Quiet-Time Rapid Subauroral Plasma Flows at High Northern Magnetic Latitudes in the Dusk Sector

Abstract

Using satellite observations and computed variables, we specified 5 Subauroral Polarization Stream (SAPS) and 28 Subauroral Ion Drift (SAID) events observed in the Northern Hemisphere by spacecraft F18 in 2013. These SAPS-SAID flows reached supersonic velocities (2400–5200 m/s), were driven by westward E × B ion drifts generated by their underlying strong poleward meridional SAPS-SAID electric (E) fields (90–190 mV/m) and northward geomagnetic B fields, and developed at high (≥68°) magnetic latitudes, in the dusk sector, sometimes on the dayside, and mostly within the downward region-2 current suggesting their previous development. Within the deepening main trough, the poleward SAPS/SAID E field increased directly with the reductions in plasma density and conductivity, suggesting positive feedback mechanisms in progress. Across the highly inclined magnetic field lines within the subauroral flow channel, the eastward/westward zonal E field E × B drifted ions equatorward/poleward and yielded large upward/downward ion drifts observed by F18. Earthward energy deposition into the SAPS and SAID channels indicates magnetospheric electromagnetic energy generations in their respective voltage generators. Conjugate observations depict the large outward SAID E field (|E_X ≈ 10 mV/m|) on 28 October 2013 and SAPS E field (|E_Z ≈ 10 mV/m|) on 14 October 2013 developed at L ≈ 10 R_E on a short timescale at dusk.

1. Introduction

The ionospheric subauroral region is located equatorward of the electron auroral oval and accommodates the subauroral main ionospheric trough (MIT) [1] situated on the same magnetic field line as the plasmapause [2,3], demonstrating their strong association. Due to the subauroral region’s low conductivity, created by the lack of electron precipitation, large poleward meridional electric (E) fields can establish themselves. By interacting with the background northward magnetic B field, these poleward meridional E fields generate electrodynamic zonal E × B drifts that drive ions horizontally westward (or sunward) alongside the dusk convection cell. Such subauroral flows have a lifetime of less than 3 h and appear within the main trough where the vertical ion drift is typically upward due to the collisional heat produced by the drifting ions and co-rotating neutrals [4]. Intense (>1000 m/s) and narrow (1–2° in magnetic latitude (MLAT)) westward ion flows are known as Subauroral Ion Drifts (SAID) [5,6,7] and Polarization Jets (PJs) [8]. Moderate (500–1000 m/s) [9] and broader (from 3–5° up to 10° in MLAT) westward ion flows are called Subauroral Polarization Streams (SAPS) [10]. According to early studies, SAID flows were most frequently observed near magnetic midnight during the substorm recovery phase [4,11], while SAPS flows were typically observed in the dusk magnetic local time (MLT) sector during geomagnetic storms and substorms [12]. However, recent studies show the first appearance of SAID during the substorm expansion phase [13] and SAPS during the main phase of severe storms [14], as well as the occurrence of SAPS in the midnight sector (00 MLT) [15] and in the extended dusk–midnight MLT range [16].

Although the development of SAID and SAPS is still highly debated, it is commonly agreed upon by the various traditional and modern generation theories that the poleward-directed SAID/SAPS E field (i) drives the plasma westward in the SAID/SAPS channel and (ii) grows and that (iii) the SAID/SAPS channel receives electromagnetic energy known as Poynting flux powering Joule heating [17,18]. Regrading characteristic (ii), such SAID/SAPS E field growth (a) is due to the positive ionospheric feedback mechanisms created by the enhanced recombination rates reducing further both the plasma density within the main trough and the already low subauroral conductivity [4,7] and (b) occurs in direct correlation with the above-mentioned conductivity reduction and via positive ionospheric feedback mechanisms [6,7]. These (a,b) lead to the development of a trough-in-the-trough feature [19,20]. By the traditional generator theories, SAID and SAPS development is explained with conventional and inherently slow mechanisms unfolding in a voltage generator [21] or current generator [6,7]. But the modern generator theories are based on unconventional mechanisms, which unfold in their respective magnetospheric voltage generators [22] where the Poynting flux is generated [23], and explain the development of SAID [24] and SAPS [25,26] as a fast-time process—occurring on a short timescale—and ~10 min after the onset of substorm expansion phase. However, SAID flows can even develop during pseudobreakups [24] or local auroral activations not followed by fully developing substorms [27]. Via field-aligned electromagnetic energy transfer, from the magnetosphere to the subauroral flow channel, the downward Poynting flux becomes transferred along the plasmapause that acts as a power plant in the magnetosphere [28]. Then, the downward Poynting flux becomes accumulated in the subauroral flow channel, where the electron temperature becomes elevated. Consequently, both SAID and SAPS have their respective emissions and optical features appearing as east–west-oriented arcs. These include the newly discovered Strong Thermal Emission Velocity Enhancement (STEVE) phenomenon [29] associated with both (1) the 400–730 nm continuous spectrum, where the peak emission is the 630 nm red line [30], and (2) the enhanced SAID flow, sometimes reaching an extreme magnitude [31,32,33,34]. Also, the long-known stable auroral red (SAR) arc [35] that is associated with 630 nm red-line emission and strong SAID/SAPS flow [36]. It is also well understood that the short-lived STEVE arc and the long-lasting SAR arc are closely related phenomena, as SAR arcs can transition to STEVE arcs and STEVE arcs can evolve into SAR arcs [37,38,39].

SAID and SAPS are prominent features of the subauroral region and have been intensively studied since their first observations by using various types of satellite measurements taken at various ionospheric altitudes and by ground-based radars. These studies show that SAID and SAPS develop (i) mainly during magnetically active times but also during magnetically quiet times and (ii) direct the electromagnetic energy from the magnetosphere to the ionosphere along the plasmapause. Poynting flux flowing into the SAPS channel is generated in the magnetospheric voltage generator (VG_M) [23] and becomes concentrated at the interface of the oppositely directed large-scale Region 1 (R1) and R2 field-aligned currents (FACs) [17]. Poynting flux flowing into the SAID channel is produced in the inner-magnetosphere fast-time voltage generator (VG_FT) [39] and is the manifestation of a new means of energy transfer triggering STEVE arc development [18].

Regarding large-scale statistical SAID studies, He et al. [40] investigated SAID flows for the first time based on an extensive database containing 18,226 SAID events observed by Defense Meteorological Satellite Program (DMSP) satellites during the time-period of 1987–2012. Their results show (a) SAID flows maximizing at 22:30 MLT and 60 MLAT, (b) the majority of SAID flows located at 56–65 MLAT and 20–23 MLT, and (c) no SAID flows detected earlier than 17 MLT and higher than 70 MLAT. Their findings are in good agreement with the previous statistical SAID investigations of Karlsson et al. [11] and Figueiredo et al. [13] but differ from the early results of Spiro et al. [5] reporting SAID flows observed by the Atmosphere Explorer-C (AE-C) satellite before 17 MLT and at latitudes higher than 70 MLAT. More recently, Laakso and Pfaff [41] studied 200 fast drift events observed by the Dynamics Explorer-2 (DE-2) satellite. Over an 18-month time-period, these drift events were observed in all MLT sectors and at latitudes higher than 50 MLAT. However, the authors specified only the high-speed flows below 55 MLAT occurring in the pre-midnight sector as SAID flows.

Focusing on SAPS, Erickson et al. [42] used Millstone Hill incoherent scatter radar measurements collected during 1979–2011 encompassing two solar cycles. For magnetically quiet and moderate times (SYM-H ≥ −100 nT), the authors documented SAPS flows at 50–68 MLAT from 22 MLT through magnetic midnight to 06 MLT. The statistical study of Kunduri et al. [43] was based on Super Dual Auroral Radar Network (SuperDARN) radar observations made between January 2011 and December 2014 during magnetically quiet times (−10 < SYM-H ≤ 10 nT; Kp < 3). Their results show that the quiet-time SAPS flows developed rarely, only 15% of the time, and preferred the midnight MLT sector in the 60–70 MLAT range, where the SAPS flows were weak. Wang et al. [44] used DMSP data recorded during the high-sunspot number years of 2002 and 2003 covering the 15–22 MLT sector. Altogether, 7760 SAPS events were specified. Quiet (Kp < 3) and active (Kp ≥ 3) times were investigated separately. In terms of MLT, up to 150 quiet-time SAPS events occurred in the 16–17 MLT sector. SAPS flows developed close to 70 MLAT at Dst ≈ 0, reached ~3200 m/s during active times and remained minimal (≤700 m/s) during quiet times. Landry et al. [45] used DMSP measurements collected during 1987–2012. Their statistical results show that under magnetically quiet conditions (AE ≤ 200 nT), moderate SAPS flows (~800 m/s) occurred from 18 MLT onwards and below 68 MLAT. He et al. [14] used DMSP measurements collected within the calendar years of 2000–2006 and investigated SAPS flows observed during 37 intense storms and 30 quiet-time substorms. Their quiet-time statistical results show that the quiet-time SAPS occurred during the substorm expansion phase’s onset, appeared first after 19 MLT, and their occurrence frequency maximized at ~21 MLT.

Statistical SAID and SAPS studies based on DMSP measurements are particularly relevant to this study. However, these SAID and SAPS statistical results are subjective because of the DMSP satellites’ mostly dawn–dusk-oriented orbits covering primarily the dawn and dusk MLT sectors. This leads to the apparent concentration of SAID and SAPS observations in the dusk sector. However, the recent study by Gallardo-Lacourt et al. [46] reported the unexpected development of STEVE arcs and enhanced SAID flows (~5000 m/s) measured by Swarm-A at ~440 km orbit altitude and at an unusually high latitude (>70 MLAT) at ~21 MLT. This unusual SAID-STEVE event occurred on 27 March 2023 and at ~10° magnetic latitude higher than the SAID-STEVE event’s usual location [47]. Furthermore, the unusual SAID-STEVE event was observed under magnetically quiet conditions and without any underlying substorm activity. These unusual features prompted new questions regarding the subauroral and magnetospheric dynamics underlying SAID and STEVE development [46,47].

In this study, we investigate quiet-time subauroral flows that were unusually strong and appeared at high latitudes. Under quiet geomagnetic conditions (−25 < SYM-H ≤ 25 nT; Kp ≤ 3), these subauroral flows were observed by DMSP spacecraft F18 close to and even higher than 70 MLAT and sometimes earlier than 18 MLT in the Northern Hemisphere during the high-sunspot-number calendar year of 2013. Our results show that the unusually strong subauroral flows (i) reached in the SAID channel up to 5000 m/s and in the SAPS channel up to 2500 m/s during quiet-time substorms and (ii) were driven by their respective strong poleward E fields, sometimes reaching 200 mV/m. Meanwhile, the earthward Poynting flux deposition locally maximized in the subauroral flow channel and often exceeded the auroral electromagnetic energy deposition. Occasionally, the unusually strong SAID flow was accompanied by an unusually strong downward drift driven by a large westward zonal E field in the SAID channel. Two sets of magnetospheric conjugate observations demonstrate the unusually strong SAID and SAPS flows’ fast-time development and their generations in their respective inner-magnetosphere voltage generators.

2. Materials and Methods

For investigating unusually strong subauroral flows and their topside-ionosphere plasma environments, we used multi-instrument data collected by the polar orbiting DMSP F18 satellite (~840 km altitude; ~101 min orbital period; ~98.7° inclination angle) along its sun-synchronous orbits [48]. These measurements were taken at high sunspot numbers (>75) during the calendar year of 2013, just one year before the 24th solar cycle peaked in 2014 (at ~170). At high sunspot numbers, when the scale height of the oxygen fraction (O+) is high, the plasma’s O+ composition at ~840 km altitude is high (>85%). Under these conditions, the Ion Drift Meter (IDM) functions well and without limitations, and the cross-track drift data are generally good quality [49]. Meanwhile, the Scintillation Meter (SM) providing ion density (Ni) data and the Langmuir Probe (LP) producing electron temperature (Te) values are not affected by the plasma’s O+ content and are regarded as good quality. According to F18 data availability, which excluded entirely the measurements of electron temperature (Te), we used the data of ion density (Ni; 1/cm³), cross-track ion drift measured in the horizontal (HOR) and vertical (VER) directions (V_HOR and V_VER; m/s), and magnetic (B) field deflection components (δB_Y and δB_Z; nT). We also included the B field components (B_X, B_Y, B_Z; nT) for computing both the vertical (X) component of the static or DC Poynting flux (S_X; mW/m²) and the meridional (MER) component of the ionospheric E field (E_MER; mV/m) in spacecraft-centered coordinate systems. We followed the equation of S_X = 1/μ₀[(V_XB_Z − V_ZB_X)dB_Z-(−V_XB_Y)dB_Y], where μ₀ = 4π × 10⁻⁷ H/m, X = downward, Y = ram direction, and Z = antisunward (see Figure 1 in Huang and Burke [50]), as reported by Huang et al. [51]. In our presentation, we marked the Poynting flux vector as S_‖ since it is a vector directed parallel (‖) to the approximately vertical (i.e., radial) magnetic field [52]. We plotted the earthward-directed S_‖ data as positive (S_‖ > 0) values. In geodetic coordinates, we computed the meridional (X) E field based on the equation of E_X = V_ZB_Y − V_YB_Z reported by Kilcommons et al. [52], where positive/negative E_X is poleward/equatorward directed in the Northern Hemisphere. We also used the values of east–west directed E_Y = V_XB_Z − V_ZB_X [52], which is a zonal directed (ZON) E field (E_ZON; mV/m), from the Madrigal Database. Since the accuracy of δB, S_X, and E field components depends on the quality of DMSP data used, the computed values reported in this study are regarded accurate and reliable because of the good-quality high-sunspot-number DMSP data used.

For investigating the conjugate magnetosphere in the regime of near-Earth plasmasheet, we used a small collection of data recorded by the Time History of Events and Macroscale Interactions during Substorms (THEMIS) A (TH-A) and TH-E satellites for the conjunction events of interest. In 2013, the three Earth-orbiting THEMIS satellites (TH-A, TH-D, and TH-E) completed their Tail Science phase with an apogee at ~12 R_E on the nightside [53]. We used various types of THEMIS data including hot (up to 700 keV) electron density (Ne; cm⁻³) from the Ground Calculated Particle Moments (GMOM) suite, and ion and electron pressure (Pi_XX, Pe_XX; eV/cm³) and ion momentum flux (Mi_XZ; eV/cm³) tensors from the On Board Moments (MOM) suite. Meanwhile, the Electrostatic Analyzer (ESA) suite [54] provided the measurements of electron flux taken at various keV channels. E field components (E_X, E_Y, E_Z; mV/m) were provided by the Electric Field Instrument (EFI) suite [55].

Geostationary Operational Environmental Satellites (GOES) orbit the Earth at 6.6 R_E, observe atmospheric and meteorological phenomena, and monitor the Earth’s space environment [56]. For observing particle injections occurring near the conjunction events of interest, we used electron flux measured in various directions by GOES-13 at 275 keV and 475 keV and by GOES-15 at 40 and 75 keV.

Orbit data included spacecraft location in Geocentric Solar Magnetospheric (GSM) coordinates, L shell (R_E), magnetic local time (MLT; Hr), Northern Hemisphere footprints in geographic [longitude (GLON), °E; latitude (GLAT), °N] coordinates, and geomagnetic latitude (MLAT, °N).

For observing the underlying ring current, geomagnetic, and auroral conditions, we employed a small collection of variables provided by the OMNI database. These include the high-resolution SYM-H (nT) index for observing ring current variations and specifying storm and non-storm time periods, the 3-h Kp index for monitoring the underlying geomagnetic activity, and the auroral AE and AL indices (nT) for specifying substorms additionally to the SuperMAG-published substorm lists of Newell and Gjerloev et al. [57] Forsyth et al. [58] that specify substorm onset times and locations.

3. Results

3.1. Rapid High-Latitude Subauroral Flows Investigated

We inspected the F18 line plots for the calendar year of 2013. Since the F18 particle spectrometer was working, we could observe electron and ion spectrogram images. Based on the equatorward electron oval boundary, we accurately located the subauroral main trough and correctly identified the intense subauroral plasma flows, which are listed in

Table 1. Rapid high-latitude westward SAPS flows and associated variables.

Westward SAPS Events					Plasma Variables
Event No.	Event Date	UT (Hr:Mn)	MLAT (°N)	MLT (Hr:Mn)	Ni 103 (cm−3)	VHOR (m/s)	EMER (mV/m)
1	10 February 2013	02:06	70.43	17:36	1.2	2000	73
2	5 October 2013	19:49	71.30	19:41	2.1	2400	89
3	5 October 2013	21:30	72.15	20:04	3.5	1400	45
4	5 October 2013	23:11	71.92	19:45	3.5	1000	30
5	14 October 2013	02:38	70.28	17:29	4.1	900	30

and

Table 2. Rapid high-latitude westward SAID flows and associated variables.

Westward SAID Events					Plasma Variables
Event No.	Event Date	UT (Hr:Mn)	MLAT (°N)	MLT (Hr:Mn)	Ni 103 (cm−3)	VHOR (m/s)	EMER (mV/m)
1	4 February 2013	23:43	70.69	19:19	0.4	5400	150
2	5 February 2013	01:25	70.63	18:09	1.0	2800	100
3	5 February 2013	03:10	71.83	16:25	1.2	3000	115
4	6 February 2013	01:13	71.39	18:15	1.5	3000	105
5	11 February 2013	20:19	67.04	19:45	0.6	4000	140
6	12 February 2013	01:40	67.58	18:16	0.7	3200	125
7	5 October 2013	18:09	70.88	19:38	1.1	4000	140
8	6 October 2013	17:55	70.01	19:36	1.0	5000	160
9	7 October 2013	19:23	65.37	19:56	1.0	4200	140
10	8 October 2013	00:27	68.96	19:19	1.9	2950	150
11	12 October 2013	18:22	66.91	19:45	1.2	5200	170
12	13 October 2013	01:07	67.65	18:58	1.0	2600	100
13	13 October 2013	18:11	68.99	19:41	1.3	2500	90
14	19 October 2013	18:32	68.32	19:48	0.6	5300	170
15	20 October 2013	21:48	68.51	20:12	0.8	3800	140
16	20 October 2013	23:29	69.06	19:48	1.1	2800	95
17	21 October 2013	03:13	70.53	17:13	1.0	5000	195
18	21 October 2013	02:53	71.45	18:48	1.5	4000	150
19	22 October 2013	12:54	67.97	18:42	2.0	4200	160
20	22 October 2013	14:51	67.47	19:00	1.0	2800	110
21	22 October 2013	16:18	67.65	19:19	1.9	3800	90
22	22 October 2013	18:12	67.83	19:41	1.5	4800	170
23	23 October 2013	22:53	72.38	19:52	2.0	5200	190
24	24 October 2013	00:35	73.18	18:50	2.0	5400	150
25	24 October 2013	19:19	70.41	19:56	1.4	5400	185
26	24 October 2013	22:41	72.00	19:57	2.5	4600	170
27	29 October 2013	01:15	71.46	18:30	3.0	3800	135
28	29 October 2013	02:58	72.12	16:56	2.5	5600	170

, appearing at high magnetic latitudes and within the main trough.

We identified 33 subauroral flow events altogether observed in the Northern Hemisphere by F18 in 2013. These observations were made during the months of February and October, in the dusk MLT sector, and near or higher than 70 MLAT. These events include 5 SAPS events (1 in February and 4 in October; see Table 1) and 28 SAID events (6 in February and 22 in October; see Table 2). All of these SAID and SAPS locations are mapped in Figure 1.

Figure 1a–f illustrates a Northern Hemisphere geographic map and a northern MLT vs. MLAT polar map where we plotted the locations of SAPS flows (square symbols in colors) and SAID flows (dot symbols in colors). For illustrating the alignment of magnetic field lines and the location of the magnetic North Pole, each Northern Hemisphere geographic map shows the modeled magnetic meridians (in blue) and the magnetic dip equator (in light magenta).

Figure 1a,b is for the month of February 2013 and covers one SAPS event and six SAID events. Figure 1c,d depicts 3 SAPS and 10 SAID events observed during 5–20 October. Figure 1e,f shows 12 SAID events observed from 21 to 29 October. We note here that all these events (i) are further illustrated in Section 3.3 with their respective line plot sets and (ii) are plotted together on the same geographic map and on the same polar plot shown in Section 3.5.

Overall, the northern geographic maps reveal that F18 tracked the intense SAPS/SAID flows mostly over the North American continent (i.e., in the longitude sector of the magnetic North Pole, where the offset between the dip and geographic equators is the largest) and in the European sector and occasionally in the Asian sector. Between these longitude sectors, the detections of intense SAPS/SAID flows were absent. This was due to the subauroral main trough’s absence on the dayside (~16 MLT) or weak presence on the nightside (~18 MLT) appearing shallow at lower latitudes and accommodating only a weak subauroral flow.

According to the MLT vs. MLAT polar plots, F18 made the SAPS/SAID observations on the duskside, between 16 and 20 MLT, and mostly close to and at 70 MLAT. But sometimes, the intense SAID/SAPS flows were observed at higher than 70 MLAT and earlier than 18 MLT. Such detections are marked by the dot symbol in green in Figure 1b, by the dot symbols in light red and light gray in Figure 1f, and by the square symbol in light blue in Figure 1d (see more details in Section 3.3).

3.2. Underlying Interplanetary and Geomagnetic Conditions

Figure 2 illustrates the underlying ring current/storm, geomagnetic, and auroral/substorm conditions with the time series of SYM-H, Kp, AE, and AL. Here, we marked the intense high-latitude SAPS/SAID detections (square/dot symbols in colors; shaded intervals in yellow). For indicating quiet geomagnetic conditions, we added the horizontal lines of SYM-H = ±25 nT (in red) and Kp = 3 (in dark blue). These reveal that the intense high-latitude SAPS/SAID flows were observed under magnetically quiet conditions. We also marked two substorm onsets (star symbols in red) occurring before their respective subauroral flows. Figure 2a covers the time-period of 3–13 February 2013, during which all the events (one SAPS and six SAID) were observed during non-storm substorms. Figure 2b covers the entire October month of 2013. Then, most of the events (four SAPS and 22 SAID) were observed during the late recovery phases of three consecutive storms and later on, under weak-storm substorm conditions. But on 29 October, the last two SAID events were observed during a non-storm weak substorm.

3.3. Characteristics of the Rapid High-Latitude SAPS and SAID Flows

In Figure 3, Figure 4 and Figure 5, the latitudinal line plots are constructed with the DMSP data of ion density (Ni), meridional E field (E_MER), ion drifts (V_HOR and V_VER), magnetic deflection component (δB_Z), and Poynting flux (S_‖). For each line plot set, we marked the region of duskside auroral zone (based on the electron particle data) along with the SAPS/SAID channel (square/dot symbol in color, shaded interval in yellow) depicted by the V_HOR line plot. We note here that both the electron and ion particle data are presented as DMSP electron and ion spectrograms in Figures S1–S33 (see Supporting Information). Based on the basic definitions (described in Section 1), we specified the broader and weaker flows as SAPS and the spikey, narrower, and stronger flows as SAID. These line plot sets depict the rapid high-latitude SAPS/SAID flows within their own ionospheric settings. These SAPS/SAID flows appeared sometimes as newly developed but mostly as previously developed along with their characteristic trough-in-the-trough (marked as T-in-T) plasma density feature.

In the δB_Z plot, where δB_Z is an east–west (antisunward–sunward) component, we specified the duskside downward (↓) R2 current (in blue) by the positive δB_Z gradient, the upward (↑) R1 current (in red) by the negative δB_Z gradient, and the poleward-directed Pedersen current (J_P; in black) within or near the SAPS/SAID channel. Since these are quiet-time SAPS/SAID events, the δB_Z change (a) reaches ~200 nT only sometimes within the previously developed (or old) SAPS/SAID channel and (b) is small (≤100 nT) within the newly developed SAPS/SAID channel.

Regarding the underlying generation mechanisms, as our first assumption, we assumed fast-time SAPS/SAID development in an inner-magnetosphere voltage generator setting, where the newly developed outward SAPS/SAID E field mapped down along the magnetic field lines (threading the plasmapause) to the subauroral ionosphere as a poleward-directed subauroral E field. Based on the above-described δB_Z diagnostic signatures (a,b), we specified the SAPS/SAID channel as newly developed or previously developed. As the newly developed SAPS/SAID channel forms in a voltage generator setting, the E_MER line plot depicts the poleward SAPS/SAID E field that is a mapped-down outward SAPS/SAID E field developed in the magnetospheric voltage generator. We know from these previous studies [22,26] that a newly formed SAPS/SAID channel, developed in a voltage generator setting, appears without any significant underlying large-scale FACs, since the large-scale FACs’ development and flow take time in a newly formed magnetospheric voltage generator [39].

However, as Figure 3, Figure 4 and Figure 5 show, most of the enhanced high-latitude SAPS/SAID flows developed previously and appeared in the regime of ↓R2 current flown into the SAPS/SAID channel. As our second assumption, we assumed that the poleward Pedersen current became established in the previously developed (or old) SAPS/SAID channel. Here, the E_MER plot depicts a net poleward SAPS/SAID E field appearing in the regime of ↓R2 current as the poleward Pedersen current-related poleward E field becomes established and adds to the poleward SAPS/SAID E field (i.e., mapped-down inner-magnetosphere outward SAPS/SAID E field) developed in a voltage generator setting. Furthermore, the Pedersen current-related poleward E field grows via positive feedback mechanisms in direct correlation with the decreasing Pedersen conductivity/plasma density [7]. Such poleward (SAPS/SAID) E field growth leads to the development of a trough-in-the-trough plasma density feature [19,20] within the main trough accommodating the previously developed (or old) SAPS/SAID flow.

Here, we describe the SAPS/SAID events shown in Figure 3, Figure 4 and Figure 5 based on the above-described two assumptions.

We start with the five SAPS events shown in Figure 3a,b (see also Figures S1–S5 in Supporting Information). In each SAPS event, both the broader SAPS flow and the underlying wider poleward E_MER (driving the plasma westward in the SAPS channel) appeared within the main trough. There, the westward SAPS flow streamed in the sunward direction, and the earthward electromagnetic energy deposition (measured by the Poynting flux) increased up to S_‖ ≈ 4.5 mW/m². At each SAPS channel’s poleward edge, the oppositely directed ↓R2–↑R1 FACs connected via poleward-directed Pedersen currents, implying that these intense SAPS flows were old and developed previously in a voltage generator setting. Meanwhile, the increasing poleward Pedersen current supported the growth of the Pedersen current-related poleward SAPS E field. Therefore, the SAPS flow was growing and deepened the main trough by locally decreasing the plasma density. By also decreasing the conductivity, the growing SAPS flow created a trough-in-the-trough feature, implying positive feedback mechanisms in progress [4,7]. Since the vertical drift in the SAPS/SAID channel is typically upward (see details in Section 1), we highlight an interesting feature, which is the underlying downward drift. The underlying physical mechanism responsible for its development is described in detail in Section 3.4. But briefly, we mention here that in the 10 February SAPS event, shown in Figure 3a, the intense SAPS flow (V_HOR ≈ 2000 m/s) was driven by a strong poleward E field (E_MER ≈ 75 mV/m) and was underlined by a strong vertical downward (−) drift reaching ~1500 m/s in magnitude. In Figure 3b, the SAPS events of 5 October show the first appearance of intense SAPS flow (V_HOR ≈ 2400 m/s) and underlying poleward E field (E_MER ≈ 90 mV/m) in their strongest forms and their respective weakening (from 1200 m/s and 45 mV/m to 1000 m/s and 30 mV/m) during the next 3 UT hours. Meanwhile, the underlying vertical drift was typically upward and weak (V_VER < 500 m/s).

Continuing with the 28 SAID events shown in Figure 3c (see also Figures S6–S11 in Supporting Information) and in Figure 4 and Figure 5 (see also Figures S12–S33 in Supporting Information), the rapid SAID flow appeared within the main trough as a latitudinally narrow (≤2°) fast westward ion drift (2800 m/s ≤ V_HOR ≤ 5200 m/s) streaming sunward and was driven by its underlying strong poleward E field (90 mV/m ≤ E_MER ≤ 190 mV/m).

In some of the intense SAID events like in the first event on 24 October shown in Figure 5b (see also Figure S29 in Supporting Information), the trough-in-the-trough feature appeared evidencing positive feedback mechanisms in progress [4,7] and mostly in the regime of ↓R2 current, implying that the SAID flow observed was old and developed previously in a voltage generator setting. However, some examples of the newly developed SAID flow are shown in Figure 4a with the 5 October event (see also Figure S12 in Supporting Information) and in Figure 5a,b with the events of 21 October (see also Figure S22 in Supporting Information) and 29 October (see also Figures S32 and S33 in Supporting Information). Within these newly developed SAID flows, their respective re-scaled δB_Z line plots (with limits of ±100 nT) show only a shallow positive gradient (δB_Z amplitude ≤ 30 nT). This is the signature of a minimal ↓R2 current flowing into the newly developed SAID channel formed in a voltage generator setting, where the strong E_MER drove the westward ion drift. The E_MER’s magnetospheric conjugate component developed in a fast-time voltage generator specified by Mishin [22,24] as VG_FT. Finally, the S_‖ plot shows that within both the enhanced SAID channel and the duskside auroral zone, the earthward-directed electromagnetic energy deposition (depicted by the Poynting flux) increased locally. We also highlight some specific features shown in Figure 3, Figure 4 and Figure 5 (see details below).

For Figure 3c, showing the six SAID flows observed during 4–12 February 2013 (see also Figures S6–S11 in Supporting Information), we draw attention to the underlying vertical drift that was directed downward and was further investigated (see more details in Section 3.4). We also note the third SAID event (marked as dot symbol in green; see also Figure S8 in Supporting Information) occurring on 5 February at 71.83 MLAT (i.e., > 70 MLAT), where the earthward Poynting flux locally increased to S_‖ ≈ 4.75 mW/m² and at 16.43 MLT (i.e., <18 MLT).

For Figure 4, showing the 10 SAID flows observed during 5–20 October 2013 (see also Figures S12–S21 in Supporting Information), we notice the underlying vertical drift that was mostly upward directed. We also note the 6 October event in Figure 4a (see also Figure S13 in Supporting Information). Then, the SAID flow (marked as dot symbol in green) reached V_HOR = 4800 m/s because of the strong poleward E field (E_MER = 170 mV/m) at 70.01 MLAT, where the earthward Poynting flux locally increased to S_‖ ≈ 9 mW/m².

For Figure 5, showing the 12 SAID flows observed during 21–29 October 2013, we highlight the events of 21 October (in Figure 5a; marked as dot symbol in light red; see also Figure S22 in Supporting Information) and 29 October (in Figure 5b; marked as dot symbol in light gray; see also Figure S33 in Supporting Information). These two SAID flows were observed at 70.53 MLAT and 72.12 MLAT (i.e., >70 MLAT) and at 17.22 MLT and 16.94 MLT (i.e., <18 MLT) and reached V_HOR ≈ 5000 m/s and V_HOR ≈ 4600 m/s, driven by their respective large poleward E fields of E_MER ≈ 190 mV/m and E_MER ≈ 170 mV/m. Meanwhile, their respective underlying earthward energy depositions measured by the Poynting flux commonly reached S_‖ ≈ 5.5 mW/m². We also note the earlier 24 October SAID event in Figure 5b (marked as dot symbol in purple; see also Figure S29 in Supporting Information). Then, the intense SAID flow and its underlying strong poleward E field reached V_HOR ≈ 5400 m/s and E_MER ≈ 185 mV/m at 70.41 MLAT (i.e., >70 MLAT) and 19.94 MLT. Meanwhile, the earthward Poynting flux increased up to S_‖ ≈ 4.5 mW/m² within the SAID channel and reached only S_‖ ≈ 2.5 mW/m² in the duskside oval.

3.4. Enhanced Downward Drifts Underlying the Rapid High-Latitude SAID Flows

Figure 6 is constructed for the six SAID events of February 2013 (see also Figures S6–S11 in Supporting Information). Then, the underlying cross-track vertical drift was downward directed and sometimes as strong as the rapid cross-track horizontal drifts (as shown in Figure 3c). Here, the line plot sets are constructed with the F18 data of Ni, V_VER, and E_ZON. As before, we marked the duskside auroral zone and the SAID flow.

Although the zonal E field (E_ZON) data coverage is not continuous at subauroral latitudes, as indicated by the data gaps at SAID latitudes (dot symbol in color, shaded interval in yellow), Figure 6 shows quite clearly the sudden increase of westward zonal E field (E_ZON < 0) within the SAID channel. By interacting with the background geomagnetic B field, this westward E_ZON drove the plasma (via E × B drift) poleward within the SAID channel and downward along the highly inclined magnetic field lines. Therefore, this poleward E × B drift had a large downward (V_VER < 0) component that was observed by F18. These E_ZON < 0 and V_VER < 0 observations are in good agreement with the early study of Isaev et al. [59] reporting the abrupt increase in the westward E field (i.e., E_ZON < 0) in the main trough causing the F region plasma E × B drifting downward (i.e., V_VER < 0). Such localized downward E × B drift increase can enhance plasma depletion within the trough/SAID channel by increasing the recombination rates [60] and by further increasing the SAID flow via positive feedback mechanisms (see details in Section 1). This is demonstrated with two exceptional SAID events, when the cross-track drifts equaled in the SAID channel: on 4 February shown in Figure 3c and Figure 6a and in Figure S6 in Supporting Information (V_HOR ≈ 5400 m/s, V_VER ≈ −5200 m/s, and E_ZON ≈ −35 mV/m) and on 12 February shown in Figure 3c and Figure 6b and in Figure S11 in Supporting Information (V_HOR ≈ 3200 m/s, V_VER ≈ −2400 m/s, and E_ZON ≈ −25 mV/m).

3.5. Statistical Results

Figure 7 shows the statistical details of the 33 SAPS/SAID events observed by F18. We investigated these events statistically in terms of location, time, and magnetic activity level by constructing column charts and plotting these variables against the number of events observed. Since most of these events are SAID events (28 out of 33), the statistical results are most characteristic of the 28 rapid high-latitude SAID flows observed.

In Figure 7a, the geographic map shows the locations of the 33 SAPS/SAID events (marked as dot symbols in red in the American longitude sector and in cyan elsewhere). There is an apparent gap between the geographic longitudes of 80°E and 220°E. Although F18 sampled all longitudes equally, the quality of measurements was not the same, and sometimes poor-quality measurements resulted in large data gaps in this section. However, the good-quality measurements clearly indicated that the intense SAPS/SAID flows were absent in this section (see details in Section 3.1). Therefore, the statistics of the SAPS/SAID events investigated are quite realistic.

Statistically, Figure 7b shows that most of the SAPS/SAID evets occurred in the (220–360)°E longitude sector, which is the northern magnetic pole’s larger longitude region. The rest of the SAPS/SAID events occurred in the (0–80)°E longitude sector, with a maximum of 3 out of 33 events at 0°E. Consequently, there was no intense SAPS/SAID flow observed within (80–220)°E at high latitudes. In terms of the Kp index, Figure 7c shows statistically that most of these SAPS/SAID events (23 out of 33 events) were observed under magnetically quiet conditions, when the Kp index was less than 1. Meanwhile, significantly less (4 out of 33) events occurred at Kp = 1 and at higher Kp such as at Kp = 2 (2 out of 33 events) and at Kp = 3 (3 out of 33 events).

In Figure 7d, the MLT vs. MLAT polar map shows the 33 SAPS/SAID locations (dot symbols in cyan and red) appearing near 70 MLAT and in the duskside: mostly on the nightside (>18 MLT) and sometimes on the dayside (<18 MLT).

Statistically and in terms of MLT, Figure 7e shows that most of the SAPS/SAID events were observed (18 out of 33 events) at 19 MLT and significantly less (8 out of 33 events) at 18 MLT. Importantly, some of the SAPS/SAID events were observed on the dayside (<18 MLT): at 16 MLT (2 out of 33 events) and at 17 MLT (3 out of 33 events).

In terms of MLAT, Figure 7f shows that the SAPS/SAID events occurred mostly at 67 MLAT (7 out of 33 events) and at 70 MLAT (8 out of 33 events) and less frequently at 68 and 72 MLAT (4 out of 33 events). Here, we indicated the L shell values, computed as L = 1/cos²(MLAT), for these magnetic latitudes. This illustrates that the SAPS/SAID events’ majority were observed at MLAT > 67° (31 out of 33 events) and thus were located at L > 6.5 R_E, up to 11.6 R_E at 73 MLAT. Therefore, the geosynchronous GOES satellites (orbiting at 6.6 R_E and earthward of the plasmapause) were not able to observe the reconnection/substorm-related particle injections. This is further demonstrated with two sets of magnetically conjugate observations covering the 29 October SAID event (shown in Figure 8; see also Figure S32 in Supporting Information) and the 14 October 2013 SAPS event (shown in Figure 9; see also Figure S5 in Supporting Information).

3.6. Conjugate Observations of the 29 October 2013 SAID Event

Figure 8 shows a set of correlated magnetically conjugate observations depicting the 29 October 2013 SAID event. These observations were made by F18 in the topside ionosphere at 1.26 UT and by TH-A in the inner magnetosphere at 0.18 UT. We augmented the TH-A observations with TH-E electron flux data covering a similar time-period.

We note here that both the SAID location and the SAID event observed by F18 are also shown in Figure 1e,f and in Figure 5b (with the second-last line plot set), respectively, and also in Figure S32 (in Supporting Information).

In Figure 8a, the geographical map covers a larger region of the North American longitude sector. We plotted the F18 pass (in cyan) tracking the intense SAID flow (dot symbol in cyan) and the TH-A footprints of interest (in orange), which crossed the F18 pass near the F18-observed SAID location, and the foot points of TH-A-observed outward SAID E field (E_SAID; dot symbol in orange) and rotating convection E field (E_C; symbol diamond in orange). At that time, the GOES-15 footprints of interest (in red) were located near and equatorward of both the intense SAID flow and the TH-A footprints. This provides observational evidence that GOES-15 traveled on the earthward side of the plasmapause, within the plasmasphere, and therefore was not able to detect the particle injections occurring on the tailward side of the plasmapause observed by TH-A and TH-E.

Figure 8b shows the DMSP F18 line plots of Ni, V_HOR, E_MER, V_VER, E_ZON, δB_Z, and S_‖. As before, we marked the region of the dusk auroral zone. These plots illustrate the intense SAID flow observed by F18 at 1.26 UT and 71.46 MLAT and its underlying E field components and plasma environment. This includes the deep (Ni ≈ 3 × 10³ cm⁻³) main trough and the plasmapause (PP) depicted by the steep Ni gradient located on the trough’s poleward wall. Because of the absence of Te data, we cannot show the plasmapause’s Te signature appearing as a locally increased Te that is generally due to the locally increased downward heat flow generated by ring current energy dissipation [61]. Within the main trough, the SAID flow was intense (V_HOR ≈ 3800 m/s) and appeared at high magnetic latitude (71.46 MLAT). In the newly formed SAID channel (shaded interval in yellow), the underlying poleward-directed meridional E field was quite strong (E_MER ≈ 135 mV/m), developed in a voltage generator setting as indicated by the small δB_Z amplitude (~25 nT), and drove the rapid westward drift (V_HOR ≈ 3800 m/s). Furthermore, the net upward drift (V_VER ≈ 500 m/s) was moderate, as the locally increased eastward zonal E field (E_ZON ≈ 7 mV/m) drove (via E × B drift) the plasma equatorward and upward along the steep magnetic field line, yielding a moderate upward drift component. This added to the typical upward drift driven by collisional heat produced by the drifting ions and co-rotating neutrals [4]. Meanwhile, the earthward electromagnetic energy deposition (measured by the Poynting flux) locally peaked within the SAID channel at S_‖ ≈ 3 mW/m².

Figure 8c illustrates the MLT vs. MLAT polar plots with the underlying two-cell polar convection. We sketched the dusk cell (in blue) and dawn cell (in red) based on the SuperDARN convection map [62] generated for the time-period of 01:14–01:16 UT (shown also in Figure S32 in Supporting Information), when F18 observed the intense SAID flow. The underlying polar convection was characterized by a dominating dusk cell that crossed the magnetic midnight meridian and extended dawnward and by a smaller dawn cell. We also indicated (in dark green) the Heppner–Maynard (H-M) boundary [63] marking the polar convection’s low-latitude limit at ~71 MLAT. However, this 71 MLAT is a modified location that was specified based on the F18 particle data. We made this modification by moving the SuperDARN-modeled H-M boundary location of 66 MLAT to 71 MLAT in order to obtain a scientifically correct presentation. It is well known that there are inaccuracies associated with the SuperDARN-estimated H-M boundary that are due to the changing ionospheric propagation conditions [64]. Furthermore, as shown in Figure S32, the SuperDARN drift data do not show the SAID signature. This reflects the limitation of the high-frequency radar created by obscuring SAID-related fast flows or by data gaps [65]. We note here that all the convection maps generated for the 33 rapid high-latitude subauroral flows investigated are shown in Figures S1–S33 (in Supporting Information). We also mapped the F18 01-02 passes (in cyan) that were oriented in the nighttime–morning (22–10 MLT) direction and crossed the dusk cell on the dayside (<18 MLT). F18 observed the SAID flow on the nightside (>18 MLT), near the dusk convection cell, at 18.5 MLT and at 71.46 MLAT (i.e., >70 MLAT). We also plotted the TH-A ground track of interest (in orange) along with the foot points of the inner-magnetosphere rotating convection E field (E_C; diamond symbol in orange) and outward E_SAID (dot symbol in orange) along with the F18-observed SAID flow (dot symbol in cyan). We also mapped the GOES-15 footprints of interest (in red) that were located at ~60 MLAT and covered the 14–19 MLT sector. All these provide further observational evidence that the inner-magnetosphere outward E_SAID (observed by TH-A) mapped down to the conjugate topside ionosphere as a poleward E field (i.e., E_MER), near the F18-observed westward SAID flow. Thus, the outward E_SAID (observed by TH-A) was the inner-magnetosphere driver of the intense SAID flow in the topside ionosphere (observed by F18).

In Figure 8d, the X vs. Z and X vs. Y orbit plots (in GSM coordinates) show the orbit sections completed by TH-A (in orange), TH-E (in magenta), and GOES-15 (in red) during their respective time intervals on 28–29 October 2013. Then, TH-A and TH-E traveled on the nightside of the Earth in the tailward direction and on the duskside, below the magnetic equatorial plane. There, the positive directions are sunward, duskward, and inward (or earthward). Along its tailward journey, TH-A observed the antisunward/duskward/outward SAID E field components (E_X, E_Y, E_Z) earlier (at 0.18 UT) and the rotating convection E field (E_C) later on (at 1.64 UT). As TH-E traveled closer to the magnetic equatorial plane and more tailward, it was able to observe the nightside dispersionless particle injections unfolding before and when the outward SAID E field was observed by TH-A. These particle injections were not observed by GOES-15, which traveled closer to Earth, in the plasmasphere, from the dayside to the nightside and on the duskside.

Figure 8e illustrates the TH-E-observed particle injections with the electron flux measurements made between 82 and 740 eV at nine different energy levels and the orbit parameters of MLT and L. For a better presentation, we used the log scale for plotting the electron flux measurements and marked the times (in UT) of precious substorm onset (star symbol in red; shaded interval in yellow) and SAID E field observation (dot symbol in orange; shaded interval in yellow) made by TH-A. By traveling at high L shells (10.5 R_E < L ≤ 11.5 R_E) in the nighttime sector (20–21 MLT), TH-E was able to observe a series of dispersionless particle injections at lower energy levels (82–740 eV). Importantly, particle injections occurred when the SAID E field was observed implying its fast-time development (see details below).

Figure 8f illustrates the TH-A observations made under weak substorm conditions and during particle injections. We show the line plots of hot (up to 700 keV) electron density (Ne), hot (up to 300 keV) ion momentum flux tensor (Mi_XZ), E field components in GSM coordinates (E_X; E_Y; E_Z), electron flux measured at nine different energy levels in the 2.97–26.83 keV range, TH-A orbit parameters (MLT and L), AL index, and electron flux measured by GOES-15 at 40 keV and 75 keV energy levels in eight directions at each energy level.

To describe the main observations, we start with the Ne plot. As TH-A traveled on the duskside tailward, it first observed multiple Ne drop-offs that could be the signatures of the multiple plasmapause crossings TH-A made in the plasmasphere. Appearing as a broad Ne enhancement, TH-A observed the near-Earth plasmasheet earthward edge (marked as shaded interval in cyan), which is a region of a locally increased hot electron population, where the hot electrons are trapped (i.e., drift many times around the Earth [66]), that is bounded by the main plasmapause (PP; shaded interval in yellow) at the earthward end and by the trapping boundary (TB; shaded interval in light green) at the tailward end, separating the plasmasheet earthward edge from the distant plasmasheet [66,67,68].

On the main plasmapause (PP), TH-A observed the outward SAID E field between the charge separation created by the earthward-traveling hot electrons and ions. Shown by the plot of hot Ne, the hot electrons stopped at the PP. Shown by the Mi_XZ plot, the hot ions passed through the PP and locally peaked more earthward. This created a charge separation, where the outward SAID E field developed in a fast-time voltage generator setting (marked as VG_FT). There, TH-A observed the net outward SAID E field’s components. These are the X and Z components pointing antisunward (−) as E_X ≈ −10 mV and outward (−) as E_Z ≈ −5 mV/m. Since the detection of the Y component was made in the dusk sector, at ~19 MLT, TH-A observed its rotation from duskward (+) as E_Y ≈ 4 mV/m to dawnward (−) as E_Y ≈ −10 mV/m at L ≈ 10 R_E. From such a large L shell, the strong net outward SAID E field mapped down along the main PP to the conjugate topside ionosphere (to ~68 MLAT, as shown in Figure 8b,c) as a poleward-directed meridional E field (E_MER ≈ 135 mV/m; observed by F18) and drove the intense westward drift (V_HOR ≈ 3800 m/s) streaming sunward along the dusk cell in the SAID flow channel (observed by F18).

On the trapping boundary (TB), TH-A observed the weak cross-tail convection E field’s rotating Y component (E_Y) from duskward (E_Y ≈ 3.5 mV/m) to dawnward (E_Y ≈ −3 mV/m) along with the sunward (E_X ≈ 7 mV/m) and inward or earthward (E_Z ≈ 5 mV/m) components. Since the plasmasheet earthward edge is located on auroral field lines, the trapping boundary/rotating E_Y coincides with the diffuse-discrete auroral boundary [68]. Between the main plasmapause (PP) and the trapping boundary (TB) and thus within the plasmasheet earthward edge (shaded interval in light blue), TH-A observed weak cross-tail convection E field variations depicted by their sunward (E_X > 0), duskward (E_Y > 0), and inward or earthward (E_Z > 0) components located on auroral field lines [66,67]. Their ionospheric conjugates drove the auroral sunward convection flows duskward and sunward.

Illustrated by the electron flux measurements of TH-A, the magnetotail reconnection-related earthward dispersionless particle injections occurred not only tailward of the main PP (some of them are marked as shaded intervals in light gray and light green) when the AL index dipped during a series of substorms but even earthward of the main PP because of the multiple plasmapause crossings made by TH-A. Evidencing fast-time SAID development, occurring on a short timescale soon after substorm onset [22,24], the outward SAID E field was observed by TH-A during the substorm recovery phase at 0.18 UT as the substorm onset started at the end of the previous day (at 23.72 UT; marked as symbol star in red; see also Figure 2). TH-A made the SAID E field detection soon (half an UT hour) after the substorm onset. Meanwhile, GOES-15 traveled (at 6.6 R_E) on the duskside but earthward of the main PP (located at ~10 R_E). Therefore, the electron flux measurements made by GOES-15 within the plasmasphere show no signatures of dispersionless particle injections, only weak undulations. Possibly, a similar scenario occurred during the 27 March 2023 SAID/STEVE event reported by Gallardo-Lacourt et al. [46], when the intense high-latitude SAID flow developed at 69 MLAT (L ≈ 7.7 R_E).

3.7. Conjugate Observations of the 14 October 2013 SAPS Event

Figure 9 is constructed the same way as Figure 8 and shows the 14 October 2013 SAPS event with conjugate TH-A and F18 observations and TH-E dispersionless particle injections. This F18-observed SAPS event can also be viewed in Figure 1c,d and Figure 3b (see also Figure S5 in Supporting Information). Because of the similarities between Figure 8 and Figure 9, we describe in detail only the new observations shown in Figure 9.

In Figure 9a, the geographic map illustrates the close locations of the F18-observed SAPS channel (square symbol in light blue) and the TH-A-observed outward SAPS E field (square symbol in orange) along with the GOES-13 ground track (in red) located just equatorward.

Figure 9b shows a moderate high-latitude SAPS flow (V_HOR ≈ 900 m/s) observed by F18 at 70.28 MLAT and 2.64 UT within the deep main trough (Ni ≈ 4.1 × 10³ cm⁻³), with the plasmapause (PP) located on its poleward wall, and the underlying poleward E field (E_MER ≈ 30 mV/m) and weak upward drift (V_VER ≈ 200 m/s). As shown by the oppositely directed ↓R2–↑R1 FACs, connected by the poleward-directed Pedersen current, the SAPS flow developed previously. This is further illustrated with the weak earthward electromagnetic deposition (S_‖ ≈ 1.8 mW/m²). Meanwhile, the small upward drift (V_VER ≈ 100 m/s) suggests declining collisional heat produced by the drifting ions and co-rotating neutrals [4].

In Figure 9c, the MLT vs. MLAT polar plot depicts the SAPS flow (square symbol in light blue) location on the dayside at 17.49 MLT and just equatorward of the dusk convection cell. Later on, TH-A observed the outward SAPS E field (E_SAPS; square symbol in orange) that mapped down to the dusk cell’s equatorward edge on the nightside to 19.91 MLT. These F18 (<18 MLT) and TH-A (>18 MLT) observations imply a continuous SAPS flow channel of inner-magnetosphere origin that had been flowing along the dusk cell for a while. Meanwhile, the GOES-13 footprints (in red) were taken equatorward of the polar convection cells, at 64.4 MLAT, in a broader region of the magnetic midnight sector. Furthermore, the previous substorm onset (star symbol in red) occurred just before magnetic midnight (23:23 MLT), within the dusk cell, at 72 MLAT. This substorm onset is from the substorm list of Forsyth et al. [58].

In Figure 9d, the orbit plots show that the tailward-traveling TH-A satellite observed the net outward SAPS E field’s outward (E_Z < 0) and duskward (E_Y > 0) components below the equatorial plane and on the duskside. There, the positive directions are duskward and inward, respectively. Meanwhile, TH-E also traveled tailward and observed nightside dispersionless particle injections closer to the magnetic equatorial plane and farther away from Earth than TH-A. However, as GOES-13 orbited closer to Earth, near the equatorial plane, and from the duskside to the dawnside, it could not observe the substorm-related particle injections.

Figure 9e shows the strong dispersionless particle injections observed by TH-E at nine energy levels (between 82 and 740 eV) during both the substorm onset and the following SAPS event observed by TH-A. Then, TH-E traveled at high L shells (11.0 R_E < L ≤ 11.5 R_E) in the nighttime sector (21.0–21.4 MLT).

In Figure 9f, the TH-A line plots cover the 5 h time-period of 2–7 UT. These line plots are constructed with the data of spacecraft potential (SC Pot), hot (up to 300 keV) ion (Pi_XX) and electron (Pe_XX) pressure tensors, SAPS E field components (E_Y and E_Z; in GSM coordinates), electron flux in the 2.97–26.83 keV range, and orbit parameters (MLT and L) along with the AE index, where the previous substorm onset is marked at 2.81 UT (star symbol in red) and the electron flux measured by GOES-13 at 275 keV and 475 keV channels in nine directions.

Based on the SC Pot plot, we marked the plasmapause (PP; shaded interval in yellow) and the trapping boundary (TB; shaded interval in light green) and between them the near-Earth plasmasheet earthward edge (shaded interval in cyan) along with the regions of the plasmasphere and distant plasmasheet (see details in Section 3.6). By mapping down to the diffuse-discrete auroral boundary, the TB separates the large-scale R1–R2 FACs [68]. This is shown with the Pi_XX plot depicting the increasing ↓R2 current flowing earthward, passing through the PP and peaking earthward of the PP, and with the Pe_XX plot showing the ↑R1 currents peaking at the TB. These Pe_XX and Pi_XX observations depict the charge separation created by the hot ions (depicted by the earthward-located hot ion pressure peak or Pi_XX peak) and hot electrons located near the PP (depicted by the hot electron pressure peak or Pe_XX peak at the PP). Between them, the SAPS E field components (E_SAPS; shaded interval in yellow) developed across the PP in a magnetospheric voltage generator (marked as VG_M) setting. We show the TH-A-observed net outward SAPS E field’s duskward (+) E_Y ≈ 6 mV/m and outward (−) E_Z ≈ −10 mV/m components across the PP (observed at 3.36 UT and 19.91 MLT and located at 9.85 R_E;) along with the weak convection E field components developed within the near-Earth plasmasheet earthward edge (see details in Section 3.6). Meanwhile, the TH-A-measured electron flux illustrates the substorm-related dispersionless particle injections occurring on the tailward side of the PP that were not observed by GOES-13 since GOES-13 was traveling on the earthward side of the PP, within the plasmapause.

4. Summary of Results

In this study, we further explored the unusual high-latitude SAID-STEVE phenomena first reported by Gallardo-Lacourt et al. [46]. We focused on intense SAPS and SAID flows, which occurred under magnetically quiet conditions and weak substorms at unusually high magnetic latitudes (>68 MLAT), and conducted a large-scale investigation. It was based on 33 subauroral flow events observed by DMSP F18 in the topside ionosphere during the high-sunspot-number calendar year of 2013 and on two magnetically conjugate events observed by TH-A in the inner magnetosphere while the earthward particle injections were observed by TH-E.

This study’s topside-ionosphere results show that F18 observed the intense 33 SAPS/SAID events at high northern magnetic latitudes (near and ≥68 MLAT) in the dusk MLT sector (sometimes on the dayside at <18 MLT) under geomagnetically quiet conditions and weak substorms. These 33 events included 5 SAPS events and 28 SAID events that allowed us to conduct statistical investigations. These documented their preferred development mostly under magnetically quiet (Kp = 0) conditions and during 18–19 MLT at 67 MLAT (6.5 R_E) and at 70–72 MLAT (8.5–10.4 R_E). With the individual multi-instrument F18 observations, we demonstrated the SAPS/SAID flows’ overall common features including:

(i)

The underlying strong poleward E field: 90–190 mV/m;

(ii)

The underlying strong zonal E field (max|E_ZON| ≈ 50 mV/m);

(iii)

The associated deep subauroral main trough (min Ni ≈ 0.4 × 10³ cm⁻³);

(iv)

The trough-in-the-trough plasma density feature implying positive feedback mechanisms in progress;

(v)

The magnitude of ↓R2 current implying previous or recent SAPS/SAID development in a voltage generator setting;

(vi)

The locally increased earthward electromagnetic energy deposition along the plasmapause into the SAPS/SAID flow channel, measured by the Poynting flux (S_‖ ≤ 12 mV/m), implying the generation of electromagnetic energy in the conjugate inner-magnetosphere voltage generator and its subsequent downward/earthward channeling.

This study’s inner-magnetosphere results present two sets of magnetically conjugate observations demonstrating the fast-time development of intense SAPS on 14 October 2013 (see Figure 9f) and intense SAID on 29 October 2013 (see Figure 8f) soon after their respective substorm onsets. By invoking the fast-time SAPS development theory put forward by Mishin [22,25] and Mishin et al. [26], the inner-magnetosphere outward SAPS E field (a) is an integral part of the substorm current wedge 2-loop (SCW2L) system created by the tailward R1 loop and earthward R2 loop and (b) develops in a magnetospheric voltage generator (VG_M). Figure 9f shows that TH-A observed the duskward and outward E_SAPS field components (of the net SAPS E field) between the pressure peaks of hot ions and hot electrons near the PP (at L ≈ 9.85 R_E). The hot ion pressure peak (depicted by the Pi_XX plot) was associated with the ↓R2 current loop and maximized earthward of the PP. The hot electron pressure (depicted by the Pe_XX plot) extended earthward to the PP, was associated with the ↑R1 current that maximized at the TB. The underlying substorm/reconnection-related earthward particle injections were observed more tailward by TH-E and at the PP and within the plasmasheet by TH-A. Near the PP, these hot ions and electrons—shown with the plots of hot electron pressure (Pe_XX) and hot ion pressure (Pi_XX)—created a charge separation wherein the net outward E_SAPS developed in a magnetospheric voltage generator (VG_M). From such a high shell (L ≈ 9.85 R_E), the strong net outward SAPS E field mapped down along the PP to the high-latitude subauroral ionosphere as a poleward E field that drove the intense westward SAPS flow observed by F18 (shown in Figure 9a–c).

By invoking the short-circuiting theory first put forward by Mishin & Puhl-Quinn [69] and further developed by Mishin [22,24], the inner-magnetosphere outward SAID E field (a) is an integral part of the short-circuiting loop set up during magnetotail reconnection/substorm-related earthward plasma injections and (b) develops in an inner-magnetosphere fast-time voltage generator (VG_FT). When the mesoscale plasma flows (MPFs) [70] are strong enough to pass through the plasma sheet earthward edge and to reach the PP, then the outward (or tailward) SAID E field develops at the PP. This is shown in Figure 8e,f with the dispersionless particle injections observed by TH-E more tailward and by TH-A near the PP and in the plasmasheet. Outward SAID E field development is due to the charge separation created by the MPF-related hot electrons that become stopped at the PP (as shown with the line plot of hot Ne) and by the MPF-related hot ions that keep traveling earthward and peak earthward of the PP (as shown with the Mi_XZ plot). Between the charge separation, the inner-magnetosphere fast-time voltage generator (VG_FT) develops, where the mixture of hot ions and cold electrons provide resistivity [22,24]. All these TH-A and TH-E observations provide evidence that the strong net outward SAID E field, which developed on a short timescale, mapped down from L ≈ 10 R_E to high magnetic latitude as a strong poleward E field that drove the unusually intense westward plasma flow (streaming sunward), as was observed by F18 (shown in Figure 8a–c).

5. Discussion

Our study is particularly relevant to the recent study of Gallardo-Lacourt et al. [46] reporting an unusual SAID-STEVE event observed on 27 March 2023 at high magnetic latitude, 68 MLAT, which is ~10° higher than the locations of SAID-STEVE events investigated by previous studies including the statistical study of Svaldi et al. [47].

Based on our findings and recently published results by other authors (cited below) and due to our improved understanding, we can both explain some of the high-latitude SAID features reported by Gallardo-Lacourt et al. [46] and answer some of the questions they put forward.

We start with the intense high-latitude SAID observation made by the Swarm-A satellite, as shown in Figure 2 of Gallardo-Lacourt et al. [46]. As also shown in our Figure 3, Figure 4 and Figure 5, the connection of ↓R2-↑R1 FACs on the SAID channel’s poleward side implies the deposition site of earthward flowing Poynting flux [17], generated in the VG_FT [71], which possibly created a novel mode of energy transfer associated with the STEVE arc [18]. Consequently, this provides observational evidence of the intense high-latitude SAID flow’s fast-time development in a voltage generator setting by short-circuiting [22,24]. Furthermore, the trough-in-the-trough plasma density feature shown by the Ne plot (in Figure 2 of Gallardo-Lacourt et al. [46]) provides observational evidence of the operational positive feedback mechanisms [6,7] that are essential to the generation of SAID-linked subauroral arcs [39].

Regarding the absence of particle injections, recorded during the high-latitude SAID-STEVE event observed on 27 March 2023 at 69 MLAT (L ≈ 7.7 R_E), this was possibly due to GOES traveling in the plasmasphere (i.e., on the earthward side of the plasmapause) at ~6.6 R_E. There, as also shown in our Figure 8, GOES could not observe the earthward particle injections associated with the fast-time development of the inner-magnetosphere SAID E field. Based on the auroral SuperMAG AL (SML) index [57], Gallardo-Lacourt et al. [46] concluded the absence of underlying substorm activity. It was correctly indicated by the SML index derived from a large number of ground-based magnetometer stations [57]. But Gallardo-Lacourt et al. [46] could not explain the intense high-latitude SAID flow’s fast-time development in the absence of underlying substorm activity. However, according to the study of Mishin [24], SAID can develop after pseudobreakups, when the earthward plasma injections (i.e., MPFs) are strong enough to reach the plasmapause and trigger SAID development via short-circuiting but not strong enough to trigger substorm development [71]. Consequently, the 27 March 2023 high-latitude SAID-STEVE event possibly occurred during a pseudobreakup.

6. Conclusions

From the observational results (listed as i–vi in Section 4), we draw the following conclusions for the 33 topside-ionosphere and 2 inner-magnetosphere SAPS/SAID events investigated in this study. During these events, the MPFs were strong enough (a) to reach the steep and well-defined plasmapause at large L shells (8.5–10.4 R_E) and (b) to trigger the emergence of the ↓R2 current loop. These (a,b) triggered the respective fast-time development of a large outward SAID E field in an inner-magnetosphere VG_FT generator (i.e., a) and a large outward SAPS E field in a VG_M (i.e., b). These large SAID and SAPS E fields mapped down to unusually high magnetic latitudes (68–72 MLAT) on the duskside (16–20 MLT) as large poleward meridional SAID and SAPS E fields and drove the unusually strong westward SAID and SAPS flows streaming sunward

Thus, the root cause of the unusually strong topside-ionosphere high-latitude SAID and SAPS flows observed by F18 during the quiet-time time-periods investigated was their respective large outward inner-magnetosphere SAID and SAPS E fields developed at high L shells. In the topside ionosphere, the deep main trough and the positive feedback mechanisms created favorable conditions for the maintenance of the rapid high-latitude subauroral flows observed by F18 during the events investigated. Under such favorable conditions, but at geomagnetically more active times, such rapid subauroral flows are expected to occur at lower latitudes.