Abstract

We study the properties of the complex network for solar proton flux (SPF) at six integral energy channels (>1, >5, >10, >30, >50, and >100 MeV). We construct a growing visibility network based on the SPF time series for proton for each channel, which observed by Geostationary Operational Environmental Satellites from April 14, 2010 to May 9, 2018. The clustering coefficient values for SPF networks are much larger than those of the equivalent random network. The logarithmic behavior of the characteristic path length refers to the characteristics of the small-world network. The degree distribution of nodes obeys the power-law behavior with certain exponents that reflects the scale-free nature of the networks. For the constructed SPF networks, the average clustering coefficient shows a jump when the large flares occurred. We show that both western solar longitude large flares (X-class, M-class) and flares associated with coronal mass ejections (with speed greater than 500 km s−1) are relatively more related to the peaks (more than the median) of both SPF time series and PageRank of the networks. We conclude that both the PageRank and clustering coefficient of the SPF network are useful indicators to identify the source of SPFs related to flares and flares associated with coronal mass ejections.