Automated Detection of Chromospheric Swirls Based on Their Morphological Characteristics

High-resolution observations have revealed that rotating structures known as “chromospheric swirls” are ubiquitous in the solar chromosphere. These structures have circular or spiral shapes, are present across a broad range of spatial and temporal scales and are considered as viable candidates for providing an alternative mechanism for the heating of the chromosphere and corona. Therefore, an accurate determination of their number and a statistical study of their physical properties are deemed necessary. In this work we present a novel, automated swirl detection method, which utilizes image pre-processing, curved structure tracing and machine learning techniques that allow for the detection of swirling events based on their morphological features as they appear in chromosphere filtergrams. The method is applied to H $\alpha $ chromospheric spectral line images obtained by the CRisp Imaging Spectropolarimeter (CRISP) at the Swedish 1-m Solar Telescope (SST). It is also tested on grayscale images of vortical sea current flows represented/visualized by synthetic streamlines from the NASA/Goddard Space Flight Center Scientific Visualization Studio. The results are rather encouraging since swirling events are successfully identified. Further improvements of the algorithm, its prospects for the detection and statistical studies of the properties of these events using a wide range of imaging data and its potential application in other scientific fields for the detection of rotating motions are discussed.