Solar Energetic Particles and Associated EIT Disturbances in Solar Cycle 23

Abstract

We explore the link between solar energetic particles (SEPs) observed at 1 AU and large-scale disturbances propagating in the solar corona, named after the Extreme ultraviolet Imaging Telescope (EIT) as EIT waves, which trace the lateral expansion of a coronal mass ejection (CME). A comprehensive search for SOHO/EIT waves was carried out for 179 SEP events during Solar Cycle 23 (1997 – 2006). 87 % of the SEP events were found to be accompanied by EIT waves. In order to test if the EIT waves play a role in the SEP acceleration, we compared their extrapolated arrival time at the footpoint of the Parker spiral with the particle onset in the 26 eastern SEP events that had no direct magnetic connection to the Earth. We find that the onset of proton events was generally consistent with this scenario. However, in a number of cases the first near-relativistic electrons were detected too early. Furthermore, the electrons had in general only weakly anisotropic pitch-angle distributions. This poses a problem for the idea that the SEPs were accelerated by the EIT wave or in any other spatially confined region in the low corona. The presence of weak electron anisotropies in SEP events from the eastern hemisphere suggests that transport processes in interplanetary space, including cross-field diffusion, play a role in giving the SEPs access to a broad range of helio-longitudes.

Appendix: EIT Waves and Their Relationship with Other Coronal Phenomena

We compared the occurrence of the EIT disturbance with the reports of shock signatures in the corona, and we found in general a high correlation, see the contingency table: Table 6. The majority of EIT waves (88 %, 112/127) are associated with signatures of a shock wave in the corona. The same percentages are found for the western and eastern sample. However, we would like to note that the data set used is only a subset of all EIT waves in Solar Cycle 23, since we started with a SEP list. Moreover, the events for which no data were found (EIT or radio) are in general dropped from the analysis.

Table 6 Contingency table for EIT waves and Type-II radio bursts.

The opposite association rate is not fully investigated. Klassen et al. (2000) presented a list of 21 shock signatures during 1997 where in 90 % of the cases the metric Type-II radio burst had an associated EIT wave. No correlation was found between the speed of the Type-II driver and the EIT wave, as the radio signature of the shock wave was found to be about three times faster than the EIT wave.

In addition, the onset time of the EIT disturbances [t _EIT,on] can be compared with the onset time of signatures of a shock wave propagating through the solar corona, i.e. metric Type-II radio bursts (Nelson and Melrose 1985). For the onset time of the western EIT waves, we performed a time shift by six minutes from the time of the first observed front. We expect an underestimation of this time only for a minority of the events (e.g. when the EIT cadence was actually longer). For the Type-II signatures, we used the reported onset times by different radio observatories at metric wavelengths.

The time difference t _II−t _EIT,on is shown in Figure 8. The number of events is represented by the length of the bars in the histograms. A larger spread than the five minutes reported by Warmuth (2010) is present. Due to the large uncertainty of t _EIT,on (of the order of the SOHO/EIT temporal cadence), the slight shift in the distributions toward positive delays (implying that the Type-II burst starts on average seven minutes after the onset of the EIT wave), is not significant. One can conclude that in a majority of the cases (88 %), at the time of the EIT disturbance a shock wave is also present in the corona.

Figure 8

Histogram for the delay between the onset of the Type-II metric radio burst and the EIT wave onset t _II−t _EIT,on for eastern (left) and western events (right).

For three events among the eastern events under study, only single (or uncertain) wave front(s) could be identified and hence no EIT wave speed could be obtained. Another three SEP events that are labeled with no EIT wave [N] may in fact be associated with a later EIT disturbance that accompanied another flare/CME. They are given with “d” in Table 2. Since no conclusion can be made, these events are dropped from the present analysis. In summary, only 26 events were found with an associated EIT disturbance in at least two subsequent images. For those, the average EIT wave speed is in the range between 170 and 670 km s⁻¹. A histogram of the EIT wave speed is given in Figure 9. The different colors indicate the events in different IP magnetic-field configuration, namely with light gray indicating the solar-wind events, with dark gray indicating the ICME events, and with black indicating particle events propagating in the vicinity of an ICME. Recently, a kinematics classification was proposed by Warmuth and Mann (2011), where all disturbances are divided into three groups: slow EIT signatures (≲ 170 km s⁻¹, due to magnetic-field reconfiguration), signatures with constant speed (in the range of 170 – 320 km s⁻¹, interpreted as linear waves traveling at the local fast-mode speed), and fast disturbances (≳ 320 km s⁻¹, probably large-amplitude waves or shocks). Under this classification the eastern EIT disturbances are all linear waves and/or shocks.

Figure 9

Histogram of the EIT wave speed (left) and scatter plots between the EIT speed and the associated flare size (right) for eastern particle events. In the histogram, the light-gray color denotes the solar-wind events, gray the ICME events, and black the event in the vicinity of an ICME. The number of the events in each category is given by the bar length of specific shades of gray. In the scatter plot, the open circles denote the solar-wind events, filled circles the ICME events, and the filled squares the events in the vicinity of an ICME.

In addition, we present the scatter plots between the propagation speed of the EIT disturbance and the properties of the associated flare SXR size (in Figure 9, right) and/or CME linear speed and/or angular width (in Figure 10). We found that the averaged EIT wave speed is neither correlated with the flare, nor with the CME properties.

Figure 10

Scatter plots between the averaged EIT speed and the associated CME linear speed (left) and angular width (right) for eastern particle events. The open circles denote the solar-wind events, filled circles the ICME events, and the filled squares the events in the vicinity of an ICME.