The flux norm, Bohr-Sommerfeld Quantization Rules and the scattering problem for h $Ψ$DO's on the real line
Abdelwaheb Ifa, Michel Rouleux
TL;DR
The paper develops a microlocal flux-norm framework for Bohr-Sommerfeld quantization of order $2$ for 1D $h$-pseudodifferential operators, showing BS rules emerge precisely when a Gram matrix built from microlocal WKB quasi-modes is non-invertible. It achieves a streamlined proof in the spatial representation, derives the generalized action through a flux-form analysis up to $\mathcal{O}(h^2)$, and expresses the BS condition as ${\cal S}_{h}(E)=2\pi n h$ with explicit $S_0,S_1,S_2$ components, including the subprincipal and higher corrections. The authors verify the Ansatz in the Schrödinger setting with analytic coefficients by reducing to Airy form and computing the order-$4$ WKB expansion, thereby establishing the phase/mode matching and connection formulas. They further extend the flux-norm method to semi-classical scattering over a compact barrier, deriving a monodromy matrix in $SU(1,1)$ and demonstrating flux conservation, while outlining how these ideas may extend to $2\times2$ systems and higher dimensions. Overall, the work provides a robust, representation-based route to BS quantization and scattering that complements traditional Weyl- and functional-calculus approaches and offers a pathway to more intricate microlocal systems.
Abstract
We revisit the well known Bohr-Sommerfeld quantization rule (BS) of order 2 for a self-adjoint 1-D h-Pseudo-differential operator within the algebraic and microlocal framework of Helffer and Sjoestrand; BS holds precisely when Gram matrix consisting of scalar products of some WKB solutions with respect to the ``flux norm'' (or microlocal Wronskian) is not invertible. We simplify somewhat our previous proof [A. Ifa H. Louati and M. Rouleux. Bohr-Sommerfeld Quantization Rules Revisited: the Method of Positive Commutators. J. Math. Sci. Univ. Tokyo, 25(2):2018] by working in spatial representation only, as in complex WKB theory for Schroedinger operator. We consider also the scattering problem.
