A New Approximation for Central and Peripheral Umbral Dots Separation

Abstract We present a comprehensive new approach for categorization of umbral dots (UDs) into two groups, namely, central umbral dots (CUDs) and peripheral umbral dots (PUDs). The approach was developed and tested by utilizing high-resolution data for 18 sunspots acquired with the Goode Solar Telescope at the Big Bear Solar Observatory. We explored the existence of a general criterion to discriminate between PUDs and CUDs in using manual and automatic approaches. Since the histogram of the average UD brightness shows a bimodal distribution, the automatic/manual approach included application of the Fuzzy C-Means (FCM) clustering method to the sunspot brightness data. Expressions for the separation criterion based on UD brightness were derived for both manual and automatic approaches. The manual approach was based on the relationship between the mean umbral brightness and the mean UD brightness of the separation region, while the automatic approach uses the relationship between the mean UD brightness inside the umbra and the mean UD brightness of the separation region. Further comparison showed that the manual approach performed better than the automatic approach, and it was concluded that both methods could be used for the purpose of CUD-PUD separation.