An information-theoretical methodology for profile parameter estimation of solar coronal spectra

Faculty of Science and Technology, Ryukoku University, Seta, Ôtu 520-2123, Japan e-mail: nakagawa@rins.ryukoku.ac.jp

Received: 24 November 2005
Accepted: 18 April 2006

Abstract

Aims.An attempt is made to establish an information-theoretical methodology for profile parameter estimation of the spectral lines of the solar corona.

Methods.In this formalism, “description lengths” of a target line profile are defined for every possible profile model and are used to identify the optimal one among possible models. To properly reflect statistical properties of count numbers representing spectral profiles, a so-called α scale is introduced and incorporated into the definition of description lengths. Formulae of error propagation from measured count numbers to derived profile parameters are also given.

Results.The advantage of this methodology, which enables mutually consistent treatment of profiles of diverse properties in particular, has been used to examine how the change in spectroscopic characteristics is associated with the coronal structure.