Evaluation of Sunspot Areas Derived by Automated Sunspot-Detection Methods

TL;DR

This study evaluates automated sunspot detection methods using data from Japanese telescopes, comparing results with established catalogs, and discusses the accuracy, discrepancies, and potential calibration of sunspot area measurements.

Contribution

It provides a validation of automated sunspot detection against a reference catalog, highlighting high correlation and calibration potential for future measurements.

Findings

High correlation coefficients (0.96-0.97) with reference data.

Detected areas are 70-83% of reference areas, indicating systematic differences.

Automated detection is promising for consistent sunspot-area measurements.

Abstract

Sunspot-area measurements using digital images captured by two telescopes at the Mitaka campus of the National Astronomical Observatory of Japan are conducted using automated sunspot detection. A comparison between sunspot areas derived from Mitaka and those from the reference data by Mandal et al. ({\it Astron. Astrophys.} {\bf 640,} A78, 2020), who compiled a cross-calibrated daily sunspot-area catalog, revealed that the correlation coefficients between them are high (0.96--0.97), whereas the areas in the Mitaka data are 70 \%--83 \% of those of Mandal et al. The correlation is limited by the differences in observation times and detection capabilities of spots near the limb, with discrepancies in areas arising from different definitions of spot outlines. Given the high correlation and the ease of calibrating area discrepancies with a correction factor, automated sunspot detection appears promising for future sunspot-area measurements. Furthermore, we addressed the measurements of brightness deficit caused by sunspots.