[Paper Review] Linear Congruences and hyperbolic Systems of conservation Laws

Pietro De Poi, Emilia Mezzetti|ArXiv.org|Apr 2, 2005

Polynomial and algebraic computation17 references3 citations

TL;DR

This paper classifies linear congruences of lines in $\mathbb{P}^5$, establishing their correspondence to $T$-systems in four variables via the Agafonov-Ferapontov construction. It identifies focal loci as Palatini threefolds—smooth or singular scrolls over cubic surfaces—and provides a sketch of classification based on the geometry of the dual Grassmannian and singular loci, offering a foundation for classifying Temple systems in four variables.

ABSTRACT

S. I. Agafonov and E. V. Ferapontov have introduced a construction that allows naturally associating to a system of partial differential equations of conservation laws a congruence of lines in an appropriate projective space. In particular hyperbolic systems of Temple class correspond to congruences of lines that place in planar pencils of lines. The language of Algebraic Geometry turns out to be very natural in the study of these systems. In this article, after recalling the definition and the basic facts on congruences of lines, Agafonov-Ferapontov's construction is illustrated and some results of classification for Temple systems are presented. In particular, we obtain the classification of linear congruences in $\mathbb{P}^5$, which correspond to some classes of $T$-systems in 4 variables.

Motivation & Objective

• To initiate a systematic classification of linear congruences of lines in $\mathbb{P}^5$, a previously unclassified case in the study of hyperbolic conservation laws.
• To establish the correspondence between such congruences and $T$-systems in four variables through the Agafonov-Ferapontov construction.
• To analyze the geometry of the focal locus, particularly Palatini threefolds, and their singularities arising from special positions of the dual $3$-space $\check{\Delta}$ relative to the dual Grassmannian $\check{\mathbb{G}}(1,5)$.
• To provide a sketch of classification based on the type of cubic surface $S$ and its embedding in $\check{\mathbb{G}}(1,5)$, including cases with isolated singularities, double lines, or reducible components.
• To lay the groundwork for a complete classification of $T$-systems in four variables using algebraic geometry techniques.

Proposed method

• The study employs the Plücker embedding to realize the Grassmannian $\mathbb{G}(1,5)$ as a cubic hypersurface in $\mathbb{P}^{14}$, with the dual $\check{\mathbb{G}}(1,5)$ being the Pfaffian hypersurface.
• Linear congruences in $\mathbb{P}^5$ are defined as $\mathbb{G}(1,5) \cap \Delta$, where $\Delta$ is a $10$-dimensional linear subspace, and their duals $\check{\Delta}$ are studied via intersection with $\check{\mathbb{G}}(1,5)$.
• The focal locus $F$ of a congruence is analyzed as the union of secant lines to the fundamental locus, with $F$ being a Palatini threefold when smooth or singular when $S$ has singularities.
• The classification relies on the geometry of the cubic surface $S = \check{\Delta} \cap \check{\mathbb{G}}(1,5)$, including cases where $S$ is smooth, has isolated singularities, is a cubic cone, or is a ruled surface with a double line.
• Special attention is given to cases where $\check{\Delta}$ is tangent to $\check{\mathbb{G}}(1,5)$ at points or along lines, leading to focal loci with linear or quadric components.
• The paper uses the classification of $6 \times 6$ skew-symmetric matrices of constant rank 4 to describe congruences arising from reducible or singular $S$, particularly in the case of parasitic components.

Experimental results

Research questions

• RQ1 How can linear congruences of lines in $\mathbb{P}^5$ be classified up to isomorphism, given the lack of a complete classification in higher dimensions?
• RQ2 What is the geometric structure of the focal locus of a linear congruence in $\mathbb{P}^5$, and how does it vary with the type of cubic surface $S$?
• RQ3 How do singularities in the cubic surface $S$—such as isolated double points or a double line—affect the structure of the associated Palatini threefold and the corresponding $T$-system?
• RQ4 What conditions on the dual $3$-space $\check{\Delta}$ lead to focal loci with irreducible components such as planes, quadric cones, or parasitic lines?
• RQ5 Can the classification of $T$-systems in four variables be systematically derived from the classification of linear congruences in $\mathbb{P}^5$?

Key findings

• Linear congruences in $\mathbb{P}^5$ correspond to $T$-systems in four variables via the Agafonov-Ferapontov construction, with the focal locus being a Palatini threefold.
• When the cubic surface $S = \check{\Delta} \cap \check{\mathbb{G}}(1,5)$ is smooth, the focal locus is a smooth Palatini scroll, which is a scroll over $S$ and has a non-trivial moduli space of dimension $4 + 5 = 9$.
• If $S$ has isolated singularities $P_i$ not lying on $\mathbb{G}(3,5)$, the focal locus becomes singular at those points, and the $3$-space $\pi_{P_i}$ becomes an irreducible component of the focal locus.
• When $S$ is a cubic cone or a ruled surface with a double line, the focal locus acquires singularities along a line or a quadric cone, respectively, and such congruences arise when $\check{\Delta}$ is tangent to $\check{\mathbb{G}}(1,5)$ along a line or point.
• In cases where $S$ is reducible, such as $\pi \cup Q$ or $C(\mathcal{V}) \cup T$, the focal locus contains parasitic components like lines or curves, and the congruence consists of secant lines meeting both components.
• The classification is sketchy due to the moduli of cubic surfaces and their embeddings, and a complete classification over $\mathbb{C}$ is already complex, with real classification requiring further refinement.