An iterative OLA method for inversion of solar spectropolarimetric data: I. Single and multiple variable inversions of thermodynamic quantities

TL;DR

This paper adapts the Optimal Localized Averaging (OLA) inversion technique for solar spectroscopic data, enabling high-resolution, minimal cross-talk inversion of thermodynamic quantities assuming LTE, with iterative refinement demonstrated on simulated spectra.

Contribution

It introduces an iterative OLA inversion method tailored for solar spectroscopic data, improving resolution and variable separation compared to existing techniques.

Findings

Successfully applied to simulated data and spectra

Addresses spectral sensitivity bias with response function amplification

Performs comparably or better than the SIR inversion scheme

Abstract

This paper describes an adaptation of the Optimal Localized Averaging (OLA) inversion technique, originally developed for geo- and helioseismological applications, to the interpretation of solar spectroscopic data. It focuses on inverting the thermodynamical properties of the solar atmosphere assuming that the atmosphere and radiation field are in Local Thermodynamic Equilibrium (LTE). We leave inversions for magnetic field and non-LTE inversions for future work. The advantage with the OLA method is that it computes solutions that are optimally resolved (in depth) with minimal cross-talk error between variables. Additionally, the method allows for direct assessment of the vertical resolution of the inverted solutions. The primary challenges faced when adapting the method to spectroscopic inversions originate with the possible large amplitude differences between the atmospheric model used to initiate the inversion and the underlying atmosphere it aims to recover, necessitating the development of an iterative scheme. Here we describe the iterative OLA method we have developed for both single and multi-variable inversions and demonstrate its performance on simulated data and synthesized spectra. We note that when carrying out multi-variable inversions, employing response function amplification factors can address the inherent spectral-sensitivity bias that makes it hard to invert for less spectrally-sensitive variables. The OLA method can, in most cases, reliably invert as well as or better than the frequently employed Stokes Inversion based on Response functions (SIR) scheme, however some difficulties remain. In particular, the method struggles to recover large-scale offsets in the atmospheric stratification. We propose future strategies to improve this aspect.