Higher Even Gaussian Maps in Algebraic Geometry

• Higher even Gaussian maps are a sequence of linear maps extending the classical Wahl map to connect quadratic relations with higher order canonical forms.
• They are constructed using explicit bases and Schiffer variations, with rank and kernel computations varying across hyperelliptic, trigonal, and complete intersection curves.
• These maps critically inform deformation and syzygy theory, influencing the structure of moduli spaces and the intrinsic geometry of projective varieties.

Higher even Gaussian maps are a rich family of linear maps arising in algebraic geometry, extending the classical Wahl (second Gaussian) map to higher even orders. They encode subtle aspects of the geometry of projective curves and varieties—especially the canonical curves—and are intricately connected to syzygies, deformation theory, and the extrinsic geometry of moduli spaces, notably through their relationship with the second fundamental form of the Torelli map. Their structure, rank, surjectivity, and kernel properties have deep consequences for the geometry of curves, surfaces, and their moduli, and are an active area of research with detailed investigations across special loci, including hyperelliptic, trigonal, and general curves, as well as higher-dimensional varieties.

1. Definitions, Constructions, and Canonical Examples

Let $C$ be a smooth projective curve of genus $g$, and $K_C$ its canonical bundle. The classical multiplication map of canonical forms,

$$\mu_0: \operatorname{Sym}^2 H^0(K_C) \to H^0(K_C^2)$$

has kernel $I_2(K_C)$, the space of quadratic relations (Petri quadrics) among holomorphic differentials on $C$. The sequence of higher Gaussian (Wahl) maps, and specifically their even-indexed members, are defined inductively as follows: $$\mu_2: I_2(K_C) \to H^0(K_C^4), \qquad \mu_{2k}: \ker(\mu_{2k-2}) \to H^0(K_C^{2k+2}), \quad k \geq 2.$$ Alternatively, for any line bundle $L$ on $C$, one considers the higher Gaussian maps

$$P_{k,L}: H^0(C \times C, L \boxtimes L(-k\Delta)) \to H^0(C, L^2 \otimes K_C^k),$$

with $\gamma_{2k}$ (or notation $\gamma^+_{2k}$, $\mu_{2k}$, etc.) denoting restriction to the kernel of the previous $P_{2k-2,L}$, yielding a filtration on symmetric powers of global sections. When $L = K_C$, these become the higher even Gaussian maps of the canonical bundle.

For explicit computations and applications, especially in loci with additional structure (hyperelliptic, trigonal), these maps are constructed using explicit bases adapted to the covering/pencil, and their action is expressed combinatorially in terms of Schiffer variations and derivative vanishing properties (Lacopo, 22 Jan 2026, Faro et al., 2024, Ortiz, 2021).

2. Rank, Kernel Structure, and Explicit Formulae

Rank and kernel computations of higher even Gaussian maps are central in understanding their geometric impact.

• For general trigonal curves of genus $g \geq 16$, the rank satisfies

$$\operatorname{rk}(\mu_{2k}) \geq 2g - 8k - 2, \qquad 2 \leq k \leq \left\lfloor \frac{g-4}{6} \right\rfloor$$

(Lacopo, 22 Jan 2026).

• For cyclic trigonal covers $C\to\mathbb{P}^1$, $g = r-2$, the decomposition

$H^0(K_C \otimes L^{-1}) = W_1 \oplus W_2$

with $n_1 = \frac{r-6}{3}$, $n_2 = \frac{2r-6}{3}$ yields for $2 \le k \le \lfloor n_i/2 \rfloor$,

$$\operatorname{rk}(\mu_{2k-1}|_{\wedge^2 W_i}) = 2 n_i - 4k + 1, \qquad \dim \ker(\mu_{2k-1}|_{\wedge^2 W_i}) = \frac{1}{2} n_i(n_i-1) - k(2 n_i - 2k - 1)$$

with vanishing for $k > \lfloor n_i/2 \rfloor$ (Lacopo, 22 Jan 2026).

For hyperelliptic curves of genus $g \geq 3$, a uniform formula holds: $$\operatorname{rk}\, \gamma^+_{2k} = 2g - (4k + 1), \qquad \dim \ker\, \gamma^+_{2k} = \frac{(g-1)(g-2)}{2} - k(2g-2k-3)$$ for $$0 \leq k \leq \left\lfloor\frac{g-1}{2}\right\rfloor$$; for $k$ above that range, the map is identically zero (Faro et al., 2024).

The explicit kernel of the second Gaussian map for trigonal or hyperelliptic curves is determined by linear equations among the expansion coefficients of quadrics with respect to canonical bases adapted to the covering or the chosen pencil (Lacopo, 22 Jan 2026, Faro et al., 2024).

3. Geometric Context: Relation to the Second Fundamental Form and Schiffer Variations

Higher even Gaussian maps have a natural geometric interpretation via their relation to the second fundamental form of the Torelli map

$j: \mathcal{M}_g \to \mathcal{A}_g$

mapping a curve to its principally polarized Jacobian. For a non-hyperelliptic curve, the conormal space $I_2(K_C)$ maps, via the Hodge–Gaussian map, to symmetric tensors in $H^0(C, K_C^2)$, and the further composition with multiplication corresponds to the classical (second) even Gaussian map $\mu_2$ (Frediani, 2022).

At higher order, higher even Gaussian maps control the vanishing of the second fundamental form on spaces generated by higher-order Schiffer variations $\sigma_n(p) \in H^1(C, T_C)$ associated to a point $p \in C$. These variations probe the behavior of the Torelli map in explicit tangent directions, with the vanishing loci and isotropic subspaces for the images of higher even Gaussian maps directly governing the existence of "asymptotic directions" and thus the geometry of specialized loci (e.g., in $\mathcal{A}_g$ or on the hyperelliptic or trigonal locus) (Lacopo, 22 Jan 2026, Faro et al., 2024, Frediani, 2022).

In particular, for cyclic trigonal or hyperelliptic curves, explicit criteria are given for when certain subspaces generated by higher Schiffer variations are totally isotropic for all quadrics in the kernel of a given even Gaussian map, with uniqueness results for isotropic directions at Weierstrass or total ramification points (Lacopo, 22 Jan 2026, Faro et al., 2024).

4. Surjectivity, Vanishing, and Rigidity Phenomena

Surjectivity and injectivity results for higher even Gaussian maps play a crucial role in syzygy theory and deformation theory.

On general curves, there is a sharp dichotomy:

• For non-hyperelliptic curves of genus $g \geq 4$, it holds that $\mu_{6g-6}$ is injective, and all higher even Gaussian maps vanish for $k > 3g-3$ ("top-degree rigidity") (Frediani, 2022).
• For curves of sufficiently high genus, constructed as hyperplane sections of a polarized K3 surface, surjectivity for the $k$-th Gaussian map holds for $g > 4(k+2)^2 + 2$, via vanishing theorems on the Hilbert square of the surface and ampleness conditions on divisors $L - (k+2)\delta$ (Ortiz, 2021).

For canonical models of non-hyperelliptic, non-trigonal, non-plane-quintic curves, and sufficiently ample line bundles, higher even Gaussian maps are surjective for all even weights above $4$ (Ballico et al., 2013).

In the context of deformations of higher spin curves, the rank of the higher even Gaussian map $\gamma_{L,h}$ controls the Zariski tangent space to the $h$-spin locus and can be explicitly related to the dimensions of symmetric powers and combinatorial quantities (Ballico et al., 2013).

5. Special Families: Trigonal, Hyperelliptic, and Complete Intersections

The structure of higher even Gaussian maps is especially tractable in families with additional structure:

• Trigonal curves: Through the splitting of the space of holomorphic differentials via the $g_3^1$, one leverages eigenbundle decompositions to obtain precise kernel and rank structures, with applications to the absence of extra asymptotic directions and restrictions on totally geodesic subvarieties of the Torelli locus (Lacopo, 22 Jan 2026).
• Hyperelliptic curves: Inductive dimension counts yield closed formulas for the rank and kernel dimensions, and explicit linear relations describe the successive kernels, directly informing the geometry of the hyperelliptic Torelli map and its second fundamental form (Faro et al., 2024).
• Complete intersection curves: For curves lying on sufficiently high-degree complete intersections in projective space, the surjectivity of higher even Gaussian maps is established via geometric and cohomological criteria, connecting to projective normality and generation of syzygies (Ballico et al., 2013).

These explicit computations determine dimensions of deformation spaces, codimensions of special loci, and provide sharpness results for theorems regarding tangent maps to moduli points of spin or rooted curves.

6. Higher Even Gauss Maps on Projective Varieties and Fundamental Form Perspective

Beyond the context of curves, higher even Gauss maps $\gamma_{2r}$ on projective varieties $X \subset \mathbb{P}^N$ are associated to higher-order osculating spaces and fundamental forms (Poi et al., 2015, Kaji, 2015). The $(2r+1)$-th fundamental form $F_{2r+1}$ cut out by these maps has pronounced geometric implications:

• If $F_{2r+1}$ is non-empty, then either the even-indexed Gauss map $\gamma_{2r}$ is generically finite or its fibers are as large as linear spaces of dimension $k-1$, corresponding to $X$ being a scroll over a lower-dimensional base and osculating spaces forming towers of cones.
• For Veronese varieties, explicit formulae for the degrees of images of higher Gauss maps, including even cases, are given in closed combinatorial form involving standard Young tableaux and partitions, with degree bounds reflecting the global geometry and complexity of the embedding (Kaji, 2015).
• The recursive structure and parity-based restrictions yield rigidity phenomena and classifications for varieties with degenerate higher Gauss maps at even indices (Poi et al., 2015).

7. Applications, Open Questions, and Directions

Higher even Gaussian maps play a central role in various geometric and deformation-theoretic problems:

• Bounds on totally geodesic subvarieties: Kernel properties of higher even Gaussian maps restrict the dimension of analytic germs lying in the Torelli locus within $\mathcal{A}_g$, refining the classical Schottky problem (Lacopo, 22 Jan 2026, Faro et al., 2024).
• Deformation theory of higher-spin structures: The cokernel of even Gaussian maps often identifies the Zariski tangent space to moduli of spin curves or higher root loci (Ballico et al., 2013).
• Syzygy theory: Surjectivity/vanishing/non-vanishing of higher even Gaussian maps is directly linked to canonical syzygies, Green's conjecture, and the geometry of special divisors (Ortiz, 2021, Frediani, 2022).

Research directions include improving bounds for surjectivity in terms of genus and order, exploring Bott vanishing for symmetric powers, extending constructions to higher-dimensional hyperkähler varieties, and clarifying the connections with higher fundamental forms and global linear systems on Hilbert schemes (Ortiz, 2021, Frediani, 2022).

---

References: (Lacopo, 22 Jan 2026, Faro et al., 2024, Frediani, 2022, Ortiz, 2021, Poi et al., 2015, Kaji, 2015, Ballico et al., 2013)