We applied subsequently the different type of methods such as Principal Component Analysis (PCA), adaptive Kalman filtering, histogram analysis and several others on large number of events. Finally we found that all EIT waves and their dimmings can be successfully described by small number of variables in special polar coordinate system introduced on the solar sphere. We could after establish the main mathematical characteristics of EIT waves and dimmings.
Based on the previous two items, we did the classification of EIT waves and dimmings depending on their morphology, dimming behavior and two different types of quantitative behavior corresponding to different solutions. We decided to restrict the detection of dimming by only EUV dimmings associated with CME. Topologically we divide EIT waves on symmetrical, asymmetrical, freely propagating and interacting with other active regions classes.
Based on previous items we finally established two criteria of EIT waves and dimmings detection. They are based on the higher order moments (HOM) technique of Fixed difference of EUV images. HOM technique is largely used for coherent structures extraction from noise and in the turbulent signals analysis.
Validation is underway using large events catalogs, but preliminary results show a high success rate including for faint events. Currently, the method is likely to fail in particular for dimmings without EIT waves and highly corrupted images. The previous steps of the method are being refined to improve detection of wider and wider classes of dimmings.
Actually space weather forecasters does nor provide daily catalogs of EIT waves because of difficulties of events creation 'by eyes'. That's why we provide the addition package of EIT wave and dimmings extraction from data. Dimmings extraction work also automatically. Full EIT wave parameters definition such as differential velocity of EIT wave front, angular velocity of EIT wave front are defined semi-automatically, and can be compared with theoretical models. |
|