Fig. 3

a Long-term variation of the radiation power of the IPT in the shorter spectral range (65–78 nm) versus day-of-year (DOY) in 2015. The brightness on dawn and dusk sides was averaged. b Time series of each emission brightness from singly (S II), doubly (S III), and triply (S IV) ionized sulfurs and oxygen ions (OII/OIII). The timing of the rise seen for singly ionized sulfur is the earliest, 5 days earlier than the volcanic eruption identified from ground-based observations in the infrared region (de Kleer and de Pater 2016b). The triangles mark the times of the brightenings correlated with the internally driven aurorae. After DOY 40, the brightenings are clear in this plot. c Aurora light curve with spectral binning (90–147 nm) applied. The circles indicate the times when the internally driven transient events of aurorae started. After DOY 40, transient events occurred more frequently and became brighter. d The same as (c) but with an enlarged scale to clearly show small transient events before DOY 40. e Simulated dynamic solar wind (SW) pressure. We extrapolated SW data measured on Earth to Jupiter with a simulation of the magnetohydrodynamic (MHD) propagation. The SW with high dynamic pressure is expected to arrive at Jupiter (e.g., DOY 10, 11, 47, 66 and 74). The red line indicates our criteria for defining quiet SW conditions for Jupiter. Note that the marks in (b)–(d) point to the period with a dynamic SW pressure below 0.1 nPa