On the prediction of geoeffectiveness of CMEs during the ascending phase of SC24 using a logistic regression method. (15th October 2019)

• Record Type: Journal Article
• Authors: Besliu-Ionescu, D.; Talpeanu, D.-C.; Mierla, M.; Muntean, G. Maris
• Is Part Of: Journal of atmospheric and solar-terrestrial physics. Volume 193(2019)
• Publication Date: 2019-10-15
• Subjects: 00-01 -- 99-00; CME -- Logistic regression; Geophysics -- Periodicals; Atmospheric physics -- Periodicals; Géophysique -- Périodiques; Météorologie physique -- Périodiques; Electronic journals; 551.51
• DOI: 10.1016/j.jastp.2019.04.017
• Languages: English
• ISSNs: 1364-6826

Abstract

Abstract: Coronal mass ejections (CMEs) are pieces of the puzzle that drive space weather. Numerous methods (theoretical, numerical and empirical) are being used to predict whether the CME will be geoeffective or not. We present here an attempt to predict the geoeffectiveness of a given CME using a modified version of logistic regression model proposed by Srivastava (2005), using only initial CME parameters. Our model attempts to forecast if the CME will be associated with geomagnetic storm defined by a minimum Dst value <−30 nT. We applied this modified logistic regression model for CMEs detected by LASCO during the ascending phase of solar cycle 24 (April 1, 2010 to June 30, 2011). Although the hit rate and proportion correctness were not promising for the training CME set, we obtained good hit rates and proportion correctness for the validation set. We expect to improve the model upon applying it to a dataset comprising an entire solar cycle. Highlights: CMEs from SC24 were more geoeffective during the ascending phase of the solar cycle. The model has provided a set of coefficients for predicting the CME geoeffectiveness. The most significant solar variables are the angular width and the central position angle. Success rates for predicting the CMEs geoeffectiveness are high.