), whence (81).

*W*is an open subset with compact closure in \(\Omega \), then (82) implies that, for sufficiently large \(J\in \mathbb {N}\), the expression \(\chi \cdot \phi ^*(k_z|_V)-\sum _{j=0}^{J-1}k_{j,z}\) is a holomorphic curve in the space of \(C^N\) sections, whence (80). \(\square \)

### Remark 4

As pointed out in the proof above, the homogeneous terms \({\hat{k}}_{j,z}\) in the asymptotic expansion (81) depend holomorphically on \(z\in \Omega \).

The characterization in Lemma 7 shows that the concept of holomorphic families considered here generalizes the definition for Heisenberg manifolds, see [43, Sect. 4].

Let us now complete the proof of Theorem 2 by establishing the following generalization of [43, Theorem 5.3.1].

### Lemma 8

The complex powers \(A^{-z}\in \Psi ^{-zr}(E)\) considered in Theorem 2 constitute a holomorphic family.

### Proof

Writing \(A^{-z}=A^kA^{-(z+k)}\) and using Lemma 6, we see that it suffices to show that the powers form a holomorphic family on \(\Omega =\{\mathfrak {R}(z)>0\}\). To prove this, we will use the characterization in Lemma 7(b) and proceed as in [43, Lemma 5.3.2].

As observed above, away from the diagonal the Schwartz kernels \(k_{A^{-z}}\), see (69), are smooth and depend holomorphically on \(z\in \mathbb {C}\). Moreover, the distributions \(K_{z,j}\) defined in (73) form a family in \(\mathcal {K}_\Omega ({\mathcal {T}}M;E,F)\) which is essentially homogeneous of order \(-rz-j\) in the sense of Definition 4, see (74). Furthermore, the \(C^N\) sections defined by the integral in (72) depend holomorphically on \(z\in \Omega \). We conclude that the asymptotic expansion (75) established above is in fact the asymptotic expansion required in Lemma 7(b). \(\square \)

## 7 Non-commutative Residue

*E*be a vector bundle over a closed filtered manifold

*M*of homogeneous dimension

*n*. We will fix a strictly positive Rockland differential operator \(D\in \mathcal {D}\mathcal {O}^r(E)\) of even Heisenberg order \(r>0\). For a pseudodifferential operator \(A\in \Psi ^k(E)\) of integer Heisenberg order \(k\in \mathbb {Z}\), the operator \(AD^{-z}\in \Psi ^{k-rz}(E)\) is trace class for \(\mathfrak R(z)>(n+k)/r\) and hence defines a holomorphic function

*M*and the restriction along the zero section of the bundle of vertical densities on \({\mathcal {T}}M\), see Sect. 3. In particular, we will use the induced identification

### Proposition 2

*E*be a vector bundle over a closed filtered manifold

*M*of homogeneous dimension

*n*. Moreover, suppose \(r>0\) and let \(R(z)\in \Psi ^{d-rz}(E)\) be a holomorphic family of Heisenberg pseudodifferential operators on

*M*with Schwartz kernels \(k_z\). Then the restriction to the diagonal, \(k_z(x,x)\), provides a holomorphic curve in \(\Gamma ^\infty ({{\,\mathrm{end}\,}}(E)\otimes |\Lambda _M|\bigr )\) for \(\mathfrak {R}(z)>(n+d)/r\). This curve can be extended meromorphically to the entire complex plane with at most simple poles located at \(z_j:=(n+d-j)/r\) with \(j\in \mathbb {N}_0\). Moreover, the residue at \(z_j\) can be computed locally by the expected formula,

^{4}

### Remark 5

*f*on \(\mathfrak t_x^*M{\setminus }\{0\}\).

### Proof of Proposition 2

*z*. The second integral converges for \(\mathfrak {R}(z)>(d+n-N)/r\) and defines a smooth section of \({{\,\mathrm{end}\,}}(E)\otimes |\Lambda _M|\) which depends holomorphically on these

*z*in view of the estimates (82). Using (88) and the homogeneity of \({\hat{k}}_{j,z}\), we rewrite the remaining terms in the form

*R*(

*z*) is continuous, and we can compute the trace by

The following corollary generalizes a result for CR manifolds obtained by R. Ponge in his thesis, see [42].

### Corollary 6

- (a)
\(\tau ([A,B])=0\) for all \(A,B\in \Psi ^\infty (E)\).

- (b)
\(\tau (A)\) does not depend on the positive Rockland differential operator

*D*. - (c)
If the order of \(A\in \Psi ^\infty (E)\) is less than \(-n\) then \(\tau (A)=0\).

- (d)If
*A*has order \(d\ge -n\), and \(\phi ^*(k_A|_V)\sim \sum _{j=0}^\infty k_j\) is an asymptotic expansion of its kernel with \(k_j\) essentially homogeneous of order \(d-j\), chosen such that \((\delta _\lambda )_*k_{d+n}=\lambda ^{-n}k_{d+n}+\lambda ^{-n}\log |\lambda |p_A\) for \(\lambda >0\) with \(p_A\in \mathcal {P}^{-n}({\mathcal {T}}M;E,E)\), then$$\begin{aligned} \tau (A)=\int _M{{\,\mathrm{tr}\,}}_E(p_A|_M). \end{aligned}$$ - (e)If \(B\in \Gamma ^\infty ({{\,\mathrm{end}\,}}(E))\) and \(j\in \mathbb {N}_0\), thenwhere \(q_j\) denotes the term in the heat kernel asymptotics of$$\begin{aligned} \tau \left( BD^{(j-n)/r}\right) =\frac{r}{\Gamma ((n-j)/r)}\int _M{{\,\mathrm{tr}\,}}_E(Bq_j), \end{aligned}$$(90)
*D*, that is \(k_{e^{-tD}}(x,x)\sim \sum _{j=0}^\infty t^{(j-n)/r}q_j(x)\), see Theorem 1. - (f)
The map \(\mu :C^\infty (M)\rightarrow \mathbb {C}\), \(\mu (f):=\tau (fD^{-\frac{n}{r}})=\frac{r}{\Gamma (n/r)}\int _Mf{{\,\mathrm{tr}\,}}_E(q_0)\) is a positive smooth measure on

*M*. - (g)
In particular, \(\mu (1)=\tau (D^{-\frac{n}{r}})=r\cdot a_0/\Gamma (\frac{n}{r})>0\), where \(a_0\) is the leading term in the heat trace expansion of

*D*, that is, \({{\,\mathrm{tr}\,}}(e^{-tD})\sim \sum _{j=0}^\infty t^{(j-n)/r}a_j\), see Corollary 1.

### Proof

*D*is assumed to be strictly positive. Hence, the holomorphic family \(A[B,D^{-z}]\) is of the form

*zR*(

*z*) for some holomorphic family

*R*(

*z*). As \({{\,\mathrm{tr}\,}}(R(z))\) can at most have a simple pole at \(z=0\), the function \({{\,\mathrm{tr}\,}}(A[B,D^{-z}])\) has no pole there and hence \(\tau ([A,B])=0\).

*A*is of order

*k*. Hence, \(AD^{-z/r}-A{\tilde{D}}^{-z/{\tilde{r}}}\) is a holomorphic family of order \(k-z\) which vanishes at \(z=0\). As before, we conclude that \({{\,\mathrm{tr}\,}}\bigl (AD^{-z/r}-A{\tilde{D}}^{-z/{\tilde{r}}}\bigr )\) is holomorphic at \(z=0\). Consequently,

To see (c), observe that for operators *A* of order less than \(-n\) the zeta function \(\zeta _A(z)\) is clearly analytic at \(z=0\) and hence has no residue there.

Part (d) follows immediately from the formula for the residue in (86).

The assertion about \(\mu (f)\) in (f) follows from (90) by specializing to the multiplication operator \(B=f\), where \(f\in C^\infty (M)\), and observing that \({{\,\mathrm{tr}\,}}_E(q_0(x))>0\) at each \(x\in M\), for we have \(q_0(x)>0\) according to Theorem 1. Hence, \(\mu \) is a positive functional.

The last statement about \(\tau (D^{-\frac{n}{r}})\) in (g) follows from (90) by specializing to \(A={{\,\mathrm{id}\,}}_E\) and using \(\int _M{{\,\mathrm{tr}\,}}_E(q_0)=a_0>0\), see Corollary 1. In particular, \(\tau \) is non-trivial. \(\square \)

### Remark 6

Using (84), we may identify the logarithmic term \(p_A=p_A|_M\) associated with a pseudodifferential operator \(A\in \Psi ^\infty (E)\), see Corollary 6(d), as an \({{\,\mathrm{end}\,}}(E)\) valued density on *M*. This Wodzicki density is intrinsic to *A*, i.e., it does not depend on the exponential coordinates used for the asymptotic expansion. The independence follows from the evident formula \(p_{BA}=Bp_A\) for all \(B\in \Gamma ^\infty ({{\,\mathrm{end}\,}}(E))\), the expression for the residue in Corollary 6(d), and the intrinsic definition of the residue using *D*.

In case of a trivially filtered manifold *M*, the trace \(\tau \) is the non-commutative residue which was introduced in Wodzicki and Guillemin [29, 58], with many applications including to index theory [18]. In Connes [17], the non-commutative residue is shown to coincide with Dixmier trace for classical pseudodifferential operators of order \(-n\). We expect such a relationship to hold also in Heisenberg calculus extending part (f) of Corollary 6.

## 8 Weyl’s Law for Rumin–Seshadri Operators

In this section, we will work out the eigenvalue asymptotics for Rumin–Seshadri operators associated with Rockland sequences of differential operators whose Heisenberg principle symbol sequence is the same at each point. More precisely, we will assume that the principle symbol sequence, the \(L^2\) inner product, and the volume density on any two osculating groups are simultaneously isometric. In this situation, the constant appearing in Weyl’s law is the product of the volume of the underlying manifold with a constant depending on the isometry class of the principal symbol sequence, see Corollary 7. Similar formulas for CR and contact geometry can be found in [43, Sect. 6].

In the second part of this section, we will specialize to a particular parabolic geometry in five dimensions associated with the exceptional Lie group \(G_2\). Every irreducible representation of \(G_2\) gives rise to a Rockland sequence of differential operators known as a (curved) BGG sequence [13]. Moreover, there is a distinguished class of fiberwise Hermitian inner products such that the Heisenberg principal symbol sequences, at any two points, are indeed isometric. In this situation, the constant in Weyl’s law is universal, depending only on the representation of \(G_2\), see Corollary 8.

### 8.1 Rumin–Seshadri Operators

*M*. Consider a sequence of differential operators

*M*. Moreover, let \(h_i\) be a fiberwise Hermitian metric on \(E_i\), and consider the associated standard \(L^2\)-inner product

### Definition 5

*Model sequence*)

- (a)A
*model sequence*consists of a finite dimensional graded nilpotent Lie algebra \(\mathfrak {n}\), a volume density \(\mu \in |\Lambda _\mathfrak {n}|\), finite dimensional Hermitian vector spaces \(F_i\), integers \(r_i\ge 1\), and \(B_i\in \mathcal {U}_{-r_i}(\mathfrak {n})\otimes \hom (F_i,F_{i+1})\). A model sequence gives rise to a sequence of left invariant homogeneous differential operators on the corresponding simply connected Lie group*N*,and provides left invariant \(L^2\) inner products on the spaces \(C^\infty (N,F_i)\) which are induced by the invariant volume density on$$\begin{aligned} \cdots \rightarrow C^\infty (N,F_{i-1})\xrightarrow {B_{i-1}}C^\infty (N,F_i)\xrightarrow {B_i}C^\infty (N,F_{i+1})\rightarrow \cdots , \end{aligned}$$*N*corresponding to \(\mu \) and the Hermitian inner products on \(F_i\), cf. (4). - (b)A model sequence is called
*Rockland*ifis exact for every non-trivial irreducible unitary representation \(\pi :N\rightarrow U(\mathcal {H})\) where \(\mathcal {H}_\infty \) denotes the subspace of smooth vectors.$$\begin{aligned} \cdots \rightarrow \mathcal {H}_\infty \otimes F_{i-1}\xrightarrow {\pi (B_{i-1})}\mathcal {H}_\infty \otimes F_i\xrightarrow {\pi (B_i)}\mathcal {H}_\infty \otimes F_{i+1}\rightarrow \cdots \end{aligned}$$ - (c)Two model sequences, \(\bigl (\mathfrak {n},\mu ,B_i\in \mathcal {U}_{-r_i}(\mathfrak {n})\otimes \hom (F_i,F_{i+1})\bigr )\) and \(\bigl (\mathfrak {n}',\mu ',B_i'\in \mathcal {U}_{-r_i'}(\mathfrak {n}')\otimes \hom (F_i',F_{i+1}')\bigr )\) are said to be
*isometric*if \(r_i=r_i'\), and there exists an isomorphism of graded Lie algebras \(\dot{\varphi }:\mathfrak {n}\rightarrow \mathfrak {n}'\) mapping \(\mu \) to \(\mu '\), and there exist unitary isomorphisms \(\phi _i:F_i\rightarrow F_i'\) such that the induced isomorphismsmap \(B_i\) to \(B_i'\), for all$$\begin{aligned} \dot{\varphi }\otimes \hom (\phi _i^{-1},\phi _{i+1}):\mathcal {U}_{-r_i}(\mathfrak {n})\otimes \hom (F_i,F_{i+1})\rightarrow \mathcal {U}_{-r_i'}(\mathfrak {n}')\otimes \hom (F_i',F_{i+1}') \end{aligned}$$*i*.

*M*equipped with a volume density, then the Heisenberg principal symbols of a sequence of differential operators \(A_i\in \mathcal {D}\mathcal {O}^{r_i}(E_i,E_{i+1})\) give rise to a model sequence

The following generalizes [43, Proposition 6.1.5].

### Corollary 7

*M*, let \(\mathrm{d}x\) be a volume density on

*M*, and suppose \(A_i\in \mathcal {D}\mathcal {O}^{r_i}(E_i,E_{i+1})\) is a sequence of differential operators such that the Heisenberg principal symbol sequence, see (95), is isometric to the model sequence at each point \(x\in M\). Moreover, fix positive integers \(s_i\) as in (93), and let \(\Delta _i\in \mathcal {D}\mathcal {O}^{2\kappa }(E_i)\) denote the associated Rumin–Seshadri operators, see (92) and (94). Then, as \(\lambda \rightarrow \infty \),

*n*denotes the homogeneous dimension of \(\mathfrak {n}\).

### Proof

### 8.2 Generic Rank-Two Distributions in Dimension Five

In the remaining part of this section, we will specialize to a particular geometry in five dimensions and discuss certain natural Rockland sequences over it, known as curved BGG sequences, which have been constructed by Čap, Slovák and Souček, see [13].

*M*be 5-manifold, and suppose \(H\subseteq TM\) is a smooth subbundle of rank two. Recall that

*H*is said to be of Cartan type [8] if it is bracket generating with growth vector (2, 3, 5). More precisely,

*H*is of Cartan type if every point in

*M*admits an open neighborhood

*U*and there exist sections \(X,Y\in \Gamma ^\infty (H|_U)\) such that

*X*,

*Y*, [

*X*,

*Y*], [

*X*, [

*X*,

*Y*]], [

*Y*, [

*X*,

*Y*]] is a frame of \(TM|_U\). Putting \(T^{-1}M:=H\) and denoting the rank-three bundle spanned by Lie brackets of sections of

*H*by \(T^{-2}M:=[H,H]\), the 5-manifold

*M*thus becomes a filtered manifold,

Rank-two distributions of Cartan type are also known as generic rank two distributions in dimension five, see [10, 31, 32, 48, 49], the condition on *H* being open with respect to the \(C^2\)-topology. Their history can be traced back to Cartan’s celebrated “five variables paper” [14]. Whether a 5-manifold admits a Cartan distribution is well understood in the open case, see [20, Theorem 2]. For closed 5-manifolds, however, this problem remains open and has served as a major motivation to develop the analysis in this paper.

*G*,

*P*) where

*G*denotes the split real form of the exceptional Lie group \(G_2\) and

*P*denotes a maximal parabolic subgroup corresponding to the shorter root, see [14, 48] and [11, Theorem 3.1.14 and Sect. 4.3.2]. The Lie algebra of

*G*admits a grading,

*i*and

*j*, such that

*P*is \(\mathfrak {p}=\mathfrak {g}_0\oplus \mathfrak {g}_1\oplus \mathfrak {g}_2\oplus \mathfrak {g}_3\). The corresponding Levi group,

*P*such that the inclusion \(G_0\subseteq P\) induces a natural isomorphism of groups, \(G_0=P/P_+\). The graded nilpotent Lie algebra \(\mathfrak {g}_-:=\mathfrak {g}_{-3}\oplus \mathfrak {g}_{-2}\oplus \mathfrak {g}_{-1}\) is isomorphic to \(\mathfrak {n}\).

The Cartan geometry consists of a principal *P*-bundle \(\mathcal {G}\rightarrow M\) and a regular Cartan connection \(\omega \in \Omega ^1(\mathcal {G};\mathfrak {g})\) satisfying a normalization condition, see [11]. The Cartan connection induces an isomorphism \(TM=\mathcal {G}\times _P\mathfrak {g}/\mathfrak {p}\). By regularity, the filtration on *TM*, see (99), coincides with the filtration induced by the filtration \(\mathfrak {g}/\mathfrak {p}=\mathfrak {g}^{-3}/\mathfrak {p}\supseteq \mathfrak {g}^{-2}/\mathfrak {p}\supseteq \mathfrak {g}^{-1}/\mathfrak {p}\supseteq \mathfrak {g}^0/\mathfrak {p}=0\). Note that \(P_+\) acts trivially on the associated graded \({{\,\mathrm{gr}\,}}(\mathfrak {g}/\mathfrak {p})\), and the inclusion \(\mathfrak {g}_-\subseteq \mathfrak {g}\) induces an isomorphism of \(G_0=P/P_+\) modules, \(\mathfrak {g}_-={{\,\mathrm{gr}\,}}(\mathfrak {g}/\mathfrak {p})\). Regularity also implies that the induced isomorphism \(\mathfrak tM={{\,\mathrm{gr}\,}}(TM)=\mathcal {G}\times _P{{\,\mathrm{gr}\,}}(\mathfrak {g}/\mathfrak {p})=\mathcal {G}_0\times _{G_0}\mathfrak {g}_-\) is an isomorphism of bundles of graded nilpotent Lie algebras. Here \(\mathcal {G}_0:=\mathcal {G}/P_+\) is considered as a principal \(G_0\)-bundle over *M*.

*M*known as curved BGG sequences, see [13] and [9, 12]. Every finite dimensional complex representation \(\rho :G\rightarrow {{\,\mathrm{GL}\,}}(\mathbb {E})\) gives rise to a sequence of natural differential operators,

*i*-th Lie algebra cohomology with coefficients in the representation of \(\mathfrak {g}_-\) obtained by restricting \(\rho \). In [21, Corollary 4.23 and Example 4.24] it has been shown that this is a Rockland sequence for every irreducible representation \(\rho \). We will denote the Heisenberg order of \(D_i\) by \(k_i\).

*i*, cf. [11, Proof of Proposition 3.3.1]. Denoting the Killing form of \(\mathfrak {g}\) by

*B*, we obtain a Euclidean inner product on \(\mathfrak {g}\), given by \(B_\theta (X,Y):=-B(X,\theta (Y))\) where \(X,Y\in \mathfrak {g}\). The summands in the decomposition (100) are orthogonal with respect to \(B_\theta \). Let \(\Theta \) denote the global Cartan involution on

*G*corresponding to \(\theta \). Then \(K:=\{g\in G:\Theta (g)=g\}\cong \bigl ({{\,\mathrm{SU}\,}}(2)\times {{\,\mathrm{SU}\,}}(2)\bigr )/\mathbb {Z}_2\) is a maximal compact subgroup of

*G*, and \(B_\theta \) is invariant under

*K*. Moreover, \(\Theta \) restricts to a Cartan involution on \(G_0\), and \(K_0:=\{g\in G_0:\Theta (g)=g\}\cong O(2,\mathbb {R})\) is a maximal compact subgroup of \(G_0\). Let

*h*be a Hermitian inner product on \(\mathbb {E}\) such that

*h*is invariant under the action of

*K*. We equip the \(G_0\)-module \(C^i(\mathfrak {g}_-;\mathbb {E})=\Lambda ^i\mathfrak {g}_-^*\otimes \mathbb {E}\) with the Hermitian inner product \(h_i\) induced by \(B_\theta \) and

*h*. Note that \(h_i\) is invariant under \(K_0\). These inner products play an important role in the construction of the curved BGG sequences since Kostant’s codifferential, \(\partial ^*:C^{i+1}(\mathfrak {g}_-;\mathbb {E})\rightarrow C^i(\mathfrak {g}_-;\mathbb {E})\), is adjoint to the Chevalley–Eilenberg codifferential \(\partial :C^i(\mathfrak {g}_-;\mathbb {E})\rightarrow C^{i+1}(\mathfrak {g}_-;\mathbb {E})\) with respect to \(h_i\), see [11, Proposition 3.3.1]. The induced \(K_0\)-invariant Hermitian inner product on \(H^i(\mathfrak {g}_-;\mathbb {E})\) will also be denoted by \(h_i\). We let \(\mu \in |\Lambda _{\mathfrak {g}_-}|\) denote the volume density associated with the Euclidean inner product on \(\mathfrak {g}_-\) induced by \(B_\theta \). Clearly, \(\mu \) is \(K_0\)-invariant.

*H*. It provides an isomorphism of principal \(G_0\)-bundles, \(\mathcal {G}_0=\mathcal {K}_0\times _{K_0}G_0\), and thus

*M*with the volume density \(\mathrm{d}x\) induced from the \(K_0\)-invariant volume density \(\mu \) on \(\mathfrak {g}_-\).

### Corollary 8

*M*be a closed 5-manifold equipped with a rank-two distribution of Cartan type. Moreover, let \(\rho \) be an irreducible complex representation of the exceptional Lie group \(G_2\), consider the associated curved BGG sequence (102), and let \(\Delta _i\) denote the corresponding Rumin–Seshadri operator, see (94), constructed using the volume density \(\mathrm{d}x\) on

*M*and the fiberwise Hermitian metrics on \(E_i\) described above. Then, as \(\lambda \rightarrow \infty \),

*M*or the distribution

*H*, nor does it depend on the Cartan involution \(\theta \) satisfying (103) or the Hermitian inner product

*h*satisfying (104) or the reduction of structure group of \(\mathcal {G}_0\) along \(K_0\subseteq G_0\), and it is also independent of the choice of positive integers \(s_i\) as in (93); it only depends on

*i*and the representation \(\rho \).

### Proof

*h*. To this end, suppose \(\tilde{\theta }\) is another Cartan involution on \(\mathfrak {g}\) such that \(\tilde{\theta }(\mathfrak {g}_i)=\mathfrak {g}_{-i}\), for all

*i*, cf. (103). Then there exists \(g\in G\) such that \(\tilde{\theta }={{\,\mathrm{Ad}\,}}_g^{-1}\circ \,\theta \circ {{\,\mathrm{Ad}\,}}_g={{\,\mathrm{Ad}\,}}_{g^{-1}\Theta (g)}\circ \,\theta \). In view of the Cartan decomposition, we may write \(g=k\exp (X)\) with \(k\in G\) and \(X\in \mathfrak {g}\) such that \(\Theta (k)=k\) and \(\theta (X)=-X\). Hence, \(g^{-1}\Theta (g)=\exp (-2X)\), and we obtain \(\tilde{\theta }={{\,\mathrm{Ad}\,}}_{\exp (-2X)}\circ \,\theta \). Since \((\tilde{\theta }\theta ^{-1})(\mathfrak {g}_i)=\mathfrak {g}_i\), we have \(\exp (-2X)\in G_0\), see (101). Using \(\theta (X)=-X\), we actually conclude \(X\in \mathfrak {g}_0\), hence \(g_0:=\exp (X)\in G_0\), and

*C*. Similarly, using (110), we see that the induced volume densities on

*M*coincide. This implies that the corresponding \(L_2\)-inner products on \(\Gamma (E_i)\) differ by a constant which is independent of

*i*, hence they give rise to the same formally adjoint operators.

Summarizing, we see that the Rumin–Seshadri operators associated with \(\tilde{\theta }\), \({\tilde{h}}\), and a reduction \(\tilde{\mathcal {K}_0}\rightarrow \mathcal {G}_0\) coincide with the Rumin–Seshadri operators associated with \(\theta \), *h*, and the reduction \(\mathcal {K}_0\rightarrow \mathcal {G}_0\) described in the previous paragraph. We conclude that the constant \(\alpha _i(\rho )\) does not depend on \(\theta \) or *h*. \(\square \)

## Footnotes

- 1.
We could use \(\Lambda =A\) and a left parametrix \(\Lambda '=B\) as above. Alternatively, we may assume \(R=0\), see [21, Lemma 3.16].

- 2.
Note that there are no non-trivial formally self-adjoint and non-negative differential operators of odd Heisenberg order.

- 3.
\(\mathcal {H}\) has to be one-dimensional in this case.

- 4.

## Notes

### Acknowledgements

Open access funding provided by Austrian Science Fund (FWF). S. D. was supported by the Austrian Science Fund (FWF): Project Number P30233. The second author gratefully acknowledges the support of the Austrian Science Fund (FWF): Project Number Y963-N35, the START-Program of Michael Eichmair.

## References

- 1.Agrachev, A., Boscain, U., Gauthier, J.-P., Rossi, F.: The intrinsic hypoelliptic Laplacian and its heat kernel on unimodular Lie groups. J. Funct. Anal.
**256**(8), 2621–2655 (2009)MathSciNetzbMATHGoogle Scholar - 2.Atiyah, M.F.: Global theory of elliptic operators. In: Proceedings of Internat. Conf. on Functional Analysis and Related Topics (Tokyo, 1969), pp. 21–30. University of Tokyo Press, Tokyo (1970)Google Scholar
- 3.Beals, R., Greiner, P.: Calculus on Heisenberg Manifolds. Annals of Mathematics Studies, vol. 119. Princeton University Press, Princeton (1988)zbMATHGoogle Scholar
- 4.Beals, R., Greiner, P.C., Stanton, N.K.: The heat equation on a CR manifold. J. Differ. Geom.
**20**(2), 343–387 (1984)MathSciNetzbMATHGoogle Scholar - 5.Beals, R., Gaveau, B., Greiner, P.C.: Hamilton-Jacobi theory and the heat kernel on Heisenberg groups. J. Math. Pures Appl. (9)
**79**(7), 633–689 (2000)MathSciNetzbMATHGoogle Scholar - 6.Blackadar, B.: \(K\)-Theory for Operator Algebras. Mathematical Sciences Research Institute Publications, vol. 5, 2nd edn. Cambridge University Press, Cambridge (1998)zbMATHGoogle Scholar
- 7.Boskèĭn, U., Got’e, Z.-P., Rossi, F.: The hypoelliptic heat kernel over three-step nilpotent Lie groups. Sovrem. Mat. Fundam. Napravl.
**42**, 48–61 (2011). Translation in J. Math. Sci.**199**, 614–628 (2014)Google Scholar - 8.Bryant, R.L., Hsu, L.: Rigidity of integral curves of rank \(2\) distributions. Invent. Math.
**114**(2), 435–461 (1993)MathSciNetzbMATHGoogle Scholar - 9.Calderbank, D.M.J., Diemer, T.: Differential invariants and curved Bernstein-Gelfand-Gelfand sequences. J. Reine Angew. Math.
**537**, 67–103 (2001)MathSciNetzbMATHGoogle Scholar - 10.Čap, A., Sagerschnig, K.: On Nurowski’s conformal structure associated to a generic rank two distribution in dimension five. J. Geom. Phys.
**59**(7), 901–912 (2009)MathSciNetzbMATHGoogle Scholar - 11.Čap, A., Slovák, J.: Parabolic geometries. I. Mathematical Surveys and Monographs, vol. 154. American Mathematical Society, Providence, RI (2009) (Background and general theory)Google Scholar
- 12.Čap, A., Souček, V.: Subcomplexes in curved BGG-sequences. Math. Ann.
**354**(1), 111–136 (2012)MathSciNetzbMATHGoogle Scholar - 13.Čap, A., Slovák, J., Souček, V.: Bernstein-Gelfand-Gelfand sequences. Ann. Math. (2)
**154**(1), 97–113 (2001)MathSciNetzbMATHGoogle Scholar - 14.Cartan, E.: Les systèmes de Pfaff, à cinq variables et les équations aux dérivées partielles du second ordre. Ann. Sci. École Norm. Sup.
**3**(27), 109–192 (1910)zbMATHGoogle Scholar - 15.Choi, W., Ponge, R.: Tangent maps and tangent groupoid for Carnot manifolds. Differ. Geom. Appl.
**62**, 136–183 (2019). To appearMathSciNetGoogle Scholar - 16.Christ, M., Geller, D., Głowacki, P., Polin, L.: Pseudodifferential operators on groups with dilations. Duke Math. J.
**68**(1), 31–65 (1992)MathSciNetzbMATHGoogle Scholar - 17.Connes, A.: The action functional in noncommutative geometry. Commun. Math. Phys.
**117**(4), 673–683 (1988)MathSciNetzbMATHGoogle Scholar - 18.Connes, A., Moscovici, H.: The local index formula in noncommutative geometry. Geom. Funct. Anal.
**5**(2), 174–243 (1995)MathSciNetzbMATHGoogle Scholar - 19.Cygan, J.: Heat kernels for class \(2\) nilpotent groups. Studia Math.
**64**(3), 227–238 (1979)MathSciNetzbMATHGoogle Scholar - 20.Dave, S., Haller, S.: On 5-manifolds admitting rank two distributions of Cartan type. Trans. Am. Math. Soc.
**371**(7), 4911–4929 (2019)Google Scholar - 21.Dave, S., Haller, S.: Graded hypoellipticity of BGG sequences (2017). arXiv:1705.01659
- 22.de Monvel, L.B.: On the index of Toeplitz operators of several complex variables. Invent. Math.
**50**(3), 249–272 (1978)MathSciNetzbMATHGoogle Scholar - 23.Debord, C., Skandalis, G.: Adiabatic groupoid, crossed product by \(\mathbb{R}_+^\ast \) and pseudodifferential calculus. Adv. Math.
**257**, 66–91 (2014)MathSciNetzbMATHGoogle Scholar - 24.Dixmier, J.: Sur les représentations unitaries des groupes de Lie nilpotents. III. Canad. J. Math.
**10**, 321–348 (1958)MathSciNetzbMATHGoogle Scholar - 25.Folland, G.B., Stein, E.M.: Estimates for the \(\bar{\partial }_{b}\) complex and analysis on the Heisenberg group. Commun. Pure Appl. Math.
**27**, 429–522 (1974)MathSciNetzbMATHGoogle Scholar - 26.Folland, G.B., Stein, E.M.: Hardy Spaces on Homogeneous Groups. Mathematical Notes, vol. 28. Princeton University Press, Princeton (1982)zbMATHGoogle Scholar
- 27.Gaveau, B.: Principe de moindre action, propagation de la chaleur et estimées sous elliptiques sur certains groupes nilpotents. Acta Math.
**139**(1–2), 95–153 (1977)MathSciNetzbMATHGoogle Scholar - 28.Gromov, M.: Carnot-Carathéodory spaces seen from within. In: Sub-Riemannian Geometry of Progress in Mathematics, vol. 144, pp. 79–323. Birkhäuser, Basel (1996)Google Scholar
- 29.Guillemin, V.: A new proof of Weyl’s formula on the asymptotic distribution of eigenvalues. Adv. Math.
**55**(2), 131–160 (1985)MathSciNetzbMATHGoogle Scholar - 30.Haj Saeedi Sadegh, A.R., Higson, N.: Euler-like vector fields, deformation spaces and manifolds with filtered structure. Doc. Math.
**23**, 293–325 (2018)MathSciNetzbMATHGoogle Scholar - 31.Hammerl, M., Sagerschnig, K.: Conformal structures associated to generic rank 2 distributions on 5-manifolds—characterization and Killing-field decomposition. In: SIGMA Symmetry Integrability and Geometry Methods and Applications, vol. 5, Paper 081, 29 (2009)Google Scholar
- 32.Hammerl, M., Sagerschnig, K.: The twistor spinors of generic 2- and 3-distributions. Ann. Global Anal. Geom.
**39**(4), 403–425 (2011)MathSciNetzbMATHGoogle Scholar - 33.Hardy, G.H.: Divergent Series. Clarendon Press, Oxford (1949)zbMATHGoogle Scholar
- 34.Helffer, B., Nourrigat, F.: Hypoellipticité pour des groupes nilpotents de rang de nilpotence \(3\). Commun. Partial Differ. Equ.
**3**(8), 643–743 (1978)zbMATHGoogle Scholar - 35.Helffer, B., Nourrigat, J.: Caracterisation des opérateurs hypoelliptiques homogènes invariants à gauche sur un groupe de Lie nilpotent gradué. Commun. Partial Differ. Equ.
**4**(8), 899–958 (1979)zbMATHGoogle Scholar - 36.Kasparov, G.G.: Topological invariants of elliptic operators. I. \(K\)-homology. Izv. Akad. Nauk SSSR Ser. Mat.
**39**((4), 796–838 (1975)MathSciNetzbMATHGoogle Scholar - 37.Kato, T.: Perturbation Theory for Linear Operators. Classics in Mathematics. Springer, Berlin (1995) (Reprint of the 1980 edition)Google Scholar
- 38.Kirillov, A.A.: Lectures on the Orbit Method. Graduate Studies in Mathematics, vol. 64. American Mathematical Society, Providence (2004)zbMATHGoogle Scholar
- 39.Morimoto, T.: Geometric structures on filtered manifolds. Hokkaido Math. J.
**22**(3), 263–347 (1993)MathSciNetzbMATHGoogle Scholar - 40.Morimoto, T.: Lie algebras, geometric structures and differential equations on filtered manifolds. In: Lie groups, geometric structures and differential equations—one hundred years after Sophus Lie (Kyoto/Nara, 1999), vol. 37 , Adv. Stud. Pure Math., pp. 205–252. Math. Soc. Japan, Tokyo (2002). http://www.mathbooks.org/aspm/
- 41.Neusser, K.: Weighted jet bundles and differential operators for parabolic geometries. PhD thesis, University of Vienna, Vienna (2010). http://othes.univie.ac.at/9768/
- 42.Ponge, R.: Calcul fonctionnel sous-elliptique et résidu non commutatif sur les variétés de Heisenberg. C. R. Acad. Sci. Paris Sér. I Math.
**332**(7), 611–614 (2001)MathSciNetzbMATHGoogle Scholar - 43.Ponge, R.S.: Heisenberg calculus and spectral theory of hypoelliptic operators on Heisenberg manifolds. Mem. Am. Math. Soc.
**194**(906), viii+ 134 (2008)Google Scholar - 44.Rockland, C.: Hypoellipticity on the Heisenberg group-representation-theoretic criteria. Trans. Am. Math. Soc.
**240**, 1–52 (1978)MathSciNetzbMATHGoogle Scholar - 45.Roe, J.: Index theory, coarse geometry, and topology of manifolds, volume 90 of CBMS Regional Conference Series in Mathematics. Published for the Conference Board of the Mathematical Sciences, Washington, DC; by the American Mathematical Society, Providence, RI (1996)Google Scholar
- 46.Rothschild, L.P., Stein, E.M.: Hypoelliptic differential operators and nilpotent groups. Acta Math.
**137**(3–4), 247–320 (1976)MathSciNetzbMATHGoogle Scholar - 47.Rumin, M., Seshadri, N.: Analytic torsions on contact manifolds. Ann. Inst. Fourier (Grenoble)
**62**(2), 727–782 (2012)MathSciNetzbMATHGoogle Scholar - 48.Sagerschnig, K.: Weyl structures for generic rank two distributions in dimension five. PhD thesis, University of Vienna, Vienna, Austria (2008). http://othes.univie.ac.at/2186/
- 49.Sagerschnig, K., Willse, T.: The geometry of almost Einstein \((2,3,5)\) distributions. In: SIGMA Symmetry Integrability and Geometry Methods and Applications, vol. 13, Paper No. 004, 56 (2017)Google Scholar
- 50.Seeley, R.T.: Complex powers of an elliptic operator. In: Singular Integrals (Proc. Sympos. Pure Math., Chicago, Ill., 1966), pp. 288–307. American Mathematical Society, Providence (1967)Google Scholar
- 51.Shubin, M.A.: Pseudodifferential operators and spectral theory, 2nd edn. Springer, Berlin (2001) (Translated from the 1978 Russian original by Stig I. Andersson)Google Scholar
- 52.Taylor, M.E.: Noncommutative microlocal analysis. I. Mem. Am. Math. Soc.
**52**(313), iv+182 (1984)MathSciNetzbMATHGoogle Scholar - 53.Trèves, F.: Topological Vector Spaces, Distributions and Kernels. Academic Press, New York (1967)zbMATHGoogle Scholar
- 54.van Erp, E.: The Atiyah-Singer index formula for subelliptic operators on contact manifolds. Part I. Ann. Math. (2)
**171**(3), 1647–1681 (2010)MathSciNetzbMATHGoogle Scholar - 55.van Erp, E.: The Atiyah-Singer index formula for subelliptic operators on contact manifolds. Part II. Ann. Math. (2)
**171**(3), 1683–1706 (2010)MathSciNetzbMATHGoogle Scholar - 56.van Erp, E., Yuncken, R.: A groupoid approach to pseudodifferential calculi. J. Reine Angew. Math. https://doi.org/10.1515/crelle-2017-0035
- 57.van Erp, E., Yuncken, R.: On the tangent groupoid of a filtered manifold. Bull. Lond. Math. Soc.
**49**, 1000–1012 (2017)MathSciNetzbMATHGoogle Scholar - 58.Wodzicki, M.: Noncommutative residue. I. Fundamentals. In: \(K\)-Theory, Arithmetic and Geometry (Moscow, 1984–1986), Lecture Notes in Mathematics, vol. 1289, pp. 320–399. Springer, Berlin (1987)Google Scholar

## Copyright information

**OpenAccess**This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.