Sparse inversion of Stokes profiles. I. Two-dimensional Milne-Eddington inversions

摘要: Inversion codes are numerical tools used for the inference of physical properties from observations. Despite their success, quality current spectropolarimetric observations and those expected in near future presents a challenge to inversion codes. The pixel-by-pixel strategy inverting data that we currently utilize needs be surpassed improved. inverted parameters have take into account spatial correlation is present contains valuable information. We concept sparsity or compressibility develop an new generation Stokes parameters. code uses optimization techniques based on idea proximal algorithms impose sparsity. In so doing, allow first time exploit presence maps Sparsity also regularizes solution by reducing number unknowns. compare results with inversions, demonstrating increase robustness solution. show how method can easily compensate effect telescope point spread function, producing solutions enhanced contrast.