Distinct Formation and Evolution Characteristics of Outer Radiation Belt Electron Butterfly Pitch Angle Distributions Observed by Van Allen Probes
Abstract
Using Van Allen Probes pitch angle-resolved electron flux data, we report intriguing events of electron butterfly pitch angle distributions (PADs) at L > ~5.5 on the nightside with distinct durations, i.e., about 9 hr for the 28–29 April 2013 event and over 2 days for the 2–5 May 2013 event. The formation and evolution of the short-duration electron butterfly PADs exhibited insignificant dependence on electron energy, probably resulting from a combination of drift shell splitting, magnetopause shadowing, and wave diffusion, but the long-duration electron butterfly evolution presented strong energy dependence, likely mainly attributed to interactions with MS waves. Both the short-duration and long-duration electron butterfly PADs disappeared closely related to enhanced substorm activities, while the responsible waves could be plasmaspheric hiss or chorus waves. Our results demonstrate that radiation belt electron butterfly PADs are naturally complex in both spatial and temporal scales and well connected to solar wind condition, substorm activity, and magnetospheric wave distribution.
Key Points
- Radiation belt electron butterfly PADs exhibited both short-duration (~ hours) and long-duration (~ days) variations along with strong dependence on energy, L-shell and MLT
- The short-duration electron butterfly PADs is probably a combined result of drift shell splitting, magnetopause shadowing, and wave diffusion, but the long-duration butterfly evolution is more likely mainly attributed to interactions with MS waves
- Both the short-duration and long-duration electron butterfly PADs disappeared closely related to enhanced substorm activities, while the responsible waves could differ
Plain Language Summary
Due to dynamic variations of magnetospheric physical processes, radiation belt electrons show different kinds of pitch angle distributions (PADs). Among them, butterfly PADs, recognized as a local flux minimum at 90° pitch angle, have received intensive attention but remain not fully understood yet. The contributing mechanisms include drift shell splitting, magnetopause shadowing and associated outward radial diffusion, and wave-particle interactions. Using high-quality Van Allen Probes Magnetic Electron Ion Spectrometer, Relativistic Electron-Proton Telescope, and Electric and Magnetic Field Instruments Suite and Integrated Science data sets, we register two interesting events of electron butterfly PADs at L > ~5.5 on the nightside with very distinct evolution profiles both in time and energy, i.e., ~9 hr duration with insignificant energy dependence for the 28–29 April 2013 event and over 2-day duration with strong energy dependence for the 2–5 May 2013 event. Detailed analyses indicate that while the short-duration butterfly PADs probably result from a combination of drift shell splitting, magnetopause shadowing, and wave diffusion, the long-duration butterfly PADs are likely to be mainly caused by interactions with MS waves. Our study therefore reports the natural complexity of outer zone electron butterfly PADs in both spatial and temporal scales and furthers current understanding of their connection with solar wind condition, substorm activity, and magnetospheric wave distribution.
1 Introduction
Earth's radiation belts, also called Van Allen belts, are the spatial regions where energetic charged particles are trapped by the geomagnetic field and consist of both inner and outer zones with the slot region devoid of ~>500 keV electrons in between (Allen & Frank, 1959). Generally, the outer radiation belt covering L = 3–8 is highly dynamic and varies largely with geomagnetic activities when interacting with a variety of magnetospheric waves (e.g., Baker et al., 2004; Reeves et al., 2012; Shprits et al., 2009; Thorne, 2010; Xiao et al., 2009).
As a significant feature of radiation belt electrons, the evolution of pitch angle distributions (PADs) provides important information about various physical processes in a specific region (e.g., Reeves et al., 2003; Xiang et al., 2020). A number of previous studies have shown extensive interest in the characteristics and evolution of electron PADs in the outer radiation belt (e.g., Chen et al., 2014; Gannon et al., 2007). In general, there are three major categories of electron PADs: 90° peaked (or normal), flattop (or pancake), and butterfly (Fritz et al., 2003; Gannon et al., 2007; Gu et al., 2011; Horne et al., 2003; Meredith et al., 2000; Zhao et al., 2014a, 2014b).
Butterfly PADs are identified by local flux minima around 90° pitch angle (e.g., Fritz et al., 2003; Ni et al., 2015, 2016; Zhao et al., 2014a). There are a number of physical mechanisms that can contribute to the occurrence of radiation belt electron butterfly PADs. It has been long established that, caused by the local azimuthal asymmetry of Earth's magnetic field, drift shell splitting acts as a common process to account for the formation of electron butterfly PADs in the outer zone where the asymmetry is relatively pronounced, e.g., L > ~5 (Fritz et al., 2003; Selesnick & Blake, 2002). Corresponding to enhanced solar wind dynamic pressures and resultant inward intrusion of the magnetopause, the effect of magnetopause shadowing occurs to result in losses of electrons on open drift shells (Turner et al., 2012). Equatorially mirroring electrons on the dayside drift farther than those at other pitch angles, hence being more likely to escape through the magnetopause and generating butterfly PADs (Sibeck et al., 1987). In association with the strong negative gradient in electron flux radial profiles caused by losses to the magnetopause, outward radial diffusion further brings electrons to higher L-shells to mitigate the gradient (Shprits et al., 2006). In addition, wave-particle interactions are essential to the formation of radiation belt electron butterfly PADs. Magnetosonic waves have been proposed to cause butterfly distributions of both subrelativistic and relativistic electrons due to parallel acceleration through Landau resonance between the waves and electrons (e.g., Li et al., 2014; Li, Bortnik et al., 2016; Li, Ni et al., 2016; Ma et al., 2015; Ni et al., 2017, 2018). Some recent studies reported that the combined effect of the Landau and bounce resonance caused by MS waves tends to reduce the net scattering rates of near equatorially mirroring electrons (Chen et al., 2015; Fu et al., 2019) but is still able to modulate the electron butterfly distribution (Maldonado et al., 2016). Scattering by plasmaspheric hiss is another candidate mechanism for radiation belt electron butterfly PADs at lower L-shells when the wave-induced cross-diffusion is included (Albert et al., 2016). Li et al. (2019) proposed that hiss waves can rapidly accelerate suprathermal electrons through Landau resonance and explain field-aligned suprathermal electron distributions measured by Van Allen Probes where parallel fluxes are an order of magnitude higher than perpendicular fluxes. Through a detailed analysis, Xiao et al. (2015) proposed that combined acceleration by chorus and magnetosonic waves can explain the electron flux evolution both in the energy and butterfly PAD. It is also reported that the formation of electron butterfly distribution can be attributed to the effect of off-equatorial chorus waves (Horne et al., 2003) or dayside chorus (Jin et al., 2018). Some other possible mechanisms responsible for the formation of electron butterfly distribution include a reduction in kinetic energy due to conservation of the first adiabatic invariant (Thorne et al., 1999) and the modulation by localized background magnetic field perturbation (Artemyev et al., 2015; Xiong et al., 2017). Intrinsically, to trigger electron butterfly PADs in geospace, the effects of drift shell splitting, magnetopause shadowing, and outward radial diffusion are energy independent, but the effects of particle interactions with various magnetospheric waves are energy dependent.
High-quality pitch angle-resolved electron flux data from Van Allen Probes Magnetic Electron Ion Spectrometer (MagEIS) and Relativistic Electron-Proton Telescope (REPT) instrument provide a unique opportunity to study the characteristics and causes of radiation belt electron butterfly PADs. While a number of statistical analyses have adopted long-term Van Allen Probes data to study the global distribution and occurrence pattern of electron butterfly PADs, this letter investigates in detail the evolution of radiation belt electron butterfly PADs observed by Van Allen Probes during the two intriguing events with distinct durations, i.e., within 2 orbits for the 28–29 April 2013 event and 7 orbits for the 2–5 May 2013 event. For these two events, there is no geomagnetic storm but moderate substorm activity in coexistence with plasma waves such as MS waves and plasmaspheric hiss. By evaluating the differences in the formation and evolution of the short- and long-duration electron butterfly distributions, we intend to understand the underlying physics that may contribute to their distinct responses to different conditions of solar wind, geomagnetic disturbance, and magnetospheric waves.
2 Data and Methodology
Launched by NASA on 30 August 2012, the Van Allen Probes mission with twin satellites has a geosynchronous transfer orbit, with a perigee of 1.09 Earth radii (RE) and an apogee of 5.77 RE with respect to the Earth's center, an inclination of 10.2°, and an orbital period of ~9 hr. The outermost L-shell reached by Van Allen Probes is L ~ 7, and two probes fly in nearly identical eccentric orbits that cover the entire radiation belt region.
In the present study, we use pitch angle-resolved electron flux data obtained by MagEIS (Blake et al., 2013) and REPT (Blake et al., 2013) onboard Van Allen Probes to examine radiation belt electron PADs. With regard to wave data acquisition, we use data obtained by Electric and Magnetic Field Instruments Suite and Integrated Science (EMFISIS) instrument (Kletzing et al., 2013) to quantify the wave amplitudes of chorus, plasmaspheric hiss and MS waves.
According to the study of Claudepierre et al. (2019), in the inner radiation belt (L < 2) and slot region (2 < L < 3), energetic electron PADs can be easily contaminated from high energy inner belt protons (e.g., Zhao et al., 2014a) and from occasionally coexisting high energetic protons in the slot region (Baker et al., 2004). Therefore, the present study focuses on investigation of electron PADs in the outer zone (3 < L < 7).
To identify an electron butterfly PAD, we follow the criteria in the studies of Zhao et al. (2014a) and Ni et al. (2016), which are outlined below:
(1) j90° < β × javg (a:b). In this inequation, we define j90° as electron flux at 90° pitch angle and javg (a:b) as the average electron flux for which the pitch angle ranges from a to b to determine the peak value of javg (a:b). Note that, for ultrarelativistic electrons measured by REPT, a = [26.5°; 37.1°; 47.6°; 58.2°; 68.8°; 79.4°; 90°] and b = 180° − a. For the electron energy channels measured by MagEIS, a = [40.9°; 57.3°; 73.6°; 90°] and b = 180° − a. We take these pitch angles by considering both the data availability and separation of a butterfly distribution from a quasi-parallel distribution. In addition, β = 0.95, denote the threshold of butterfly PAD index, which means that a butterfly PAD exists when the β-index < 0.95.
(2) In order to avoid contamination from the background noise, we have discarded some ultrarelativistic electron flux data with the level less than 10−2/cm2/s/sr/MeV (Baker et al., 2013). For electrons fluxes measured by MagEIS, we assign the threshold flux value as 102/cm2/s/sr/MeV.
We subsequently follow the above criteria to investigate the occurrence and evolution of outer zone electron butterfly distributions for five representative energy channels, i.e., 143 keV, 597 keV, 909 keV, 1.65 MeV, and 3.4 MeV.
3 Two-event Analysis
3.1 The 28–29 April 2013 Event: Short-duration Electron Butterfly PADs
Figure 1 displays the first event of radiation belt electron butterfly PADs from 00 UT on 28 April 2013 to 10 UT on 29 April 2013 observed by Van Allen Probe A. Figure 1a presents the time series of AE (blue curve) and Dst (red curve) indices. They varied from 70 to 630 nT and from −20 to 0 nT, respectively, indicating that there is no geomagnetic storm but moderate substorm activity. Figure 1b shows the temporal variations of solar wind dynamic pressure (Pdyn, red curve) and magnetopause standoff distance (Lmp, blue curve) calculated following the model of Shue et al. (1997). The magnetopause was generally located at L > 10 but moved inwards to L ~ 9 during the period of enhanced Pdyn. Figure 1c displays the integrated amplitude of chorus (blue curve), plasmaspheric hiss (green curve) and MS waves (red curve) for the considered event. We note that the corresponding wave frequency spectrogram, wave normal angle, and ellipticity are shown in Figure S1 in the supplementary information. It is seen from Figure 1c that MS waves occurred occasionally and were generally weak but reached a peak amplitude ~300 pT at L ~ 5.8 at 11:32 UT on 28 April 2013. Similarly, chorus waves were primarily weak outside the plasmasphere but reached a peak amplitude >200 pT around 14 UT on 28 April. In contrast, hiss emissions occurred more frequently with the wave amplitude varying from a few pT to ~50 pT. Figures 1d–1h illustrate the color-coded differential fluxes of radiation belt electrons measured by MagEIS and REPT for the indicated five energies as a function of pitch angle and time. While the electron distributions from 143 keV to 3.4 MeV exhibited the 90°-peaked profiles during the first orbit, electron butterfly PADs formed quickly during the second orbit, lasted but weakened during the third orbit, and completely disappeared during the fourth orbit. Interestingly, during the second orbit, the electron fluxes increased sooner or later for all the five energies, in parallel with the enhanced substorm activity indicated by the AE index (Figure 1a). Overall, the electron butterfly PADs occurred following the enhancement of Pdyn, inward magnetopause intrusion, and intensified wave activities (especially MS and chorus waves) and existed for a rather short time duration of ~9 hr. According to the simultaneous EMFISIS, REPT, and MagEIS measurements from Van Allen Probe B (see Figures S2 and S3 in the supplementary information), the electron butterfly distribution for this event lasted ~9 hr and disappeared closely corresponding to the 29 April substorm activity. In addition, the strong activity of MS waves was also captured by Van Allen Probe B, in synchronization with the occurrence of electron butterfly distribution, further supporting the scenario that MS waves play an important role in the formation and evolution of the latter for this short-duration event.
A more detailed check is conducted regarding the evolution characteristics of electron butterfly PADs for the 28–29 April 2013 event, the results of which are shown in Figure 2. Figures 2a–2e present the ratio j90°/javg for the five energies at L ≥ 3 during the four orbits corresponding to Figure 1. In each panel, the horizontal black solid line denotes the threshold value β = 0.95. Therefore, the red points correspond to β < 0.95 and present butterfly PADs, while the blue points denote no butterfly PADs. Figures 2f and 2g provide the corresponding temporal variations of L-shell and magnetic local time (MLT) so that the spatial location of electron butterfly distribution can be exactly identified. There are a number of interesting features of the formation and evolution electron butterfly PADs for this short-duration event: (1) the electron butterfly PADs for all the five energies formed almost simultaneously at L > ~5.7 on the nightside with MLT ~22–02; (2) evolving distinctly with time, the butterfly distribution for 143 keV electrons switched into the 90°-peaked distribution right after one orbit (orbit 3), and the butterfly PADs for the other four higher energy electrons weakened in the following orbit and then disappeared during the fourth orbit; (3) indicated by the j90°/javg ratio, the most pronounced butterfly distributions were constantly present for 3.4 MeV electrons. The electron butterfly distributions became weaker with decreasing electron energies and/or lower L-shells.
3.2 The 2–5 May 2013 Event: Long-duration Electron Butterfly PADs
Figure 3 depicts the 2–5 May 2013 event of radiation belt electron butterfly PADs observed by Van Allen Probe A for a much longer time period of over 2 days. Figures 3a and 3b indicate that similar to the first event, there was no geomagnetic storm but certain substorm activity. In addition, the solar wind dynamic pressure remained very weak so that the magnetopause was nominally located well above L ~ 10. Figure 3c shows that intense MS waves with the peak amplitude ~400 pT merely occurred at L = 4.8 around 19:00 UT on 4 May 2013 throughout the event. Chorus waves remained at a weak level during the majority period of the event. By comparison, hiss waves occurred more frequently and were normally weak (i.e., <~25 pT) except around 12 UT on 2 May 2013 when the waves reached ~150 pT. Similar to Figure 1, the corresponding wave frequency spectrogram, wave normal angle, and ellipticity are shown in Figure S4 of the supplementary information.
As shown in Figures 3d–3h, the radiation belt electron PADs were quite dynamic during this 3.5-day period. In association with the substorm injections during the first orbit, electron fluxes increased for all the five energies during the second orbit. Similar to the first event, the electron butterfly PADs occurred on the nightside of MLT ~22–02. While there was no clear signature of butterfly PADs for 143 keV electrons during the entire event, electron butterfly distributions formed earlier or later and lasted shorter or longer for the other four higher energies. There exists a clear tendency that butterfly PADs were more likely to occur at higher energies for this long-duration event. Specifically, electron butterfly PADs occurred slightly during the second and eighth orbits for 597 keV, but they became increasingly apparent during the second, seventh, and eighth orbits for 909 keV and during the second, fifth, sixth, seventh, and eighth orbits for 1.65 MeV. More strikingly, 3.4 MeV electrons exhibited pronounced butterfly PADs continuously for the second to eighth orbits, lasting over 2 days. All the electron butterfly PADs disappeared during the ninth orbit in association with considerable flux decreases for all the five electron energies. It is interesting to see that the deepening of 597 keV to 3.4 MeV electron butterfly PADs was synchronized with the occurrence of strong MS wave activity (~400 pT) during the time interval of ~19–20 UT on 4 May 2013 (orbit 8), implying a good connection between them. A similar synchronization also exists between the more pronounced butterfly PADs for 1.65 and 3.4 MeV electrons and enhanced MS waves ~18 UT on 3 May 2013 (orbit 5). Complementary to Van Allen Probe A observations, Van Allen Probe B also captured the duration of electron butterfly distribution over 2 days (see Figure S5 in the supplementary information), in a quite similar manner of evolution. In contrast, Van Allen Probe B observed strong chorus waves with the amplitudes up to ~300 pT around 21 UT on 2 May (see Figures S5 and S6), which may contribute to the disappearance or flattening of electron butterfly distribution at this time interval (Yang et al., 2016).
Figure 4 displays further analysis results of the formation and spatiotemporal evolution of electron butterfly PADs for this long-duration event, similar as in Figure 2. Distinct from the first event, this second event is remarkable in that the electron butterfly PADs lasted for seven orbital periods, i.e., ~2.5 days, and were continuously pronounced for 3.4 MeV only. The butterfly j90°/javg ratio for 3.4 MeV electrons varied between 0.28 and 0.95, with two peak values during the fifth and eighth orbits, respectively. The butterfly PADs for 1.65 MeV electrons occurred less frequently but also evidently, exhibiting a rather weak profile during the second orbit but increasingly deepened structures from the sixth to eighth orbit. According to the j90°/javg ratios for the three lower energy electrons, butterfly PADs were present occasionally for 597 and 909 keV but absent for 143 keV. In contrast to the spatial confinement of 597–909 keV electron butterfly PADs at L ~ 5.6–6.4, the butterfly PADs extended to broader L-shell coverage for higher energy electrons, i.e., L ~ 5.6–7.0 for 1.65 MeV and L ~ 5.3–7 for 3.4 MeV, while the electron butterfly PADs became less pronounced with decreasing L-shells.
4 Discussions
In order to understand the underlying physics responsible for the formation and evolution of radiation belt electron butterfly PADs during the presented short-duration and long-duration events, Table 1 provides a comparative summary of the occurrence and evolution characteristics of the two electron butterfly PAD events in terms of solar wind conditions, geomagnetic activity level, and plasma wave activity, based on which the possible contributing mechanisms are reasonably explored.
| Short-duration butterfly PAD event (28–29 April 2013) | Long-duration butterfly PAD event (2–5 May 2013) | |
|---|---|---|
| Solar wind conditions | Vsw: 400–600 km/s | Vsw: 350–500 km/s |
| IMF BZ: −4–4 nT | IMF BZ: −5–5 nT | |
| Pdyn: 0–3 nPa (increased at first and then decreased) | Pdyn: 0–2 nPa (decreased) | |
| Geomagnetic conditions | Dst: −20 – 0 nT (no geomagnetic storm) | Dst: −30–0 nT (no geomagnetic storm) |
| AE: 70–630 nT (substorms) | AE: 50–700 nT (substorms) | |
| Magnetopause location | ~9–12 RE (inward magnetopause movement) | ~10–12 RE |
| Wave activities | MS waves: occurred occasionally and generally weak; strongest amplitude ~300 pT at L ~ 5.8 at 11 UT on 28 April | MS waves: occurred infrequently and weakly during the entire period; intense at L ~ 5.6 near 19 UT on 4 May 2013, reaching ~400 pT |
| Chorus waves: primarily weak outside the plasmasphere but reached a peak amplitude >200 pT around 14 UT on 28 April | Chorus waves: remained at a weak level during the majority period of the event but reached up to ~300 pT around 21 UT on 2 May | |
| Hiss waves: frequently present and varied from a few pT to ~50 pT; intensified associated with the substorm activity around 00 UT on 29 April | Hiss waves: occurred much more frequently and were normally weak except around 12 UT on 2 May 2013 when the waves reached ~150 pT | |
| Outer zone electron butterfly PAD features | • Formed simultaneously for all five electron energies at L > ~5.6 on the nightside of MLT ~21–01 | • No butterfly for 143 keV electrons; formed for the four higher electron energies (597 keV to 3.4 MeV) at L > ~5.3 at MLT ~20–01 |
| • Lasted for ~9 hr and became more pronounced with increasing electron energy and L-shell | • Lasted for >2 days for 3.4 MeV, >1 day for 1.65 MeV, but quite occasionally for 597 and 909 keV; extended to broader L-shell coverage for higher energy electrons and became more pronounced with increasing electron energy and L-shell | |
| • Exhibited the strongest butterfly distribution for 3.4 MeV electrons with the j90°/javg ratio reaching down to ~0.5 | ||
| • Related closely to the MS wave activity and gradually weakened with time | • Exhibited the strongest butterfly distribution for 3.4 MeV electrons with the j90°/javg ratio reaching down to ~0.28 | |
| • Disappeared for all the four higher energies following the substorm occurrence ~00 UT on 29 April | • Became deepened along with enhanced MS waves and absent for all the four higher energies following enhanced substorm activity ~03 UT on 5 May | |
| Possible mechanisms | Combination of drift shell splitting, magnetopause shadowing and interactions with MS waves | Wave-particle interactions dominantly with MS waves |
The 28–29 April 2013 event illustrates the occurrence and evolution of electron butterfly PADs from subrelativistic to ultrarelativistic energies at L > ~5.5 on the nightside within short duration of ~9 hr. The simultaneous formation of electron butterfly PADs at all the five energies, along with the inward magnetopause movement to L ~ 9 due to the solar wind dynamic pressure increase, supports that drift shell splitting and outward radial diffusion in association with magnetopause shadowing can dominantly contribute to the butterfly PAD distribution at the initial phase. However, the varying depth of the formed butterfly PADs with electron energy suggests that wave-particle interactions also play a nonnegligible role, especially when the coexistence of strong MS waves with the electron butterfly PADs is taken into account. Overall, it is likely that the formation of electron butterfly PADs for this short-duration event is a result of the combination of all these three physical processes. They may take effect individually or jointly but all in a short time interval. After the formation, electron butterfly PADs lasted for ~9 hr and then completely disappeared for all the five energies, possibly due to the enhanced substorm activity at the beginning of 29 April and ensuing elevation of hiss wave intensity (Figures 1a and 1c).
For the 2–5 May 2013 event, since the solar wind dynamic pressure was relatively weak, the effect of magnetopause shadowing and outward radial diffusion can be excluded for interpretation of the formation of electron butterfly PADs at L < ~ 7. The strong energy dependence indicated by the formation and evolution of electron butterfly PADs for this event complies favorably with the scenario of wave-particle interactions. There are a number of evident features supporting the dominance of this mechanism: (1) the formation of electron butterfly PADs during the second orbit and their deepening during the fifth and eighth orbits was well synchronized with the intensification of MS waves; (2) there was a tendency that more intense MS waves could produce more pronounced electron butterfly profiles (Lei et al., 2017), particularly when the MS wave amplitude was strongest ~400 pT around 19–20 UT on 4 May; (3) electron butterfly PADs were more likely to occur at ultrarelativistic energies (i.e., 3.4 MeV) than lower energies (i.e., 597 keV to 1.65 MeV), consistent with the theoretical analysis that outside the plasmasphere, MS wave-induced electron diffusion tends to be stronger at higher energies (e.g., Li et al., 2014). It is worthwhile to note that the common effect of drift shell splitting due to the natural day-night asymmetry of the geomagnetic field cannot be ruled out but its contribution may be small compared to the effect of wave-particle interactions. After the electron butterfly PADs evolved and deepened for the four higher energies during the eighth orbit, they all disappeared in the following orbit. Such a behavior may result from enhanced substorm activities around 03 UT on 5 May, in a manner similar to that for the first event, but the mechanism responsible for the butterfly disappearance may be some other wave mode rather than plasmaspheric hiss, e.g., whistler-mode chorus (Xiao et al., 2015; Yang et al., 2016).
It is worthwhile to note that magnetosonic waves can occur more frequently than measured by Van Allen Probes because the twin satellites are mostly confined near the geomagnetic equator, in other words, often off the equator where MS waves are characteristically absent. In addition, the distinct features of the solar wind dynamic pressure for the two events result in large difference in the magnetopause location, which is significant for distinguishing the contribution of drift shell splitting and magnetopause shadowing from that of wave-particle interactions. While the present study has focused on the formation and evolution characteristics of electron butterfly PADs in the outer radiation belt (L ≥ 3), electron butterfly distributions can also occur in the inner belt and slot region (e.g., Zhao et al., 2014a, 2014b), predominantly due to wave-particle interactions, which however is outside the scope of the present study.
5 Conclusions
Using Van Allen Probes high quality wave and particle data, in this study, we have investigated in detail the formation and evolution of two distinct radiation belt electron butterfly PAD events, i.e., one featured by short duration within ~9 hr and relatively weak energy dependence and the other featured by long-duration ~2.5 days and pronounced energy dependence. By evaluating the characteristics of the two electron butterfly PAD events, we have explored the underlying physics likely responsible for such dynamic and diverse variations of radiation belt electron butterfly distributions.
- Radiation belt electron butterfly PADs can exhibit short-duration (~ hours) and long-duration (~ days) variations, both showing strong dependence on energy, L-shell, and MLT but distinct spatiotemporal evolution processes.
- Featured by the simultaneous formation of electron butterfly PADs at 143 keV to 3.4 MeV along with the inward magnetopause movement to L ~ 9 and the evolving depth of the butterfly PADs with electron energy, the short-duration electron butterfly PADs is probably a combined result of drift shell splitting, magnetopause shadowing, and wave diffusion.
- For the long-duration event, there exists a clear tendency that electron butterfly PADs were more likely to occur at higher energies (1.65–3.4 MeV) and a good correlation between the deepening of 597 keV to 3.4 MeV electron butterfly PADs and enhanced MS wave activities, supporting that the long-duration butterfly evolution is more likely mainly attributed to electron interactions with MS waves.
- Both the short-duration and long-duration electron butterfly PADs can disappeared closely related to enhanced substorm activities, while the responsible mechanism can be whistler-mode plasmaspheric hiss or chorus waves.
Acknowledgments
This work was supported by the NSFC grants 41674163, 41704162, 41974186, 41904143,and 41574160, the China Postdoctoral Science Foundation Funded Projects 2018T110792 and 2018M632908, and the Hubei Province Natural Science Excellent Youth Foundation (2016CFA044). We acknowledge the Van Allen Probes mission, particularly the ECT and EMFISIS team for providing particle and wave data. The electron fluxes data are obtained from http://www.rbsp-ect.lanl.gov/data_pub/, and the wave data are obtained from http://emfisis.physics.uiowa.edu/data/index. The solar wind parameters and geomagnetic indices are obtained from the online OMNIWeb (http://omniweb.gsfc.nasa.gov/).
