to cover a semicontinental scale, which is critical for account for the entire substorm region. Figure 3.3 shows 2‐D energy flux during a substorm at 7 UT on 26 March 2014. Technical details are beyond the scope of this chapter and will be reported more in detail elsewhere (Nishimura et al., 2018c) but, in brief, the THEMIS ASI data were used to convert white light images to red‐green‐blue colors by comparing to the nearest NORSTAR meridian scanning photometers, and then the color ratios were converted to energy fluxes and characteristic energies using the Rees and Luckey (1974) formulas. While the data were mapped to 110 km altitude, mapping and viewing angle effects provide uncertainties in interpretation of derived parameters.
Figure 3.3 Energy flux distribution and its temporal variation during the 7 UT 26 March 2014 substorm detected by THEMIS ASIs. (a) H‐component of ground magnetic field at Fort Smith (FSMI); (b) THEMIS ASI keogram at FSMI; (c) energy flux in the same format as Panel (b); (d) total energy flux integrated over the five imagers (black: total; colored: mesoscale component detrended over 500 (red), 300 (green), and 100 (blue) km size); (e) the ratio between the mesoscale and total energy fluxes; (f) averaged energy over the five imagers (energy flux‐weighted average); and (g–l) selected snapshots of energy flux maps.
This is a classical isolated substorm, whose onset was right after 7 UT followed by a ~ ‐300 nT magnetic bay, auroral breakup, and poleward expansion (Fig. 3.3a,b). During the growth phase, a quiet arc was located at ~67 deg MLAT with 1–2 erg/cm2/s energy flux (Fig. 3.3c,g). The total energy flux and averaged characteristic energy integrated over the available imager FOVs were ~30 GW and ~700 eV (Fig. 3.3d,f). Then the auroral intensity and energy flux showed a rapid two‐step increase during the substorm expansion phase, reaching ~50 erg/cm2/s along the poleward expanding arc and streamers. The total energy flux and averaged energy reached 120 GW and 2.5 keV. It is interesting that the auroral structures during the expansion phase involve numerous mesoscale auroras of the order of 100 km size (Fig. 3.3i–k). Those correspond to repetitive auroral intensifications (PBIs and streamers). To quantify the contribution of such mesoscale precipitation, 2‐D energy flux distributions were smoothed over 100, 300, and 500 km at each time, and the difference from the original data was defined as the mesoscale precipitation below the specified scale (shown in colors in Fig. 3.3d,e). Mesoscale precipitation repetitively increased the total energy flux by ~20–40 GW for a duration of ~10–20 min each. Their contribution relative to the total energy flux reached ~25 (<100 km), 40 (<300 km), and 50% (<500 km) during the expansion phase. This indicates that mesoscale precipitation is critically important to describe the total precipitation energy input. The average energy stayed almost constant after the initial rise.
While the imager array can be used as in Figure 3.3 to potentially improve specification of instantaneous mesoscale precipitation over a regional scale, it is currently difficult to specify convection and currents with a similar level of resolution and coverage. Figure 3.4 shows a comparison of the energy flux, SuperDARN convection map, and vertical and horizontal currents from ground magnetometers at the same region and time of Figure 3.3j. The currents were obtained by the spherical elementary current systems technique (Weygand et al., 2011), and the vertical currents are a proxy of FACs. Generally coherent radar echoes in the nightside auroral oval are sparse. In this example, a good amount of radar echoes exists, but the flow pattern does not reproduce mesoscale structures corresponding to the mesoscale energy flux structures. Instead, the flows are much smoother, only showing the large‐scale flow pattern. This is likely due to the spherical harmonic fitting of the radar data and to the limited spatial and temporal resolution of the data. Line‐of‐sight velocity measurements (not shown) show more structured flows, and techniques, such as divergence free fitting (Amm et al., 2010; Bristow et al., 2016), may be able to provide 2‐D flow structures near the echo areas, although it is not possible to reproduce flows where radar echoes are sparse. The vertical currents (FACs) show enhanced upward currents at the auroral structures extending over > ~500 km (westward traveling surge at ~22 MLT and a group of bright streamers at ~0–1 MLT). However, many of the mesoscale auroral structures are missed. Similarly, the horizontal currents highlight an intense electrojet along the poleward‐expanding arc, while the resolution is not sufficiently high to resolve currents associated with the surge and streamers. This comparison signifies limitation of our current capability of capturing mesoscale structures, and further advances are necessary to properly specify mesoscale structures.
Figure 3.4 (a) Energy flux; (b) SuperDARN fitted convection map; (c) vertical current (red = upward, blue = downward) technique; and (d) horizontal current at 7:15 UT on 26 March 2014. Panels (b–d) use THEMIS ASI counts as the black‐white background.
3.5 CROSS‐REGIONAL AND GLOBAL INTERACTION PROCESSES
While many dayside flow channels decay in the vicinity of the cusp, a portion of them propagates over much longer distances and have a major impact on nightside processes (day‐night interaction) (Nishimura et al., 2014a; Lyons et al., 2016b). A flow channel initiated in the cusp (cusp auroral brightening and PMAF) can propagate into the polar cap (seen as polar cap patches and arcs) (Lockwood, 1991), and even reach the nightside auroral oval. When reaching the nightside auroral poleward boundary, the flow channel drives a PBI and then streamer (de la Beaujardière et al., 1994; Lorentzen et al., 2004; Moen et al., 2007; Zou et al., 2014; Ohtani & Yoshikawa, 2016), indicating triggering of nightside reconnection. If it occurs during the substorm growth phase, the streamer may trigger a substorm by making the near‐Earth plasma sheet unstable (Oguti, 1973; Kepko et al., 2009; Nishimura et al., 2010b; Lyons et al., 2011; Kornilova & Kornilov, 2012). Streamers/auroral flow channels further propagate into the subauroral ionosphere and drive subauroral processes (subauroral polarization streams (SAPS) (Gallardo‐Lacourt et al., 2017; Mishin et al., 2017) and proton aurora (Nishimura et al., 2014b). The day‐night interaction process can also be seen as enhanced plasma density drifting from the dayside to nightside and into the auroral oval, and then further propagating back to the dayside (Zhang et al., 2013b).
Localized ionosphere structures can affect large‐scale magnetosphere‐ionosphere dynamics. As mentioned in section 3.1, global ionosphere conductance can increase due to small‐scale density structures (Oppenheim & Dimant, 2013), and this provides a net positive increase in large‐scale conductance. Global MHD simulations have shown that such small‐scale conductance enhancements can have a global impact by affecting FACs and storm development (Wiltberger et al., 2017). Such studies indicate the importance of mesoscale/small‐scale structures for understanding global processes, and call for studies of characterizing properties of mesoscale/small‐scale structures. Figure 3.5 shows representative global distributions of mesoscale/small‐scale structures of selected quantities.