Pseudo-Spherical Focal Sets

• Pseudo-Spherical Focal Sets are loci of focal points derived from families of lines in pseudo-Riemannian spaces, capturing key singularity behaviors.
• The construction employs distance-squared functions, nullcone fronts, and curvature invariants to classify singularities such as cuspidal edges and swallowtails.
• This framework aids in analyzing wavefronts and degeneracies in constant curvature models, linking submanifold geometry with focal set properties.

A pseudo-spherical focal set arises as the locus of focal points associated to a congruence of lines (normal, tangent, or more generally isotropic) in an ambient space of constant curvature or pseudo-Riemannian signature. These sets generalize the classical notion of the focal set (evolute or caustic) from Euclidean differential geometry, extending it to pseudo-Riemannian manifolds—most notably, Minkowski, de Sitter, anti-de Sitter, and related model spaces. Pseudo-spherical focal sets encode the geometric and singularity-theoretic behavior of the underlying submanifolds, with canonical constructions involving distance-squared functions, wavefronts, curvature invariants, and the classification of singularities.

1. Definitions and General Framework

Let $(M,g)$ be a pseudo-Riemannian manifold of constant sectional curvature $\kappa$ (examples include Minkowski space $\mathbb{R}_1^n$ and the de Sitter/anti-de Sitter spaces $S^n_1(r)$, $\text{AdS}^3$). Consider a regular curve or hypersurface $\gamma \colon I \to M$. The focal set of $\gamma$, sometimes referred to in the literature as a pseudo-spherical focal set, is defined as follows (Nabarro et al., 2015):

For a family of "distance-squared" functions

$$f(s, p) := \langle \gamma(s) - p,\, \gamma(s) - p\rangle_M,\quad f_p(s) = f(s, p),$$

the bifurcation set

$$\mathrm{Bif}(f) = \{\,p \in M \mid \exists\,s\in I:\: f_p'(s)= f_p''(s) = 0\,\}$$

is the focal set. Geometrically, $\mathrm{Bif}(f)$ consists of points $p \in M$ at which the pseudo-sphere centered at $p$ has second-order contact with $\gamma$.

In higher dimensions, an oriented affine line in $\mathbb{R}^{n+1}$ associates to a point in the unit tangent bundle $TS^n$, so families of lines can be studied as submanifolds $\Gamma\subset TS^n$. The focal set of such a family is defined as the set of $p = \Psi((\xi,\eta), r) = \eta + r\xi$ for which the differential $d\Psi$ fails to be of maximal rank; i.e., the projection from the space of lines to points has singularities at the focal locus (Georgiou et al., 24 Jan 2026). This construction generalizes naturally to isotropic congruences and submanifolds.

2. Pseudo-Spherical Focal Sets of Curves and Hypersurfaces

In Minkowski, de Sitter, or anti-de Sitter spaces, the classical evolute and focal set constructions must be adapted for the pseudo-Riemannian context due to possible causal degeneracies (spacelike, timelike, lightlike) (Nabarro et al., 2015). For a regular curve $\gamma(s)$ (with non-lightlike tangent), the focal set parametrizes as:

$$B(s, \mu) = \gamma(s) + \frac{\epsilon}{\delta_n k(s)} n(s) + \mu b(s),\qquad \mu \in \mathbb{R},$$

where $$\epsilon = \langle \gamma', \gamma' \rangle = \pm 1$$, $\delta_n = \text{sign}\langle n, n\rangle$, $k(s)$ is the (pseudo-)curvature, and $\{t, n, b\}$ is a Frenet–Serret frame. The locus $$\mu = \mu(s) = \dfrac{k'(s)}{\epsilon\delta k(s)^2\tau(s)}$$ is the set of cuspidal edges (A$_3$-type singularities) on the focal surface.

For a strictly convex hypersurface $\Sigma\subset\mathbb{R}^{n+1}$ with simple principal radii $\rho_1,\dots,\rho_n$, the $i$th focal sheet is

$\Y_i(q) = q - \rho_i(q)\nu(q),$

with $\nu$ the unit normal. Gaussian curvature of the focal sheets in principal directions is given by

$K = \frac{d s_{ij}(X_i) - 1}{s_{ij}^2},$

where $s_{ij} = \rho_i-\rho_j$ is the astigmatism. If $s_{ij}$ is constant, each focal sheet has constant negative curvature: the focal sheet is a "pseudo-sphere" of curvature $-1/s_{ij}^2$ (Georgiou et al., 24 Jan 2026).

3. Pseudo-Spherical Focal Sets in Pseudo-Riemannian Model Spaces

In pseudo-spherical spaces such as anti-de Sitter 3-space ($\text{AdS}^3$), focal sets acquire further structure due to the ambient signature. For a pseudo-spherical framed curve $(\gamma,v_1,v_2)$, with $\gamma\colon I\rightarrow \text{AdS}^3$, and $v_i(s)$ are normal vector fields, the associated null surfaces (nullcone fronts) are

$$\mathcal{NF}^\pm_\gamma(s, \lambda) = \gamma(s) + \lambda \bigl(v_1(s) \pm v_2(s)\bigr),\quad \lambda\in\mathbb{R},$$

with $v_1(s)\pm v_2(s)$ lightlike (null vectors). These nullcone fronts are interpreted as focal sets formed by the light rays (null geodesics) emanating from points on the framed curve.

The singular locus of the nullcone front, corresponding to focal (envelope) points, satisfies

$\alpha(s) + \lambda (m(s) \pm n(s)) = 0,$

where $(\alpha,\ell,m,n)$ are curvature invariants arising from the Frenet–Serret equations of the frame.

The singularity type at each focal point is determined by a curvature-dependent invariant $\sigma(s)$:

• A cuspidal edge occurs if $\sigma(s_0)\neq0$.
• A swallowtail occurs if $\sigma(s_0)=0$ but $\sigma'(s_0)\neq0$ (Tuncer, 2023).

Evolutes and focal surfaces in these settings can be interpreted as discriminant and secondary discriminant sets of families of height functions, with the evolute set being the locus of singularities of the focal surface (Tuncer, 2023).

4. Singularity Theory and Classification

Singularity theory provides the conceptual and technical framework for understanding focal locus degeneracies. Singularities of the distance-squared function $f(s,p)$ are classified as:

• $A_2$ (Morse) points: Generic smooth points where the focal set is a regular surface.
• $A_3$: Cusp edge (cuspidal edge) singularities.
• $A_4$: Swallowtail points at isolated values (higher-order singularity) (Nabarro et al., 2015).

In the anti-de Sitter case, the contact properties of the wavefront (nullcone front or envelope of null hyperplanes) reflect the Legendrian singularity class:

• At regular parameter values, the focal surface is smooth.
• At values where the curvature invariant vanishes to appropriate order, the singularity is a cusp or swallowtail as detected by the behavior of $\sigma(s)$ or the speed of the evolute (Tuncer, 2023, Tuncer, 2023).

5. Metric and Causal Properties Near Degeneracies

At lightlike points ($\langle\gamma',\gamma'\rangle=0$), the standard focal set parametrization fails. The focal set is replaced by a bifurcation set $\mathcal{B}$, a smooth $2$-surface in the ambient pseudo-Riemannian space, intersecting the original curve only at the degenerate point. The unique lightlike direction in the tangent space to $\mathcal{B}$ is the tangent direction of the curve. The locus of degeneracy of the induced metric on $\mathcal{B}$ is the normal line through the basepoint, separating regions where the induced metric is Riemannian (spacelike tangent planes) from those where it is Lorentzian (timelike) (Nabarro et al., 2015).

In higher codimension, the induced sections of the focal submanifold may possess multi-Kleinian curvature properties rather than a uniform pseudo-spherical structure.

6. Classical Results and Generalizations

Bianchi’s theorem (1874) for surfaces in $\mathbb{R}^3$ states: if a strictly convex surface $\Sigma$ has constant astigmatism $s = \rho_1 - \rho_2$, then each focal sheet is a pseudo-sphere (i.e., has constant negative Gaussian curvature $K = -1/s^2$). This result generalizes to higher dimensions, where the focal set possesses constant negative sectional curvature in directions where corresponding astigmatisms $s_{ij}$ are constant (Georgiou et al., 24 Jan 2026).

For general isotropic families of lines in $\mathbb{R}^{n+1}$, the focal set’s curvature is determined by signed distances between focal points along common normals, establishing a direct link between the geometry of the original submanifold and the curvature of its associated focal set.

7. Wavefront Viewpoint and Legendrian Geometry

Pseudo-spherical focal sets can be interpreted as wavefronts or discriminant sets of families of distance-squared or height functions, viewed as generating families in Legendrian singularity theory. In anti-de Sitter geometry, the discriminant set of the distance-squared function yields the nullcone front, while its "secondary" discriminant set gives the evolute as the singular locus of the focal surface (Tuncer, 2023, Tuncer, 2023). The envelope of null hyperplanes, or the wavefront, provides a geometric realization of the singularities and structure of the focal set, generalizing the Euclidean constructions to the pseudo-Riemannian context.

---

Key references include (Nabarro et al., 2015) for focal sets near lightlike points, (Tuncer, 2023) for nullcone fronts and singularity classification in anti-de Sitter space, (Tuncer, 2023) for the Legendrian viewpoint and higher singularity structures, and (Georgiou et al., 24 Jan 2026) for the curvature and inverse problem, with generalization of Bianchi’s theorem.