The Poleward Shift of the Equatorial Ionization Anomaly During the Main Phase of the Superstorm on 10 May 2024

Abstract

On 10 May 2024, a super geomagnetic storm with a minimum Dst index of less than −400 nT occurred. It has attracted a significant amount of attention in the literature. Using total electron content (TEC) observations from a global navigation satellite system (GNSS), in situ electron density data from the Swarm satellite, and corresponding simulations from the thermosphere–ionosphere–electrodynamics general circulation model (TIEGCM), the dynamic poleward shift of the equatorial ionization anomaly (EIA) during the main phase of the super geomagnetic storm has been explored. The results show that the EIA crests moved poleward from ±15° magnetic latitude (MLat) to ±20° MLat at around 19.6 UT, to ±25° MLat at 21.2 UT, and to ±31° MLat at 22.7 UT. This poleward shift was primarily driven by the enhanced eastward electric field, neutral winds, and ambipolar diffusion. Storm-induced meridional winds can move ionospheric plasma upward/downward along geomagnetic field lines, causing the poleward movement of EIA crests, with minor contributions from zonal winds. Ambipolar diffusion contributes/prevents the formation of EIA crests at most EIA latitudes/the equatorward edge.

1. Introduction

In the equatorial ionosphere, there are a number of physical phenomena, such as the thermospheric wind jet that is confined to the dip equator [1,2], the equatorial electrojet and its counter above the dip equator [3,4], and the equatorial plasma fountain and its associated equatorial ionization anomaly (EIA) [5,6,7,8]. EIAs are among the most important physical phenomena and have received a large amount of attention. They are large-scale structures of electron density, with a trough at the dip equator and two crests at low latitudes of approximately ±15° magnetic latitude (MLat) during geomagnetic quiet time. The primary driver for EIAs is the daytime eastward electric field [8,9]. An upward E × B drift is thus generated and moves the plasma upward to a higher altitude above the dip equator. Then, the plasma moves along the geomagnetic field lines to a lower altitude but higher latitudes. Two crests with high plasma density will be formed at ±15° MLat.

In the literature, the behaviors of quiet-time EIAs have been widely explored [3,5,9,10,11,12,13,14,15,16,17,18,19,20]. Using a consistent set of observations from IMAGE FUV, TIMED GUVI, and OGO D 12, England et al. [5] found that the four-peak longitudinal structure was present in the post-sunset ionosphere and lasted until at least midnight. They suggested that atmospheric tides were responsible for the longitudinal pattern of EIAs since there was a good correlation between the tidally modulated winds and temperature in the thermosphere and the ionospheric plasma. This longitudinal wave-4 pattern was found to be the result of the diurnal eastward-propagating non-migrating tide with zonal wavenumber 3 (DE3) via the equatorial fountain effects. This finding was also confirmed by Lühr et al. [3,17]. Sudden stratospheric warming (SSW) involves rapid temperature increases of tens of degrees—reaching up to ~50 °C within days—at an altitude of 10–50 km. During SSWs, enhancement/suppression was found in morning/afternoon EIAs [12].

The behaviors of EIAs are also affected by geomagnetic storms, e.g., [21,22,23,24,25,26,27]. Dungey [28] demonstrated that the interaction between the interplanetary magnetic field and the geomagnetic field can trigger geomagnetic storms. When geomagnetic storms occur, the energy and momentum carried by solar wind can enter the upper atmosphere at high latitudes. During storms, disturbances in EIAs can be generated by the prompt penetrating electric field (PPEF), disturbance wind dynamo electric field (DDEF), and traveling atmospheric/ionospheric disturbances (TADs/TIDs). PPEF is a result of imbalance between Region 1 (R1) and Region 2 (R2) field-aligned currents (FACs) [29]. It is related to IMF Bz, which will be described in Section 3.1. It is eastward/westward during daytime for undershielding/overshielding when R1 FACs are stronger/weaker than R2 FACs. The eastward/westward PPEF can be immediately superposed into the ionospheric electric field at middle and low latitudes. The eastward/westward PPEF can lead to the enhancement/suppression of EIAs [22,23,26]. Energy deposition at the polar atmosphere can enhance the Joule heating, thereby increasing the neutral temperature [28]. Disturbances in neutral winds at high latitudes can travel to middle and low latitudes in 2 to 3 h; these are known as TADs/TIDs. A westward DDEF can be established by the dynamo effects of disturbed winds and weaken the EIA [30]. During storms, the generated equatorward winds can move the ionospheric plasma upward to a higher altitude with lower chemical recombination, enhancing the EIA [31].

On 10–14 May 2024, a super geomagnetic storm occurred, which has been termed the Mother’s Day storm. It had a minimum Dst index of less than −400 nT. This superstorm was the second strongest magnetic storm recorded since the study of geomagnetic storms began. The responses of the ionosphere and thermosphere to this superstorm have been studied in a series of studies, e.g., [24,25,26,27,32,33,34,35]. For instance, based on COSMIC-2, Spire, and FengYun-3 radio occultation observations, Lee et al. [35] found that the broadening of the EIA width during the main phase of the superstorm on 10 May 2024 was related to the electric field from the PPEF. Subsequently, the EIA crests merged and remained for about 21 hr at the end of the main phase and the recovery phase. Using joint observations from GNSS receivers, Swarm, DMSP, and TIMED, the dynamic expansion and merging of EIAs during the superstorm were explored by Aa et al. [26]. During the main phase, TECs at low latitudes were relatively enhanced up to 50–100%, and the EIA had an obvious poleward expansion to ±35° MLat [26,34]. This was driven by the enhanced equatorial fountain effects associated with the eastward PPEF, with contributions from ambipolar diffusion.

Although a number of studies have discussed the observations of this superstorm and EIAs during storms, the major differences and new findings in our work are given as follows. Our study found that the effects of the eastward PPEF, ambipolar diffusion, and neutral winds competed with each other, depending on the latitudes. This has never been described before, and it was the aim of this work. Moreover, the question of which forcing plays the dominant role needs clarification. It was addressed using TIEGCM in this work. This superstorm produced super fountain effects, hence the poleward movement of the EIA crests. With the help of the thermosphere–ionosphere–electrodynamics general circulation model (TIEGCM), detailed plasma transport from E × B drifts, ambipolar diffusion, and horizontal neutral winds (zonal and meridional) will be analyzed based on the momentum equation.

2. Data and Model

In this work, the observations from Swarm A, and GNSS receivers are processed to investigate the dynamic poleward movement of EIA crests. The Swarm mission was launched by the European Space Agency on 22 November 2013. The orbit of Swarm is near-polar. It comprises three identical satellites: Alpha, Bravo, and Charlie (A, B, and C) [36]. Swarm A and C fly side by side with a longitude gap of 1.4°. Swarm B orbits at a higher altitude. In this work, the in-situ electron density from Swarm A on 8–10 May 2024 is selected. On 10 May 2024, the altitude of Swarm A and B was 480–490 km and 520 km, respectively. The temporal and spatial resolution of Swarm observations is 1 s and approximately 7.6 km, respectively. Moreover, the data from the Langmuir probe (including Ne) are also available at 2 Hz as the basic product.

TEC is defined as the integral of the electron density along the ray path between the GNSS satellite and ground-based receivers. It has a unit of TECU, which equals 10¹⁶ electrons/m². Worldwide, the GNSS has thousands of receivers. TEC observations from the receivers provide an opportunity to understand the variability of the ionosphere. The temporal and horizontal resolution of the vertical TEC product from Madrigal is 5 min and 1° GLat × 1° GLon, respectively.

TIEGCM is a three-dimensional time-dependent model of the coupled ionosphere-thermosphere system. It was developed by the High Altitude Observatory at the National Center for Atmospheric Research. TIEGCM is driven by the high-latitude electric field solved by the Heelis or Weimer model [37,38], and the solar extreme ultraviolet and ultraviolet spectral fluxes parameterized by the F_10.7 index [39]. The low boundary of TIEGCM is specified by the diurnal and semidiurnal migrating and nonmigrating tides from the global-scale wave model (GSWM) [40,41]. The horizontal and vertical resolution of the model are 2.5° GLon × 2.5° GLat, and a quarter scale height, respectively. In this work, the high-latitude electric field is calculated by the empirical Weimer model based on the IMF data obtained from the OMNI website. The observed solar wind speed and density (Figure 1) are also imposed. On 10 May 2024, the solar activity index (F_10.7) was 227 sfu, and was used to parameterize the solar extreme ultraviolet and ultraviolet spectral fluxes. The lower boundary of TIEGCM is specified by the nonmigrating and migrating tides from the empirical GSWM. It is not the same on all days. The configuration of TIEGCM follows that described in Xia et al. [42]. The responses of the ionosphere and thermosphere can be achieved by the differences between simulations during storm time (10 May) and under quiet conditions (9 May).

3. Results

3.1. The Geomagnetic Condition

Figure 1 depicts the temporal variations in IMF Bx, By, and Bz, solar wind speed, solar wind density, and Dst index on 10–13 May 2024. In Figure 1f, the Dst index can be used to identify the main phase and the recovery phase of superstorm. At 00-16 UT on 10 May, the Dst index ranges from 00 to 18 nT. This is the pre-storm quiet-time interval. Then, the Dst index decreases rapidly to −412 nT by 02 UT on 11 May. The time interval from 17 UT on 10 May to 02 UT on 11 May indicates the main phase of superstorm. After the main phase, the Dst index slowly recovers from the trough to −56 nT by the end of 13 May. This time coverage is the recovery phase of a superstorm. In this work, the poleward movement of EIA crests is identified during the main phase. In Figure 1a, at the time interval of the main phase, IMF Bx is significantly perturbed, with several peaks and troughs. IMF Bz and to some extent IMF By are relevant to the dynamics observed [43]. The most important aspect for the dynamics investigated is that IMF Bz turns southward and remains strongly negative for an extended period. It had a consequence on PPEF-driven dynamics in the main phase. In Figure 1c, when IMF Bz varies rapidly, it is almost always southward. On 10 May, the peak of southward Bz is −18.0 and −35.0 nT at 18 and 21 UT, respectively. On 11 May, the southward Bz peaks at 00, 09, and 13 UT, with magnitudes of −35.3, −35.2, and −22.3 nT, respectively. In Figure 1b, during the period of significant changes, IMF By has two peaks. One is located at 23 UT on 10 May, with a magnitude of 67.6 nT. In Figure 1d, it is found that the speed increases significantly from around 400 km/s to 713.7 km/s within minutes at 17.4 UT on 10 May. Note here that the solar wind is observed at the L1 point and takes tens of minutes to reach the Earth. Then, the speed continues to increase slowly to a peak of 989.55 km/s at 1.7 UT on 12 May. In Figure 1e, during the main phase, most of the density is higher than 20 N/cm³. In Figure 1g, the Kp index is less than 3 before the superstorm (8–9 May, and 00-12 UT on 10 May), indicating the quiet condition. During the main phase, most of the Kp indices are larger than 8.0. The above results confirm the quiet-time condition on 8–9 May 2024. This can support the background selection in the Results subsection.

3.2. Data-Model Comparison

Figure 2 shows the geographic latitude (GLat) and geographic longitude (GLon) variations in GNSS TEC at 17, 19, 21, and 23 UT on 10 May 2024. The poleward extension of EIA is clearly shown at the main phase of the superstorm, which has been captured in Figure 3 as well. At 17 UT, the storm does not occur and is chosen as the pre-storm period. After that, we present a whole sequence (every 1–2 h) until the end of 10 May. This is the justification for why 17–23 UT was chosen in this work. In Figure 2a, the high TEC larger than 50 TECU is only found at equatorial latitudes less than 30° MLat. This feature is consistent with previous studies, e.g., [9,44,45,46,47]. Liang [47] demonstrated that the quiet-time EIA is the high plasma density crest at ±15° MLat. In other sectors, such as Europe, South Africa, and Australia, the observed TEC is low, with a magnitude close to zero, as they are on the nightside. During storm time, the spatial distributions of TEC are significantly different from those during quiet time. In Figure 2b,c, the peak TEC over America gradually increases, reaching approximately 70 and 85 TECU. The coverage of the high-density crest is wider in Figure 2c than in Figure 2a. For instance, the poleward edge seems to be located at around 50° Mlat at −120°~−60° GLon in the northern hemisphere in Figure 2c. Meanwhile, in Figure 2a, the northern poleward edge is around 30° MLat at −120°~−60° GLon. The poleward movement of EIA is our major focus and will be diagnosed singularly. In Figure 2d, the north poleward boundary of the high-density crest can be found around 45° MLat at −120°~−60° GLon in the northern hemisphere. The maximum TEC is close to 100 TECU.

On 10 May 2024, the local time of Swarm A is 19.2 LT. In Figure 3, at 19.2 LT, the observed MLat of quiet-time EIA crests at 00-16 UT remains at around 15° MLat and −13° MLat in northern and southern hemispheres, respectively. During the storm time of 17-00 UT, the MLat of EIA crests starts to quickly move poleward. For instance, in the northern hemisphere, the center MLat in the northern/southern hemisphere increases from 14.6°/−14.5° MLat to 30.3°/−25.0° MLat by the end of 10 May. This temporal evolution has been well reproduced by TIEGCM. Because the observations and simulations are similar to each other. The center MLat of EIA crests also increases significantly after the onset of superstorm.

In Figure 4a, the TIEGCM-modeled quiet-time TEC at 17 UT has two obvious peaks on both sides of the dip equator. Similar to the observations in Figure 2a, the high-density crest is located at low latitudes over America at around ±15° Mlat; however, with greater magnitudes (97.5/82.3 TECU on the North/South Hemisphere, respectively, vs. the observed ~70 TECU). Note here that the MLat information is obtained from IGRF, as indicated by the magenta lines. A similar result is found in the comparison between Figure 2b–d and Figure 4b–d. For instance, in Figure 4d, the poleward and equatorward edges of the north (south) crest are approximately 15° and 55° (−15° and −60°) MLat at −180°~−60° GLon, respectively. A comparison between Figure 2 and Figure 4 reveals that the model accurately reproduces the poleward extension of the EIA crests.

In Figure 5a, at 19.6 UT, Swarm A flies across the middle and low latitudes at approximately −6.0° GLon. This indicates that Swarm A has an LT of 19.2 LT. The in-situ observations can be used to diagnose the potential mechanism. In Figure 5a, on 8 (9) May 2024, the observed Ne shows two peaks at around ±10° MLat, with an almost equal density of 1.5 × 10¹² (1.7 × 10¹²) m⁻³. However, on 10 May 2024, the in-situ Ne has two more outstanding peaks than those on 8–9 May. The peaks on 10 May are located at 18° and −16° MLat, with a density of 2.3 × 10¹² and 2.5 × 10¹² m⁻³, respectively. In Figure 5b, at 21.2 UT, the two peaks of Ne on 10 May are 2.3 × 10¹² at −20° Mlat and 1.1 × 10¹² at 20° MLat. In Figure 5c, at 22.7 UT, the storm-time EIA is found at −25° MLat and 30° MLat, with a magnitude of 2.9 × 10¹² and 2.8 × 10¹² m⁻³, respectively. Based on Figure 5, it can be found that EIA moves significantly poleward from 19.6 UT to 22.7 UT.

In Figure 6a, on 8 May 2024, the modeled EIA at 20.6 UT has two peaks at around ±15° MLat, with a maximum density of 2.3 × 10¹² m⁻³. Swarm A flies through the nightside of 19.2 LT. The modeled Ne at the Swarm orbit has two peaks as well, with a maximum magnitude of approximately 1.5 × 10¹² m⁻³. This is close to the observations in Figure 5a. The same results could be achieved at the following two orbits on 8 May 2024 in Figure 6d,g, and the three consecutive orbits at 20.1–23.2 UT on 9 May 2024 in Figure 6b,e,h. Figure 6c,f,i give the global variations in storm-time Ne from TIEGCM at 450 km at 19.6 UT, 21.2 UT, and 22.7 UT, respectively. In Figure 6c, the storm-time Ne also shows a two-peak structure, but at a higher latitude than that in Figure 6a,b. In Figure 6c, the center of the two peaks is located at ±20° MLat. They have a maximum density of 3.0 × 10¹² and 2.8 × 10¹² in the northern and southern hemispheres, respectively. At the Swarm orbit, the density of observed Ne is significantly higher in the southern EIA crest (2.5 × 10¹² m⁻³) than in the northern EIA crest (1.0 × 10¹² m³). This hemispheric asymmetry is also found in Figure 6f. The density of southern and northern EIA crest at 21.2 UT is about 2.6 × 10¹² and 1.7 × 10¹² m⁻³, respectively. In Figure 6f,i, both the northern and southern EIA crest moves poleward significantly. The center of northern and southern EIA is close to ±30° MLat, which has also been pictured in the Swarm A observations. In summary, the reliability and reasonability of TIEGCM in capturing the poleward extension of the EIA have been achieved here based on Figure 2, Figure 3, Figure 4, Figure 5 and Figure 6.

To show the dynamic evolution of EIA during the main phase, the temporal variations in the Ne disturbances (ΔNe) at low latitudes at fixed LT of 19.2 h on 10 May 2024 are shown in Figure 7. Note here that ΔNe is obtained by the differences between Ne on 10 May and 09 May. In Figure 7a, as indicated by the upper black arrow, after the onset of the superstorm, the Swarm A-observed ΔNe is enhanced and shifts poleward at low latitudes. For instance, two areas of the enhanced ΔNe are found at 20 UT and 20° MLat, and 22 UT and 31° MLat. In the southern hemisphere, the enhanced ΔNe shows a band structure that moves from −15° MLat at 17 UT to 32° MLat at 22 UT. This is a poleward shift of EIA crests in ΔNe. The strong depletion of ΔNe is found at the equatorial and low latitudes. This is reasonable because of the super fountain effects during the main phase [27]. The super fountain effects could move more plasma into higher altitudes and form the crests at higher latitudes. In Figure 7b, the orbit-by-orbit simulations are binned. The modeled ΔNe shows a large degree of similarity with that in Figure 7a. For instance, as indicated by the two black arrows, the poleward shift of EIA crests in ΔNe is captured by TIEGCM. However, the magnitude of the modeled ΔNe is larger than the observed. The former is larger than 1.5 × 10¹² m⁻³, and the latter is larger than 1.2 × 10¹² m⁻³. Here, the day-to-day variability in TIEGCM should be discussed. In the literature, it has been reported that the ionospheric plasma, and E × B drifts, atmospheric tides, etc., have significant day-to-day variations, e.g., [17]. However, this has been removed because the data during storm time were subtracted from that on 9 May.

The horizontal and vertical distributions of equatorial ionospheric plasma could be found in the responses of hmF₂ and NmF₂ to the superstorm. Figure 8 shows the displacement of hmF₂ and the deviation of NmF₂ with respect to their quiet-time values. In Figure 8a, at 19.6 UT, it can be found that the hmF₂ in the equatorial ionosphere at −180°~−60° GLon decreases by 80 km on average. However, at low latitudes and −180°~−60° GLon, the F₂ layer is pushed upward by more than 100 km. The same goes for that at 21.2 UT (Figure 8b), except for that at the equatorial latitudes and −180°~−60° GLon, where the vertical displacement is reduced to almost zero. In Figure 8c,d, the poleward extension of EIA crests is also found. In Figure 8c or Figure 8d, the equatorward edge of EIA crests locates at ±15°/±18° (±15°/±18°) MLat at −180°~−60° GLon. The poleward edge of EIA crests at −180°~−60° GLon expands from 45°/−30° MLat at 19.6 UT to 60°/−60° MLat at 21.2 UT in the northern/southern hemisphere.

4. Discussion

As reported in the literature, the dynamic evolution of ionospheric plasma is controlled by three drivers: chemical production, chemical recombination, and plasma transport, e.g., [9,48]. This can be expressed by the following equation: $O^{+}=Q-L+\nabla \left(nV\right)$, where O⁺, Q, L, and $\nabla \left(nV\right)$ are the ionospheric ions, and the contributions from chemical production, loss, and plasma transport, respectively, and where n and V are the density and velocity, respectively. All the data of the O⁺ due to plasma transport (including E × B drifts, neutral winds, and ambipolar diffusion), chemical production, and loss are extracted from TIEGCM.

To identify the effects of different forces, δ∆O⁺ in association with the total forcings should be determined, as shown in Figure 9. Note here that in TIEGCM, the initial O⁺ due to total forcings is the change rate in electron density per unit time. At this place, ∆O⁺ from the onset of the superstorm to the current UT is integrated to make the conclusion more reasonable, which is termed δ∆O⁺. The same integration applies for that due to ambipolar diffusion, E × B drifts, and neutral winds. Since in the ionosphere, the major ion is the O⁺ [9], thus, only O⁺ is considered here. In Figure 9a, the total δ∆O⁺ at 19.6 UT is reduced at the dip equator but enhanced at northern and southern low- and mid-latitudes, with an average density variation of −5.22 × 10¹¹, 2.51 × 10¹¹, and 5.42 × 10¹¹ m⁻³. The peak δ∆O⁺ at dayside EIA latitudes is 6.79 × 10¹¹ and 14.33 × 10¹¹ m⁻³ at 24° and −25° MLat, respectively. δ∆O⁺ forms an interesting pattern with two bands of enhanced δ∆O⁺ and two bands of δ∆O⁺ reduction. The enhanced δ∆O⁺ at ±18°~±50° MLat indicates the advance of EIA crests toward higher latitudes at 19.6 UT. The MLat gap between the equatorward edge of EIA in Figure 6c and Figure 9a is about 3° MLat. The movement of the EIA crests indicates the reduction (enhancement) in Ne at current EIA latitudes (higher latitudes than the current EIA). The same conclusion can be achieved in Figure 9b.

δ∆O⁺ due to different factors, including the chemical processes (production and loss) and plasma transport, can be output in the model. This method has been widely accepted in the study of ionospheric plasma through physical models, e.g., [49]. However, in this work, the roles of both chemical production and recombination are not outstanding in the EIA crests during superstorm (figures not shown). However, it should be noted that as reported in the literature, the chemical recombination has some role in the EIA formation due to its altitudinal dependence [9]. In this work, only the contributions from the plasma transport should be considered. The plasma transport can be separated into three components: neutral winds, E × B drifts, and ambipolar diffusion.

4.1. δ∆O⁺ Due to E × B Drifts

E × B drift-driven transport includes all electric field effects (PPEF + DDEF). Though DDEF is triggered by neutral winds, it manifests as an electric field in the ionosphere. Thus, its transport effect is captured in the E × B term, not in the neutral wind transport term. Figure 10a,b illustrate the temporal and spatial evolutions of the integrated δ∆O⁺ due to the E × B drifts at 19.6 and 21.2 UT, respectively. In Figure 10a, it can be found that at the dayside, the latitudinal profile of ionospheric δ∆O⁺ shows an interesting structure, that is, enhancement–reduction–weak enhancement–reduction–enhancement. At −150°~−60° GLon, the enhanced δ∆O⁺ can be found at middle latitudes that are higher than 30° MLat and less than −20° MLat. The average δ∆O⁺ is 3.03 × 10¹¹, and 10.67 × 10¹¹ m⁻³ at northern and southern middle latitudes, respectively. Between the bands of the enhanced δ∆O⁺, the significantly reduced δ∆O⁺ can be found at low latitudes. The average reduction in δ∆O⁺ at northern and southern EIA latitudes is approximately −11.54 × 10¹¹ and −14.93 × 10¹¹ m⁻³, respectively. A comparison between Figure 9a and Figure 10a shows that the poleward extension of EIA is deeply influenced by the ionospheric E × B drifts. At 15°~30° and −15°~−18°MLat (equatorward edge of EIA), δ∆O⁺ due to total E × B in Figure 10a is opposite to that due to total forcings. The same conclusion is arrived at by a comparison between Figure 10b and Figure 9b. Thus, the poleward extension of EIA is deeply affected by E × B drifts during the main phase. At the equatorward boundary, the effects of E × B drifts are offset by unknown forcings, which will be discussed later.

At 19.6 UT, PPEF through magnetosphere–ionosphere coupling achieves instantaneous global penetration, not limited to the dayside. Theory and observations confirm PPEF can generate strong polarization electric fields in dawn/dusk sectors at the equatorial latitudes, e.g., [28]. At the main phase of the superstorm, IMF Bz is persistently southward (Figure 1c), providing a continuous driving force for eastward PPEF. To analyze the δ∆O⁺ due to E × B drifts at 19.6 UT (Figure 10a), Figure 10c depicts the vertical and horizontal E × B drift disturbances at middle and low latitudes. It can be found that during the superstorm on 10 May 2024, the vertical E × B drifts were significantly disturbed, with an average speed of 30 and 60 m/s at the equatorial and middle latitudes, respectively. A similar result is found at 21.2 UT in Figure 10d. This might be related to the eastward PPEF due to the imbalance between R1 and R2 FACs, e.g., [9,28,29]. During the main phase, PPEF is eastward at the dayside, and promptly established in the middle and equatorial ionosphere. This eastward PPEF can generate the upward E × B drifts. The upward E × B drifts can move the plasma upward, causing the enhanced electron density at ~450 km. During this superstorm, the southward IMF Bz has a peak magnitude larger than 30 nT (Figure 1a). This indicates that the generated PPEF and hence the upward E × B drifts are quite strong, as shown in Figure 10c. This is significantly stronger than that during the other storms, e.g., [50]. Elvira et al. [50] reported that the extreme upward lift was generated during the superstorm on 10 May 2024. A significant super equatorial fountain effect was induced. Then, the two crests of plasma can be found not only at the Swarm altitudes of 450 km but also at the DMSP orbits at ~800 km [50]. The plasma above the dip equator can be moved to a higher altitude, causing the enhanced δ∆O⁺ in Figure 9a.

4.2. δ∆O⁺ Due to Neutral Wind-Generated Vertical Plasma Velocity

Previous studies have demonstrated that the plasma can be moved along the geomagnetic field lines to a higher or lower altitude, through the ion-neutral collision, e.g., [9,31]. Considering the inclination and declination of the geomagnetic field, a vertical plasma drift could be generated by the horizontal winds. At different altitudes, the chemical recombination varies, causing the Ne disturbances, because the neutral density and the effects of vertical E × B drift greatly depend on the altitudes. Figure 11a gives the GLon and GLat distributions of the neutral wind-linked integrated δ∆O⁺ at 19.6 UT. It can be found that two bands of enhanced δ∆O⁺ and two bands of reduced δ∆O⁺ are found at low and equatorial latitudes in the dayside ionosphere. The average δ∆O⁺ is −4.05 × 10¹¹, 4.54 × 10¹¹, 5.92 × 10¹¹, and −4.34 × 10¹¹ m⁻³ at 25°~40°, 10°~25°, −10°~−20°, and −20°~−35° MLat at −150°~−30° GLon, respectively. A comparison between Figure 9a and Figure 11a shows that the pattern of δ∆O⁺ associated with the neutral winds shares a large degree of similarity with that due to the total forcings at the equatorward edge of EIA crests, but opposite at latitudes higher than ±25° MLat. Therefore, it can be concluded that the neutral winds play a critical role in the formation of the poleward shift of EIA crests. It promotes/prevents the formation of the density crest at lower/higher EIA latitudes. A similar conclusion can be obtained at 21.2 UT (Figure 11b).

The neutral wind-driven transport is only the generated field-aligned vertical plasma velocity. While DDEF is caused by the neutral winds, its electrodynamic consequences (the induced electric field) are solved separately in the model’s current equation and fed into the E × B term. The horizontal winds can be separated into two components, that is, zonal and meridional winds. The plasma could be pushed upward or downward by the horizontal winds along the geomagnetic field lines, e.g., [31,32]. The vertical plasma velocity (VV) driven by the neutral winds can be expressed by the following equation: $VV=VN·cosD·\mathrm{s}\mathrm{i}\mathrm{n}\left|I\right|\mathrm{c}\mathrm{o}\mathrm{s}\left|I\right|\mp UN·sinD·\mathrm{s}\mathrm{i}\mathrm{n}\left|I\right|\mathrm{c}\mathrm{o}\mathrm{s}\left|I\right|$, where UN, VN, D, and I are the zonal winds, meridional winds, the declination, and inclination of the geomagnetic field, respectively. The negative and positive signs correspond to the northern and southern hemispheres, respectively.

A comparison between Figure 11c,d and Figure 11a,b can identify the roles of zonal winds. At dayside (−120° GLon) EIA latitude of ±15°~±30° MLat, ΔVV is close to zero at 19.6 UT (Figure 11c). This is related to the weak westward winds as indicated by the arrows. In Figure 11d, at 20°~35° and −20°~−35°MLat, ΔVV is generally upward and downward at −120° GLon in the northern and southern hemispheres, with an average speed of 10 and −30 m/s, respectively. This indicates that the plasma is moved upward/downward by the westward winds in the northern/southern hemisphere, as indicated by the arrows. It is related to the declination of the geomagnetic field, e.g., [31]. At low-latitudes, the downward/upward plasma velocity causes the reduction/enhancement of the ions because of the high/low recombination at lower/higher altitudes. However, δΔO⁺ due to winds at dayside EIA latitudes are strongly negative and positive at 19.6 and 21.2 UT, respectively. This indicates that the roles of zonal winds are ignorable at 19.6 UT, but negative/positive in the southern/northern hemisphere at 21.2 UT.

The equatorward/poleward winds could push the ionosphere upward/downward because of the generated upward/downward plasma velocity [9]. As indicated by the arrows in Figure 11e, the disturbed meridional winds at dayside low latitudes are generally equatorward. The generated upward plasma velocity at low latitude and dayside is upward at 19.6 UT with an average speed of 30 m/s. Thus, the meridional winds in equatorward direction contribute to the accumulation of plasma at the EIA crest in both hemispheres. In Figure 11f, as indicated by the arrows, the meridional winds at −150°~−60° GLon are poleward at 30°~60° MLat and equatorward at −60°~−30° MLat. The generated plasma velocity is generally downward at an average speed of −20 m/s at daytime low latitudes. This could lead to the reduction in plasma at the EIA crest, and the development of asymmetric crests. Moreover, the poleward surge could enhance the ambipolar diffusion and produce the poleward extension of EIA [26]. This will be discussed later. At −60°~−30° MLat and daytime, the equatorward winds lead to the upward plasma velocity. This can also contribute to the high density at the southern EIA crest.

Based on Section 4.1 and Section 4.2, Figure 11a,b indeed share a lot of features with Figure 10a,b but in an opposite sense. This anticorrelation demonstrates that neutral wind-driven plasma transport systematically counteracts E × B drift-driven transport across most low- and mid-latitude regions. Where upward E × B drift generates plasma depletion at the EIA equatorward edge (e.g., ±15°~±25° MLat), the neutral winds, especially the meridional winds, produce plasma enhancement in the same regions. Conversely, the plasma enhancement by E × B drift at higher EIA latitudes (>±25° MLat) is partially offset by neutral wind-driven reduction.

4.3. δ∆O⁺ Due to Ambipolar Diffusion

Ambipolar diffusion depends on gravity and the pressure gradient force [49]. The ambipolar diffusion is interdependent with E × B and neutral wind transport. Diffusion (along the geomagnetic field lines) responds to plasma gradients created by the other processes. It is not an isolated driver but a redistributive mechanism. Figure 12a,b show the GLon and GLat variations in the accumulated δΔO⁺ due to ambipolar diffusion at 19.6 and 21.2 UT, respectively. At 19.6 UT, the ambipolar diffusion-induced δΔO⁺ is strongly positive at both northern and southern EIA latitudes, with an average density of 11.73 × 10¹¹, and 8.46 × 10¹¹ m⁻³, respectively. However, the equatorward edge of enhanced δΔO⁺ due to ambipolar diffusion is lower than that due to total forcings. This indicates that the ambipolar diffusion prevents/promotes the poleward shift of EIA crests at lower latitudes. As shown in Figure 11f, the strong poleward wind surge at the equatorial latitudes would greatly strengthen the ambipolar diffusion via neutral drag, consistent with Aa et al. [26]. This thereby contributes to some extent to the poleward expansion of EIA crests at 21.2 UT.

5. Conclusions

Using the observations from Swarm, GNSS receivers, and corresponding simulations from TIEGCM, the poleward shift of EIA crests during the main phase of a superstorm on 10 May 2024 is investigated. The poleward shift of EIA crests is observed in GNSS-observed TEC and Swarm A-observed in situ electron density. EIA crests move poleward from about ±19° MLat at 19.6 UT to 30.3°/−25.0° MLat by the end of 10 May.

In this work, the contributions from the plasma transport due to neutral winds, E × B drifts, and ambipolar diffusion are investigated. The simulations and observations show that the effects of E × B drift, ambipolar diffusion, and neutral winds vary with latitude. The plasma transport due to E × B drifts is primarily controlled by the eastward PPEF. It contributes to the formation of crests at higher EIA latitudes. However, at the equatorward edge, the density increases in the new crest position, it decreases where it was previously located. The roles of neutral winds are mainly from the meridional winds, with minor contributions from zonal winds. The plasma in the equatorward/poleward EIA region is enhanced/reduced by neutral winds. However, the effects of ambipolar diffusion are complicated. At 19.6 UT and 21.2 UT, it is strongly positive at latitudes larger than ±14° MLat. Thus, it contributes/prevents the formation of EIA crests at most EIA latitudes/equatorward edge.