Search papers, labs, and topics across Lattice.
This paper addresses the inverse problem of recovering a spatiotemporal potential $q$ in the wave equation $(\Box_g+q)u=0$ on a Lorentzian manifold. By proving Lipschitz stability for a formally determined version of this problem under a null-cone foliation assumption, the authors establish uniqueness for the Lorentzian Calderón problem. The core technique involves adapting the Bukhgeim--Klibanov method with distorted plane waves.
Uniqueness for the Lorentzian Calderón problem is achieved by proving Lipschitz stability for a formally determined inverse problem under a null-cone foliation assumption.
We consider the inverse problem of recovering a potential $q$, depending on space and time, in the wave equation $(\Box_g+q)u=0$ on a Lorentzian manifold. We prove Lipschitz stability for a formally determined version of this problem under a \textit{null-cone foliation} assumption. As a consequence, we obtain uniqueness in the Lorentzian Calderón problem for $q$ under this assumption. Our approach is based on a modification of the Bukhgeim--Klibanov method using distorted plane waves.