4 Discussion
Based on 15-years of ASI data at Athabasca, we present a detailed comparative study on the occurrence characteristics of arc detachment from the main auroral oval for three different types of subauroral arc: pure red arc, red+green emission arc, and STEVE. Using monochromatic images at six different wavelengths, we identified 139 red arcs, 42 red+green arcs, and 26 STEVE events. The pure reds arcs observed in this study may be closely related to traditional SAR arcs (Takagi et al., 2018). However, we are not referring pure red arcs as SAR arc because of the association of red arcs with subauroral ionospheric drift (SAID) (Sazykin et al., 2002) and particle precipitation from the plasma sheet particularly in the midnight sector (Yadav et al., 2021b; under communication).
Covering two solar minimum periods, our results revealed that the detachment rate of red arcs was maximum during solar minimum and higher geomagnetic activity conditions. The geomagnetic activity (both solar wind and Ap index) were higher in 2019-2020 as compared to 2008-2009. Likewise, detachment rate of red arcs was higher in 2019-2020 as compared to 2008-2009. In comparison to red+green arcs, the detachment rate of red arcs was higher even for the lowest Ap years (2009-2010). The detachment rate of STEVE also maximized in solar minimum and high-geomagnetic activity year (2019). The red+green arc showed no occurrence or had minimum occurrence during low Ap and low solar wind speed years (2009-2011 and 2020). The detachment rate of both red and red+green arcs was low during the years of highest geomagnetic activity (2016-2017). Takagi et al. (2018) reported that the occurrence rate of SAR arc detachment was low in the solar maximum and minimum and high during the period of increase or decrease of solar activity. Alexeyev et al. (2009) also showed that the occurrence rate of SAR arcs has a maximum in the declining phase of solar activity. These studies also reported that the occurrence rate of SAR arcs correlated well with the geomagnetic activities represented by the Ap index. Note that both of these studies were based on nearly one-solar cycle data, whereas, 15-years of data used in the present study provides us a distinct opportunity to compare two solar minimum periods. Our results indicate that the detachment rate of red and red+green arc has dependence not only on geomagnetic activity but also on the solar flux.
It is well-known that the boundary of auroral oval expands with increasing geomagnetic activity, that is, the auroral zone expands equatorward at times of enhanced geomagnetic period. Thus, during high geomagnetic activity period, observed in 2016-2017, there is a high possibility that the auroral oval expands to reach latitudes of Athabasca. The magnetic latitudes of SAID is also known to shift equatorward with increasing geomagnetic activities (Karlsson et al., 1998; He et al., 2014). This implies that the probability of appearance of subauroral arc should be reduced at the latitudes of Athabasca during higher geomagnetic activity, but at the same time, should have higher occurrence at lower latitudes. Mendillo et al. (2016) by analyzing 27-years of ASI data at Millstone Hill [42.6°N, 288.5°E, 56° invariant latitude] reported that occurrence of SAR arcs was minimum during solar minimum years and maximum during solar maximum years, opposite to what we observe at Athabasca.
We presented the magnetic local time (MLT) distribution of red, red+green arcs, and STEVE. STEVE is found to occur predominantly in the premidnight sector (highest occurrence at 22-24 MLT, and then at 20-22 MLT). Using THEMIS ASI and REGO database Gallardo‐Lacourt et al. (2018) identified 28 STEVE events between 22 and 01 MLT. A large majority of SAID occurred in the premidnight sector (2000-2300 MLT) with strength being higher for events close to 2200 MLT (Karlsson et al., 1998; Figueiredo et al., 2004). Being an optical manifestation of SAID (MacDonald et al., 2018; Archer et al., 2019), STEVE showed good correlation with the occurrence characteristics of SAID. Red and red+green arcs exhibit maximum detachment rate around the midnight sector (highest occurrence at 00-02 MLT, and then at 22-24 MLT). Using Athabasca ASI (2006-2016), Takagi et al. (2018) reported that the occurrence rate of SAR arc detachment was highest in the premidnight sector (20–22 MLT). They suggested that the detachment of SAR arcs may correspond to the dynamical injection of ring current ion populations into the inner magnetosphere. Note that the study of Takagi et al. (2018) is different from the present study because they considered the total occurrence rate (from the time of detachment till until the arc remained detectable in the ASI images) of SAR arcs, whereas we have focused solely on the detachment rate. In addition, Takagi et al. (2018) perceived STEVE events as the SAR arc. In the present study, highest detachment rate of red and red+green arcs is observed around the midnight sector, suggesting that the low-energy particle (<10 keV) precipitation from plasma sheet might also play a role in the formation of subauroral arcs around the midnight sector (Yadav et al., 2021b; under communication).
In section 3.2, we performed superposed epoch analysis to compare various geomagnetic activity indices during the detachment of red, red+green arcs, and STEVE. Results reveal that STEVE occurred during higher geomagnetic activity as compared to that observed for red and red+green arcs. Gallardo-Lacourt et al. (2018) reported that STEVE occurred at the end of a prolonged substorm expansion phase (~60 min). However, our results show that the substorm expansion phase persisted for ~60 mins (both in AL index and X-component magnetogram) not only for STEVE but also for red and red+green arcs. The notable point is that substorm intensities estimated by the magnetic field variations were ~2-3 times higher for STEVE as compared to red and red+green arcs.
In general, all arcs occurred towards the end of expansion phase and beginning of recovery phase, as already shown in the earlier studies for STEVE (Gallardo-Lacourt et al. 2018) and SAR arcs (Takagi et al., 2018). The X-component magnetogram at Fort Smith indicated that the detachment of STEVE coincided with the sharp recovery from the expansion phase and the beginning of additional activity in the recovery phase of a substorm. Based on three STEVE events, Yadav et al. (2021a) first highlighted the association of STEVE detachment and brightness with the additional activities in the recovery phase of a substorm. Our results, based on 26 STEVE events, provide further credence to the findings of Yadav et al. (2021a) that triggering of STEVE requires additional energization, observed in terms of additional activities in the recovery phase of a substorm. Gallardo-Lacourt et al. (2018) identified streamers within the auroral oval for STEVE events, indicating the association of STEVE with substorm activity.
Our results, for the first time, show the association of STEVE with ASY-H index in terms of a prominent bay-like enhancement just prior to the detachment of STEVE. The detachment of STEVE occurred immediately after the peak in ASY-H index. A sharp bay-like enhancement is also observed for X-component magnetogram at Boulder, a mid-latitude station in the longitude zone of Athabasca. The ASY-H index is associated with an intensification of asymmetric ring current in the dusk sector (Nishida, 1978). The bay-like variations of H- and D-components might indicate the onset of substorm expansion (Rostoker et al., 1980). Although the bay-like enhancement is also observed for the red and red+green arcs, it was weak (5 nT) and broad compared to STEVE (20 nT). The association of SAR arc with a bay-like structure in the ASY-H index is shown in the past studies (e.g., Ievenko et al., 2008). Unlike red arcs, a clear bay-like structure prior to the detachment of arc is not observed for green+red arc, indicating that the mechanism for the formation of red+green arc may be different from the red arc. It is notable that positive bay enhancement in ASY-H index for STEVE was 4 times stronger than red and red+green arcs. This suggests the strong association of STEVE with the asymmetry in ring current.
Lastly, in order to further provide evidence to the association of STEVE with substorms, we analyzed GOES particle flux. Results unveil that majority (20/26) of the STEVE events were accompanied by the dispersionless injection of both electron and proton flux at geosynchronous orbit, a ubiquitous characteristic of substorms. Such enhancements were observed ~30 mins prior to the detachment of STEVE into the nightside. Note that such particle injection events at the geostationary orbit were also observed often for red and red+green arcs. The dispersionless injection of both electron and proton flux at geosynchronous orbit for SAR arc was reported in the past (e.g., Ievenko et al., 2008). However, the peculiar features in the magnetic field was observed only for STEVE, for example, i) both ASY-H index and X-component magnetogram at Boulder showed an abrupt enhancement ~30-mins prior to the STEVE arc detachment, forming a feature of prominent positive bay-like enhancement before the detachment of STEVE arc ii) the SYM-H index and X-component magnetogram at Athabasca showed small enhancement ~30 mins prior to the detachment of STEVE, and iii) the X-component magnetogram at Fort Smith shows the presence of additional activities in the recovery phase of substorms. Such features were not observed for red arcs and red+green arcs. These results unambiguously indicate that STEVE develops after the substorm associated energy injection and consequent intensification of asymmetric ring current.