Rumin’s Differential Forms

• Rumin’s differential forms are a reorganization of the classical de Rham complex adapted to the anisotropic structure of contact and Carnot manifolds.
• They enable sharp hypoelliptic analysis, Hodge theory refinements, and Poincaré–Sobolev inequalities by preserving underlying scaling properties.
• Their construction on models like the Heisenberg group extends cohomological and analytic tools to sub-Riemannian geometry.

Searching arXiv for recent and foundational papers on the Rumin complex, Heisenberg groups, and related analytic developments. Rumin’s differential forms are differential forms organized into a complex adapted to contact, Carnot, and related sub-Riemannian geometries, where the ordinary de Rham complex fails to respect the underlying anisotropic filtration. In this setting, the basic object is the Rumin complex, denoted in the Heisenberg case by $(E_0^\bullet,d_c)$ and more generally by $(E^\bullet,d_c)$ or $(R^\bullet,d)$ depending on the formulation. Its defining feature is that the differential $d_c$ preserves the contact or Carnot grading, is homogeneous under Carnot dilations, is first order away from a middle degree, and becomes second order at that middle degree. The complex is homotopy-equivalent to the de Rham complex and therefore computes the same cohomology, while being better suited to hypoelliptic analysis, Hodge theory, CR geometry, and sharp Poincaré–Sobolev estimates (Pansu, 2016, Baldi et al., 2017, Kitaoka, 2022).

1. Geometric motivation and failure of the de Rham complex

A Carnot, or sub-Riemannian, manifold is a smooth manifold $M$ endowed with a bracket-generating subbundle

$$H = H^1 \subset H^2 \subset \cdots \subset H^r = TM,$$

where $H^{k+1}=H^k+[H,H^k]$. The Carnot–Carathéodory distance is defined using curves tangent to $H$. Under the anisotropic dilations of the tangent cone, vectors in higher layers $H^k$ scale like $\epsilon^k$, whereas ordinary differential forms all scale like $(E^\bullet,d_c)$0. As a consequence, the de Rham differential does not separate interactions between layers, the Hodge–de Rham Laplacian does not reflect the sub-Riemannian homogeneities, and geometric inequalities become awkward in the full de Rham complex (Pansu, 2016).

Rumin’s idea is to replace the de Rham complex by a smaller, homotopy-equivalent complex adapted to the filtration. In the Heisenberg group $(E^\bullet,d_c)$1, this adaptation is especially transparent because the Lie algebra is stratified as $(E^\bullet,d_c)$2, with $(E^\bullet,d_c)$3 and $(E^\bullet,d_c)$4. The dilations act by weight $(E^\bullet,d_c)$5 on $(E^\bullet,d_c)$6 and weight $(E^\bullet,d_c)$7 on $(E^\bullet,d_c)$8, so the usual de Rham complex is not homogeneous under these dilations, whereas Rumin’s complex is left-invariant and homogeneous (Baldi et al., 2021, Baldi et al., 2024).

On contact manifolds, the same issue appears through the splitting induced by a contact form $(E^\bullet,d_c)$9. If $(R^\bullet,d)$0 and $(R^\bullet,d)$1 is the Reeb vector field determined by

$(R^\bullet,d)$2

then horizontal forms and vertical $(R^\bullet,d)$3-components behave differently. Rumin’s construction isolates the forms and differentials compatible with that structure (Kitaoka, 2022).

2. Construction of the complex

On an equiregular Carnot manifold, one considers the associated graded bundle

$(R^\bullet,d)$4

whose dual induces a weight decomposition on forms. The exterior differential preserves the weight filtration and therefore induces an algebraic operator

$(R^\bullet,d)$5

which on the graded model is the Lie algebra cohomology differential. Under an equihomological condition, one chooses smooth complements, defines $(R^\bullet,d)$6 on $(R^\bullet,d)$7, and introduces the retraction

$(R^\bullet,d)$8

Iterating $(R^\bullet,d)$9 yields a projection $d_c$0 onto a subcomplex $d_c$1, and the Rumin differential is

$d_c$2

The resulting complex is homotopy-equivalent to the de Rham complex (Pansu, 2016).

In the contact case of dimension $d_c$3, one introduces

$d_c$4

For $d_c$5, the complex is modeled on $d_c$6; for $d_c$7, on $d_c$8. In the middle degree $d_c$9, one introduces a second-order operator

$M$0

defined by extending a representative $M$1 to $M$2 so that its $M$3-image vanishes on $M$4, and then taking that image. This produces the original Rumin complex in the contact setting (Pansu, 2016).

A closely related formulation on a contact manifold $M$5 uses the bundles

$M$6

where $M$7 and $M$8. For $M$9, $$H = H^1 \subset H^2 \subset \cdots \subset H^r = TM,$$0 is the horizontal projection of $$H = H^1 \subset H^2 \subset \cdots \subset H^r = TM,$$1; for $$H = H^1 \subset H^2 \subset \cdots \subset H^r = TM,$$2, it becomes a second-order operator $$H = H^1 \subset H^2 \subset \cdots \subset H^r = TM,$$3; and for $$H = H^1 \subset H^2 \subset \cdots \subset H^r = TM,$$4, it is the quotient differential induced by $$H = H^1 \subset H^2 \subset \cdots \subset H^r = TM,$$5. Endowed with $$H = H^1 \subset H^2 \subset \cdots \subset H^r = TM,$$6, the sequence

$$H = H^1 \subset H^2 \subset \cdots \subset H^r = TM,$$7

is the Rumin complex, and its cohomology is canonically isomorphic to de Rham cohomology (Kitaoka, 2022).

3. Heisenberg group model and explicit formulas

The Heisenberg group $$H = H^1 \subset H^2 \subset \cdots \subset H^r = TM,$$8 may be identified with $$H = H^1 \subset H^2 \subset \cdots \subset H^r = TM,$$9 with coordinates $H^{k+1}=H^k+[H,H^k]$0 and group law

$H^{k+1}=H^k+[H,H^k]$1

A standard left-invariant frame is

$H^{k+1}=H^k+[H,H^k]$2

and the contact form is

$H^{k+1}=H^k+[H,H^k]$3

The horizontal bundle is $H^{k+1}=H^k+[H,H^k]$4 (Baldi et al., 2017).

In this model, the Rumin bundles $H^{k+1}=H^k+[H,H^k]$5 are left-invariant subbundles of $H^{k+1}=H^k+[H,H^k]$6. One description is

$H^{k+1}=H^k+[H,H^k]$7

where $H^{k+1}=H^k+[H,H^k]$8 is the algebraic map obtained by wedging with $H^{k+1}=H^k+[H,H^k]$9. For $H$0, $H$1 consists of primitive horizontal $H$2-covectors; for $H$3, it is identified with $H$4 (Baldi et al., 2021). Another contact formulation gives

$H$5

There is then a projection $H$6 and

$H$7

(Baldi et al., 2017).

The differential has a degree-dependent order. If $H$8, then $H$9 is a homogeneous horizontal differential operator of order $H^k$0; if $H^k$1, then it is of horizontal order $H^k$2 (Baldi et al., 2021, Baldi et al., 2024). In local frames, if $H^k$3, then for $H^k$4,

$H^k$5

while for $H^k$6,

$H^k$7

with $H^k$8 (Baldi et al., 2021).

For $H^k$9, one also has the concrete formula that $\epsilon^k$0 is the horizontal differential: $\epsilon^k$1 or, in components,

$\epsilon^k$2

At the middle degree $\epsilon^k$3, one instead extracts the vertical part of $\epsilon^k$4 and obtains a genuine second-order horizontal operator (Baldi et al., 2017).

For $\epsilon^k$5, the contact example can be written particularly explicitly. On functions, $\epsilon^k$6 is the horizontal differential $\epsilon^k$7. On a horizontal $\epsilon^k$8-form $\epsilon^k$9, one defines

$(E^\bullet,d_c)$00

hence

$(E^\bullet,d_c)$01

For $(E^\bullet,d_c)$02, $(E^\bullet,d_c)$03 coincides with the restriction of the ordinary $(E^\bullet,d_c)$04 to $(E^\bullet,d_c)$05 (Pansu, 2016).

4. Cohomology, Hodge theory, and Laplacians

A central property is that the Rumin complex is homotopy-equivalent to the de Rham complex. Therefore

$(E^\bullet,d_c)$06

on Carnot manifolds (Pansu, 2016), and in the Heisenberg case

$(E^\bullet,d_c)$07

(Baldi et al., 2021). The Poincaré lemma holds on small charts: on a contractible Carnot ball the Rumin complex is exact except in degree zero (Pansu, 2016).

The formal adjoint is defined using an $(E^\bullet,d_c)$08 structure. In one formulation, if the weight splitting is orthogonal, then

$(E^\bullet,d_c)$09

up to the sign conventions stated in the source (Pansu, 2016). In the Sasakian contact setting, for $(E^\bullet,d_c)$10,

$(E^\bullet,d_c)$11

while in the middle degree one takes the adjoint $(E^\bullet,d_c)$12 of $(E^\bullet,d_c)$13 (Kitaoka, 2022). On the Heisenberg group, one likewise has

$(E^\bullet,d_c)$14

(Baldi et al., 2017).

The associated Laplacian is adapted to the degree. On a contact manifold,

$(E^\bullet,d_c)$15

while on the middle two degrees,

$(E^\bullet,d_c)$16

Thus the Rumin Laplacian is second order away from the middle degrees and fourth order at the middle degrees (Kitaoka, 2022). In the CR formulation, the Rumin Laplacian is written as

$(E^\bullet,d_c)$17

when $(E^\bullet,d_c)$18, and it is fourth order in degrees $(E^\bullet,d_c)$19 and $(E^\bullet,d_c)$20. It is nonnegative, self-adjoint, and maximally hypoelliptic, with partial inverse $(E^\bullet,d_c)$21 and projection $(E^\bullet,d_c)$22 onto $(E^\bullet,d_c)$23, giving the decomposition

$(E^\bullet,d_c)$24

(Case, 2021).

On compact Sasakian manifolds, the harmonic theory of the Rumin complex agrees with the ordinary Hodge theory: if $(E^\bullet,d_c)$25 is the Hodge Laplacian of the Sasakian metric, then

$(E^\bullet,d_c)$26

An immediate corollary is the primitiveness of Sasakian harmonic forms (Kitaoka, 2022). The same work also describes the adiabatic family

$(E^\bullet,d_c)$27

for which the low-lying spectrum and harmonic forms of $(E^\bullet,d_c)$28 converge to those of $(E^\bullet,d_c)$29, so the Rumin complex appears as the sub-Riemannian limit of de Rham theory (Kitaoka, 2022).

5. Analysis on the complex: homotopy, kernels, and reproducing formulas

A major advantage of the complex is that one can develop sub-Riemannian Hodge and heat-kernel theory directly on Rumin forms (Pansu, 2016). On the Heisenberg group, one inverts the Rumin Laplacian by convolution with homogeneous kernels and obtains homotopy operators. In one formulation, if $(E^\bullet,d_c)$30, there exist convolution operators $(E^\bullet,d_c)$31 with kernels homogeneous of degree $(E^\bullet,d_c)$32 or $(E^\bullet,d_c)$33 such that

$(E^\bullet,d_c)$34

The choice between type $(E^\bullet,d_c)$35 and type $(E^\bullet,d_c)$36 depends on whether the degree is away from or at the middle degrees (Baldi et al., 2017).

A related $(E^\bullet,d_c)$37 homotopy statement asserts that on all of $(E^\bullet,d_c)$38, any compactly supported $(E^\bullet,d_c)$39 satisfies

$(E^\bullet,d_c)$40

where $(E^\bullet,d_c)$41 is convolution by a kernel of homogeneous type $(E^\bullet,d_c)$42 or $(E^\bullet,d_c)$43 (Baldi et al., 2021). There are also local operators

$(E^\bullet,d_c)$44

such that

$(E^\bullet,d_c)$45

on $(E^\bullet,d_c)$46, with $(E^\bullet,d_c)$47 bounded $(E^\bullet,d_c)$48 and $(E^\bullet,d_c)$49 gaining arbitrary smoothness (Baldi et al., 2021). On bounded-geometry contact manifolds, these local constructions can be transported by contact charts and patched to obtain a global homotopy operator and a Hodge-type decomposition (Baldi et al., 2017).

The heat-kernel approach gives a more explicitly spectral reproducing formula. On the Heisenberg group, for the Rumin Laplacian

$(E^\bullet,d_c)$50

on $(E^\bullet,d_c)$51, let $(E^\bullet,d_c)$52 be the heat semigroup with kernel $(E^\bullet,d_c)$53. If $(E^\bullet,d_c)$54 satisfies $(E^\bullet,d_c)$55, define

$(E^\bullet,d_c)$56

Then

$(E^\bullet,d_c)$57

as an identity of currents. Equivalently,

$(E^\bullet,d_c)$58

which recovers the usual Calderón reproducing formula with $(E^\bullet,d_c)$59 and normalization $(E^\bullet,d_c)$60 (Ciatti et al., 2023). The proof uses duality against test forms, integrability of the heat-kernel term, and the identity

$(E^\bullet,d_c)$61

so convolution by $(E^\bullet,d_c)$62 is a two-sided inverse of $(E^\bullet,d_c)$63 (Ciatti et al., 2023).

6. Inequalities, regularity, and endpoint estimates

The Rumin complex supports sharp Poincaré and Sobolev inequalities that reflect the sub-Riemannian grading. On a Heisenberg ball, the interior $(E^\bullet,d_c)$64-Poincaré inequality for Rumin $(E^\bullet,d_c)$65-forms asserts that for every closed $(E^\bullet,d_c)$66 there exists $(E^\bullet,d_c)$67 with $(E^\bullet,d_c)$68 and

$(E^\bullet,d_c)$69

The sharp scaling condition is

$(E^\bullet,d_c)$70

Under the equality condition one has global strong inequalities on all of $(E^\bullet,d_c)$71, while the non-strict inequality governs interior estimates on Korányi balls (Baldi et al., 2017).

The endpoint case $(E^\bullet,d_c)$72 is especially notable. For $(E^\bullet,d_c)$73 and degrees $(E^\bullet,d_c)$74,

• if $(E^\bullet,d_c)$75, then interior $(E^\bullet,d_c)$76 holds;
• if $(E^\bullet,d_c)$77, then interior $(E^\bullet,d_c)$78 holds.

The same exponents give the corresponding interior Sobolev inequalities (Baldi et al., 2021). In particular, for $(E^\bullet,d_c)$79,

$(E^\bullet,d_c)$80

and similarly for $(E^\bullet,d_c)$81 with $(E^\bullet,d_c)$82 (Baldi et al., 2021). By contrast, in degree $(E^\bullet,d_c)$83 the endpoint $(E^\bullet,d_c)$84 fails on $(E^\bullet,d_c)$85 (Baldi et al., 2021).

A global continuous-primitive theorem sharpens this picture. If $(E^\bullet,d_c)$86 and $(E^\bullet,d_c)$87, every $(E^\bullet,d_c)$88-exact $(E^\bullet,d_c)$89 admits $(E^\bullet,d_c)$90 such that

$(E^\bullet,d_c)$91

If $(E^\bullet,d_c)$92, every $(E^\bullet,d_c)$93-exact $(E^\bullet,d_c)$94 admits $(E^\bullet,d_c)$95 with

$(E^\bullet,d_c)$96

Analogous local and compact-manifold versions hold in the contact sub-Riemannian setting (Baldi et al., 2024).

These estimates are proved by combining functional analysis, right inverses for $(E^\bullet,d_c)$97, kernel representations of $(E^\bullet,d_c)$98, sub-Riemannian Hardy–Littlewood–Sobolev estimates, and degree-sensitive arguments in the middle degree (Baldi et al., 2024). This suggests that the analytic strength of the Rumin complex is not merely cohomological; it is tightly linked to the scaling of kernels and the order jump at the middle degree.

7. CR and Sasakian extensions

Beyond contact and Heisenberg settings, the Rumin complex admits CR-invariant and bigraded refinements. On a strictly pseudoconvex CR manifold of dimension $(E^\bullet,d_c)$99, one has bundles $(R^\bullet,d)$00 on the contact distribution and a decomposition

$(R^\bullet,d)$01

The Rumin sheaves $(R^\bullet,d)$02 are defined by contact conditions involving $(R^\bullet,d)$03, $(R^\bullet,d)$04, and $(R^\bullet,d)$05, with the differential preserving the sheaves. There is a CR-invariant projection

$(R^\bullet,d)$06

and a bigraded version $(R^\bullet,d)$07 in which $(R^\bullet,d)$08 splits as $(R^\bullet,d)$09 away from the middle degree, and as $(R^\bullet,d)$10 at the middle degree (Case, 2021).

This bigraded complex carries a balanced $(R^\bullet,d)$11-structure. On the Rumin complex one sets

$(R^\bullet,d)$12

with $(R^\bullet,d)$13 for $(R^\bullet,d)$14. The $(R^\bullet,d)$15 identities hold, and the shuffle-symmetrizations kill $(R^\bullet,d)$16 and $(R^\bullet,d)$17, so the structure is balanced (Case, 2021).

The CR and Sasakian frameworks also lead to Hodge decomposition theorems, Serre–Poincaré duality, Frölicher-type inequalities, Kohn–Rossi dimension bounds, and topological applications such as cup-length bounds, Chern-class vanishing statements, Hard Lefschetz isomorphisms, $(R^\bullet,d)$18-type lemmas, and the Lee class formalism for pseudo-Einstein contact forms (Case, 2021). In Sasakian geometry, the analytic torsion function of the Rumin complex is expressed in terms of the Reeb flow. If $(R^\bullet,d)$19 is the nonzero spectrum of $(R^\bullet,d)$20 on $(R^\bullet,d)$21, then

$(R^\bullet,d)$22

and under the compact Sasakian assumptions described in the source, $(R^\bullet,d)$23 is expressed through spectral zeta functions of operators built from $(R^\bullet,d)$24 (Kitaoka, 2022).

Taken together, these developments place Rumin’s differential forms at the intersection of sub-Riemannian geometry, hypoelliptic analysis, CR geometry, and nonlinear functional inequalities. Their defining principle is structural adaptation: the complex discards the components that are invisible or poorly scaled for the contact or Carnot filtration, while preserving cohomology and sharpening analysis.