A wealth of statistical studies has indicated the dependence of the probability of occurrence of Solar Energetic Particle (SEP) events on the magnitude and the longitude of the solar flare (SF) (Kurt et al., 2004; Belov et al., 2005; Belov, 2009), and the relation between the peak proton flux and the velocity of the coronal mass ejection (CME) (Kahler, 2001), as well as the magnitude of the SF (Cane, Richardson, and Von Rosenvinge, 2010). It has also been shown that SEP events are related to both type II and type III radio bursts (Miteva, Samwel, and Krupar, 2017). However, most studies are limited to two dimensional (2-D) correlations. In addition, similar coefficients are identifeded for the pair-wise correlation of the SEP peak intensity (at E>10 MeV) to both the SF magnitude and the CME speed, while the situation is further complicated by the fact that the solar parameters are not independent. To this end, Trottet et al. (2014) performed an analysis with partial correlation coefficients in order to disentangle the effects of correlations between the solar parameters themselves. The next step was to investigate possible 3-D relationships among three numeric variables projected in two dimensions. From such a study it was verified that the combination of strong SFs and fast CMEs result in enhanced radiation storms. Furthermore it was shown that strong SFs result in enhanced radiation effects even when associated with moderate CMEs. In addition, these strong SFs can lead to major radiation storms even when they are not situated on the west part of the visible solar disk (Papaioannou et al., 2016). Therefore, aiming at a higher dimensional order correlations seems to be the way forward. Given the complexity of the parent solar events of SEPs (e.g. SFs, CMEs) and the different variables (e.g. GOES peak photon flux, longitude of the SF, velocity and width of the CME) that give rise to their peak proton flux, possible new methods for the nowcasting of SEP events have to be associated with more accurate mathematical methods of statistical analysis.

One method that can be used is the Principal Component Analysis (PCA); a multivariate statistical technique being used to examine the interrelations among a set of variables (e.g. a dataset) aiming to identify the underlying structure of those variables (Jolliffe, 2014).

A. Papaioannou, A. Anastasiadis, A. Kouloumvakos, M. Paassilta, R. Vainio, E. Valtonen, A. Belov, E. Eroshenko, M. Abunina, A. Abunin: Nowcasting Solar Energetic Particle (SEP) Events using Principal Components Analysis (PCA), Solar Physics, DOI:10.1007/s11207-018-1320-7, 293:100, 2018 | available here

A. Papaioannou, A. Anastasiadis, A. Kouloumvakos, M. Paassilta, R. Vainio, E. Valtonen, A. Belov, E. Eroshenko, M. Abunina, A. Abunin: Nowcasting Solar Energetic Particle (SEP) Events using Principal Components Analysis (PCA), European Geosciences Union (EGU) General Assembly, 08 - 13 April 2018, Vienna, Austria [abstract] & [presentation]

A. Papaioannou, A. Anastasiadis, A. Kouloumvakos, M. Paassilta, R. Vainio, E. Valtonen, A. Belov, E. Eroshenko, M. Abunina, A. Abunin: Creating an index for Solar Energetic Particle (SEP) events using multivariate analysis, European Space Weather Week 14, 27 November - 01 December 2017, Oostende, Belgium [presentation]

A. Papaioannou, A. Anastasiadis, A. Kouloumvakos, M. Paassilta, R. Vainio, E. Valtonen, A. Belov, E. Eroshenko, M. Abunina, A. Abunin: On the usage of Principal Components Analysis (PCA) for the Prediction of Solar Energetic Particle (SEP) events, European Space Weather Week 15, 05 -09 November 2018, Leuven, Belgium [presentation]