Coronal Electron Densities derived with Images acquired during the 21 August 2017 Total Solar Eclipse

TL;DR

This study used eclipse images to derive the 2D electron density distribution of the solar corona from 1.1 to 3 solar radii, employing a novel image combination and calibration method, revealing discrepancies with existing models.

Contribution

Developed a new method to combine multi-exposure eclipse images for detailed coronal electron density mapping and performed absolute calibration, highlighting differences with prior measurements and models.

Findings

Derived 2D electron density distribution from 1.1 to 3 solar radii.

Identified a factor of two discrepancy in absolute calibration compared to MLSO data.

Highlighted steep intensity gradients in the inner corona affecting polarization analysis.

Abstract

The total solar eclipse of August 21st, 2017 was observed with a Digital Single Lens Reflex (DSLR) camera equipped with a linear polarizing filter. A method was developed to combine images acquired with 15 different exposure times (from 1/4000 sec to 4 sec), identifying in each pixel the best interval of detector linearity. The resulting mosaic image of the solar corona extends up to more than 5 solar radii, with a projected pixel size by 3.7 arcsec/pixel, and an effective image resolution by 10.2 arcsecs, as determined with visible $\alpha-$Leo and $\nu-$Leo stars. Image analysis shows that in the inner corona the intensity gradients are so steep, that nearby pixels shows a relative intensity difference by up to $\sim 10 \%$; this implies that careful must be taken when analyzing single exposures acquired with polarization cameras. Images acquired with two different orientations of the polarizer have been analyzed to derive the degree of linear polarization, and the polarized brightness $pB$ in the solar corona. After inter-calibration with $pB$ measurements by the K-Cor instrument on Mauna Loa Solar Observatory (MLSO), data analysis provided the 2D coronal electron density distribution from 1.1 up to $\sim 3$ solar radii. The absolute radiometric calibration was also performed, with the full sun image, and with magnitudes of visible stars. The resulting absolute calibrations show a disagreement by a factor $\sim 2$ with respect to MLSO; interestingly, this is the same disagreement recently found with eclipse predictions provided by MHD numerical simulations.