The effect of non‐storm time substorms on the ring current dynamics

During geomagnetically active times such as geomagnetic storms, large amounts of energy can be released into the Earth’s magnetosphere and change the ring current intensity. Previous studies showed that significant enhancement of the ring current was related to geomagnetic storms, while few studies have examined substorm effects on ring current dynamics. In this study, we examine the ring current variation during non‐storm time (SYM‐H > −50 nT) substorms, especially during super‐substorms ( AE > 1000 nT). We perform a statistical analysis of ring current plasma pressure and number flux of various ion species under different substorm conditions, based on Van Allen Probe observations. The plasma pressure and ion fluxes of the ring current increased dramatically during super‐substorms, while little change was observed for substorms with AE < 1000 nT. The results shown in this study indicate that a non‐storm time super‐substorm may also have a significant contribution to the ring current.


Introduction
The ring current contains mainly energetic ion species: hydrogen, helium, and oxygen, with an energy of several keV to several hundreds of keV, generated by the westward magnetic gradient and curvature drifts of these energetic ions. The location of the ring current is between ~2−7R E (Earth radii) (e.g., Le et al., 2004;Yue C et al., 2018Yue C et al., , 2019aSandhu et al., 2019). During quiet times, the main source of ring current particles is supplied by the ionosphere and the solar wind. While the main carriers of the ring current are protons (H + ) (Daglis, 2006), oxygen (O + ) contributes to the ring current significantly during geomagnetically active times (e.g., Hamilton et al., 1988;Korth et al., 2000;Fu SY et al., 2001;Kistler et al., 2016;Yue C et al., 2018Yue C et al., , 2019aYue C et al., , 2020aHuang Z et al., 2020). In addition, the ring current would decay when loss processes are dominant over the acceleration and source processes; ring current particles are continuously lost due to charge exchange and coulomb collisions, as well as wave-particle interactions (Fok et al., 1991;Yuan ZG et al., 2012;Yue C et al., 2019bYue C et al., , 2020bChen A et al., 2021). Three-dimensional ring current models were also constructed to investigate the variations of ring current ions (Fok et al., 1995;Jordanova et al., 1996).
The relationship between the ring current and geomagnetic storms has already been discussed extensively. During geomagnetic storms, the number of particles in the ring current increases and it becomes strongly intensified; Chapman and Ferraro (1941) described the formation of ring current during the storm's main phase. Greenspan and Hamilton (2002) estimated the relative contributions of H + and O + to the ring current energy density, close to the magnetic storm maximum. Daglis (2001) reviewed recent studies about the storm time ring current. According to previous studies, the storm-time substorm can also contribute to the ring current (e.g., Fok et al., 1999;Daglis et al., 2004;Boakes et al., 2011). The ions enter directly into the ring current during substorm injections, showing the large effect of ions on the ring current during storm-time substorms (Yue C et al., 2019a). It has been suggested that storm-time substorms had an effect on the rise of the O + /H + energy density ratio and also led to the increase of plasma sheet density . By examining the 12 August 2000 storm event, Ohtani et al. (2005) inferred that ring current ions were de-energized during the growth phase and energetic particles were injected into the ring current during the expansion phase. Their result was the storm-time substorm contribution to the storm-time ring current intensification. Furthermore, Sandhu et al. (2018) conducted a statistical analysis of ring current energy variations with different substorm phases and spatial distributions, without distinguishing whether it occurred during storm time or non-storm time. It has been shown that there is a significant enhancement of ring current energy content in the substorm expansion phase, or in the premidnight magnetic local time (MLT) sector.
Previous works have shown the contribution of ions to the ring current during geomagnetic substorms (e.g., Sandhu et al., 2018), while no study has discussed the non-storm time substorms' contributions to the ring current. In this study, we focus on the dynamics of the ring current in non-storm time substorms. We perform a superposed analysis to examine the time evolution of plasma pressure and ion flux for different L-shells during nonstorm time substorms. We determine the contribution of nonstorm time substorms to the ring current and find that super-substorms with AE > 1000 nT have a significant contribution to the ring current.

Observation and Methods
The Van Allen Probes (RBSP) consists of two identical satellites (A and B), launched on 30 August 2012 into a highly elliptical orbit, with perigee and apogee around ~1.1R E and ~5.8R E , respectively (Mauk et al., 2013). In this study, we used the combined ion flux data of different species from the Helium, Oxygen, Proton, and Electron (HOPE) Mass Spectrometer (Funsten et al., 2013), and the Radiation Belt Storm Probes Ion Composition Experiment (RB-SPICE) instrument (Mitchell et al., 2013). These are level 3 differential ion flux data. Using this combined database (Yue C et al., 2018), we get the proton (H + ), oxygen (O + ) and helium (He + ) ions energy flux from several eV to several hundred keV. More details about the data processing of ion flux can be found in Yue C et al. (2018).
For this study, we used the total plasma pressure, which is calculated for the plasma pressure of different ions in the parallel and perpendicular directions using the unidirectional fluxes (Yue C et al., 2018). The plasma pressure (P i ) of different species is and the total plasma pressure (P t ) is the sum of the plasma pressures from different species, . More details about the plasma pressure calculation can be found in Yue C et al. (2018).
The substorm event list (Newell and Gjerloev, 2011a) from the years 2012 to 2018 is obtained from the SuperMAG database (Gjerloev, 2012), which defined a substorm using a simple automated algorithm to identify substorm expansion phase onsets from the SML index (the SuperMAG equivalent of the well-known AL index). We used the substorm list of Gjerloev (2011a, 2011b) which can be downloaded from the SuperMAG website (https://supermag.jhuapl.edu/substorms/). The Super-MAG database is a collaboration of ~110 ground magnetometer stations. In order to assess the intensity of each substorm, we search for the maximum value of AE index (AE max ) within 2 hr after substorm onset. To identify the non-storm time substorm, we use the 1-min time resolution OMNI dataset of magnetic indices: Dst and AE index. We require that the minimum Dst index be larger than −50 nT for two hours prior and two hours after the substorm onset, to exclude the time interval of geomagnetic storms. We have a total of 1188 non-storm time substorm events with maximum AE (AE max ) values larger than 1000 nT (super-substorms), and 5786 events with AE max > 300 nT and <1000 nT, respectively. In addition, we use the 1-min time resolution SMR data from the Super-MAG database to show the ring current variations. SMR (Newell and Gjerloev, 2012) is a ring current index similar to Dst or SYM-H index. SMR also provides local time resolution of 4 sectors, to investigate local time dependence of the ring current during substorm events.
By using the list of non-storm time substorms, we performed a superposed epoch analysis of plasma pressures and different energies of ion fluxes for various species (H + , O + , He + ) during nonstorm time substorms. Epoch time runs from two hours before to three hours after the substorm onset. The epoch time is divided into 5-min bins, and L-shell is divided into 0.5L bins from L = 2 to 6. To describe the pressure and flux variations before and after onset, each data point was normalized by the average value of two hours before substorm onset at each 0.5 L-shell bin. Therefore, if a value is higher than 1, it is colored red and means an enhancement at that time. Besides the storm time enhancement of ion fluxes, there is also a comparable non-storm time increase of ion fluxes, as illustrated by the vertical blue shaded areas which mark the non-storm time super-substorm (AE max > 1000 nT) cases. This demonstrates that the non-storm time substorm may also play an important role in increasing ring current intensity.

Results
To examine the effects of substorms on the ring current, we provide in Figure 2 the superposed epoch results of plasma pressures of various ions for the non-storm time substorm events with AE max > 1000 nT (1188 events) and AE max < 1000 nT (5786 events). The left plots (Figure 2 (a−f)) depict the statistical results of AE max > 1000 nT, while the right plots (Figure 2 (g−l)) demonstrate the results of AE max < 1000 nT. The vertical dashed line marks the substorm onset time, while the plot time duration is from 2 hours before to 3 hours after the substorm onset. In Figure 2, we show the median value of AE and the normalized plasma pressure variation as a function of epoch time and L-shell. Figure 2 (a and g) shows the satellite dwelling time (in minutes) in each bin and Figure 2 (b and h) shows the AE index distributions. As shown, the dwelling time at each bin is larger than 50 mins, which ensures statistical significance, although it is smaller at smaller L shells. This is due to the fact that the spacecraft moves faster near perigee and slower near apogee. According to Figure 2 (b and h), the AE index distributions are very similar at different L shells and it is dramatically enhanced after substorm onset time, indicating that there is large amount of magnetic energy released after substorm onset. Figure 2 (c−f) and Figure 2 (i−l) show the total plasma pressure and the plasma pressure of H + , O + and He + , respectively, during different substorm conditions. As shown, the plasma pressures were significantly enhanced in a large range of L-shells during the super-substorm (AE max > 1000 nT) while almost no change is found for substorms with AE max < 1000 nT. During the super-substorms, the total plasma pressure and proton pressure increases immediately at L = 2−4 after the substorm onset. Meanwhile, the oxygen and helium pressures show large enhancements after half an hour of super-substorm onset, demonstrating different transportation processes. After substorm onset, the oxygen pressure ( Figure 2e) has increased about 1.5 times, while the changes of other ions' pressure are less than that of oxygen ions, resulting in a rapid increase of O + ring current during the substorm (Fok et al., 2006). On the other hand, the pressures show almost no enhancement after the substorm onset during AE max < 1000 nT substorms.
These features suggest that a large number of ions are injected when non-storm time super-substorms occur, as opposed to normal substorms (AE max < 1000 nT). The red color means that the flux is relatively higher than the average value for the two hours before onset at a specific L-shell. Figure 3 (a−d) shows the proton flux variation during super-substorms, whereas Figure 3 (e−h) is for normal substorms. During super-substorms, proton flux enhancement occurs immediately after the substorm onset. It is shown in Figure 3 (a−d) that the different energy flux results from different enhancements in each L-shell dur-   ing super-substorms: the < 1 keV proton flux has strong enhancement in L = 3−5 ( Figure 3a) and a small decrease of flux at low L shells; the several keV protons are significantly increased in L > 3 ( Figure 3b); while the tens of keV protons are significantly increased in a large range of L-shells (Figure 3c). In contrast, the high energy protons (Figure 3d) are increased in low L-shells (L < 3) and decrease in high L shells during super-substorms. Compared with Figure 3 (a−d), Figure 3 (e−h) show very small variation after the onset of substorms. Figures 4 and 5 show the flux variation of oxygen (O + ) and helium (He + ) ions with the same format as Figure 3. We plotted the O + flux of 1.1 keV, 5.2 keV, 38.1 keV and 142 keV in Figure 4, and the He + flux of 2.4 keV, 5.2 keV, 28.1 keV and 142 keV in Figure 5. There are significant differences between super-substorms and normal substorms. During super-substorms, the low-energy oxygen ions (Figure 4 (a−c)) show similar variation trends with the low energy protons. The oxygen ions with 1.1 keV and 5.2 keV first had the flux enhancement at L > 3.5 and gradually enhanced towards the lower L-shells (Figure 4 (a and b)). In contrast, the several tens of keV oxygen ions (Figure 4c) increased immediately at L > 3 after the super-substorm onset. The flux in the higher energy channels (> 100 keV) started to increase across a large range of Lshells (Figure 4d). Compared with the case of super-substorms, the fluxes of different energy oxygen ions during normal substorms does not exhibit much change (Figure 4 (e−h)). However, we found that the flux variation of helium is slightly different from the previous ions, shown in Figure 5. During super-substorms, the near 2 keV helium increased after the substorm onset ( Figure 5a) and the several to tens of keV helium started to increase after 2 hours of onset ( Figure 5 (b and c)). Further, similar to Figure 4d, the flux of > 100 keV helium (Figure 5d) also increased after the onset for a large range of L-shells. Similar to the proton and oxygen flux variation, the helium fluxes are nearly unchanged during normal substorms. According to the above three Figures (3-5), when super-substorms occur, most of the energetic ions would be injected deep in the inner magnetosphere and contribute to the ring current enhancement.  Figure 6 (a and b) show the statistical SMR variations of AE max > 1000 nT and AE max < 1000 nT, respectively, following the standard features as demonstrated in Newell and Gjerloev (2012). After the substorm onset, the SMR index of 12 LT and 18 LT dropped while the SMR index of 0 LT increased, and the SMR index of 6 LT dropped the least in magnitude, indicating different variations of the ring current. However, there is a disparity in the magnitude of SMR variation between Figure 6a and Figure 6b. First, during super-substorms (Figure 6a), SMR of 0 LT is smaller than −10 nT during the entire interval, although it had a rise of about 5 nT after the substorm onset. Besides, SMR of 12 LT and 18 LT shifted more negative during super-substorms than during normal-substorms (Figure 6b). The largest change is at 18 LT with ΔSMR of 8 nT within 50 min after the super-substorm onset, while it is only 3 nT within 50 min for normal substorms, indicating that the super-substorm may contribute to the ring current intensity significantly.

Discussion and Conclusions
In this study, we have conducted a superposed epoch analysis of ring current fluxes and plasma pressures for three ion species during different conditions of non-storm time substorms, by using HOPE and RBSPICE measurements onboard the Van Allen Probe mission. Based on the SuperMAG substorm list from 2012 to 2018, we separated the non-storm time (SYM-H > −50 nT) substorms into two categories, AE max > 1000 nT and AE max < 1000 nT. AE max denotes the maximum AE index value within 2 hr after substorm onset. The main results are as follows: (1) The plasma pressures have significant enhancement after the super-substorm onset while there are almost no variations for substorms with AE max < 1000 nT.
(2) The flux of ions with various energies was enhanced at different L-shells during non-storm time super-substorms.
(3) The non-storm time super-substorms have a significant contribution to the ring current.
Compared with AE max < 1000 nT of normal substorms, there are significant plasma pressure enhancements at a large range of Lshells during super-substorms. Especially, the oxygen pressure has the most significant variation, which suggests that the O + ions are related to the ring current during super-substorms (e.g., Ohtani et al., 2005;Nose´ et al., 2005;Fok et al., 2006;Yue C et al., 2019a;Zong QG et al., 2021), whereas, during normal substorms, the plasma pressures have no obvious changes.
The energy flux variations of various ion species have different characteristics across L-shells. The hundreds of eV to several tens of keV H + and O + fluxes are enhanced by nearly double after super-substorm onset; in contrast, the He + flux with several tens of keV is strongly enhanced after 2 hours of super-substorm onset. Additionally, the enhancements of O + and He + flux with hundreds of keV are in all ranges of L-shells, meanwhile, H + flux is increased only in lower L-shells. In general, O + ions in the ring current are increased during active times (e.g., Hamilton et al., 1988). During geomagnetic storms, O + ions increase dramatically and there are two source regions of ring current O + ions. The lower energy O + ions (< 1 keV) are from the dayside cusp and transported to the nightside plasma sheet, or they can access the inner magnetosphere through the nightside aurora region (e.g., Kistler et al., 2016). On the other hand, the tens of keV O + ions may be directly from the plasma sheet (e.g., Hall et al., 1998). Modeling work has confirmed that the source regions of the energetic O + ions during substorms are also consistent with sources during storms (e.g., Nakayama et al., 2017). According to our results, O + ions are increased in high L-shells during non-storm time super-substorms, which means that the enhancement of O + ions originated from the plasma sheet and appeared in the ring current region due to super-substorm injection. According to the result of high energy ions, during super-substorms the peak of the ring current moved to lower L-shells; however, there is no significant change in the flux during normal substorms.
The magnitude of the decrease in SMR index during super-substorms is much larger (9.6 nT/hr) than during normal substorms (3.6/hr). From Newell and Gjerloev (2012), we have found that the perturbation of SMR during geomagnetic storms is almost 11 nT/hr. Therefore, the variation rate of SMR is comparable between geomagnetic storms and the non-storm time super-substorms, indicating that the latter also have a significant contribution to the ring current. This is due to the fact that, during super-substorms, there are more energetic particles that could be injected deeper into the inner magnetosphere, thus contributing more significantly to the ring current. In addition, during the super-substorms, the plasma pressure and fluxes are larger compared with normal substorms. Overall, the ions contribute to the ring current not only during storm time substorms but also during non-storm time super-substorms.