https://doi.org/10.

5194/angeo-2022-18
Preprint. Discussion started: 5 September 2022
c Author(s) 2022. CC BY 4.0 License.

1 Latitudinal variations of ionospheric-thermospheric

2 responses to Geomagnetic Storms from Multi-
3 Instruments
4 Rasim Shahzada, Munawar Shaha*, Ayesha Abbasb, Amna Hafeeza, Andres Calabiac, Angela
5 Melgarejo-Moralesd, Najam Abbas Naqvib

a
7 Department of Space Sciences, Space Education and GNSS Lab, National Center of GIS and
8 Space Application, Institute of Space Technology, IST, Islamabad, 44000, Pakistan
b
9 Department of Petroleum Engineering, NED University of Engineering and Technology,
10 Karachi 75270, Pakistan
c
11 School of Remote Sensing and Geomatics Engineering, Nanjing University Information Science
12 Technology, China.
d
13 Department of Earth and Space Science, Autonomous University of Sinaloa, Culiacan, Sinaloa,
14 80040, Mexico, Email:
15
16
17
*
18 Corresponding author: [email protected]
19

20 Abstract

21 Scintillations of transionospheric satellite signals during geomagnetic storms can severely threaten
22 navigation accuracy and the integrity of space assets. We analyze vertical Total Electron Content
23 (vTEC) variations from the Global Navigation Satellite System (GNSS) at different latitudes
24 around the world during the geomagnetic storms of June 2015 and August 2018. The resulting
25 ionospheric perturbations at the low-and mid-latitudes are investigated in terms of the prompt
26 penetration electric field (PPEF), the equatorial electrojet (EEJ), and the magnetic H component
27 from INTERMAGNET stations near the equator. East and South-East Asia, Russia, and Oceania
28 exhibited positive vTEC disturbances, while South American stations showed negative vTEC
29 disturbances during both storms. We also analyzed the vTEC from the Swarm satellites and found
30 similar results to the GNSS retrieved vTEC during different phases of both geomagnetic storms.
31 Moreover, we observed that ionospheric plasma tended to increase rapidly during the afternoon in

1
https://doi.org/10.5194/angeo-2022-18
Preprint. Discussion started: 5 September 2022
c Author(s) 2022. CC BY 4.0 License.

32 the main phase of the storms. At nighttime, the ionosphere depicted an opposite behavior under
33 similar conditions. The equatorial ionization anomaly (EIA) crest expansion to mid and high
34 latitudes is driven by PPEF during daytime at the main and recovery phases of the storms. The
35 magnetic H component exhibits a longitudinal behavior along with the EEJ enhancement near the
36 magnetic equator.

37 Keywords: Ionosphere, Geomagnetic Storms, Total Electron Content, Prompt Penetration Electric
38 Field

39

40 1. Introduction

41 The Sun trigger space weather events such as geomagnetic storms that can cause negative impacts
42 on communication and navigation through transionospheric electromagnetic signals on the Earth.
43 Geomagnetic storms result from large-scale disturbances of the Earth's magnetosphere under
44 variable solar activity, leading to anomalous ionosphere variability. These disturbances occur at
45 short-term scales (hours to a few days) and are usually triggered by Coronal Mass Ejection (CME),
46 Co-rotating Interaction Regions (CIRs), or fast-moving solar wind streams. Anomalous
47 ionospheric variations are observed during geomagnetic storms from plasma content variability
48 during the geomagnetic storms of 6 April and 29 May 2010 (Joshua et al., 2011; Adebiyi et al.,
49 2012). Several studies have investigated the ionospheric variations during storms at different
50 latitudes from satellite data (e.g., Fang et al. 2012; Adebesin et al. 2013; Calabia et al., 2022).
51 Moreover, the seasonal variations and hemispherical ionospheric irregularities are also presented
52 during geomagnetic storms (e.g., Tsurutani et al. 2004, Mannucci et al. 2005, Gao et al. 2008,
53 Stankov et al. 2010). However, the overall perception of storm-time ionospheric variations across
54 the different latitude ranges in both hemispheres is still uncertain.

55 Moreover, GNSS based TEC and in-situ data from multi-instruments describe the ionosphere
56 abnormalities in different spatial and temporal resolution at different latitudes during solar and
57 geomagnetic conditions (Chartier et al. 2018). The ionospheric irregularities have significant
58 effects on GNSS signals in the low latitudes during main phase of the storm; however, the
59 triggering reasons are unknown (Buchert et al. 2015; Xiong et al. 2016a). Transionospheric signal
60 delay during storm conditions results in unacceptable GNSS positioning errors for practical

2
https://doi.org/10.5194/angeo-2022-18
Preprint. Discussion started: 5 September 2022
c Author(s) 2022. CC BY 4.0 License.

61 applications (e.g., Stankov et al. 2007, 2009; Warnant et al., 2007). Since the ionospheric delay in
62 GNSS signal is not yet corrected, the Global Ionospheric Maps (GIMs) of TEC from the
63 International GNSS Service (IGS) are an exceptional product to calibrate the ionospheric
64 correction and eliminate discrepancies from GNSS signal with the help of other multi-instrument
65 data.

66 Geomagnetic storms induce effects in the ionosphere at different latitudes and longitudes in the
67 form of electric field penetration from high to low latitudes due to PPEF. Furthermore, the
68 perturbations of global thermospheric circulation in high latitude induced joule-heating
69 enhancement during geomagnetic activity leading to Disturbed Dynamo Electric Fields (DDEF).
70 In the equatorial and low latitudes, the electrodynamics in the ionospheric E and F regions
71 influences the plasma distribution (Heelis, 2004). Field Aligned Current System (FACS) controls
72 the transfer of energy and momentum from the magnetosphere to the ionosphere in the form of
73 two clear shells (Binod et al. 2017). These two shells include regions 1 and 2 for higher and lower
74 latitudes connected through the ionosphere around the Earth, respectively. The neutral wind
75 dynamo induced polarized electric fields in the low latitude during dayside (night side) in eastward
76 (westward) direction (Fuller‐Rowell, 2011). The horizontal component of magnetic field
77 corresponds to zonal electric field generates electrons upwelling due to E×B effect. As a result,
78 negatively and positively charged particles form on top and bottom of the ionospheric E region,
79 respectively. At an altitude of 90-130 km, the migration of electrons produces an electric current
80 known as the equatorial electrojet (EEJ).

81 Sharma et al. (2011) presented two enhanced peaks in TEC with twice in intensity as compared to
82 quiet days in low latitude region as storm-time responses of August 25, 2005. They showed that
83 the first peak in ionospheric TEC is due to PPEF and the second peak occurred due to plasma
84 fountain. Moreover, the PPEF influences along the longitudes showed nearly homogeneous effects
85 in the storm of August 25, 2005. On the other hand, the southward shifted interplanetary magnetic
86 field (IMF) Bz component induced the increased activity in the high-latitude convection. Previous
87 researches have provided insights on mid-latitude TEC enhancements during the initial phase of
88 geomagnetic storms as compared to main phase (Hargreaves, 1992; Araujo-Pradere et al. 2006).
89 For example, Astafyeva et al. (2016) showed the equatorial-and mid-latitudinal ionospheric TEC
90 during main phase of the storm at different part of the world from multi-instrument satellite data.

3
https://doi.org/10.5194/angeo-2022-18
Preprint. Discussion started: 5 September 2022
c Author(s) 2022. CC BY 4.0 License.

91 Similarly, Astafyeva et al. (2017) also assessed the effects of the June 2015 geomagnetic storm
92 with a comprehensive study using multiple satellite observations. They further demonstrated that
93 the storm had major effects on the ionosphere due to thermospheric winds in the low-and mid-
94 latitude regions. They also showed that dayside neutral mass density enhancement during storms
95 exceeded the quiet period in the thermosphere due to strong and robust PPEF influences the
96 ionosphere with significant variability. Moreover, Adebiyi et al. (2012) and Joshua et al. (2011)
97 reported an enhanced electron density in the African equatorial region during the geomagnetic
98 storms of 6 April and 29 May 2010.

99 Apart from above reports, positive and negative ionospheric anomalies due to geomagnetic storms
100 can significantly vary depending on the duration of the solar activity, season, latitude, local solar
101 time, etc., and each storm showed different characteristics. Clearly, we need to observe satellites
102 with multiple instruments in order to find the missing drivers (Araujo-Pradere et al. 2006;
103 Mannucci et al. 2008). This study comprises the understanding of the probable latitudinal
104 mechanisms that influence the variable ionosphere by studying the geomagnetic storms of June
105 2015 and August 2018 using multi-instrumental data. In the following section, we present a brief
106 description of the data and methods used in this study. Section 3 describes deeply the results, and
107 section 4 discuss the observed magnetosphere-thermosphere-ionosphere (MIT) coupling during
108 the storm. The last section summarizes the conclusions.

109 2. Data and Methods

110 In this paper, we study ionospheric response to the geomagnetic storms of 2015 and 2018 on global
111 scale to find out the source triggered the ionospheric variations. In particular, we analyze the 3-
112 hourly geomagnetic Kp index, the 1-min averaged electric Ey field, the IMF Bz component, the
113 solar wind velocity Vsw, the aurora AE index, the geomagnetic disturbance storm time index (Dst),
114 and the solar flux F10.7 index. The data is available at the Omni Web of NASA at
115 http://omniweb.gsfc.nasa.gov/. The beginning of a geomagnetic storm usually exhibits a prompt
116 decrease in the Dst index. The AE index can be used to study the energy transmitted to the auroral
117 ionosphere during the storm. The Kp index can provide a good description of the magnitude of the
118 storm; the range of Kp is between 0 to 9. The PPEF data is obtained from the real-time model of
119 the Cooperative Institute for Research in Environmental Sciences website
120 https://geomag.colorado.edu/real-time-model-of-the-ionospheric-electric-fields.html.

4
https://doi.org/10.5194/angeo-2022-18
Preprint. Discussion started: 5 September 2022
c Author(s) 2022. CC BY 4.0 License.

121 The Global Ultraviolet Imager (GUVI) onboard the Thermosphere-Ionosphere-Mesosphere

122 Energetics and Dynamics (TIMED) satellite senses far-ultraviolet emissions and provides
123 thermospheric [O/N2] ratio maps (Christensen et al., 2003). These maps are obtained from
124 https://guvitimed.jhuapl.edu/. The [O/N2] ratio is a measure of the electron density at the
125 ionospheric F region; increases in N2 decreases electron density (Prölss and Bird, 2010).

126 The TEC data from 15 different GNSS stations at low-mid and high-latitude regions were retrieved
127 from the IONOLAB website, https://www.ionolab.org/. Fig. 1 shows the location of the GNSS
128 stations used in this study and Table 1 details them. Slant TEC (STEC) is estimated as the number
129 of free electrons in a square meter section along the line of sight between a GNSS satellite and
130 receiver. The STEC units are TEC Units (TECU), where 1 TECU = 1016 electron/m2. The STEC
131 is obtained from IONOLAB and is processed by below equations (Arikan et al. 2008).

𝑓 2 𝑓2
1 2
132 STEC = 40.28(𝑓 2 −𝑓2 ) (𝐿1 − 𝐿2 + 𝜆1 (𝑁1 + 𝑏1 ) − 𝜆2 (𝑁2 + 𝑏2 ) + 𝜖) (1)
1 2

𝑓12 𝑓22
133 STEC = (𝑃1 − 𝑃2 − (𝑑1 − 𝑑2 ) + 𝜖) (2)
40.28(𝑓12 −𝑓22 )

134 In this equation, carrier phase frequencies are presented by f1 and f2, pseudo-range is denoted as L,
135 the delay path of the signal of carrier phase observations is P, the signal wavelength is λ, and the
136 ray path uncertainty is N. Here, d and b denote the biases of consequent signal pseudo-range and
137 instrumental carrier phase, and ϵ is the random error in the signal. The STEC is converted to VTEC
138 using the following equation (Shah et al. 2020):

𝑅𝑠𝑖𝑛𝑍
139 VTEC = STEC × cos(𝑎𝑟𝑐𝑠𝑖𝑛𝑒 ( 𝑅+𝐻 )) (3)

140 In this equation, Z is the elevation angle of the satellite, and R and H are the Earth’s radius and the
141 ionosphere height, respectively (Klobuchar 1987).

142 Table 1. The details of GNSS stations used to study ionospheric variations.

5
https://doi.org/10.5194/angeo-2022-18
Preprint. Discussion started: 5 September 2022
c Author(s) 2022. CC BY 4.0 License.

Geographic Geomagnetic Latitude

Region Station Receiver Latitude (Longitude)
(Longitude) 2015 2018
Australia SEPT 12.188°S 21.62°S 21.46°S
South East (COCO) POLARX5 (96.834°E) (168.89°E) (168.95°E)
Asia Indonesia LEICA 6.49°S 16.13° S 15.97°S
(BAKO) GR50 (106.85°E) (179.44°E) (179.49°E)
LEICA
India 17.417°N 8.77°N 8.92°N
GRX1200G
(HYDE) (78.551°E) (152.23°E) (152.26°E)
GPRO
SEPT 13.021°N 4.50° N 4.64°N
South Asia India (IISC)
POLARX5 (77.570°E) (150.92°E) (150.9°E)
Low
New
Latitude TRIMBLE 20.559°S 25.48°S 25.40°S
Oceania Caledonia
NETR9 (164.287°E) (119.59°W) (119.61°W)
(KOUC)
Ecuador JAVAD 0.743°S 8.49°N 8.33°N
(GLPS) TRE_G3TH (90.304°W) (17.89°W) (17.84°W)
French SEPT
South 5.252°N 14.31°N 14.15°
Guiana POLARX5
America (52.640°W) (20.55°E) (20.58°E)
(KOUR) TR
Ecuador TRIMBLE 1.651°S 7.99N 7.83°N
(RIOP) NETRS (78.651°W) (6.09W) (6.05°W)
New
TRIMBLE 36.6.3°S 39.58°S 39.53°S
Oceania Zealand
ALLOY (174.834°E) (105.37°W) (105.47°W)
(AUCK)
Japan TRIMBLE 43.529°N 35.14°N 35.29°N
(STK2) ALLOY (141.845°E) (149.78°W) (149.69°W)
East Asia
Mid Japan SEPT 36.133°N 27.51°N 27.66°N
Latitude (USUD) POLARX5 (138.362°E) (151.98°W) (151.91°W)
Eastern
Russia JAVAD 47.030°N 38.69°N 38.84°N
Europe and
(YSSK) TRE_3N (142.717°E) (149.55°W) (149.45°W)
Russia
South Chile SEPT -33.150°S -23.29°S -23.46°S
America (SANT) POLARX5 (70.669°W) (1.78°E) (1.81°E)
Sweden SEPT 67.878°N 65.26°N 65.33°N
High Western (KIR0) POLARX5 (21.060°E) (115.42°E) (115.13°E)
Latitude Europe Sweden SEPT 60.595°N 59.04°N 59.08°N
143 (MAR6) POLARX5 (17.259°E) (106.40°E) (106.17°E)

144 Moreover, we also study the ionospheric indices from Swarm satellites to provide more evidence
145 to vTEC variations from GNSS. The Swarm mission is comprised of three identical satellites,
146 where Swarm A and C orbit at 440-460 km height and Swarm B track is at 520-530 km height.
147 These satellites carry sophisticated magnetometers, an Electric Field instrument to measure
148 electron density (Ne), and a GNSS receiver to provide vTEC. Moreover, all satellites have polar
149 orbits with inclination angle of 87°- 88°. The vTEC data from Swarm is available at
150 https://vires.services. The Swarm data was also analyzed during the different phases of both the
151 storms.

6
https://doi.org/10.5194/angeo-2022-18
Preprint. Discussion started: 5 September 2022
c Author(s) 2022. CC BY 4.0 License.

152
153 Fig. 1. Geographic location of GNSS and INTERMAGNET stations used in this study. The
154 yellow line represents the magnetic equator. The corresponding coordinates are given in Table 1.

155 Furthermore, vTEC and dTEC from GNSS of IGS network is analyzed in bi-hourly temporal
156 resolution and spatial resolution of 2.5° by 5° in latitude and longitude, respectively (Hernández-
157 Pajares et al. 1999; Roma-Dollase et al., 2018). The maps are available in the IONEX
158 (IONosphere map Exchange ) format at the Crustal Dynamics Data Information System (CDDIS)
159 Goddard Space Flight Center (GSFC) National Aeronautics and Space Administration (NASA)
160 website https://cddis.nasa.gov/index.html.

161 In order to investigate the abrupt TEC anomalies during geomagnetic storms, the new empirical
162 vTEC model of Calabia and Jin (2020, 2019) is used as quiet-time background. In this model,
163 vTEC observables from 2003 to 2018 were reduced to a lower-dimensional through the principal
164 component analysis, and the resulting time-expansion coefficients were parameterized in terms of
165 solar and magnetospheric forcing, annual, and LST cycles. The quiet magnetospheric forcing is
166 set during the geomagnetic index condition at Am=6. In this scheme, the diurnal, annual, and solar
167 cycle variations are eliminated, and the residuals mainly show the short-term variations due to
168 magnetospheric forcing; i.e., those variations mainly caused due to geomagnetic storms. The
169 Calabia and Jin model is available at https://zenodo.org/record/3563463.

170 The Earth’s magnetic field components are obtained from the magnetometer stations near the
171 magnetic equator. This data aims to help investigate the E region response during various phases

7
https://doi.org/10.5194/angeo-2022-18
Preprint. Discussion started: 5 September 2022
c Author(s) 2022. CC BY 4.0 License.

172 of the geomagnetic storms. The data at 1-min resolution is available at the INTERMAGNET
173 network http://intermagnet.org. We employ data from the stations at HUA (America), GUA
174 (Pacific Ocean), and MBO (Africa). The geographic and geomagnetic coordinates of the
175 magnetometer stations are listed in Table 2, and their locations are shown in Figure 1. According
176 to Biot and Savart’s law, ground magnetic field perturbations can be an integral part of ionospheric
177 and magnetospheric electric current (Shao et al., 2002; Le and Amoray-Mazaudier, 2005). The
178 horizontal component (H) of geomagnetic field can be computed using the north (X) and east (Y)
179 components of the magnetic field (i.e., 𝐻 = √𝑋 2 + 𝑌 2 ). The observed H component corresponds
180 to the current flow into the magnetosphere-ionosphere systems (Cole, 1966). The equation is as
181 follows:

182 H = SR + D (4)

183 In this equation, SR and D represent the solar regular variations of Earth’s magnetic field due to
184 regular ionospheric dynamo and the combined effect of various current systems flowing in the MI
185 system, respectively (Zaourar et al., 2017). According to Le and Amoray-Mazaudier (2005), the
186 H component can be rewritten as follows:

187 H = H0 + SR + DM + Diono (5)

188 In this equation, H0 and SR are Earth’s core induced baseline magnetic field and regular variation
189 of Earth’s magnetic field on a given day, respectively. Whereas, magnetic field variations
190 associated to magnetosphere and ionosphere currents are represented as DM and Diono, respectively.
191 The Diono is estimated as follows (Le and Amoray-Mazaudier, 2005):

192 Diono = H - DM – Sq (6a)

193 Diono = DP2 + Ddyn (6b)

194 In this equation, Diono consists on the combined effect caused by ionospheric disturbance due to
195 polar currents (DP2) and dynamo currents (Ddyn) at low latitudes, DP2 is associated with PPEF,
196 and Ddyn is associated with DDEF (Nishida, 1968; Le and Amoray-Mazaudier, 2005). DM is
197 calculated using the SYM-H index and the dip angle Φ as follows:

198 DM = SYM-H * Cos (Φ) (7)

8
https://doi.org/10.5194/angeo-2022-18
Preprint. Discussion started: 5 September 2022
c Author(s) 2022. CC BY 4.0 License.

199 In equation 6a, Sq represents the average of selective quiet days (SR). Here, 5 quiet days are
200 considered to compute Sq. We average the H component as suggested by the German Research
201 Center of Geosciences (GFZ) (ftp://ftp.gfz-potsdam.de/pub/home/obs/kp-ap/quietdst/). The results
202 are shown in Table 3, and the equation of Sq is as follows:

1
203 Sq = 〈H quiet 〉 = ∑ni=1 Hiquiet (8)
n

204 The EEJ at each station is computed by differences of the H component inside and outside the EEJ
205 region at similar longitudes. These differences are related to the contribution of the EEJ current
206 (Anderson et al., 2004):

207 EEJ = H1 – H2 (9)

208 In equation 9, H1 and H2 are the average of H components inside and outside the EEJ region,
209 respectively.

210

9
https://doi.org/10.5194/angeo-2022-18
Preprint. Discussion started: 5 September 2022
c Author(s) 2022. CC BY 4.0 License.

211

212 Table 2. The geographic and geomagnetic locations and magnetic dip angle
213 of Magnetometer stations.

Geographic Geomagnetic Latitude

Station Dip Angle
Region Latitude (Longitude)
Code
(Longitude) 2015 2018 2015 2018
12.0686°S 2.31°S 2.48°S -0.3612° -0.8384°
America HUA
(75.2103°W) (2.54°W) (2.50°W)

13.4443°N 5.61°N 5.74°N 12.4583° 12.3219°

Pacific
GUA (144.7937°E
Ocean (143.57°W)
) (143.52°W)
14.4228°N 19.63°N 19.54°N 7.0608° 6.6283°
Africa MBO
(16.9654°W) (58.13°E) (58.12°E)

2.18°N 2.34°N 11.230°

11.9404°N 11.6661°
Asia DLT
(108.4583°E
(178.95°W)
214 ) (178.91°W)

215

216 Table 3. The selected magnetic quiet day to calculate Sq during

217 June 2015 and August 2018 storms.

Q1 Q2 Q3 Q4 Q5
June, 2015 20 5 2 4 3

218 August, 2018 6 14 10 13 23

219

220

221 3. Results
222 3.1 The Geomagnetic Storms of June 2015 and August 2018

223 The geomagnetic storm of June 2015 occurred during the solar cycle 24, and it was the second
224 largest known storm after the St. Patrick’s storm. On 22 June 2015, two CMEs hit the Earth’s
225 magnetosphere at 05:45 UT and 18:35 UT. Figures 2-3 shows the Sudden Storm Commencement
226 (SSC), where the different phases are classified on the basis of different storm indices. The IMF
227 Bz component shows a sharp southward turning immediately after the SSC, followed by a second

10
https://doi.org/10.5194/angeo-2022-18
Preprint. Discussion started: 5 September 2022
c Author(s) 2022. CC BY 4.0 License.

228 southward IMF Bz before the main phase. These IMF Bz turnings are associated with more than
229 720 km/s speed of solar wind after the second SSC.

230

231 Fig. 2. Space weather indices for the storm of 22 June 2015. The SSCs are marked with red arrows
232 and different phases of the storm are marked with vertical dashed lines.

233 The storm of August 2018 occurred due to a large CME ejection from the Sun on 20 August 2018
234 (Figure 3). Formerly, scientist from the National Oceanic Atmospheric Administration (NOAA)
235 called it a minor storm due slow speed stream but later on a G4 severe geomagnetic storm evolved
236 as long term southward IMF Bz; i.e., from 15:55 h UT on 25 August to 09:45 h UT on 26 August,
237 thus allowing a large number of particles entering the Earth’s magnetosphere. The SSC initiated
238 at 09:00 UT on 25August 2018 and, after 3 hours of the SSC, at 09:00 h UT, a rapid drop in Dst
239 index was observed until 23:00 h UT on 23 August. The lowest Dst value was -203nT around
240 07:00 h UT on 26 August.

11
https://doi.org/10.5194/angeo-2022-18
Preprint. Discussion started: 5 September 2022
c Author(s) 2022. CC BY 4.0 License.

241

242 Fig. 3. Space weather indices during the storm of 26 August 2018 from OMNI web NASA. The
243 SSC is marked with a red arrow and the different phases of the storm are marked with different
244 color vertical dashed lines.

245 3.2 Ionospheric-Thermospheric Irregularities

246 The vTEC variations occurred at low-latitude stations in South America, South Asia, South East
247 Asia, and Oceania region effect during the 2 geomagnetic storms (Figure 4). During the initial
248 phases of both the storms, no clear enhancements occurred at the low latitude GNSS stations.
249 However, for both storms, the GNSS stations at South East Asia showed significant vTEC
250 variations during the main phase. For the South American stations, only the KOUR showed
251 significant variability. Although both storms are of similar intensity, VTEC enhancements of > 50
252 TECU, 42<TECU<50, and 40< TECU<45 occurred in the low latitude stations of South East Asia,
253 South Asia and American stations, respectively, during main phase of June 2015 storm. On the
254 other hand, vTEC variations occurred in the range of 18 <TECU<20, 42<TECU<50, 40<
255 TECU<45, and 18 <TECU<20 for COCO, BAKO, South Asia, and KOUR GNSS stations in 2018
256 during the main phase of the storm, respectively. Moreover, there were not significant variations
257 in vTEC during the recovery phases of 2015 and 2018 storms; only a minor depletion in the South

12
https://doi.org/10.5194/angeo-2022-18
Preprint. Discussion started: 5 September 2022
c Author(s) 2022. CC BY 4.0 License.

258 American stations. During the recovery phase, TEC depletions in South American stations were
259 more prominent in 2015 than in 2018.

260

261 Fig. 4. vTEC variation at the low-latitude stations in different longitudinal sectors for the
262 geomagnetic storms of 2015 and 2018. The locations of the stations are showed in Figure 1. The
263 different phases of the storm are marked with vertical dashed lines.

264 The vTEC variations at the mid-latitude GNSS stations are shown in Figure 5. We employ the
265 AUCK station in New Zealand, the STK2 and USSD stations in East Asia, the YSSK station in
266 Russian, and the SANT station in Chile. During the initial phases of both storms, no clear vTEC
267 variations occurred in any of the stations; only the SANT station showed a weak variation. During
268 the main phases of both storms, the sharp enhancements are shown for all stations; except for the
269 SANT station. The vTEC during the main phase of the 2015 geomagnetic storm at Oceania, East
270 Asia, and Russia is 30<TECU<40, 30<TECU<40, 20<TECU<30, respectively. On the other hand,
271 Oceania, East Asia, and Russia exhibited 10<TECU<20, 10<TECU<20, 10<TECU<20 during the
272 2018 geomagnetic storm, respectively. All the mid-latitude stations showed no significant
273 anomalies during the recovery phases of both storms.

13
https://doi.org/10.5194/angeo-2022-18
Preprint. Discussion started: 5 September 2022
c Author(s) 2022. CC BY 4.0 License.

274

275 Fig. 5. vTEC variation at the mid-latitude stations in different longitudinal sectors for the
276 geomagnetic storms of 2015 and 2018. The locations of the stations are showed in Figure 1. The
277 different phases of the storm are marked with vertical dashed lines.

278

279 The vTEC at the high-latitude stations of KIR0 and MAR6 in Sweden and Europe are shown in
280 Figure 6. In this Figure, enhancements of 2 TECU in KIR0 are shown within 2 hours after the SSC
281 of the storm of 2015; Then, a sudden depletion until the main phase of the storm occurred.
282 Similarly, the MAR6 station increases 4 TECU after the SSC, and then a depletion in the main
283 phase occurred. In the recovery phase, no increases were seen for both stations.

14
https://doi.org/10.5194/angeo-2022-18
Preprint. Discussion started: 5 September 2022
c Author(s) 2022. CC BY 4.0 License.

284

285 Fig. 6. vTEC variation at the high-latitude stations in different longitudinal sectors for the
286 geomagnetic storms of 2015 and 2018. The location of the stations is shown in Figure 1. The
287 different phases of the storm are marked with vertical dashed lines.

288

289 The TEC variations in GIMs for both storms are shown in Figures 7-8. During the storm of 2015,
290 all 3 American, African, and Asian sectors showed a moderate high-latitude enhancement after the
291 SCC at the southern latitudes. Then, the American and African sectors showed strong
292 enhancements at the low-latitude regions above 15 dTECU, whereas the Asian sector showed
293 depletion of similar magnitude. For this storm, the high-latitude regions showed a clear depletion
294 during the main phase for all the 3 longitudes. During the main phase of the 2018 storm, vTEC
295 enhancements were very prominent in the American and African sectors, in comparison to that in
296 the Asian sector. As compared to the storm of 2015, no clear depletions were seen at any location.

15
https://doi.org/10.5194/angeo-2022-18
Preprint. Discussion started: 5 September 2022
c Author(s) 2022. CC BY 4.0 License.

297

298 Fig. 7. GIM TEC maps at different longitudinal sectors during the June 2015 geomagnetic storm
299 where; a-c) are TEC maps of America, Africa and Asia region, and a`-c`) are dTEC maps of
300 America, Africa and Asia.

16
https://doi.org/10.5194/angeo-2022-18
Preprint. Discussion started: 5 September 2022
c Author(s) 2022. CC BY 4.0 License.

301

302 Fig. 8. GIM TEC maps response to geomagnetic storm of August 2018 where a-a`) TEC and
303 dTEC maps of America, b-b`) TEC and dTEC map of Africa and c-c`) TEC and dTEC map of
304 Asia region

305 Figure 9 shows the O/N2 ratio during the storms of 2015 and 2018. The African sector showed
306 reductions of O/N2 ratio in the low-and mid-latitudes during the main phase of the 2015 storm.
307 This resulted in the increment of vTEC in the African region. The Asian, Australia, and Oceania
308 regions also showed significant enhancements in O/N2 ratio during the main phase of the 2015
309 geomagnetic storm and it result in vTEC depletion in the above mentioned regions. On the other
310 hand, we also observe enhancement in O/N2 ratio (depletions in vTEC) in South American and
311 Asian regions during the main phase of the 2018 geomagnetic storm. There are several reports in
312 enhancement/depletion in O/N2 ratio (reduction/enrichment in vTEC) in different part of the world
313 through thermospheric O/N2 variability (Martinis et al. 2005; Buresova et al. 2014; Kassa &
314 Damite, 2017).

17
https://doi.org/10.5194/angeo-2022-18
Preprint. Discussion started: 5 September 2022
c Author(s) 2022. CC BY 4.0 License.

315

316 Fig. 9. The O/N2 ratio from GUVI during the storms of June 2018 and August 2015.

317 The vTEC from Swarm satellites for both the storms is shown in Figures 10-11. Clear
318 enhancements were seen in the American region at the low-and mid-latitudes during the initial
319 phase of both storms; no clear variations were observed for the Asian region during the initial
320 phase of both storms. The low-and mid-latitudes of the Asian and African regions depicted larger
321 VTEC variations than those in the American sector during the main phase of both storms. During
322 the recovery phase of both storms, larger variations were observed at the American region than
323 those at the Asian sector. The VTEC values from Swarm during the main and recovery phases
324 were different than those from the GNSS stations in Asia, Australia and Russia.

18
https://doi.org/10.5194/angeo-2022-18
Preprint. Discussion started: 5 September 2022
c Author(s) 2022. CC BY 4.0 License.

325

326 Fig. 10. The VTEC from Swarm during the geomagnetic storm of June 2015.

327

328 Fig. 11. The VTEC from Swarm during the geomagnetic storm of August 2018.

19
https://doi.org/10.5194/angeo-2022-18
Preprint. Discussion started: 5 September 2022
c Author(s) 2022. CC BY 4.0 License.

329 The PPEF variations at the low-and mid-latitude regions during both storms are shown in Figure
330 12. The PPEF variations during the 2018 storm were smaller than those during the storm of 2015.
331 This is different from the results obtained through IGS GIMs VTEC (Figures 7 and 8). Moreover,
332 strong PPEF occurred at all longitudes during the main phase of the storm of 2015, while the PPEF
333 peak during the 2018 storm occurred in the far East and West. During the main phase of both
334 storms, stronger PPEF occurred in comparison to that seen during the other phases.

335

336

337 Fig. 12. PPEF behavior during geomagnetic storms where a) is June 2015 geomagnetic storm
338 and b) is geomagnetic storm of August 2018.

339

340 3.3 Earth’s Magnetic Field Variations

341 The variations in the Earth’s magnetic field during the storms of June 2015 and August 2018 are
342 shown in Figure 13. We investigate the variations in the H-component of Earth’s magnetic field
343 and the EEJ estimated from INTERMAGNET stations near the magnetic equator. This shows
344 significant variations at the SSC events during both storms, followed by a considerable decrease
345 in Earth horizontal component during the recovery phases. The largest disturbances of the H
346 component in the American region (HUA station) reached 259.92 nT on 22 June 2015 at 20:49 h
347 UT. Moreover, on 25 August 2018 at 23:55 h UT the initial phase of the storm reached -123.91
348 nT. The Dion exhibited a decrease in the initial phase, followed by an increase in the main phase,
349 this due to H minima during nighttime in the South American region. Two negative peaks in the
350 H component were observed during the storm of June 2015 in the Pacific region, one during the
351 beginning of the initial phase, and other during the main phase. Similarly, only one negative peak

20
https://doi.org/10.5194/angeo-2022-18
Preprint. Discussion started: 5 September 2022
c Author(s) 2022. CC BY 4.0 License.

352 was observed in the main phase during the storm of August 2018. The values of Dion exhibited
353 abrupt variations for both storms after each respective SSC, corresponding with the variations in
354 the H component. The MBO station in Africa and the DLT station in Asia showed prominent
355 decreases in the H component for both storms. The lowest values were -207.12 nT for the storm
356 of June 2015 and -107.78 nT for the storm of August 2018. The VTEC variations triggered by Dion
357 were prominent at different longitudes, specifically during the SSC and the main phase of the storm
358 of June 2015. No clear variations were seen in the HUA station during the storm of August 2018.

359

360

361 Fig. 13. Magnetic field variability during the storms of June 2015 and August 2018. a and a`) have
362 the variation of SYM-H and ASY-H index, in (b-d) we show the H component in 2015 at HUA,
363 GUA and MBO stations, in (b`-d`) we show the H component in 2018 at HUA, GUA and DLT

21
https://doi.org/10.5194/angeo-2022-18
Preprint. Discussion started: 5 September 2022
c Author(s) 2022. CC BY 4.0 License.

364 stations, and in (e & e`) we show the EEJ responses. The SSC is marked with a red arrow. The
365 different phases of the storm are marked with vertical dashed lines.

366 4. Discussion

367 The VTEC enhancements during the storms of June 2015 and August 2018 initiated approximately
368 4 hours after the SSC events at the low-latitude regions in East Asia, South East Asia and Oceania.
369 All the 3 sources of VTEC data used in this study, i.e. GNSS, Swarm, and IGS GIM TEC, have
370 provided similar results with minor differences, specifically between GNSS and IGS GIM TEC,
371 most likely due to local anomalies not well represented by GIM TEC (Lisa et al. 2020). All the
372 VTEC enhancements occurred during the main phase for all 3 datasets. The variations at different
373 geographical coordinates followed the PPEF and thermospheric O/N2 variations. Moreover, the
374 PPEF enhancements started at the SCC in Asia and Oceania, along with the O/N2 enhancements
375 leading to clear effects during the main phase at the low-latitude regions (Figs. 9 & 12). These
376 positive enhancements are due to PPEF and the increment of oxygen (Klimenko et al. 2011). On
377 the other side, no prominent enhancements or depletions occurred in the South American sector,
378 most likely due to Dst minimum, along with depletion in the recovery phase due to a drop in O/N2
379 ratio (Figs. 4 & 9).

380 At the mid-latitudes, the Asia-Oceania region exhibited peak values during the main phase of the
381 storms, coinciding with Dst minima. The station from South America exhibited depletions at the
382 night side during the storm of June 2015; the storm of August 2018 lacked this feature. The EIA
383 expansion from equatorial regions to mid-latitude regions was responsible for VTEC
384 enhancements at all longitudes. Both storms analyzed here revealed that Dion and O/N2 drivers
385 control these fluctuations. In fact, Fuller-Rowell et al. (1994), Mannucci et al. (2005), and
386 Vankadara et al. (2022) presented similar results. The PPEF plays a vital role in VTEC
387 enhancements through plasma diffusion along magnetic field lines, thus creating the fountain
388 effect during the daytime (Mannucci et al. 2005). The depletions seen in VTEC were due to
389 variations in thermospheric composition such as those generated by recombination processes
390 creating N2. These depletions were observed in the recovery phase of both storms in the low-
391 latitude South American regions (Figs. 4 & 9).

392 External electric field can penetrate into equator to disturb low and mid latitudes, as they are
393 connected to inner magnetosphere through closed magnetic field lines. External sources should
22
https://doi.org/10.5194/angeo-2022-18
Preprint. Discussion started: 5 September 2022
c Author(s) 2022. CC BY 4.0 License.

394 also be considered, taking into account the fundamental forces that drive the penetration of electric
395 field, such as solar wind drivers. Nishida (1968) compared the north-south oscillation in IMF with
396 the geomagnetic fluctuations, and Jaggi and Wolf (1973) considered PPEF as a temporary failure
397 mechanism of shielding. PPEF can exhibit multiple pulses, as it is the direct consequence of IEF
398 fluctuations (Kelley et al. 1979). Magnetic reconnection is an important parameter for dusk-ward
399 PPEF processes (lasting < 3h), and the dawn-ward IEF shows the opposite behavior, as long IMF
400 Bz oscillates between northward and southward polarity; dawn-ward IEF rarely does. Wei et al.
401 (2010) pointed out that the shielding effect would not fully develop under these circumstances,
402 and would not cancel the PPEF during the short pulse of dusk-ward IEF. Nevertheless, this is not
403 always the case since the transition to the northward IMF Bz component does not necessarily
404 generate over-shielding. However, the reduction in the convective electric field can transit to over-
405 shielding status (Wei et al. 2010). The magnetosphere under sustained pressure due to dense solar
406 winds can suppress the development of electric field shielding during multiple PPEF events. PPEF
407 can exhibit long-duration patterns as long as the magnetic activity is being strengthened under
408 storm conditions (Huang et al. 2005). In this work, PPEF has demonstrated to generate variations
409 in VTEC throughout the globe, except for the South America region, which was more prominent
410 during the storm of June 2015. The max PPEF was confined to only the far East and West regions
411 during the storm of August 2018, depicting clear variations in Oceania and not in the American
412 sector. As the storm commenced, Asia, Oceania, and Russia exhibited VTEC enhancements at the
413 low- and mid-latitudes due to PPEF. Storm time variations at the low- and mid-latitudes were
414 generated by a large fountain effect, creating a stronger EIA. In fact, many researchers (Manucci
415 et al. 2005; Abdu et al. 2007; Sharma et al. 2011; Lu et al. 2013) have reported these effects. The
416 ionosphere exhibited a variable response along different longitudes. This has also been confirmed
417 by different magnitudes of PPEF and satellite data (Figs. 4-8 & 10-12). Fagundes et al. (2016)
418 demonstrated that the influence of PPEF in EIA shows significant longitudinal differences during
419 geomagnetic storms.

420 During these geomagnetic storms, the Earth’s magnetic field observations at different longitudes
421 make possible to comprehend the processes of large-scale ionosphere electric currents. There are
422 2 main types of disturbances, namely DP2 and Ddyn, which are associated with PPEF and DDEF,
423 respectively. Ddyn exhibits a more dynamic variation in comparison to DP2, which only lasts for 2
424 to 3 hours (Nishida et al., 1968; Le and Amory-Mazaudier (2005). During these geomagnetic

23
https://doi.org/10.5194/angeo-2022-18
Preprint. Discussion started: 5 September 2022
c Author(s) 2022. CC BY 4.0 License.

425 storms, normal circulation of thermospheric winds are perturbed due to moment transfer and
426 energy inputs at high latitudes, giving eastward and westward electric field at the nightside and
427 dayside, respectively (Blanc & Richmond, 1980; Fuller-Rowell et al. 2002). Diono exhibited large
428 nighttime enhancements at the low-latitude stations. These variations are associated with PRC, as
429 indicated by ASYM-H (Fig. 13a & 13a`). The anti-Sq signatures observed during the recovery
430 phase in the magnetic data are due to the orientation of electric fields (Yamazaki & Kosch 2015).
431 Vankadara et al. (2022) did a similar study, where the authors showed Dion minima at different
432 Local Solar Time (LST) locations, leading to equatorial plasma bubble developments. Our results
433 have shown differences in longitude because of magnetospheric convection processes and electric
434 field penetration (Fejer et al. 2008). In this scheme, all three American regions have shown clear
435 variations in the initial phases, but none in the main phases of both storms (Figs. 4 - 5, -7 - 8, and
436 10 - 11). On the other side, Asia, Oceania, and Russia have shown VTEC enhancements during
437 the main phases (Figs. 4, 5, 7, 8, 10 & 11). Various authors have shown latitudinal and longitudinal
438 ionosphere responses due to PPEF (Kikuchi et al. 2000; Mene et al. 2011; Kashcheyev et al. 2018).
439 The EEJ variations at different longitudes are due to the underlining effects of local winds, which
440 are responsible for EEJ driving (Stening 1985, 1995). In addition, longitudinal differences in EEJ
441 are caused by the different nature of the propagating diurnal tides, the meridional winds, and the
442 dynamics of the migratory tides (Luhr et al. 2004; Rabiu et al. 2011). In this study, clear EEJ
443 enhancement has been observed at the beginning of both the 2015 and 2018 storms. In the
444 American region (Fig. 13e), EEJ resulted in VTEC variability in the initial phase, but no clear
445 variability along the main phase. In the Asian region, the EEJ increment has been more prominent
446 during the main phase (Fig. 13e`), leading to VTEC enhancements in the low-latitude stations (Fig.
447 4a-4c, 4a`-4c`). Our results demonstrate the existence of longitudinal variability due to EEJ during
448 storm-time conditions. According to Lühr et al. (2004), dependencies of EEJ strength can be
449 explained by varying the cross-section area of the longitudinal Cowling channel.

450

451 5. Conclusions

452 The upper atmospheric responses to 22-23 June 2015 and 25-26 August 2018 geomagnetic storms
453 have been investigated for different regions of the world. The ionospheric variations during the

24
https://doi.org/10.5194/angeo-2022-18
Preprint. Discussion started: 5 September 2022
c Author(s) 2022. CC BY 4.0 License.

454 storms are also showed in the context of different drivers at global and regional scales during the
455 two storms. The main conclusions are as follows:

456  Different regions have exhibited variable patterns of vTEC enhancements/depletions

457 depending on thermospheric O/N2 ratio reduction/enrichment. In low latitude, the GNSS
458 stations of East Asia (HYDE & IISC), South East Asia (COCO & BAKO), and Oceania
459 (KOUC) have shown vTEC enhancement at the main phases of the storms. On the other
460 side, the stations in South America (GLPS, KOUR and RIOP) registered no such
461 enhancements. vTEC enhancement in the Asian and Oceania regions were approximately
462 double the value as that during quite days. At the mid-latitudes of Oceania, East Asia, and
463 Russia, the GNSS stations exhibited enhancements during both storms.

464  The Swarm satellites vTEC confirmed the low-and mid-latitude ionospheric irregularities
465 during main phase of both the storms.

466  The GIM-TEC from IGS has also shown clear agreement with the GNSS-derived vTEC at
467 most part of world during main phase of both the storms. These ionospheric variations at
468 low-and mid-latitude regions during main phases of the both the storms are mainly driven
469 by thermospheric O/N2 ratio, PPEF and EEJ.

470  The PPEF variations at different longitudes provided different vTEC responses. These
471 variations were clearly present in the low-and mid-latitude regions of Asia, Africa, Russia,
472 and Oceania. The southward-northward oscillation of the IMF Bz component drives this
473 variability along with interactions with Earth’s Magnetosphere and solar wind. vTEC
474 enhancements at different longitudes were mainly attributed to PPEF variability. vTEC
475 depletions were mainly due to the enriched thermospheric winds composition, as seen by
476 changes in the O/N2 density ratio.

477  The Dion from H-component of the Earth’s magnetic field has exhibited clear variations
478 during the 2015 storm as compared to 2018 storm. Moreover, significant EEJ is also noted
479 in the low-latitude American and African stations during main phase of both the storms,
480 that induced clear ionospheric variations.

481 Author Contributions

25
https://doi.org/10.5194/angeo-2022-18
Preprint. Discussion started: 5 September 2022
c Author(s) 2022. CC BY 4.0 License.

482 RS and MS did analysis and paper writing, MS designed the idea, AA, AH, AMM, NAN helped
483 in paper revision, AH helped in the improvement of figures and paper writing, AC provided writing
484 review and editing. M.A.H., I.K., A.M., E.A.A. reviewed the final draft. All authors have read and
485 agreed to the published version of the manuscript.

486 Acknowledgments

487 The TIMED/GUVI data are available at https://guvitimed.jhuapl.edu/. The space weather indices
488 are available at the NASA OMNI website http://omniweb.gsfc.nasa.gov/. The magnetometer data
489 are available at https://www.intermagnet.org/. GNSS vTEC data of multiple stations is available
490 at http://www.ionolab.org/. Swarm satellite data is available at https://vires.services/. Authors are
491 appreciative of all the above sources for providing valuable datasets.

492 Conflicts of Interest

493 The authors declare no conflict of interest

494 References

495 Abdu, M. A., Maruyama, T., Batista, I. S., Saito, S., Nakamura, M., 2007. Ionospheric responses
496 to the October 2003 superstorm: Longitude/local time effects over equatorial-low and mid-
497 latitudes. Journal of Geophysical Research. 112:A10306. 10.1029/2006JA012228.
498 Adebesin, B. O., Ikubanni, S. O., Adebiyi, S.J., Joshua, B. W., 2013. Multi-station observation of
499 ionospheric disturbance of March 9 2012 and comparison with IRI-model, Advances in Space
500 Research. 52:604–613.
501 Adebiyi, S.J., Adeniyi, J.O., Adimula, I.A., Joshua, B., Gwani, M., 2012. Effect of the
502 Geomagnetic Storm of April 5th to 7th, 2010, on the F2-Layer of the Ionosphere of Ilorin,
503 Nigeria. World Journal of Engineering and Pure and Applied Science. 2(2):56–62.
504 Anderson, D., Anghel, A., Chau, J. L., Veliz, O., 2004. Daytime vertical E × B drift velocities
505 inferred from ground-based magnetometer observations at low latitudes. Space Weather.
506 2:S11001. 10.1029/2004SW000095.
507 Araujo-Pradere, E. A., Fuller-Rowell, T. J., Spencer, P. S. J., 2006. Consistent features of the TEC
508 changes during ionospheric storms. Journal of Atmospheric and Solar-Terrestrial Physics.
509 68(16):1834–1842. doi:10.1016/j.jastp.2006.06.
510 Arikan, F., Nayir, H., Sezen, U., Arikan, O., 2008. Estimation of single station interfrequency
511 receiver bias using GPS-TEC. Radio Science. 43(4):1-13. 10.1029/2007RS003785.
512 Astafyeva, E., Zakharenkova, I., Alken, P., 2016. Prompt penetration electric fields and the
513 extreme topside ionospheric response to the June 22–23, 2015 geomagnetic storm as seen by
514 the Swarm constellation. Earth, Planets and Space. 152.
515 Astafyeva, E., Zakharenkova, I., Huba, J. D., Doornbos, E., Van den IJssel, J., 2017. Global
516 ionospheric and thermospheric effects of the June 2015 geomagnetic disturbances: Multi-

26
https://doi.org/10.5194/angeo-2022-18
Preprint. Discussion started: 5 September 2022
c Author(s) 2022. CC BY 4.0 License.

517 instrumental observations and modeling. Journal of Geophysical Research: Space Physics.
518 122(11):716-742. https://doi.org/10.1002/2017JA024174.
519 Blanc, M., Richmond, A. D. 1980. The ionospheric disturbance dynamo. Journal of Geophysical
520 Research. 85(A4):1669. https://doi.org/10.1029/ja085ia04p01669.
521 Binod, A., Dahal, S., and Chapagain, N. P. 2017. Study of field-aligned current (FAC),
522 interplanetary electric field component (Ey), interplanetary magnetic field component (Bz), and
523 northward (x) and eastward (y) components of geomagnetic field during supersubstorm, Earth
524 and Space Science. 4:257–274. 10.1002/2017EA000258.
525 Buchert, S., Zangerl, F., Sust, M., André, M., Eriksson, A., Wahlund, J-E., Opgenoorth, H., 2015.
526 SWARM observations of equatorial electron densities and topside GPS track losses.
527 Geophysical Research Letters. 42:2088–2092. 10.1002/2015GL063121.
528 Buresova, D., Lastovicka, J., Hejda, P., Bochnicek, J., 2014. Ionospheric disturbances under low
529 solar activity conditions. Advances in Space Research. 54(2):185-196.
530 https://doi.org/10.1016/j.asr.2014.04.007.
531 Calabia, A., Jin, S., 2019. Supporting Information for "New modes and mechanisms of long-term
532 ionospheric TEC variations from Global Ionosphere Maps". 10.1029/2019JA027703. Zenodo.
533 https://doi.org/10.5281/zenodo.3563463
534 Calabia, A, Jin, S., 2020. New modes and mechanisms of long-term ionospheric TEC variations
535 from Global Ionosphere Maps", Journal of Geophysical Research. Space Physics. 125(6).
536 10.1029/2019JA027703.
537 Calabia, A, C Anoruo, S Munawar, C Amory-Mazaudier, Y Yasyukevich, C Owolabi, and S Jin
538 (2022). Low-Latitude Ionospheric Responses and Coupling to the February 2014 Multiphase
539 Geomagnetic Storm from GNSS, Magnetometers, and Space Weather Data, Atmosphere, 13,
540 518. doi:10.3390/atmos13040518.
541 Chartier, A. T., Mitchell, C. N., Miller, E. S., 2018. Annual occurrence rates of ionospheric polar
542 cap patches observed using Swarm. Journal of Geophysical Research: Space Physics.
543 123:2327–2335. https://doi.org/10.1002/2017JA024811.
544 Christensen, A. B., et al. (2003), Initial observations with the Global Ultraviolet Imager (GUVI)
545 in the NASA TIMED satellite mission, J. Geophys. Res., 108(A12), 1451,
546 doi:10.1029/2003JA009918.
547 Cole, K. D., 1966. Magnetic storms and associated phenomena. Space Science Reviews. 5:699–
548 770.
549 Fang, H., Weng, L., Sheng, Z., 2012. Variations in the thermosphere and ionosphere response to
550 the 17–20 April 2002 geomagnetic storms. Advances in Space Research. 49:1529–1536.
551 Fagundes, P. R., Cardoso, F. A., Fejer, B. G., Venkatesh, K., Ribeiro, B. A. G., Pillat, V. G., 2016.
552 Positive and negative GPS-TEC ionospheric storm effects during the extreme space weather
553 event of March 2015 over the Brazilian sector. Journal of Geophysical Research: Space Physics.
554 121:5613–5625. 10.1002/2015JA022214.
555 Fejer, B. G., Jensen, J. W., Su, S., 2008. Quiet time equatorial F region vertical plasma drift model
556 derived from ROCSAT-1 observations. Journal of Geophysical Research: Space Physics.
557 113:A5. https://doi.org/10.1029/2007JA012801.
558 Fuller-Rowell,T.J., Codrescu, M. V., Moffelt, R.J., Quegan. S., 1994. Response of thermosphere
559 and ionosphere to geomagnetic storms. Journal of Geophysical Research. 99:3893-391.
560 Fuller-Rowell, T. J., Millward, G. H., Richmond, A. D., Codrescu, M. V., 2002. Storm-time
561 changes in the upper atmosphere at low latitudes. Journal of Atmospheric and Solar-Terrestrial
562 Physics. 64(12–14):1383–1391. https://doi.org/10.1016/s1364-6826(02)001013.

27
https://doi.org/10.5194/angeo-2022-18
Preprint. Discussion started: 5 September 2022
c Author(s) 2022. CC BY 4.0 License.

563 Fuller-Rowell, T. J., 2011. Storm-time response of the thermosphere-ionosphere system.

564 Aeronomy of the Earth’s Atmosphere and Ionosphere. 2:419–435).
565 https://doi.org/10.1007/978-94-007-0326-1.
566 Gao, Q., Liu, L., Zhao, B., Wan, W., Zhang, M., Ning, B., 2008. Statistical study of the storm
567 effects in middle and low latitude ionosphere in the East-Asian sector, Chinese Journal of
568 Geophysics. 51:435–443.
569 Hargreaves, J. K., 1992. The Solar-Terrestrial Environment. Cambridge University Press.
570 https://doi.org/10.1017/CBO9780511628924.
571 Heelis, R. A., 2004. Electrodynamics in the low and middle latitude ionosphere. Journal of
572 Atmospheric and Solar - Terrestrial Physics. 66(10):825–838.
573 https://doi.org/10.1016/j.jastp.2004.01.034.
574 Hernández-Pajares, M., Juan, J., Sanz, J., 1999. New approaches in global ionospheric
575 determination using ground GPS data. Journal of Atmospheric and Solar-Terrestrial Physics.
576 61(16):1237–1247. 0.1016/S1364-6826(99)00054-1.
577 Huang, C. S., Foster., J. C., Kelley, M. C., 2005. Long-durationpenetration of the interplanetary
578 electric field to the low-latitudeionosphere during the main phase of magnetic storms. Journal
579 of Geophysical Research. 110:A11309
580 Jaggi, R. K., Wolf, R. A., 1973. Self-consistent calculation of the motion of a sheet of ions in the
581 magnetosphere. Journal of Geophysical Research. 78:2852.
582 Joshua, B., Adeniyi, J.O., Adimula, I. A., Abbas, M., Adebiyi, S.J., 2011. The Effect of Magnetic
583 Storm of May 2010 on the F2 Layer over the Ilorin Ionosphere. World Journal Young
584 Researchers. 1(5):71–78.
585 Kashcheyev, A., Migoya-Orué, Y., Amory-Mazaudier, C., Fleury, R., Nava, B., Alazo-Cuartas,
586 K., Radicella, S. M., 2018. Multivariable Comprehensive Analysis of Two Great Geomagnetic
587 Storms of 2015. Journal of Geophysical Research: Space Physics. 123(6):5000–5018.
588 https://doi.org/10.1029/2017ja024900.
589 Kassa, T., Damtie, B., 2017.Ionospheric irregularities over Bahir Dar, Ethiopia during selected
590 geomagnetic storms. Advances in Space Research. 60(1):121-129.
591 https://doi.org/10.1016/j.asr.2017.03.036.
592 Kelley, M. C., Fejer, B. G., Gonzales, C. A., 1979. An explanation foranomalous ionospheric
593 electric fields associated with a northwardturning of the interplanetary magnetic field.
594 Geophysical Research Letters. 6:301–304.
595 Klimenko, M.V., Klimenko, V.V., Ratovsky, K.G., Goncharenko, L.P.,2011. Disturbances in the
596 ionospheric F-region peak heights in the American longitudinal sector during geomagnetic
597 storms of September 2005, Advances in Space Research. 48(7):1184-1195.
598 https://doi.org/10.1016/j.asr.2011.06.002.
599 Klobuchar, J. A., 1987. Ionospheric Time-Delay Algorithm for Single-Frequency GPS Users.
600 IEEE Transactions on Aerospace and Electronic Systems. AES-23(3):325–31.
601 10.1109/TAES.1987.310829.
602 Kikuchi, T., Lühr, H., Schlegel, K., Tachihara, H., Shinohara, M., Kitamura, T.-I., 2000.
603 Penetration of auroral electric fields to the equator during a substorm. Journal of Geophysical
604 Research: Space Physics. 105(A10):23251–23261. https://doi.org/10.1029/2000ja900016.
605 Le H., Amory-Mazaudier, C., 2005. Magnetic signature of the ionospheric disturbance dynamo at
606 equatorial latitudes: “Ddyn.” Journal of Geophysical Research. 110(A10).
607 https://doi.org/10.1029/2004ja010578.

28
https://doi.org/10.5194/angeo-2022-18
Preprint. Discussion started: 5 September 2022
c Author(s) 2022. CC BY 4.0 License.

608 Lissa, D., Srinivasu, V. K. D., Prasad, D. S. V. V. D., Niranjan, K., 2020. Ionospheric response to
609 the 26 August 2018 geomagnetic storm using GPS-TEC observations along 80° E and 120° E
610 longitudes in the Asian sector. Advances in Space Research. 66:1427-1440.
611 Lu. G., 2006. High-Speed Streams, Coronal Mass Ejections, and Interplanetary Shocks: A
612 Comparative Study of Geoeffectiveness. Recurrent Magnetic Storms: Corotating Solar Wind
613 Streams. 167. https://doi.org/10.1029/167GM10.
614 Luhr, H., Maus, S., Rother, M., 2004. Noon-time equatorial electrojet: Its spatial features as
615 determined by the CHAMP satellite. Journal of Geophysical Research. 109:A01306.
616 10.1029/2002JA009656.
617 Mannucci, A.J., Tsurutani, B.T., Iijima, B.A., Saito, A., Gonzalez, W.D., Guarnieri, F.L., Kozyra,
618 J.U., Skoug, R., 2005. Dayside global ionospheric response to the major interplanetary events
619 of October 29–30, 2003 ‘‘Halloween Storms’’. Geophysical Research Letters. 32:L12S02.
620 10.1029/2004GL021467.
621 Mannucci, A.J., Tsurutani, B.T., Abdu, M.A., Gonzalez, W.D., Komjathy, A., Echer, E., Iijima,
622 B.A., Crowley, G., Anderson, D., 2008. Superposed epoch analysis of the dayside ionospheric
623 response to four intense geomagnetic storms. Journal of Geophysical Research. 113:A00A02.
624 10.1029/2007JA012732.
625 Martinis, C. R., Mendillo, M. J., and Aarons, J., 2005. Toward a synthesis of equatorial spread F
626 onset and suppression during geomagnetic storms. Journal of Geophysical Research.
627 110(A07306). 10.1029/2003JA010362.
628 Mene, N. M., Kobéa, A. T., Obrou, O. K., Zaka, K. Z., Boka, K., Amory-Mazaudier C., Assamoi.
629 P., 2011. Statistical study of the DP2 enhancement at the dayside dip-equator compared to low
630 latitudes. Annales Geophysicae. 2225-2233.
631 Mayaud, P. N., 1965. Analyse morphologique de la variabilité jour‐à‐jour de la variation
632 journalière régulière SR du champ magnétique terrestre, I‐ le système de courants Cp (régions
633 polaires et subpolaires). Annales Geophysicae. 21:369–401.
634 Nishida, A., 1968. Coherence of geomagnetic DP2 fluctuationswith interplanetary magnetic
635 variations. Journal of Geophysical Research. 73:5549–5559.
636 Prölss, G.W., and M. K. Bird (2010), "Physics of the Earth's Space Environment", Springer Verlag,
637 Heidelberg.
638 Rabiu, A. B., Yumoto, K., Falayi, E. O., Bello, O. R., MAGDAS/CPMN Group., 2011. Ionosphere
639 over Africa: Results from geomagnetic field measurement during international Heliophysical
640 Year IHY. Journal of Sun and Geosphere. 6:61–64.
641 Roma-Dollase, D., Hernández-Pajares, M., Krankowski, A., Kotulak, K., Ghoddousi-Fard, R.,
642 Yuan, Y., Li, Z., Zhang, H., Shi, C., Wang, C., Feltens, J., Vergados, P., Komjathy, A., Schaer,
643 S., García-Rigo, A., Gómez-Cama, J. M., 2018 .Consistency of seven different GNSS global
644 ionospheric mapping techniques during one solar cycle. Journal of Geodesy. 92(6):691–706.
645 https://doi.org/10.1007/s00190-017-1088-9.
646 Shah, M., Calabia, A. G., Tariq, M. A., Ahmed, J., Ahmed, A., 2020. Possible ionosphere and
647 atmosphere precursory analysis related to Mw > 6.0 earthquakes in Japan. Remote Sensing of
648 Environment. 239:111620. https://doi.org/10.1016/j.rse.2019.111620.
649 Sharma, S., P. Galav, N., Dashora, S., Alex, R. S., Dabas, R., Pandey. 2011. Response of low-
650 latitude ionospheric total electron content to the geomagnetic storm of 24 August 2005. Journal
651 of Geophysical Research. 116:A05317. 10.1029/2010JA016368.
652 Shao, X., Guzdar, P. N., Milikh, G. M., Papadopoulos, K., Goodrich, C. C., Sharma, A.,
653 Wiltberger, M. J., Lyon, J. G., 2002. Comparing ground magnetic field perturbations from

29
https://doi.org/10.5194/angeo-2022-18
Preprint. Discussion started: 5 September 2022
c Author(s) 2022. CC BY 4.0 License.

654 global MHD simulations with magnetometer data for the 10 January 1997 magnetic storm
655 event. Journal of Geophysical Research: Earth Surface. 107.
656 Stankov, S.M., Jakowski, N., 2007. Ionospheric effects on GNSS reference network integrity.
657 Journal of Atmospheric and Solar-Terrestrial Physics. 69(4–5):485–499.
658 10.1016/j.jastp.2006.08.008.
659 Stankov, S.M., Warnant, R., Stegen, K., 2009. Trans-ionospheric GPS signal delay gradients
660 observed over mid-latitude Europe during the geomagnetic storms of October–November 2003.
661 Advances in Space Research. 43(9):1314–1324. 10.1016/j.asr.2008.12.012.
662 Stankov, S.M., Stegen, K., Warnant, R., 2010. Seasonal variations of storm-time TEC at European
663 middle latitudes. Advances in Space Research. 48:1318–1325.
664 Stening, R. J., 1985. Modeling the equatorial electrojet, Journal of Geophysical Research.
665 90:1705–1719. 10.1029/JA090iA02p01705.
666 Stening, R. J., 1995. Variations in the strength of the Sq current system. Annals of Geophysics.
667 13:627–632.
668 Tsurutani, B. T., Mannucci, A. J., Iijima, B. A., Sobral, J. H., Gonzalez, W. D., Gaurnieri, F. L.,
669 Tsuda, T., Saito, A., Yumoto, K., Fejer, B., Fuller-Rowell, T., Kozyra, J. U., Foster, J. C.,
670 Coster, A. J., Vasyliuas, V. M., 2004. Global dayside ionospheric uplift and enhancement
671 associated with interplanetary electric fields. Journal of Geophysical Research. 109:A08302.
672 Vankadara, R. K., Panda, S. K., Amory-Mazaudier, C., Fleury, R., Devanaboyina, V. R., Pant, T.
673 K., Jamjareegulgarn, P., Haq, M. A., Okoh, D., Seemala, G. K., 2022. Signatures of Equatorial
674 Plasma Bubbles and Ionospheric Scintillations from Magnetometer and GNSS Observations in
675 the Indian Longitudes during the Space Weather Events of Early September 2017. Remote
676 Sensing. 14:652. https://doi.org/10.3390/rs14030652.
677 Warnant, R., Lejeune, S., Bavier, M., 2007. Space Weather influence on satellite based navigation
678 and precise positioning. Space Weather – Research towards Applications in Europe. 129–146.
679 Wei, Y., Pu, Z., Hong, M., 2010. Long-lasting goodshielding at the equatorial ionosphere.
680 Geophysical Research Letters. 115:A12256.
681 Xiong, C., Stolle, C., Luhr, H., Park, J., Fejer, B. G., Kervalishvili, G. N., 2016a. Scale analysis of
682 equatorial plasma irregularities derived from Swarm constellation. Earth, Planets and Space.
683 68(1). https://doi.org/10.1186/s40623-016-0502-5.
684 Yamazaki, Y., Kosch, M. J., 2015. The equatorial electrojet during geomagnetic storms and
685 substorms. Journal of Geophysical Research: Space Physics. 120(3). 10.1002/2014JA020773.
686 Zaourar, N., Amory‐Mazaudier, C., Fleury, R., 2017. Hemispheric asymmetries in the ionosphere
687 response observed during the high‐speed solar wind streams of the 24–28 August 2010.
688 Advances in Space Research. 59(9):2229–2247. https://doi.org/10.1016/j.asr.2017.01.04.
689

690

691

692

693

694

30