Cauchy problem for noneffectively hyperbolic operators (Q354018)

This book, written with great clarity by a renowned expert of the field, is an introduction to the study of the well-posedness in \(C^\infty\) (and in Gevrey classes) of the Cauchy problem for weakly hyperbolic operators with double characteristics that are of noneffectively hyperbolic type. Let \(P\) be a second-order partial differential operator with smooth coefficients, and let \(p_2\) be its principal symbol. At a double characteristic point, that is a point at which \(p_2=0\) and \(dp_2=0\), one can consider the linearization of the Hamilton vector field \(H_p\), called the Hamilton map \(F_p\), and depending on whether \(F_p\) possesses non-zero real eigenvalues or not, respectively, \(P\) is said to be effectively hyperbolic or noneffectively hyperbolic, respectively. The two kinds of operators exhibit different behaviors in regard to the well-posedness: the Cauchy problem for effectively hyperbolic operators is well-posed regardless the lower order terms in the symbol of \(P\), whereas in case of effectively hyperbolic operators V.~Ivrii and V.~Petkov, and L.~Hörmander, proved that in order to have \(C^\infty\) well-posedness, at the double characteristic points the subprincipal symbol \(p_1^s\) of the operator (\(p_1^s=p_1+(i/2)\langle\partial_x,\partial_\xi\rangle p_2\)) must necessarily be real and be bounded, in absolute value, by the sum of the modulus of the purely imaginary eigenvalues of \(F_p\). But crucial to the \(C^\infty\) well-posedness is not only the spectrum of the Hamilton map, but also the behavior of the integral curves of \(H_p\), the bicharacteristic curves. In particular, if one has bicharacteristics that issue from simply characteristic points and land tangentially on the double-characteristic set, then the Cauchy problem is not \(C^\infty\) well-posed, and is well-posed in the smaller Gevrey spaces of order \(s\) only for \(1\leq s<5\), but not for \(s> 5\). In this monograph, the author considers noneffectively hyperbolic operators and gives a general picture of the Cauchy problem in this case, by providing all the necessary background and detailed proofs, the main point being that the Hamilton map and the geometry of the bicharacteristic orbits completely characterize the \(C^\infty\) well-posedness/non-well-posedness of the Cauchy problem. More precisely: Chapter 1 contains a discussion of the general necessary conditions for the \(C^\infty\) well-posedness and the proof of the necessary condition of Ivrii-Petkov-Hörmander; in Chapter 2 the author considers double characteristic points and gives the classification of effectively/noneffectively hyperbolic operators; in Chapter 3 one finds a detailed study of the behavior of a noneffectively hyperbolic symbol near the double-characteristic manifold, the main point being the existence of an elementary decomposition (which is crucial to energy estimates); Chapters 4 and 5 deal with the \(C^\infty\)-well posedness and in particular in Chapter 5 one finds the very important method of the parametrix with finite propagation speed of wave front sets, which allows one to pass from microlocal well-posedness to well-posedness; Chapter 6 deals with the well-posedness in Gevrey classes and Chapter 7 with the behavior of bicharacteristics; Chapter 8 deals with a discussion of the optimality of the Gevrey index and Chapter 9 with non-well-posedness results; a final Appendix gathers various results from symplectic geometry that are necessary throughout the book. To conclude, this book is a thorough and up-to-date introduction to the the study of the problem of the well-posedness of the Cauchy problem of the important class of noneffectively hyperbolic operators, and may be recommended to everyone who is willing to learn some deep and fascinating aspects of the geometric analysis of weakly hyperbolic PDEs.