Recent results in nonlinear geometric optics (Q4247618)

The objective of the authors is to give a review of works related to high frequency oscillatory nonlinear waves done during last twenty years. Their previous works (published or not), which aim to extend the well-known Lüneberg-Kline type asymptotic expansions connected with the classical geometric optics to the nonlinear waves, constitute an important basis of the review. The starting expansions are of the forms \[ u^\varepsilon (t,x) \sim \varepsilon^p \sum_{n\geq 0}\varepsilon^n U_n\bigl(t,x, \varphi(t,x)/ \varepsilon \bigr) \] or \[ u^\varepsilon (t,x)\sim \varepsilon^p \sum_{n\geq 0} \varepsilon^n U_n\bigl(t,x, \psi_1(t,x)/ \varepsilon,\dots, \psi_m (t,x)/ \varepsilon\bigr), \] where \(\varepsilon\) stands for the wavelength. The first case is the so-called `one phase expansion' while the second is the `multiphase expansion'. The order of magnitude \(p\) along with the phase functions \(\varphi\) and \(\psi_j\) as well as the functions \(U_n(t,x, \theta_1, \dots, \theta_m)\) are determined by plugging these expressions into the given wave equation. The functions \(U_n(t,x, \theta_1, \dots, \theta_m)\) are supposed to be periodic (or quasi/almost periodic) with respect to \(\theta_j\). Basic problems connected with the above-mentioned expansions concern both the justification of the existence of waves of these types and the examination of their propagation, interactions, reflections etc. The review involves also the so-called `three scales expansion' \[ u^\varepsilon (t,x)\sim \varepsilon^p \sum_{n\geq 0} \varepsilon^n U_0\bigl(\varepsilon t,t,x, \varphi(t,x)/ \varepsilon\bigr), \] which is useful to describe long time propagation and diffraction effects. The paper involves a reference list of 55 works (published papers, preprints, papers to appear as well as books). For the entire collection see [Zbl 0911.00029].