The Response of Auroral-Oval Waves to CIR-Driven Recurrent Storms: FY-3E/ACMag Observations

Abstract

Alfven-branch waves provide an efficient means for transporting energy into the auroral oval. Here, we report observations of these waves obtained by the Fengyun-3E (FY-3E)/ACMag instruments, which are designed to detect three-dimensional AC magnetic fields in the 0.05–25 Hz band. The observations suggest that broadband waves are a permanent feature of the auroral oval, although their amplitude and locations vary with the global state of the magnetosphere. We primarily focus on the data obtained from 10 July 2021 to 26 August 2021, during which a series of recurrent storms driven by solar wind corotating interaction regions (CIRs) occurred. Analysis of the observations shows that the auroral-oval waves grow in amplitude (1–3 orders of magnitude) and shift to lower latitude (∼10°) immediately following the decrease in the SYM-H index in each storm. Further investigation reveals the response of the auroral-oval waves has a time scale equal to or less than FY-3E’s effective revisiting time, which is about 45 min. The observations presented in this paper confirm that the FY-3E/ACMag instruments provide a high-resolution monitor of the auroral-oval waves and could further our understanding of auroral physics.

1. Introduction

How geomagnetic storms affect the Earth’s magnetosphere has served as a fundamental question since the coming of the space age. Until now, it has been demonstrated that storms are capable of disturbing the whole magnetosphere [1]. Among various manifestations of the storm effects, the enhanced aurora [2] serves as the most obvious and spectacular one.

A number of models have been proposed to explain the development of the aurora during storms and other geomagnetic activities (see [3,4,5] and references therein), with one class of models relying on the role of Alfven-branch waves, e.g., [3,5,6]. The significance of these waves has been well-documented by several spacecraft missions, such as Cluster [7], FAST [8], Freja [9], Polar [10,11], Swarm [12], THEMIS [13], and Van Allen Probes [14,15]. These waves could accelerate charged particles [14,16,17], precipitate magnetospheric particles into the ionosphere and atmosphere [18], and cause ionospheric ion outflows [19,20]. Regarding the scope of our study, we specifically focus on the role played by Alfven-branch waves in the dynamics of the aurora. Ground-based and spacecraft observations on field lines threading the auroral oval have consistently shown the presence of these waves with frequencies ranging from the Pc 5 band (∼mHz) to >10 Hz. These waves, which are usually generated in the equatorial magnetospheric regions (including both the inner magnetosphere and distant magnetotail) [21,22] and then propagate to the ionosphere with a phase speed close to the Alfven velocity (ranging from ∼${10}^3$ km/s at the equator to ∼${10}^5$ km/s at the topside ionosphere), are considered to be a crucial factor in the efficient transfer of energy from the magnetosphere to the auroral oval [8,10,11]. Interestingly, many similarities have been found between Alfven-branch waves and the aurora [10,11], e.g., they both can be observed all the time but are stronger during geomagnetic activities, and are most likely to be observed inside an oval extending from 65° to 75° in the nightside and from 70° to 80° in the dayside (namely, the so-called auroral oval) [23]. Because of these similarities, Alfven-branch waves in the auroral oval (hereinafter called auroral-oval waves for brevity) are believed to be one of the causes of the aurora.

The main focus of this paper is the response of auroral-oval waves to storms driven by corotating interaction regions (CIR). Although CIR-driven storms are weaker than those driven by coronal mass ejection (CME) in general, the former are usually of longer duration and are more harmful to spacecraft in terms of, for example, spacecraft charging [24]. Thus, it is also very critical to elucidate the effects of CIR-driven storms on the magnetosphere and space weather. Here, with the first two months of data obtained by the AC Vector Magnetometer (ACMag) instruments [25] onboard the Fengyun-3E (FY-3E) satellite, we show that the auroral-oval waves would grow in magnitude and shift to lower latitude during CIR-driven storms, which is very similar to the case of CME-driven storms [11]. Moreover, the observations indicate that the response is very quick, with a time scale equal to or less than the effective revisiting period of the FY-3E satellite, ∼45 min. Based on these observations, we suggest that the FY-3E/ACMag observations of the auroral-oval waves provide a sensitive monitor of the global state of the magnetosphere and space weather.

The rest of this paper is organized as follows. We first briefly describe the instruments and dataset in Section 2. Then, we show in Section 3 the auroral-oval waves detected by FY-3E/ACMag. In Section 4, we investigate the response of the auroral-oval waves to CIR-driven recurrent storms that occurred between 10 July and 26 August 2021. Finally, we summarize our findings and give a brief discussion in Section 5.

2. Materials and Methods

The primary dataset used in this study was obtained by the ACMag instruments [25] onboard the FY-3E satellite, which was inserted into a dawn–dusk Sun-synchronous orbit in July of 2021 with the following characteristics: altitude ∼820 km, inclination ∼90°, orbit speed ∼7 km/s, and period ∼90 min (thus, the effective revisiting period of the satellite, which can be taken as the half of the orbit period, is about 45 min). The ACMag instruments, developed by the Peking University instrument development team, are designed to detect three-dimensional AC magnetic field oscillations within 0.05–25 Hz (termed as AC magnetic fields hereinafter). The original data are in waveform and thus provide both the amplitude and phase of the AC magnetic fields. Here, we mainly focus on the integrated total power (i.e., $B_w^2=P_x+P_y+P_z$, where $P_i$ with $i=x,y,z$ represents the integrated power in the corresponding direction) within the frequency range from 0.1 Hz to 10 Hz (in the rest frame of the satellite). The lowest frequency limit is set so to avoid the influence of satellite motion, whereas the highest frequency limit is approximately the Nyquist frequency of the instruments. In this paper, the power is derived from a windowed fast Fourier transformation (wFFT) with a window length of 20 s.

In addition to the AC magnetic fields, we also used the DC magnetic field data taken by the FY-3E/DCMag instruments [26] and the energetic electron data obtained by the FY-3E/IES instruments [27]. The two datasets provide information on the location of the satellite in the magnetosphere. We also used the observations obtained by the WIND satellite to determine the solar wind conditions and used the SYM-H index to represent the level of the geomagnetic activities.

Before delving into the specifics, we would like to issue a caveat about a potential ambiguity regarding the nature of the fluctuations examined in this work. Throughout this paper, we will refer to them as “waves”, given their spatial scale of approximately 0.7–70 km as derived from the satellite’s velocity and the frequency range recorded. Previous research has suggested that such small-scale (<150 km) fluctuations observed in and around the auroral oval can be classified as Alfven-branch waves [28,29,30]. However, we note that there is an alternative interpretation of these fluctuations as quasi-static fractal current filaments [31,32]. We do not discount this possibility, and without additional measurements such as electric fields and plasma, we cannot determine the true nature of the fluctuations with certainty. Nonetheless, our analysis does not rely on any theoretical framework, and we merely present what was observed by FY-3E/ACMag. Therefore, our primary findings regarding the observational features of these fluctuations within the auroral oval and their response to storms should be trustworthy.

3. The Auroral-Oval Waves Observed by FY-3E/ACMag

Figure 1 shows the auroral-oval waves observed by FY-3E/ACMag in the northern hemisphere during a storm on 14 July 2021, with the left and right panels presenting observations obtained before and during the main phase, respectively. From top to bottom, Figure 1 shows the SYM-H index, the DC magnetic fields in the Geographic (GEO) coordinates, the deviations of the azimuthal component (Ba) of the DC magnetic fields from the International Geomagnetic Reference Field (IGRF) model, the x, y, and z components of the AC magnetic fields in the GEO coordinates, and the fluxes of energetic electrons.

As shown in Figure 1d,e, both before and in the main phase, a Ba peak and valley are observed around the poleward boundary of the duskside and dawnside outer radiation belt (Figure 1i), respectively. These signatures represent the Birkeland field-aligned current (FAC) system [28,33]. Just within the FACs, waves are observed in the AC magnetic fields (Figure 1f–k). The waves are dominated by their transverse components, as the oscillations in the z component are much weaker than those in the x and y components. Furthermore, the waves are broadband. Figure 2 gives the dynamic spectra derived from a wFFT, with Figure 2a,b showing Ba again for reference. It is clear that all the spectra extend continuously from 0.1 Hz to 10 Hz, with the power decreasing as frequency increases. No local peaks can be identified clearly. Figure 2 also suggests a latitude-dependence of the waves. As they approach the center of the FACs, the power spectra increase in amplitude and become harder (i.e., extending to higher frequency), leading to a bell-like latitude profile. This latitude profile suggests that the waves contain more energy in the FAC center regions and thus could lead to stronger charged particle energization there. We note that the latitude extent in which the waves are significant is overlapped with the auroral oval, indicating these waves can be classified as the aforementioned auroral-oval waves.

To further illustrate the morphology of the waves observed by FY-3E/ACMag, Figure 3 shows the spatial distributions of the average wave power obtained during the first two months of the operation of FY-3E/ACMag. Figure 3a,b give geographic maps of the northern and southern hemispheres, respectively, with the solid curves representing the 70° latitude in the Altitude-Adjusted Corrected Geomagnetic (AACGM) coordinates [34]. Ovals of intense waves can be observed in both hemispheres. They correspond to the wave power enhancements shown in Figure 1 and Figure 2. Strong hemispheric asymmetries can be noted if one compares the two figures, with the waves in the southern hemisphere extending to lower geographic latitudes. Figure 3c,d shows the wave power as a function of the magnetic local time and latitude, together with the statistical location of the auroral oval given by [23]. Obviously, the most intense waves are observed in the auroral oval, confirming our suggestion that the intense waves observed by FY-3E/ACMag are associated with the aurora. Figure 3c,d also shows that FY-3E covers different local times in different hemispheres: dawnside, noonside, and duskside in the northern hemisphere, and dawnside, nightside, and duskside in the southern hemisphere. One should keep this in mind when considering the north–south asymmetries shown below.

Although the auroral-oval waves can be observed almost all the time, Figure 1 and Figure 2 indicate that the storm-time waves differ from the non-storm-time ones in many ways. First, the storm-time waves are much stronger. As shown in Figure 1f–i, the maximum amplitude observed during the storm is about ∼300 nT, whereas the amplitude observed before the storm is generally less than 200 nT. Second, the storm-time waves appear at a lower magnetic latitude, compared with the pre-storm waves (65° vs. 80°). Finally, the spectra of the storm-time waves are harder than those of the non-storm-time waves. As shown by what follows, the features obtained here can be generalized to other storms.

4. Response of the Auroral-Oval Waves to CIR-Driven Storms

4.1. Overview

From 10 July 2021 to 26 August 2021, a series of moderate storms were recorded by the SYM-H index, as shown in Figure 4d. It is worth noting that these storms have a quasi-periodic nature with a period of approximately 5–7 days, roughly a quarter of the solar spin period. This periodicity can also be observed in the solar wind data collected by the WIND spacecraft. Figure 4a–c shows the solar wind speed (V${}_{sw}$), the solar wind proton number density (N${}_{sw}$), and the z-component of the interplanetary magnetic field in the GSM coordinates (IMF B${}_z$). One can see that the solar wind speed increases and decreases quasi-periodically. At the leading edge of each V${}_{sw}$ enhancement, there is a corresponding N${}_{sw}$ peak. These characteristics, particularly the quasi-periodicity, indicate that the observed V${}_{sw}$ enhancements represent CIRs [36]. To confirm that these enhancements are not caused by interplanetary CMEs (ICME), we have checked an ICME list established using the methods described in [37] (https://izw1.caltech.edu/ACE/ASC/DATA/level3/icmetable2.htm accessed on 1 January 2023), and found that there were no ICMEs during the studied period. Interestingly, there is a rough one-to-one correspondence between the CIRs observed by WIND and the storms recorded in the SYM-H index, suggesting that the former triggers the latter.

The last two panels of Figure 4 show the power of the auroral-oval waves, with the upper and lower panel corresponding to the northern and southern hemispheres, respectively. The vertical axes represent magnetic latitude, with the upper and lower half corresponding to the dawnside and duskside, respectively. The color code represents the integrated power (hereinafter directly referred to as “power”) of the auroral-oval waves. The white background seen before 10 August is caused by noises and does not affect our analysis much. One can see that the auroral-oval waves are observed whenever the satellite with the ACMag instruments switching on flies by the auroral oval (i.e., ∼70° magnetic latitude) at an altitude of about 820 km. Moreover, the power is larger for the northern hemisphere than for the southern hemisphere, agreeing with previous studies, e.g., [12,38], that found that the waves in the northern hemisphere are always stronger than those in the southern hemisphere because of the off-set between the geographic and magnetic poles.

The auroral-oval waves presented in Figure 4 are continuous but not steady. They respond to the CIR-driven storms in a way similar to the cases shown in Section 3. Namely, when the SYM-H index decreases, the wave power increases, and the region of large power shifts to lower latitudes. Furthermore, Figure 4 suggests that the response is very quick. First, the auroral-oval waves start to grow immediately as the SYM-H index decreases. Second and more surprisingly, the latitude of the auroral-oval waves also follows the variation of the SYM-H index well. This is most clearly illustrated in the 14 July, and 7, 25, and 28 August storms. Both in the northern and southern hemispheres, and at the dawnside and duskside, the auroral-oval waves shift equatorward as the SYM-H index decreases. All of the observations indicate that the auroral-oval waves are able to respond to the SYM-H index (in other words, CIR-driven storms) with a time scale less than or at least equal to the effective revisiting period of the FY-3E satellite, which is about 45 min. In what follows, we analyze the auroral-oval waves and their response to CIR-driven storms more quantitatively.

4.2. The Power of the Auroral-Oval Waves

We first investigate the power of the auroral-oval waves. Figure 5a gives the SYM-H index for reference. Figure 5b shows the Akasofu $ϵ$ parameter [39] representing the amount of energy transported from the solar wind into the magnetosphere (based on WIND observations). From top to bottom, the other four panels of Figure 5 show the maximum power in the northern dawnside hemisphere, northern duskside hemisphere, southern dawnside hemisphere, and southern duskside hemisphere. Again, Figure 5 indicates that the auroral-oval waves in the northern hemisphere are stronger than those in the southern hemisphere. For example, in the 14 July storm, the maximum power reached in the northern dawnside hemisphere is ∼3500 nT${}^2$, about 3 times larger than that in the southern dawnside hemisphere (∼1200 nT${}^2$). Figure 5 also shows that the power at the dawnside is larger than that at the duskside. As an example, we note that the maximum power reached in the northern dawnside hemisphere is about 2 times larger than that reached in the northern duskside hemisphere (∼1600 nT${}^2$) in the 14 July storm. Thus, combining the north–south and dawn–dusk asymmetry together, we have the following order: northern dawnside hemisphere > northern duskside hemisphere > souther dawnside hemisphere > souther duskside hemisphere.

Furthermore, as shown in Figure 5, the maximum power grows significantly during the storms both in the northern and southern hemispheres, and at the dawnside and duskside. The maximum power obtained when the SYM-H index is less than −25 nT (storm time) is generally about 1–3 orders of magnitude larger than that obtained when the SYM-H index is larger than −25 nT (quiet time). Besides the dependence on the SYM-H index, we show in Figure 6 how the maximum power varies with the $ϵ$ parameter. As indicated by the red curves representing the mean values of the power in each $ϵ$ bin, there is generally a positive relationship between the two variables. This relationship is most clearly illustrated in Figure 6a, where the maximum power increases by about three orders of magnitude as the $ϵ$ parameter increases by the same order of magnitude. This positive relationship indicates that the power of the auroral-oval waves can be considered a proxy for the level of energy entering the magnetosphere from the solar wind.

However, the relationship identified above is not of the one-to-one type. First, the larger $ϵ$ parameter does not necessarily indicate larger power. For example, as shown in Figure 4b,c, the $ϵ$ parameter of the 22 July storm (∼900 GW) is smaller than that of the August 3 storm (∼1300 GW). In contrast, the maximum power obtained during the former storm (∼5000 nT${}^2$ in the northern dawnside hemisphere) is larger than that during the latter storm (∼2500 nT${}^2$ in the northern dawnside hemisphere). Second, there seems to be a threshold effect. As shown in Figure 6, the maximum power generally does not change much with the $ϵ$ parameter until the latter becomes greater than 100 GW. These results suggest that the auroral-oval waves might be related to the $ϵ$ parameter (or in other words, the energy input into the magnetosphere) in a nonlinear way.

4.3. The Latitude Profiles of the Auroral-Oval Waves

As discussed above, the auroral-oval waves would shift equatorward during storms. Figure 7 shows this feature quantitatively. To obtain this figure, we first divided the SYM-H index into 3 intervals: SYM-H > 0 nT, −25 nT < SYM-H < 0 nT, and SYM-H < −25 nT (see Figure 7a). Then, we classified the auroral-oval waves according to the corresponding SYM-H index (the thin curves in Figure 7b–g). Finally, we calculated the median values in each SYM-H interval, and show the resulting curves as thick black curves in Figure 7b–g. As one can see, when the SYM-H index is greater than −25 nT, the peak of the power of the auroral-oval waves is located at about 75°–80° magnetic latitude. In contrast, when SYM-H < −25 nT, the peak is located at about 65°–70° magnetic latitude in general. It is noted that this variation can be seen both in the northern and southern hemispheres, and at the dawnside and duskside, although it is most clear in the northern dawnside hemisphere.

5. Discussion and Summary

In this paper, we investigate the first two months of FY-3E/ACMag observations of the waves detected within the auroral oval. We find the auroral-oval waves have the following characteristics both in quiet time and geomagnetically active time.

• The auroral-oval waves are a permanent feature of the auroral oval, although they vary significantly with time.
• The auroral-oval waves are broadband in general, with higher power and harder spectra seen at the center of the FACs.
• The auroral-oval waves are stronger in the northern hemisphere than in the southern hemisphere (∼3 times). Moreover, the auroral-oval waves are found to be stronger at the dawnside than at the duskside (∼2 times).
• The auroral-oval waves appear at a higher magnetic latitude in the northern hemisphere than in the southern hemisphere (∼3°), and appear at higher magnetic latitude at the dawnside than at the duskside (∼5°).

The auroral-oval waves are continuous but not steady. They vary with the global state of the magnetosphere. Here, we primarily study how they respond to CIR-driven recurrent storms. The main findings are as follows:

• The power of the auroral-oval waves increases significantly during the CIR-driven storms. In general, the power integrated over 0.1–10 Hz is ∼10${}^3$ nT${}^3$ during the storms, whereas the quiet-time value is ∼10${}^2$ nT${}^2$.
• The auroral-oval waves shift equatorward during CIR-driven storms. The magnetic latitude of the peak of the auroral-oval waves is 65°–70° during storms, and 75°–80° in quiet time.
• The response of the auroral-oval waves to CIR-driven storms is very quick. The corresponding time scale is equal to or even less than the effective revisiting period of the FY-3E satellite, ∼45 min. As a result, as observed by FY-3E/ACMag, the auroral-oval waves grow immediately as the SYM-H index decreases. Furthermore, as the SYM-H index gradually decreases, the auroral-oval waves shift to lower latitudes in a gradual way.

Interestingly, the auroral-oval waves respond to storms in a way very similar to the aurora. The latter, as has been broadly established in the literature, e.g., [2], also becomes brighter and shifts equatorward during geomagnetic activities such as storms. The similarities, together with their spatial overlap, indicate that the auroral-oval waves should be partially responsible for the brightening of the aurora. Indeed, many authors have suggested that the auroral-oval waves can efficiently accelerate electrons (e.g., via Landau resonance) and consequently lead to the aurora, e.g., [8,40,41]. Fortunately, besides the ACMag instruments, the FY-3E satellite is also equipped with instruments (FY3E/IES) capable of detecting ∼10–1000 keV electrons with an angle resolution of ∼10°. Future studies combining field and electron observations together can provide us with further information on the role of the auroral-oval waves in aurora formation.

In addition, we note that the auroral-oval waves observed by FY-3E/ACMag can provide us with a monitor of the magnetosphere in state and space weather. As we have shown, the auroral-oval waves would increase in amplitude dramatically and shift to lower latitudes during storms. The responses are very sensitive. In our study, most of the storms are weak (the minimum SYM-H index is generally greater than −50 nT). However, the response of the auroral-oval waves is rather significant. The wave power can even increase by about 1–3 orders of magnitude. Furthermore, the response of the auroral-oval waves is very quick. The power of the auroral-oval waves increases almost immediately as the SYM-H index decreases. Besides the sensitivity and quick response, we also note that the time resolution of this “monitor” is high. The effective revisiting period of the FY-3E satellite is ∼45 min, meaning that indices established based on its observations can have a time resolution as small as ∼45 min. Therefore, the FY-3E/ACMag observations of the auroral-oval waves can offer us a good monitor of the magnetosphere state and space weather.