Linear toroidal-inertial waves on a differentially rotating sphere with application to helioseismology: Modeling, forward and inverse problems
Tram Thi Ngoc Nguyen, Damien Fournier, Laurent Gizon, Thorsten Hohage
TL;DR
The paper develops a rigorous framework for modeling solar inertial waves in a purely toroidal, linear regime on a differentially rotating sphere, deriving a time-harmonic fourth-order Orr-Sommerfeld-type equation for the stream function and its separated-m$^m$ counterparts. It proves well-posedness under explicit smallness or regularity conditions on the differential rotation and viscosity, and formulates an inverse problem to jointly identify $\gamma$ and $\Omega$ from surface observations, establishing a lifted regularity approach and tangential cone condition for convergence of iterative regularization methods. Local identifiability results are shown (with viscosity known or unknown) under full measurements, and partial-data results are obtained under additional Cauchy-data assumptions. Numerical experiments with an accelerated Nesterov-Landweber scheme demonstrate robustness of reconstructions across observation strategies and noise levels, validating the practical potential for helioseismic inferences of internal solar rotation and viscosity.
Abstract
This paper develops a mathematical framework for interpreting observations of solar inertial waves in an idealized setting. Under the assumption of purely toroidal linear waves on the sphere, the stream function of the flow satisfies a fourth-order scalar equation. We prove well-posedness of wave solutions under explicit conditions on differential rotation. Moreover, we study the inverse problem of simultaneously reconstructing viscosity and differential rotation parameters from either complete or partial surface data. We establish convergence guarantee of iterative regularization methods by verifying the tangential cone condition, and prove local unique identifiability of the unknown parameters. Numerical experiments with Nesterov-Landweber iteration confirm reconstruction robustness across different observation strategies and noise levels.
