Cohomology of Mapping Class Groups

Updated 30 December 2025

Cohomology of MCGs is the study of algebraic invariants derived from mapping class groups that capture global topological properties of surfaces.
It employs methods like spectral sequences, graph homology, and combinatorial models to compute Poincaré series and Betti numbers across various surface types.
The approach has practical implications in understanding moduli spaces, classifying topological phases, and analyzing subgroup structures such as Torelli and spin-hyperelliptic groups.

The cohomology of mapping class groups (MCGs) is a central topic in algebraic and geometric topology, connecting the structure of diffeomorphism groups of surfaces with graph complexes, representation theory, and configuration spaces. MCGs—groups of isotopy classes of diffeomorphisms—encode the global topological and geometric properties of surfaces, with applications ranging from moduli spaces of curves to the classification of topological phases in condensed matter physics. Cohomological investigations cover a range of variants: orientable and non-orientable surfaces, punctured surfaces, and subgroups such as the Torelli group and spin-hyperelliptic mapping class groups. The theory leverages methods from spectral sequences, combinatorics of partitions, graph homology, and representation theory.

1. Cohomological Frameworks: Classical, Configuration Spaces, and Graph Complexes

Classical presentations of the MCG cohomology for orientable surfaces focus on the stable range, leveraging homological stability and fiber bundle decompositions. Harer's stability theorem and the Madsen–Weiss proof of Mumford's conjecture provide the stable cohomology as a polynomial algebra, with the Poincaré polynomial realized as the plethystic exponential of $1/(1-t^2)$ . For the closed genus- $g$ surface $S_g$ , the (oriented) mapping class group $\mathrm{Mod}_g$ is known to act as a virtual duality group, with virtual cohomological dimension $4g - 5$ and dualizing module given by the integral homology of the curve complex $\mathcal{C}_g$ (Church et al., 2011).

Configuration space methods encode the cohomology of mapping class groups of punctured and non-orientable surfaces in terms of homology of (punctured) configuration spaces and associated braid groups. For the punctured projective plane, the mod-2 cohomology of the mapping class group is expressed as the tensor product of the cohomology of $BSO(3)$ and the unordered configuration space $F_k(\mathbb{R}P^2)/\Sigma_k$ , with the latter admitting an additive description in terms of classical braid group homology (Maldonado et al., 2017).

Graph complex approaches furnish powerful tools for understanding the deep algebraic structures in the stable cohomology of MCGs, especially for the Torelli group and its prounipotent completion. The use of trivalent graphs and their Koszul duals enables the computation of the symplectic character and full Poincaré series for MCG cohomology with coefficients in symplectic representations (Garoufalidis et al., 2017, Felder et al., 2021). In the stable limit $g\to\infty$ , the Malcev completion of the mapping class group relative to the symplectic group is shown to be Koszul (Felder et al., 2021).

2. Explicit Cohomological Calculations: Orientable and Non-orientable Cases

For non-orientable surfaces such as the punctured projective plane, the cohomology of the mapping class group $\Gamma^k(\mathbb{R}P^2)$ with mod-2 coefficients is given by

$H^*(\Gamma^k(\mathbb{R}P^2); \mathbb{Z}/2) \cong \mathbb{Z}/2[w_2,w_3] \otimes H^*(F_k(\mathbb{R}P^2)/\Sigma_k; \mathbb{Z}/2)$

with $w_2, w_3$ the Stiefel–Whitney classes in $H^*(BSO(3); \mathbb{Z}/2)$ , and $H^*(F_k(\mathbb{R}P^2)/\Sigma_k; \mathbb{Z}/2)$ expressible as direct sums of braid group homologies (Maldonado et al., 2017). The Serre spectral sequence associated to the universal bundle collapses at $E_2$ due to the trivial action of $\pi_1(BSO(3))$ on the fiber, and there are no multiplicative extensions between the cohomology of $BSO(3)$ and the configuration space factors.

For orientable surfaces, the top rational cohomology $H^{4g-5}(\mathrm{Mod}_g; \mathbb{Q})$ vanishes, as shown by Broaddus’s chord diagram techniques and virtual duality with the coinvariants of the Steinberg module ( $\mathrm{St}_g$ ) under the mapping class group action. Specifically, combinatorial arguments using filling systems and boundaries yield $[X]=0$ in $\mathcal{U}_0/\partial(\mathcal{U}_1)$ , leading to the vanishing of the corresponding cohomology group (Church et al., 2011). This resolves the open question regarding top-degree rational cohomology and aligns with analogous results for $SL(n, \mathbb{Z})$ .

3. Subgroup Cohomology: Torelli, Spin, and Hyperelliptic Mapping Class Groups

The Torelli group $\mathcal{I}_g$ —the kernel of the symplectic representation of the MCG—admits a deep cohomological and representation-theoretic analysis. Hain established that its graded Lie algebra is quadratic, generated by the standard third-exterior-power representation and quadratic relations, enabling computation of the quadratic dual and, in the stable range, identification with a graph complex algebra (Garoufalidis et al., 2017). The stable Poincaré series and symplectic character of the quadratic dual are derived as explicit plethystic generating functions, contingent on the Koszulity of the quadratic algebra.

Spin-hyperelliptic mapping class groups, stabilizers of spin structures under the MCG, have rational cohomology realized as invariants under symmetric group actions on braid group cohomology. Explicit combinatorial models using partitions $\lambda$ and cycle data $d$ generate canonical bases for $H^*(P_n; \mathbb{Q})^G$ , where $G$ encapsulates both block permutation and involutive symmetries (via $\omega$ ), with survival of classes determined by explicit signs $\epsilon(\lambda,d)$ (Fu et al., 31 Mar 2025, Wang, 6 Jan 2024). This machinery yields closed expressions and Poincaré series for Betti numbers in all degrees. The dimension formulas and explicit Betti numbers in low genus cases validate and extend classical calculations, revealing nontrivial classes up to degree $g+1$ for spin-hyperelliptic groups, which contrasts with the stable range vanishing for the full MCG (Fu et al., 31 Mar 2025).

4. Graph Complexes and Stable Koszul Properties

Graph complexes—such as $GC^{(g),1}$ , $GC^{(g)}$ , and their “extended” variants—encode the cohomology of MCGs and related diffeomorphism groups via decorated trivalent graphs. These complexes support weight gradings and admit differentials splitting vertices ( $\partial_{split}$ ) or gluing decorations ( $\partial_{glue}$ ), reflecting the algebraic structure of the corresponding Lie algebras (Felder et al., 2021). Key results include:

Concentration and vanishing in specific degrees for each fixed weight $W$ as $g\to\infty$ .
Formality and Koszulity: For $g\ge 3W$ , the Chevalley–Eilenberg complex is quasi-isomorphic to its cohomology algebra, ensuring a Koszul dual pair and quadratic presentation of the Malcev Lie algebra associated to the MCG.
Stable isomorphism: The inverse limit of cohomologies ( $\lim_{g\to\infty} H^*(GC^{(g)})$ ) coincides with the stable (prounipotent) cohomology of the mapping class group.

In particular, these results connect the classical and elliptic Grothendieck–Teichmüller Lie algebras to the cohomology of mapping class groups and their decorated variants (Felder et al., 2021).

5. Representation Stability and Combinatorial Models

Representation-theoretic approaches, particularly those based on the Lehrer–Solomon description of pure braid group cohomology, lead to explicit bases and dimension formulas for invariant cohomology under symmetric group and product-involution symmetries. The rational cohomology algebra $H^*(P_n; \mathbb{Q})^G$ of spin-hyperelliptic groups is structured by partitions and counts of full invariant cycles, with the group invariants determined by sign computations and symmetry criteria (Fu et al., 31 Mar 2025). Cohomological stability results demonstrate that the invariant subalgebras stabilize in fixed degree once the subgroup size (e.g., $q$ for spin-hyperelliptic groups) exceeds the degree threshold, mirroring but refining classical MCG stability phenomena (Wang, 6 Jan 2024). Closed-form formulas for Betti numbers in low genus cases (e.g., $g=1,2$ ) are combinatorially classified and consistent with geometric expectations (Fu et al., 31 Mar 2025).

6. Extensions to Non-standard and Crystallographic MCGs

Recent developments extend the cohomological perspective on MCGs to settings such as momentum-space crystallographic groups in the classification of band topology and topological phases. In this context, the second cohomology $H^2(\Gamma_F, \mathbb{Z})$ (for an MCG $\Gamma_F$ ) classifies all abelian crystalline topological phases, while $H^3(\Gamma_F, \mathbb{Z})$ captures all possible twistings for equivariant $K$ -theory (Liu et al., 26 Dec 2025). Key technical results include isomorphisms between higher group cohomology and cohomology with values in the space of momentum-space $U(1)$ -valued functions, facilitating a purely algebraic classification of topological invariants. The theory unifies known classifications in both mathematical and physical contexts, and is constructed via group extensions, Lyndon–Hochschild–Serre spectral sequences, and group-cohomological methods.

Summary Table: Betti Numbers for Spin-Hyperelliptic Mapping Class Groups, Low Genus (Fu et al., 31 Mar 2025)

Genus $g$	Degrees with $b_k\neq0$	Betti numbers $b_k$
$1$	$0,1,2$	$1,1,1$
$2$	$0,1,2,3,4$	$1,2,3,2,1$

7. Open Questions and Future Directions

Open problems include the full determination of the unstable cohomology of MCGs near their virtual cohomological dimension, as well as systematic identification of torsion phenomena, especially for non-orientable and spin-hyperelliptic mapping class groups (Church et al., 2011, Fu et al., 31 Mar 2025). For the latter, the extension to higher-order spin analogues and the detection of torsion classes in integral cohomology remain unresolved. There is ongoing interest in the geometric realization of combinatorial cycles as configuration-space submanifolds, their connection to string topology, and the role of explicit cohomology calculations in the study of moduli spaces of spin curves and Picard groups. Developments in graph complex formality and “Koszulness” suggest deep connections between the algebraic structures of MCG cohomology and moduli space geometry (Felder et al., 2021, Garoufalidis et al., 2017). The connection between MCG cohomology and group cohomologies appearing in crystalline symmetry classification points toward further unification across geometric, algebraic, and physical applications.