The EUV Sun as the superposition of elementary Suns

TL;DR

This paper presents a Bayesian method to decompose EUV solar irradiance spectra into three elementary spectra, revealing they correspond to different atmospheric layers and capturing spectral variability without prior assumptions.

Contribution

It introduces a novel Bayesian source separation approach to infer elementary spectra from EUV measurements, avoiding predefined component assumptions.

Findings

Spectral irradiance decomposes into three elementary spectra.

Elementary spectra relate to different solar atmospheric layers.

Method captures spectral variability effectively.

Abstract

Many studies assume that the solar irradiance in the EUV can be decomposed into different contributions, which makes the modelling of the spectral variability considerably easier. We consider a different approach, in which these contributions are not imposed a priori but are effectively and robustly inferred from spectral irradiance measurements. This is a source separation problem with a positivity constraint, for which we use a Bayesian solution. Using five years of daily EUV spectra recorded by the TIMED/SEE satellite, we show that the spectral irradiance can be decomposed into three elementary spectra. Our results suggest that they describe different layers of the solar atmosphere rather than specific regions. The temporal variability of these spectra is discussed.