Energy estimate and fundamental solution for degenerate hyperbolic Cauchy problems (Q2575067)

The autors are interested in the Cauchy problem for the second-order operator \[ P=D^2_t-a(t,x,D_x)+b(t,x,D_x)+c(t,x) \] under the hyperbolicity condition \[ a(t, x,\xi)\geq 0\text{ for }(t,x,\xi)\in(-T,T)\times\mathbb R^{2n}_{x,\xi}. \] The authors apply a unified approach to study degenerate hyperbolic problems. Degeneracies are either produced by higher multiplicity of characteristic roots or by a non-Lipschitz behaviour of coefficients of \(a(t,x,D_x)\) with respect to \(t\). The approach consists of three steps: -- Factorization of the principal part of \(P\) by means of regularized characteristic roots. -- Reduction to an equivalent \(2\times 2\) system, where the ``remainder'' satisfies a suitable property -- Proof of energy estimates by using sharp Gårding's inequality. The degenerate hyperbolic operators have the special form \[ P = D^2_t-\alpha(t) Q(x,D_x) +b(t,x,D_x)+c(t,x). \] In the weakly hyperbolic case a condition between effective hyperbolicity and \(C^\infty\)-type Levi condition for \(\alpha(t)\) is used. In the strictly hyperbolic case \(\alpha(t)\) is supposed to belong to LogLip (global irregular behaviour) or to satisfy an unbounded behaviour of \(\alpha'\). In the second case the parametrix is constructed and the propagation of singularities along the broken bicharacteristics is explained.