A statistical study of poleward boundary intensifications (PBIs) associated with the arrival of polar cap patches

Poleward Boundary Intensifications (PBIs) are transient auroral brightenings at the poleward edge of the auroral oval that serve as optical tracers of nightside plasma transport and energy deposition. However, their statistical properties and coupling with solar wind and geomagnetic activity remain insufficiently understanding. Using the observations from Yellow River Station auroral All-Sky Imager during 2012–2024, we identified 1,225 PBIs that were specifically associated with nightside polar cap patches touching the poleward boundary of the auroral oval. This study investigated the spatiotemporal characteristics of PBIs and their relationship with interplanetary magnetic field (IMF) conditions and geomagnetic indexes. The key findings include: (1) PBIs predominantly occur between 21 and 01 magnetic local time (MLT), with a clear pre-midnight distribution; (2) Due to the modification by IMF By, PBIs occurrence exhibits a dawn-dusk asymmetry near 23 MLT, indicating a systematic longitudinal shift of the nightside ionospheric convection; (3) relative to substorm onset, the PBIs occurrence rate peaks within the 20–40 min window thereafter. These results provide quantitative evidence of IMF By modulating the MLT distribution of PBIs and establish PBIs as optical tracers of nightside plasma transport and energy deposition.

1 Introduction

Poleward Boundary Intensifications (PBIs) are transient, localized brightenings that occur along the poleward edge of the auroral oval. They dynamically evolve into auroral streamers and represent the ionospheric manifestation of magnetotail reconnection–driven bursty bulk flows (BBFs). PBIs thus serve as key indicators of nightside plasma transport and energy deposition (Lyons et al., 2011; Mende et al., 2011; Milan et al., 2019; Wilkins et al., 2023).

A series of studies has highlighted the role of localized flow enhancements in driving PBIs. Coordinated radar and optical observations have shown that narrow channels of fast flow can cross the open-close boundary (OCB), couple to bursty bulk flows (BBFs) in the magnetotail, and be associated with trigger transient auroral brightenings and streamers at or inside the auroral boundary (Nishimura et al., 2010; Lyons et al., 2011; Zou et al., 2015a). Statistical studies further demonstrate that the longitudinal and temporal evolution of polar cap flow and the subsequent PBIs along the auroral boundary are closely correlated, establishing a direct linkage from polar cap flows to PBIs (Zou et al., 2015b).

The polar cap patches are generated primarily on the dayside cusp regions, typically twice or more exceeding the background electron density. They transport along the large-scale convection (Dungey, 1961; Zhang et al., 2013), across the polar cap to the nightside auroral oval. Observations show that polar cap patches are usually accompanied by narrow, enhanced mesoscale polar cap flows (Nishimura et al., 2014; Zou et al., 2015a; Zou et al., 2015b). Ground-based auroral imaging of polar cap patch arrival locations is therefore an effective approach to infer when and where flow–boundary interactions and associated energy deposition are likely to occur. Patch trajectories thus provide valuable diagnostics of BBFs and associated PBIs, offering insight into their spatiotemporal distribution and triggering conditions (Moen and Gulbrandsen, 2007; Moen and Hosokawa, 2015). IMF conditions are also critical: the east–west drift of polar cap patches is strongly modulated by IMF By condition (Hosokawa et al., 2009), while Zou et al., 2015a further reported that nearly 67% of patch events were accompanied by localized flow channels, typically with velocities around 600 m/s. When such flows traverse the OCB, they generate strong polarization electric fields and enhanced field-aligned currents, directly triggering PBIs and promoting auroral streamer formation. Collectively, these results underscore the critical role of IMF conditions in regulating nightside auroral dynamics.

Despite these advances, several critical questions remain. First, many studies examined patch–OCB crossings as a whole, rather than distinguishing those specifically associated with PBIs, leaving the exact flow and IMF conditions required for PBI onset poorly constrained (Moen and Gulbrandsen, 2007; Moen and Hosokawa, 2015; Zou et al., 2015b). Second, due to a limited number of samples, most analyses could not explicitly disentangle the coupled influences of IMF Bz and By (Moen and Gulbrandsen, 2007; Moen and Hosokawa, 2015). Since magnetotail reconnection occurs predominantly under southward IMF Bz, neglecting its role may introduce biases into interpretations of By-driven dawn–dusk asymmetries. Finally, the relationship between PBIs and substorm onset had been explored just through case studies or short-term statistics (Chu et al., 2015; Lyons et al., 2011). Hence, a systematic, long-term statistical study is still required.

We utilize high temporal and spatial resolution all-sky auroral imager at the Chinese Yellow River Station observation from 2012 to 2024, identifying 1,225 nightside PBIs associated with polar cap patches reaching the poleward boundary of the auroral oval. We examine their occurrences dependence on substorm onset and IMF variations, aiming to (1) quantify the modification of IMF By on PBI occurances, including potential dawn–dusk asymmetries near 23 MLT, as well as the substorm initiation; and (2) determine the range of propagation delays for IMF variations from the dayside magnetopause to the nightside ionosphere by tracking the PBIs. This comprehensive analysis provides statistical evidence for the coupling mechanisms of polar cap patch transport and nightside energy release, improving the understanding of nightside magnetosphere-ionosphere coupling. The paper is organized as follows: Section 2 describes the data and methodology, Section 3 presents the statistical results, Section 4 provides the discussion, and Section 5 concludes the study.

2 Data and methodology

2.1 Data

All-sky images acquired at the Yellow River Station (78.92° N, 11.93° E, 76.4° MLAT) provide auroral observation in this study. The 630.0 nm all-sky imager, corresponding to red-line auroral emissions at roughly 250 km altitude, has a wide field of view (with a radius of 1,000 km) and 10-s temporal resolution, capable to track the evolution of polar cap patches and PBIs. Moreover, distinctive located at near the poleward edge of the auroral oval during the nighttime, the YRS enables continuous optical observations of nightside polar cap and auroral oval phenomena during the polar night. This strategic location, combined with the instrumental capability, makes YRS especially effective for monitoring the transport of plasma patches across the polar cap and capturing the associated PBI processes as they occur near the nightside auroral boundary.

To evaluate solar wind driving effects alongside ground-based optical observations, we utilized the OMNI data in a cadence of 1-min, including solar wind velocity, dynamic pressure, proton density, and IMF components (Bx, By, Bz), and geomagnetic activity indices such as AE and SYM-H (King and Papitashvili). Considering the typical propagation time on solar wind measurements from the nose of bow shock at the dayside to their geoeffective impact on the magnetosphere–ionosphere system, we uniformly approximated a 7-min time delay to all solar wind and IMF parameters from OMNI data throughout the analysis (Liou and Newell, 1998; Lockwood et al., 1989).

In addition, we incorporated a substorm onset list developed by Newell and Gjerloev, 2011 to examine the temporal connection between PBIs and substorm initiation. This list is derived from the SuperMAG SML index, which extends the traditional AL index by including over one hundred magnetometer stations globally. Substorm onsets are identified through a set of objective criteria requiring a sharp and sustained decrease in SML: (1) a drop of more than 15 nT within 1 min, (2) a cumulative drop exceeding 30 nT within 2 min, (3) a cumulative drop exceeding 45 nT within 3 min, and (4) the average SML value over the subsequent 26 min remaining at least 100 nT lower than its value at onset. The onset time is defined as the last minute before the initial 15 nT drop, and a minimum separation of 20 min is enforced between consecutive events. This objective definition provides a consistent reference for identifying abrupt and persistent SML disturbances commonly associated with substorm activity, providing a reliable reference for assessing the time difference between PBI occurrence and substorm initiation.

2.2 Methodology

This study focuses on nighttime auroral observations from all-sky images at YRS during winter months from 2012 to 2024. To visualize the transport of plasma patches across the polar cap and their interaction with the nightside auroral oval, we constructed magnetic-meridian keograms from all-sky image sequences. Each keogram was generated by extracting a vertical column of pixels along the magnetic meridian from individual frames and stacking these columns on time, thereby capturing the temporal evolution and latitudinal motion of individual plasma structures.

Only clear-sky, quality-controlled all-sky imager observations were used, and each keogram of event was also checked by the corresponding all-sky images to avoid contamination from clouds or instrumental artifacts. The reported annual, monthly, and MLT distributions represent the relative occurrence of PBIs under valid observational conditions, and the influence of cloud coverage and seasonal availability is minimized.

On each keogram, we record (i) the time and corresponding magnetic latitude when a plasma polar cap patch (identified as a structure with significantly lower counts than the background) passes the station’s zenith and (ii) the time and latitude at which the patch propagates poleward to the auroral oval’s poleward boundary, accompanied by a PBI (the distinct bright spot on the poleward boundary of the auroral oval in the keogram). An event was classified as a valid PBI only if it satisfied the following criteria: (1) the plasma patch reached the auroral oval poleward boundary, and produce a distinct localized brightening, clearly exceeding the background emission level on the keogram, and (2) the brightening exhibited a clear poleward bulge or structural enhancement at the boundary region (Forsyth et al., 2020).

The event time was defined as the moment when the patch contacted the poleward boundary of the auroral oval and the associated PBI became visible in the keograms. Following this selection procedure, we identified a total of 1,225 PBI events, which serve as the basis for the detailed statistical analyses in Section 3.

Figures 1a,b display a representative keogram during 19:00–24:00 UT on 2 January 2022, demonstrating the arrival of polar cap patches at the auroral oval’s poleward boundary and the associated onset of a PBI. In order to enhance the visibility of auroral structures (Figure 1a) and highlight regions of enhanced emission (Figure 1b), different intensity scales were used in the top two panels, Figure 1a for 0-1,200 counts, and Figure 1b for 0-2000 counts, respectively. Dashed lines mark the PBI region along the poleward boundary. In all observed cases, the interaction of a polar cap plasma patch with the poleward boundary coincides with the emergence of a discrete PBI.

Figure 1

Figure 1. (a,b) Keogram of 630.0 nm auroral emissions observed at the Yellow River Station on 2 January 2022 from 19:00 to 24:00 UT. The solid black line denotes the zenith angle, the dashed black lines indicate the evolution of polar cap patches, and the dashed white lines correspond to the appearance of PBI under different count values. (c-h) Solar wind, interplanetary magnetic field, and geomagnetic condition during the same period. (c) IMF components in the geocentric solar magnetospheric (GSM) coordinate system: Bx (blue), By (black) and Bz (red); (d) Solar wind density (N_SW); (e) Solar wind Velocity (V); (f) Solar wind dynamic pressure (P_Dyn); (g) SYM-H geomagnetic index; (h) Auroral electrojet indices: SME (red), SMU (blue) and SML (green).

Figures 1c–h show the concurrent IMF conditions and solar wind parameters, and geomagnetic indices. The events occurred under a persistently southward IMF Bz, with IMF By remaining negative after 19:40 UT. During this interval, the solar wind velocity was approximately 500 km/s, while both proton density and dynamic pressure stayed relatively stable. The SYM-H index stayed low (Figure 1g), indicating a quiet geomagnetic background, favorable for identifying PBI–substorm relationships.

Furthermore, during the two intervals when multiple consecutive PBIs occurred (19:40–20:40 UT and 21:30–22:30 UT), Figure 1h reveals a concurrent increase in the auroral electrojet SME index and a sharp decrease in the westward electrojet SML index, corresponding to two successive substorm onsets. This case exemplifies that sustained PBI activity can span the entire substorm cycle, reinforcing the coupling between magnetotail processes and ionospheric auroral responses.

3 Statistical results

Based on the dataset described above, we conducted a statistical analysis to examine the solar wind and IMF conditions associated with the arrival of polar cap patches and the subsequent occurrence of PBIs. For each PBI event, the associated solar wind and IMF conditions were defined as the averages over the 60–90 min interval from the OMNI dataset immediately preceding the PBI onset (Rong et al., 2015; Han et al., 2020; Feng et al., 2021; Lu et al., 2022). As a reference “background” distribution, we compiled all available OMNI data from periods when the Yellow River station was operational in the nightside (18–06 MLT), from winter 2012 to spring 2024. A consecutive 30-min window was used to calculate mean solar wind and IMF parameters for each day, yielding a totally 14,376 background samples.

All statistics results are calculated using only time intervals when the ASI was operational and provided usable nightside observations (18–06 MLT). Therefore, the reported distributions represent conditional occurrence rates based on observational availability, rather than absolute occurrence frequencies.

3.1 Solar–geophysical conditions

Figure 2a shows the annual occurrence of the PBIs overlapped with the sunspot number. It indicates a clear correlation between the PBI occurrence and the sunspot number, with the number of PBI events increasing significantly during high solar activity years. Compared with the background dataset, PBIs exhibit a pronounced solar-cycle variation. Figures 2b,c show the monthly and MLT distributions of the observed PBI events from 2012 to 2024, indicating that the PBIs occur mainly in winter and more frequently during 20-01 MLT.

Figure 2

Figure 2. Dependence of poleward boundary intensification (PBI) events from winter 2012 to spring 2024 on solar cycle variation, interplanetary magnetic field (IMF) conditions, solar wind parameters, and geomagnetic activity. Panels (a–c) show the distributions of PBI events in year, month, and magnetic local time (MLT), respectively. Panels (d–f) present the distributions of the IMF components Bx, By, and Bz. Panels (g–i) show the distributions of solar wind flow speed, dynamic pressure, and proton density. Panels (j) and (k) display the distributions of the AE and SYM-H geomagnetic indices, respectively. Panel (l) shows the distribution of the IMF clock angle, calculated as arctan(By/Bz). Blue bars represent background conditions, while red bars indicate conditions during PBI events.

Figures 2d–k present comparative histograms for IMF components (Bx, By, Bz), solar wind parameters (speed, dynamic pressure, proton density), and geomagnetic indices (SME, SYM-H), with red bars representing PBI-associated intervals and blue bars showing the background distributions. As shown in Figures 2d–f, it is evident that PBIs tend to occur preferentially during southward IMF Bz conditions, with a slight preference for positive IMF By. And PBIs are independent on IMF Bx. Furthermore, the effect of IMF By on the PBIs MLT distribution will be explored in Section 4.

In Figures 2g–i, dynamic pressure shows no significant difference between PBIs and background, while higher solar wind speed and lower proton density are slightly more prevalent during PBI occurrences. Under conditions of steady solar wind dynamic pressure, a combination of higher solar wind speed and reduced proton density indicates that the magnetosphere is driven predominantly by electromagnetic forcing rather than by dynamic compression (Dorelli, 2019; Madelaire et al., 2022). Such a condition enhances the efficiency of magnetic reconnection and strengthens solar wind-magnetosphere energy coupling, frequently leading to more substorm activity and intensified auroral emissions (Grandin et al., 2019). The AE index is significantly elevated during PBI-associated intervals, suggesting that most PBIs occur during or shortly after substorm activity. PBIs preferentially occur during intervals of more negative SYM-H, indicating enhanced magnetospheric activity and increased energy loading in the magnetotail, which is commonly associated with stronger magnetotail convection and more frequent reconnection-driven plasma transport. Such conditions are favorable for the generation of bursty bulk flows in the magnetotail and their ionospheric signatures, including PBIs.

The distribution of IMF clock angles further emphasizes this relationship: compared with background periods, PBIs predominantly occur under southward IMF condition (clock angle between 90° and 270°), with a preference for positive IMF By. These results highlight the important role of southward Bz in enhancing magnetotail reconnection and auroral particle precipitation processes that increase the likelihood of auroral boundary disturbances such as PBIs.

3.2 Dawn-dusk asymmetry of PBIs modulated by IMF by

To evaluate how plasma patches transport from the dayside influencing the spatial pattern of PBIs onset on the nightside, it is essential to consider the distinct temporal scales involved: plasma patches typically take 45 min to 2 h to traverse the polar cap (Eriksen et al., 2023; Moen et al., 2013; Spicher et al., 2015), whereas variations in solar wind–magnetosphere coupling—particularly changes in the IMF—can propagate to the polar cap in just 10–20 min. Furthermore, the reconfiguration of the twin-cell convection pattern is influenced by IMF conditions over the preceding 1–2 h and typically exhibits response times longer than 10–20 min (Fiori et al., 2012; Snekvik et al., 2017; Shore et al., 2019; Lockwood and Cowley, 2022). Thus, examining only a 10–20 min window is insufficient to characterize the convection-cell exit location. To capture the combined nightside substorm activity and sustained IMF effects on the nightside convection exit, it is essential to include IMF conditions over a 30–90 min interval (Rong et al., 2015). Within this time window, the variation trends of the IMF generally exhibit consistent characteristics. Thus, in this study, we adopt a 30–60 min IMF history window and use the 30-min average within this window to reflect the relevant response time of the high-latitude ionosphere to dayside changes.

To investigate how IMF evolution modulates nightside polar cap patch arrivals and subsequent PBI onset, we classify all events into three IMF clock-angle variability classes—Quasi-steady, Oscillating, and Turning (Zou et al., 2015b). An event is classified as Quasi-steady if the IMF clock angle fluctuates within ±30° over the 30 min preceding the PBI, indicating a persistently stable ambient field. It is categorized as oscillating if the clock angle undergoes multiple shifts exceeding 30° within the same 30-min response interval, reflecting sustained solar wind disturbances. Finally, an event is labeled Turning if only a single clock-angle shift greater than 30° occurs, signifying a discrete transition in IMF orientation. To ensure statistical reliability, MLT bins containing fewer than six events are excluded from the occurrence-rate analysis.

Based on the above classification, Figure 3 explores how IMF By and Bz modulated PBI occurrence across different MLT sectors. For the total event set (Figure 3e), PBI occurrence under southward IMF Bz is relatively uniform across MLT bins. Although the total number of PBIs under positive and negative IMF By is comparable, their occurrences exhibit a dawn–dusk asymmetry around 23 MLT: events before 23 MLT are more common under + By, while those after 23 MLT favor -By (Figure 3i). On average, polar cap patches exit the convection system at 22:22 MLT under + By condition and at 23:07 MLT under -By condition.

Figure 3

Figure 3. (a) Frequency distribution of IMF clock angles (arctan (By/Bz)) under all kinds of event conditions. The numbers 20, 40, 60, and 80 indicate the percentage of PBI events within each 30° clock angle bin. (b–d) Event classification by clock angle variability: (b) Quasi-steady, (c) Oscillating, and (d) Turning. (e–h) Relative occurrence rates of IMF Bz as a function of MLT for all events (e), Quasi-steady (f), Oscillating (g), and Turning (h). (i–l) Relative occurrence rates of IMF By as a function of MLT for all events (e), Quasi-steady (f), Oscillating (g), and Turning (h). In each panel, the gray histograms show the background distribution of all OMNI samples across MLT bins under the corresponding IMF condition. Error bars in (e–l) represent binomial uncertainties (±1σ).

Under quasi-steady IMF conditions (Figure 3f), which account for 53% of all events, PBI occurrences exhibit a clear dependence on southward IMF Bz, and occurrence exceeds 75% across all MLT sectors. A modest asymmetry is evident around 23 MLT. PBIs are more prevalent during positive By conditions prior to midnight.

In contrast, under oscillating IMF conditions (Figure 3g), the dependence on southward Bz largely disappears, and occurrences under positive and negative IMF Bz conditions are comparable. Nevertheless, the dawn–dusk tendency in PBI occurrence corresponding to the sign of IMF By, particularly around 23 MLT (Figure 3k), implies that IMF By continues to influence the nightside convection-cell exit even during periods of strong solar wind variability.

For Turning-type cases, defined by a significant change in IMF clock angle (Figures 3h,l), the dependence on southward Bz becomes even more prominent, accompanied by an enhanced PBI occurrence under positive IMF By conditions before local midnight. These results provide compelling evidence that transient IMF polarity shifts can enhance magnetotail reconnection and modulate asymmetric convection patterns (Ohma et al., 2022; Pitkänen et al., 2021), thereby governing the locations and timing of PBI occurrence.

As shown in Figure 3, the dawn–dusk asymmetry around ∼23 MLT exceeds the binomial uncertainties in the steady and turning IMF cases, demonstrating that the observed difference is statistically robust rather than arising from sampling variability.

3.3 MLT - MLAT distribution of PBI events

To further quantify the spatial and temporal distribution of PBI occurrence, we analyzed the magnetic latitude (MLAT) and MLT coordinates of all events derived from auroral keograms. Figure 4 displays the normalized probability of PBIs across MLAT–MLT bins, with normalization by the total number of events under each IMF condition.

Figure 4

Figure 4. Distribution of PBI events in MLAT–MLT coordinates under different IMF conditions. Panel (a) shows the distribution of all PBI events. Panels (b–d) show the distributions of PBI events under steady, oscillating, and turning IMF conditions, respectively. The grid resolution is 2° in MLAT and 30 min in MLT. Colorbar values denote the percentage of the number of events relative to the sample size of each case.

As shown in Figure 4a, PBIs occur predominantly in the pre-midnight sector, with the highest occurrence concentrated between 21 and 01 MLT and within a magnetic latitude range of 70°–76°. This spatial preference overlaps with the typical nightside reconnection region and suggests enhanced convection toward the poleward boundary in this sector. After midnight, during the substorm recovery phase, the auroral oval expands equatorward beyond the station’s field of view, leading to a noticeable drop in observed PBI frequency. It should be noted that the statistical results presented in this study are further constrained by the field of view of the YRS, whose lowest observable geomagnetic latitude is approximately 70° MLAT. Therefore, the results in Figure 4 represent conditional occurrence within the observable nightside polar-cap–auroral boundary region, rather than the full latitudinal extent of all PBIs.

3.4 Association between PBIs and substorm activity

To investigate the temporal coupling between PBIs and substorm activity, we compared the occurrence of PBIs with substorm onsets, which were identified using the SML threshold method proposed by Newell and Gjerloev, 2011. According to this method, a substorm onset is defined as a sustained SML decrease of 45 nT over 3 min, lasting at least 30 min. For each PBI, the earliest onset within ∼30 min is adopted to reduce contamination from subsequent intensifications.

As shown in Figure 5, PBIs occurrence exhibits a gradual increase before substorm onset, following a sharp rise immediately after substorm onset. Within 20 min after onset, the occurrence is more than twice that 20 min before onset. The occurrence decreases from the ∼40 min after onset and recovers to the pre-onset level approximately in 2 h.

Figure 5

Figure 5. Distribution of PBI occurrence times relative to substorm onset (t = 0). Substorm onsets are identified using the SML index following the threshold criteria of Newell and Gjerloev, 2011.

The temporal variation of PBI occurrence is consistent with the findings of Yadav et al. (2024), who reported that sustained sequences of PBIs and their associated poleward-traveling auroral forms are characteristic features of the substorm expansion phase. These results suggest that PBIs are statistically enhanced prior to substorm onset and remain prevalent throughout the subsequent substorm phases, demonstrating a close temporal association between PBIs and substorm evolution.

Notably, beyond the total number of PBIs, the timing of the first PBI associated with each substorm (pink bars in Figure 5) also exhibits a strong correlation with substorm onset. Specifically, the occurrence of the first PBI begins to rise distinctly approximately 20 min before onset, surges sharply at the onset time, and reaches its peak within 20 min after onset. This distribution highlights the close temporal coupling between PBI initiation and substorm onset.

4 Discussion

This study investigates how the variations in the IMF influence the spatiotemporal distribution of PBIs. As shown in Figure 3i, PBI occurrence shows a clear dependence on the sign of IMF By across MLT sectors: under positive/negative IMF By conditions, increased PBI occurrence is observed before/after 23 MLT, respectively. Statistical study is conducted to results reflect the influence of IMF By on the twin-cell convection morphology, under quasi-steady clock-angle conditions. Specifically, when IMF By is positive, magnetic tension forces stretch the dayside dawn-cell convection duskward (Tenfjord et al., 2015), shifting the poleward-boundary convection exit—and hence PBI occurrence—toward earlier MLTs. Conversely, for negative IMF By, the dusk-cell dominates, leading to PBIs occurrence peaks in post-23 MLT dawn sectors.

Under quasi-steady IMF conditions, the spatiotemporal distribution of PBIs closely matches the total event set. A pronounced concentration is observed in 22–01 MLT sector (Figure 4b). Most pre-midnight PBIs during this interval correlate with +By, consistent with reports of more frequent and intense substorms under such conditions (Liou et al., 2001). This interval also features enhanced magnetotail reconnection rates and intensified bursty bulk flows (BBFs), which IMF By modulates. These observations indicate that PBIs are optical signatures of magnetotail energy release and intensified reconnection (Lyons et al., 2011; Tenfjord et al., 2015; Wilkins et al., 2023). Furthermore, electrons with energies ≥50 keV predominantly precipitate in the 21–01 MLT sector (Wilkins et al., 2023). This precipitation enhances field-aligned conductance and convection, thereby shaping the observed PBIs distribution.

In contrast, oscillating IMF conditions—characterized by frequent IMF reversals and large fluctuations—significantly weaken the cumulative influence of IMF By on the convection-cell exit. Consequently, PBIs distribute more broadly across MLT, with the peak occurrence shifting modestly to post-midnight hours (Figure 4c). This pattern reveals a more extensive spatiotemporal distribution than the quasi-steady case (Figure 4b). The 30-min pre-event averaging window under such oscillating conditions often contains multiple By reversals, which can obscure clearer By/Bz dependencies. Nevertheless, the dominance of quasi-steady intervals in the overall dataset ensures robust key statistical trends.

For turning-type events, which involve a single discrete shift in the IMF clock angle (Figures 3h,l), PBI activity again concentrates pre-midnight (Figure 4d). A southward reversal of Bz in these cases facilitates enhanced magnetic reconnection and plasma entry into the ionosphere, elevating geomagnetic activity. PBIs observed during these intervals most often associate with southward Bz (Figure 3h) and show a concurrent increase under positive By (Figure 3l).

In this study, PBIs are observed throughout the entire substorm cycle (Figure 5), indicating their close connection to substorm evolution. In Figure 5, both the frequency and occurrence rate of PBIs increase sharply around substorm onset. This increase coincides with enhanced frequency and amplitude of BBFs in the near-Earth magnetotail after onset (Juusola et al., 2011; Merkin et al., 2019), appearing as intensified PBIs activity. On the one hand, PBIs are closely associated with substorm initiation and serve as ionospheric footprints of bursty bulk flows, which are generated by nightside magnetotail reconnection (Mende et al., 2011; Elhawary et al., 2022). Specifically, nightside magnetotail reconnection drives the formation of these localized high-speed flows, which propagate earthward and deposit energy in the polar ionosphere-magnetosphere coupling region along the auroral oval’s poleward boundary, ultimately triggering the onset of PBIs. On the other hand, they can also be regarded as optical signatures of energy release after onset. During the expansion phase, the magnetotail releases stored energy explosively. Fast earthward flows transport plasma and magnetic flux into the inner magnetosphere while simultaneously driving PBIs along the auroral oval (Zesta et al., 2002; Mende et al., 2011; Forsyth et al., 2020).

Another notable feature is that the PBI occurrence rate does not immediately return to background levels after onset. Instead, it takes nearly 2 h for the frequency to decline to the pre-onset baseline—defined as the level measured 20 min before onset. This delay matches the characteristic substorm timescale. After energy release, the magnetotail undergoes a 1–2 h reloading phase. During this time, plasma and magnetic flux replenish, restoring conditions that favor reconnection and BBF activity. Ionospheric conductance also recovers gradually as precipitation weakens (Baker et al., 1996; Ebihara, 2019).

As discussed above, IMF strongly modulates the spatial distribution of PBIs around midnight by shifting the ionospheric convection exit. Statistically, PBI occurrence rises markedly on both sides of midnight, but the pre-midnight rate is nearly triple the post-midnight rate (Figure 4a). This pattern reflects not only the By-controlled dawn–dusk asymmetry but also the spatiotemporal distribution of substorm activity. Specifically, about 80% of substorm onsets occur between 22 and 24 MLT (Liou et al., 2001; Nishimura et al., 2010), an interval that largely overlaps with the pre-midnight PBIs enhancement. Moreover, over 95% of PBIs are accompanied by substorm processes. During the expansion phase, magnetotail reconnection injects fast earthward plasma flows (e.g., BBFs or bursty jets) into the polar ionosphere. These flows transiently enhance one of the convection cells near midnight and amplify dawn–dusk asymmetry (Grocott et al., 2003; Ruohoniemi and Greenwald, 2005; Reistad et al., 2018; Walach et al., 2022). During the growth and recovery phases, redistribution of magnetotail flux and reconfiguration of ionospheric convection maintain and deepen this asymmetry (Walach et al., 2022).

Taken together, our results indicate that the enhanced pre-midnight PBI occurrence results from the combined effects of an IMF By-controlled background asymmetry and a higher probability of substorm onset associated with strong per-midnight flow injections. Together, these factors produce the pronounced dawn–dusk asymmetry in ionospheric PBIs.

It should be noted that this study tracked only those PBIs associated with polar cap patches reaching the poleward auroral boundary, rather than all forms of poleward-boundary brightening. While this selection may exclude some PBIs unrelated to patch arrivals, it ensures a physical context linking PBIs to plasma transport processes.

Although patch-associated PBIs therefore provide an effective ionospheric indicator of enhanced nightside plasma transport and boundary interactions within the coupled magnetosphere–ionosphere system, our use of two-dimensional keograms from the single-site Yellow River Station limited our ability to capture the global dynamics at both dawn and dusk convection exits. Moreover, the two-dimensional nature of keogram tracking may underestimate the true modulation of three-dimensional convection reconfiguration by IMF By.

While the present statistics are restricted to patch-associated PBIs observable within the field of view of the Yellow River Station, previous studies indicate that such events constitute a dynamically important subset of PBIs that are directly linked to enhanced plasma transport and reconnection-driven flows. Therefore, our results are not intended to represent all PBIs, but rather to quantify the spatiotemporal behavior of PBIs occurring in a well-defined plasma transport context.

5 Conclusion

Based on high spatiotemporal resolution observations from the all-sky imager at the Yellow River Station, this study focuses on PBIs triggered by polar cap patches arriving auroral oval, and statistically analyzes the dependence of PBIs occurrence on solar wind, IMF, and geomagnetic indices during an entire solar cycle. The results reveal a significantly enhanced occurrence frequency in the 22–01 MLT sector. PBI events are strongly associated with substorms, peaking in the expansion phase within 20–40 min of onset, suggesting a close temporal association with substorm evolution. The distribution of these events is also controlled by IMF By, shifting systematically near 23 MLT. Consequently, PBIs serve as effective tracers for quantifying how polar cap plasma enters the nightside auroral oval and responds to solar wind conditions. These findings may improve our understanding of nightside plasma transport associated with the polar cap patches and auroral dynamics.

Data availability statement

The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author.

Author contributions

Y-LF: Formal Analysis, Investigation, Methodology, Writing – original draft, Writing – review and editing. Z-YX: Data curation, Funding acquisition, Project administration, Resources, Writing – review and editing. Q-HZ: Writing – original draft. H-XF: Writing – review and editing. H-GY: Data curation, Formal Analysis, Methodology, Writing – review and editing. Z-JH: Data curation, Formal Analysis, Writing – review and editing. YW: Formal Analysis, Funding acquisition, Writing – review and editing. Y-ZM: Formal Analysis, Methodology, Project administration, Writing – review and editing. SL: Writing – original draft. Z-FX: Methodology, Resources, Writing – original draft. B-LZ: Formal Analysis, Investigation, Methodology, Writing – review and editing. X-YW: Conceptualization, Formal Analysis, Methodology, Writing – review and editing. DZ: Data curation, Investigation, Methodology, Writing – review and editing. X-MC: Formal Analysis, Methodology, Writing – review and editing.

Funding

The author(s) declared that financial support was received for this work and/or its publication. This work is supported by the National Natural Science Foundation of China (grants No. 42325404, 42120104003, 42474219, 42204164, 42441828), Shandong Provincial Natural Science Foundation (Grant ZR2022MD034), the Stable-Support Scientific Project of China Research Institute of Radiowave Propagation (A241204230), the Basic Research Project (DLCGZIYY-2024-038-02), the Chinese Meridian Project, and the Xiaomi Young Talents Program. The authors thank the International Space Science Institute (ISSI) in Beijing, through ISSI International Team project #511 (Multi-Scale Magnetosphere-Ionosphere-Thermosphere Interaction).

Acknowledgements

This work is supported by the National Natural Science Foundation of China (grants No. 42325404, 42120104003, 42474219, 42204164, 42441828), Shandong Provincial Natural Science Foundation (Grant ZR2022MD034), the Stable-Support Scientific Project of China Research Institute of Radiowave Propagation (A241204230), the Basic Research Project (DLCGZIYY-2024-038-02), the Chinese Meridian Project, and the Xiaomi Young Talents Program. The authors thank the International Space Science Institute (ISSI) in Beijing, through ISSI International Team project #511 (Multi-Scale Magnetosphere-Ionosphere-Thermosphere Interaction). We would like to thank the China Polar Research Center for operating the Yellow River Station and providing the all-sky auroral imaging data. We also acknowledge NASA’s OMNIWeb service for making the solar wind and IMF data available. Furthermore, we acknowledge the substorm timing list identified using the Newell and Gjerloev technique (Newell and Gjerloev, 2011), the SMU and SML indices (Newell and Gjerloev, 2011); and the SuperMAG collaboration for providing ground-based magnetometer data (Gjerloev et al., 2012).

Conflict of interest

The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fspas.2026.1725682/full#supplementary-material

References

Baker, D. N., Pulkkinen, T. I., Angelopoulos, V., Baumjohann, W., and McPherron, R. L. (1996). Neutral line model of substorms: past results and present view. J. Geophys. Res. Space Phys. 101 (A6), 12975–13010. doi:10.1029/95JA03753

CrossRef Full Text | Google Scholar

Chu, X., McPherron, R. L., Hsu, T., and Angelopoulos, V. (2015). Solar cycle dependence of substorm occurrence and duration: implications for onset. J. Geophys. Res. Space Phys. 120 (4), 2808–2818. doi:10.1002/2015ja021104

CrossRef Full Text | Google Scholar

Dorelli, J. C. (2019). Does the solar wind electric field control the reconnection rate at earth’s subsolar magnetopause? J. Geophys. Res. Space Phys. 124 (4), 2668–2681. doi:10.1029/2018JA025868

CrossRef Full Text | Google Scholar

Dungey, J. W. (1961). Interplanetary magnetic field and the auroral zones. Phys. Rev. Lett. 6 (2), 47–48. doi:10.1103/PhysRevLett.6.47

CrossRef Full Text | Google Scholar

Ebihara, Y. (2019). Simulation study of near-earth space disturbances: 2. Auroral substorms. Prog. Earth Planet. Sci. 6 (1), 24. doi:10.1186/s40645-019-0273-2

CrossRef Full Text | Google Scholar

Elhawary, R., Laundal, K. M., Reistad, J. P., and Hatch, S. M. (2022). Possible ionospheric influence on substorm onset location. Geophys. Res. Lett. 49 (4), e2021GL096691. doi:10.1029/2021gl096691

CrossRef Full Text | Google Scholar

Eriksen, N. K., Lorentzen, D. A., Oksavik, K., Baddeley, L., Hosokawa, K., Shiokawa, K., et al. (2023). On the creation, depletion, and end of life of polar cap patches. J. Geophys. Res. Space Phys. 128 (12), e2023JA031739. doi:10.1029/2023JA031739

CrossRef Full Text | Google Scholar

Feng, H.-T., Han, D., Qiu, H., Shi, R., Yang, H., and Zhang, Y. (2021). Observational properties of 15MLT-PCA in the Southern hemisphere and the switching effects of IMF B_y on 15MLT-PCA occurrence. J. Geophys. Res. Space Phys. 126 (12), e2021JA029140. doi:10.1029/2021JA029140

CrossRef Full Text | Google Scholar

Fiori, R. A. D., Boteler, D. H., and Koustov, A. V. (2012). Response of ionospheric convection to sharp southward IMF turnings inferred from magnetometer and radar data. J. Geophys. Res. Space Phys. 117 (A9), 2012JA017755. doi:10.1029/2012JA017755

CrossRef Full Text | Google Scholar

Forsyth, C., Sergeev, V. A., Henderson, M. G., Nishimura, Y., and Gallardo-Lacourt, B. (2020). Physical processes of Meso-Scale, dynamic auroral forms. Space Sci. Rev. 216 (4), 46. doi:10.1007/s11214-020-00665-y

CrossRef Full Text | Google Scholar

Gjerloev, J. W. (2012). The SuperMAG data processing technique. J. Geophys. Res. 117, A09213. doi:10.1029/2012JA017683

CrossRef Full Text | Google Scholar

Grandin, M., Aikio, A. T., and Kozlovsky, A. (2019). Properties and Geoeffectiveness of solar wind high-speed streams and stream interaction regions during solar cycles 23 and 24. J. Geophys. Res. Space Phys. 124 (6), 3871–3892. doi:10.1029/2018JA026396

CrossRef Full Text | Google Scholar

Grocott, A., Cowley, S. W. H., and Sigwarth, J. B. (2003). Ionospheric flow during extended intervals of northward but BY-Dominated IMF. Ann. Geophys. 21 (2), 509–538. doi:10.5194/angeo-21-509-2003

CrossRef Full Text | Google Scholar

Han, D.-S., Feng, H., Zhang, H., Zhou, S., and Zhang, Y. (2020). A new type of polar cap arc observed in the ∼1500 MLT sector: 1. Northern hemisphere observations. Geophys. Res. Lett. 47 (20), e2020GL090261. doi:10.1029/2020GL090261

CrossRef Full Text | Google Scholar

Hosokawa, K., Kashimoto, T., Suzuki, S., Shiokawa, K., Otsuka, Y., and Ogawa, T. (2009). Motion of polar cap patches: a statistical study with all-sky airglow imager at resolute Bay, Canada. J. Geophys. Res. Space Phys. 114 (A4), 2008JA014020. doi:10.1029/2008JA014020

CrossRef Full Text | Google Scholar

Juusola, L., Østgaard, N., Tanskanen, E., Partamies, N., and Snekvik, K. (2011). Earthward plasma sheet flows during substorm phases: PS FLOWS DURING SUBSTORMS. J. Geophys. Res. Space Phys. 116 (A10). doi:10.1029/2011JA016852

CrossRef Full Text | Google Scholar

Liou, K., Newell, P. T., Meng, C., Brittnacher, M., and Parks, G. (1998). Characteristics of the solar wind controlled auroral emissions. J. Geophys. Res. Space Phys. 103 (A8), 17543–17557. doi:10.1029/98JA01388

CrossRef Full Text | Google Scholar

Liou, K., Meng, C.-I., Newell, P. T., Lui, A. T. Y., Reeves, G. D., and Belian, R. D. (2001). Particle injections with auroral expansions. J. Geophys. Res. Space Phys. 106 (A4), 5873–5881. doi:10.1029/2000ja003003

CrossRef Full Text | Google Scholar

Lockwood, M., and Cowley, S. W. H. (2022). Magnetosphere-ionosphere coupling: implications of non-equilibrium conditions. Front. Astronomy Space Sci. 9, 908571. doi:10.3389/fspas.2022.908571

CrossRef Full Text | Google Scholar

Lockwood, M., Sandholt, P. E., and Cowley, S. W. H. (1989). Dayside auroral activity and magnetic flux transfer from the solar wind. Geophys. Res. Lett. 16 (1), 33–36. doi:10.1029/GL016i001p00033

CrossRef Full Text | Google Scholar

Lu, S., Xing, Z. Y., Zhang, Q. H., Zhang, Y. L., Ma, Y. Z., Wang, X. Y., et al. (2022). A statistical study of space hurricanes in the Northern hemisphere. Front. Astronomy Space Sci. 9, 1047982. doi:10.3389/fspas.2022.1047982

CrossRef Full Text | Google Scholar

Lyons, L. R., Nishimura, Y., Kim, H. J., Donovan, E., Angelopoulos, V., Sofko, G., et al. (2011). “Possible connection of polar cap flows to Pre- and post-substorm onset PBIs and streamers: POLAR CAP FLOWS AND SUBSTORM ACTIVITY. J. Geophys. Res. Space Phys. 116. A12 doi:10.1029/2011ja016850

CrossRef Full Text | Google Scholar

Madelaire, M., Laundal, K. M., Reistad, J. P., Hatch, S. M., Ohma, A., and Haaland, S. (2022). Geomagnetic response to rapid increases in solar wind dynamic pressure: event detection and large scale response. Front. Astronomy Space Sci. 9, 904620. doi:10.3389/fspas.2022.904620

CrossRef Full Text | Google Scholar

Mende, S. B., Frey, H. U., Angelopoulos, V., and Nishimura, Y. (2011). Substorm triggering by poleward boundary intensification and related equatorward propagation. J. Geophys. Res. Space Phys. 116 (A5). doi:10.1029/2010ja015733

CrossRef Full Text | Google Scholar

Merkin, V. G., Panov, E. V., Sorathia, K. A., and Ukhorskiy, A. Y. (2019). Contribution of bursty bulk flows to the global dipolarization of the magnetotail during an isolated substorm. J. Geophys. Res. Space Phys. 124 (11), 8647–8668. doi:10.1029/2019JA026872

PubMed Abstract | CrossRef Full Text | Google Scholar

Milan, S. E., Walach, M., Carter, J. A., Sangha, H., and Anderson, B. J. (2019). Substorm onset latitude and the steadiness of magnetospheric convection. J. Geophys. Res. Space Phys. 124 (3), 1738–1752. doi:10.1029/2018ja025969

CrossRef Full Text | Google Scholar

Moen, J., Gulbrandsen, N., Lorentzen, D. A., and Carlson, H. C. (2007). On the MLT distribution of F region polar cap patches at night. Geophys. Res. Lett. 34 (14). doi:10.1029/2007gl029632

CrossRef Full Text | Google Scholar

Moen, J., Oksavik, K., Alfonsi, L., Daabakk, Y., Romano, V., and Spogli, L. (2013). Space weather challenges of the polar cap ionosphere. J. Space Weather Space Clim. 3, A02. doi:10.1051/swsc/2013025

CrossRef Full Text | Google Scholar

Moen, J., Hosokawa, K., Gulbrandsen, N., and Clausen, L. B. N. (2015). On the symmetry of ionospheric polar cap patch exits around magnetic midnight. J. Geophys. Res. Space Phys. 120 (9), 7785–7797. doi:10.1002/2014ja020914

CrossRef Full Text | Google Scholar

Newell, P. T., and Gjerloev, J. W. (2011). Evaluation of SuperMAG auroral electrojet indices as indicators of substorms and auroral power. J. Geophys. Res. Space Phys. 116 (A12), 2011JA016779. doi:10.1029/2011JA016779

CrossRef Full Text | Google Scholar

Nishimura, Y., Lyons, L. R., Zou, S., Xing, X., Angelopoulos, V., Mende, S. B., et al. (2010). Preonset time sequence of auroral substorms: coordinated observations by all-sky imagers, satellites, and radars. J. Geophys. Res. Space Phys. 115 (A5). doi:10.1029/2010ja015832

CrossRef Full Text | Google Scholar

Nishimura, Y., Lyons, L. R., Zou, Y., Oksavik, K., Moen, J. I., Clausen, L. B., et al. (2014). Day-night coupling by a localized flow channel visualized by polar cap patch propagation. Geophys. Res. Lett. 41 (11), 3701–3709. doi:10.1002/2014gl060301

CrossRef Full Text | Google Scholar

Ohma, A., Laundal, K. M., Reistad, J. P., and Østgaard, N. (2022). Evolution of IMF by induced asymmetries during substorms: superposed Epoch analysis at geosynchronous orbit. Front. Astronomy Space Sci. 9. doi:10.3389/fspas.2022.958749

CrossRef Full Text | Google Scholar

Pitkänen, T., Kullen, A., Cai, L., Park, J. S., Vanhamäki, H., Hamrin, M., et al. (2021). Asymmetry in the earth’s magnetotail neutral sheet rotation due to IMF $$B_y$$ sign? Geosci. Lett. 8 (1), 3. doi:10.1186/s40562-020-00171-7

CrossRef Full Text | Google Scholar

Reistad, J. P., Østgaard, N., Laundal, K. M., Ohma, A., Snekvik, K., Tenfjord, P., et al. (2018). Observations of asymmetries in ionospheric return flow during different levels of geomagnetic activity. J. Geophys. Res. Space Phys. 123 (6), 4638–4651. doi:10.1029/2017JA025051

CrossRef Full Text | Google Scholar

Rong, Z. J., Lui, A. T. Y., Wan, W. X., Yang, Y. Y., Shen, C., Petrukovich, A. A., et al. (2015). Time delay of interplanetary magnetic field penetration into Earth’s magnetotail. J. Geophys. Res. Space Phys. 120 (5), 3406–3414. doi:10.1002/2014JA020452

CrossRef Full Text | Google Scholar

Ruohoniemi, J. M., and Greenwald, R. A. (2005). Dependencies of high-latitude plasma convection: consideration of interplanetary magnetic field, seasonal, and universal time factors in statistical patterns. J. Geophys. Res. Space Phys. 110 (A9), 2004JA010815. doi:10.1029/2004JA010815

CrossRef Full Text | Google Scholar

Shore, R. M., Freeman, M. P., and Gjerloev, J. W. (2019). Interplanetary magnetic field control of polar ionospheric equivalent current system modes. Space Weather 17 (7), 976–988. doi:10.1029/2019SW002161

CrossRef Full Text | Google Scholar

Snekvik, K., Østgaard, N., Tenfjord, P., Reistad, J. P., Laundal, K. M., Milan, S. E., et al. (2017). Dayside and nightside magnetic field responses at 780 Km altitude to dayside reconnection. J. Geophys. Res. Space Phys. 122 (2), 1670–1689. doi:10.1002/2016JA023177

CrossRef Full Text | Google Scholar

Spicher, A., Cameron, T., Grono, E. M., Yakymenko, K. N., Buchert, S. C., Clausen, L. B. N., et al. (2015). Observation of polar cap patches and calculation of gradient drift instability growth times: a swarm case study. Geophys. Res. Lett. 42 (2), 201–206. doi:10.1002/2014GL062590

CrossRef Full Text | Google Scholar

Tenfjord, P., Østgaard, N., Snekvik, K., Laundal, K. M., Reistad, J. P., Haaland, S., et al. (2015). How the IMF B_y induces a B_y component in the closed magnetosphere and how it leads to asymmetric currents and convection patterns in the two hemispheres. J. Geophys. Res. Space Phys. 120 (11), 9368–9384. doi:10.1002/2015ja021579

CrossRef Full Text | Google Scholar

Walach, M.-T., Grocott, A., Thomas, E. G., and Staples, F. (2022). Dusk-dawn asymmetries in SuperDARN convection maps. J. Geophys. Res. Space Phys. 127 (12), e2022JA030906. doi:10.1029/2022JA030906

PubMed Abstract | CrossRef Full Text | Google Scholar

Wilkins, C., Angelopoulos, V., Runov, A., Artemyev, A., Zhang, X., Liu, J., et al. (2023). Statistical characteristics of the electron isotropy boundary. J. Geophys. Res. Space Phys. 128 (10), e2023JA031774. doi:10.1029/2023ja031774

CrossRef Full Text | Google Scholar

Yadav, S., Lyons, L. R., Nishimura, Y., Liu, J., Tian, S., and Zou, Y. (2024). Investigating the spatiotemporal development of substorm expansion phase aurora: successive onsets or poleward boundary intensifications?. J. Geophys. Res. Space Phys.. doi:10.1029/2024JA033086

CrossRef Full Text | Google Scholar

Zesta, E., Donovan, E., Lyons, L., Enno, G., Murphree, J. S., and Cogger, L. (2002). “Two-dimensional structure of auroral poleward boundary intensifications. J. Geophys. Res. Space Phys. 107. A11 doi:10.1029/2001ja000260

CrossRef Full Text | Google Scholar

Zhang, Q.-He, Zhang, B. C., Lockwood, M., Hu, H. Q., Moen, J., Ruohoniemi, J. M., et al. (2013). Direct observations of the evolution of polar cap ionization patches. Science 339 (6127), 1597–1600. doi:10.1126/science.1231487

PubMed Abstract | CrossRef Full Text | Google Scholar

Zou, Y., Nishimura, Y., Lyons, L. R., Shiokawa, K., Donovan, E. F., Ruohoniemi, J. M., et al. (2015a). Localized polar cap flow enhancement tracing using airglow patches: statistical properties, IMF dependence, and contribution to polar cap convection. J. Geophys. Res. Space Phys. 120 (5), 4064–4078. doi:10.1002/2014ja020946

CrossRef Full Text | Google Scholar

Zou, Y., Nishimura, Y., Lyons, L. R., Donovan, E. F., Shiokawa, K., Ruohoniemi, J. M., et al. (2015b). Polar cap precursor of nightside auroral oval intensifications using polar cap arcs. J. Geophys. Res. Space Phys. 120 (12). doi:10.1002/2015JA021816

CrossRef Full Text | Google Scholar