Hyperbolic Equation Classification

Updated 21 December 2025

Hyperbolic equations are partial differential equations defined by real characteristic directions that model wave propagation and discontinuities.
They are classified using analytic discriminants, Lie symmetry analyses, and geometric interpretations to distinguish strict, degenerate, and mixed types.
This framework reveals integrability conditions and discrete analogues that enhance the practical analysis and modeling of hyperbolic systems.

A hyperbolic equation is a partial differential equation (PDE) whose principal symbol has signature admitting real characteristic directions, reflecting the propagation of waves. The classification of hyperbolic equations encompasses analytic, algebraic, geometric, and symmetry-based methodologies across both scalar and system forms, including continuous, semidiscrete, and non-strictly hyperbolic frameworks. This article presents a systematic exposition of established classification principles, symmetry-integrable exceptions, differentiating invariants, and exemplary canonical forms with reference to recent foundational research.

1. Fundamental Types and Analytic Criteria

The principal analytic criterion for hyperbolicity is the positivity of the discriminant $D_{\mathrm{pr}} = B^2 - AC$ in second-order PDEs of the general form

$A(x,u)\,u_{tt} + 2 B(x,u)\,u_{tx} + C(x,u)\,u_{xx} + \text{l.o.t.} = 0$

where $A$ , $B$ , $C$ are smooth coefficients. The sign and vanishing properties of $A$ , $B$ , $C$ , and $D_{\mathrm{pr}}$ divide equations into major classes (Makarenko et al., 2017):

Subclass	Discriminant Condition	Characteristic Structure
Strictly hyperbolic	$D_{\mathrm{pr}} > 0$	Two distinct real characteristics
Weakly hyperbolic (double)	$D_{\mathrm{pr}} = 0$	Coincident characteristics
Degenerate hyperbolic	$A=0$ or $C=0$ , $D_{\mathrm{pr}}>0$	Loss of one wave direction
Mixed type	$D_{\mathrm{pr}}$ changes sign	Hyperbolic/elliptic regions

Canonical normal forms can be achieved via local coordinate changes: strictly hyperbolic equations reduce to wave-type operators, while the weakly hyperbolic case yields forms analogous to squared-transport operators. Example: the quasilinear hyperbolic modification of the Burgers equation (QHMB) is strictly hyperbolic for $\tau,\mu>0$ (Makarenko et al., 2017).

2. Symmetry-Based and Lie Group Classifications

Extensive Lie symmetry analysis partitions scalar second-order hyperbolic equations,

$u_{xy} = F(u, u_x)$

into distinct symmetry classes using the structure and dimension of the admitted Lie algebras (Ndogmo, 2020):

Canonical $F(u)$ or $F(u_x)$	Algebra dimension	Integrability/Remarks
arbitrary $F(u)$	$3$	Only translation-dilation symmetry
$F(u)=e^u$	$\infty$	Liouville, integrable
$F(u)=u$	$\infty$	Linear wave
Power/exponential forms $F$	$4$	Enhanced symmetry
arbitrary $F(u_x)$	$\infty$	Large symmetry, open integrability

Specific symmetry-integrable forms are rare; for instance, $u_{xy}=e^{u}$ and $u_{xy}=e^{u}+e^{-u}$ or $u_{xy}=e^u+e^{-2u}$ (Tzitzeica type) are the prototypical integrable cases in this class. The sine-Gordon equation $u_{xy}=\sin u$ does not belong to any of the maximal algebra cases but is integrable via the inverse-scattering method (Ndogmo, 2020).

3. Higher Symmetry and Integrability Constraints

The existence of higher (third, fifth, etc.) order autonomous symmetries imposes extremely rigid conditions, restricting the form of admissible hyperbolic equations (Garifullin, 14 Dec 2025, Garifullin, 2022):

For scalar equations of the form $u_{xy} = F(u_x,u_y,u)$ , requiring a fifth-order symmetry generates a functional overdetermined system whose only non-Darboux-integrable solutions are:

The Tzitzeica equation: $u_{xy} = e^u + e^{-2u}$
Quadratic potential: $u_{xy} = 2 f(u_x)\,u$ with $f$ satisfying $(f+p)^2(2f-p)+1=0$
Square-root case: $u_{xy} = 2 f(u_x)\sqrt{u_y}$ , $f$ as above
Weierstrass-elliptic case: $u_{xy} = -2f(u_x)f(u_y)\frac{\omega'(u)}{\omega(u)}$ , $\omega'^2=4\omega^3+c$

No other scalar non-Darboux-integrable equation of this type admits a fifth-order autonomous symmetry (Garifullin, 14 Dec 2025). These forms are all symmetry-integrable and generally admit Lax representations, infinite symmetry hierarchies, and connections to classical integrable models, such as Toda, Volterra, and Tzitzeica systems.

The semidiscrete analog for chains $u_{n+1,x}=f(u_{n,x},u_{n+1},u_n)$ admitting third-order continuous and discrete symmetries yields precisely thirteen canonical families, each linked via Bäcklund-type transformations to continuous integrable hierarchies (Garifullin, 2022).

4. Geometric and System-Theoretic Classifications

The system approach relies on analytic and geometric criteria, particularly in relation to conservation of pseudo- or spherical surface geometries. For $2\times2$ systems of the form

$u_{xt} = F(u,u_x,v,v_x),\quad v_{xt} = G(u,u_x,v,v_x)$

classification is based on the ability to realize these PDEs as Maurer–Cartan structure equations for coframes inducing constant curvature metrics, i.e., they are the compatibility conditions for $sl(2,\mathbb{R})$ (pss, $K=-1$ ) or $su(2)$ (ss, $K=+1$ ) zero-curvature representations (Kelmer et al., 2021).

Admissible systems, up to coframe parametrization and under suitable genericity/nondegeneracy, fall into two main classes: the "phi-family" parameterized by arbitrary monotone functions and the "linear-p family" defined via arbitrary non-proportional functions, each determined by solving the corresponding Maurer–Cartan compatibility system. Principal examples include the Pohlmeyer–Lund–Regge and Konno–Oono systems, with explicit 7-parameter and infinite-parameter extensions describing a variety of integrable geometric flows.

The geometric interpretation is that every such system encodes a metric of constant curvature; their classification is equivalent to enumerating all local pseudo- and spherical surface parametrizations via PDE systems with compatible Lax pairs (Kelmer et al., 2021).

5. Nonstrictly Hyperbolic and Degenerate Cases

Recent frameworks systematically classify nonstrictly hyperbolic $2\times2$ inhomogeneous systems (e.g., compressible Euler–Poisson equations) by analyzing the spectral structure of the source matrix $Q$ and tracking the decisive function $q(t)$ whose zero signifies derivative blow-up (Turzynsky, 2024). There are three eigenvalue regimes (real-distinct, real-repeated (Jordan), complex-conjugate), each enabling closed-form classification of singular versus global smoothness scenarios directly in terms of initial data.

The critical threshold theory developed in this context provides explicit necessary and sufficient inequalities for gradient catastrophe or global-in-time $C^1$ regularity for any initial profile and applies directly to classical and model Euler–Poisson systems, unifying various blow-up and boundedness results.

Invariant domains in the phase plane correspond to regions in which the flow remains regular. Additionally, all such systems admit simple-wave reductions, enabling a further reduction in complexity and context-specific analysis (Turzynsky, 2024).

6. Discrete, Semidiscrete, and Nonlocal Taxonomies

Hybrid cases include semidiscrete and fully discrete analogues, with classification governed by the existence of higher-order (e.g., third-order) symmetries in both discrete and continuous directions. The explicit list of thirteen canonical families in the semidiscrete case encapsulates all nondegenerate autonomous chains admitting such symmetries, each corresponding to a known integrable continuous flow in the continuum limit (Garifullin, 2022).

These discrete and semidiscrete taxonomies reveal deep connections between difference and differential equations under symmetry constraints, integrable structure propagation, and Bäcklund transformations interlinking hierarchies. The continuous limits of these families map directly onto hyperbolic PDEs classified via Darboux and symmetry methods.

In summary, the classification of hyperbolic equations is multi-faceted, governed by analytic (discriminant), algebraic (Lie symmetry), geometric (surface/curvature), and integrability (higher symmetry) principles. The resulting taxonomy delineates a landscape dominated by structurally rigid exceptional cases possessing full symmetry integrability, along with rich families of geometric and system-theoretic significance, complemented by sharp singularity criteria in non-strictly hyperbolic settings and discrete analogues (Garifullin, 14 Dec 2025, Kelmer et al., 2021, Ndogmo, 2020, Turzynsky, 2024, Makarenko et al., 2017, Garifullin, 2022).