Variability of Ionosphere Over Indian Longitudes to a Variety of Space Weather Events During December 2006

This paper highlights the impact of intense solar events over India during 3–20 December 2006. Ionospheric effects of a major solar flare (X9) on December 5 (10:35 UT) have been investigated by using dayside and nightside magnetometer data, dayside ionosondes, and dayside GPS vTEC observations. On the next day, a stream of fast solar wind hits the magnetosphere, causing a HILDCAA (High Intensity Long Duration Continuous Auroral Activity) preceded by moderate geomagnetic storm. The origin and characteristics of a positive ionospheric storm which occurred over Tirunelveli (TIR, geomagnetic latitude: −0.18°N) in the recovery phase of storm due to simultaneous presence of enhanced O/N2 and WEJ or weakened EEJ during the HILDCAA (7th and 8th of December) is investigated. Subsequently, on December 14, the most powerful CME since the Halloween event impacts the Earth, and three SSCs are recorded on December 14, 16, and 18. The variability of the ionosphere over the Indian longitude sector due to these intense space weather fluctuations is presented by utilizing the magnetometers, ionosonde, GPS vTEC, and satellite‐based observations in the same region. This study reports the influence of prompt penetration of the magnetospheric convection electric field and the disturbance dynamo on several key ionospheric and magnetic parameters within the Indian longitude sector.

convection in high latitude regions during geomagnetic events result in strong equatorward neutral winds, prompt penetration of electric fields (PPEF) (Vasyliunas, 1970), that can also modify the electrodynamics and compositions of the equatorial to low-latitude ionosphere.Large scale traveling ionospheric disturbances (LSTIDs), and disturbance dynamo electric field (DDEF) can be two of the primary by-products of these disturbances in equatorial to low latitude regions (Blanc & Richmond, 1980;Fujiwara et al., 1996;Mendillo, 2006).A combination of storm-induced compositional changes (O/N 2 ratio), ionospheric electric field perturbations (PPEF and DDEF), thermospheric wind, and TID can affect the distribution of ionospheric plasma density by generating positive and negative ionospheric storms over equatorial to mid-latitude regions (Fuller-Rowell et al., 1994;Prölss, 1993;Ramsingh et al., 2015;Rishbeth, 1998;Rout et al., 2019).Exchange of plasma between the ionosphere and plasmasphere due to ionospheric upflow during geomagnetic storms can result in a decrease in the loss rate of electrons, which can also perturb the ionospheric electron density (Lemaire & Gringauz, 1998).
Multiple X-class solar flares (5,6,13,and 14 December) events occurred between the period of 3 and 20 December 2006.Response of the GPS (Global Positioning system) signals to the powerful solar radio bursts (SRBs) associated with these X-class solar flares has been investigated in the past (Afraimovich et al., 2009;Carrano et al., 2009;Cerruti et al., 2008).If flare flux is large enough in L-band frequency range with Right-Hand Circular Polarization (RHCP), SRBs associated with the flare might have an impact on GPS performance in terms of reduced carrier to noise (C/N) ratio of GPS signals in sunlit hemisphere.
An additional atmospheric ionization in the sunlit hemisphere during the solar flare events also causes geomagnetic effects.Solar flare effects (Sfe) are sudden variations in Earth's magnetic field (crochet-like shape), which is mainly attributed to the enhanced conductivity of D-region or boundary between D and E-region (below the quiet time dynamo region).However, it was calculated that Sfe is not a simple increase in diurnal variation (Sq).The center of the Sfe current system was found to be a bit higher in latitude (3°-4°N) and 1 hr of local time eastward in comparison to Sq (Curto, 2020;Curto et al., 1994;Gaya-Piqué et al., 2008).In some cases, Sfe can cause magnetic changes in the opposite sense of Sq (reversed Sfe) due to displacement in the ionospheric current system along with the center of its location in both the hemispheres.It was also found that reversed Sfe happens mostly during equinox between 10 and 12 LT with a weaker correlation with solar activity compared to the Sfe.
The progression of December 13 CME in terms of Sun-Earth connection has been studied using MHD-models, in situ, and remote measurements (Kataoka et al., 2009;Liu et al., 2008).The CME was also associated with X3.5 flare with a speed of about 1,775 km/s near the Sun.It produced SRBs, an interplanetary shock (about 1,030 km/s at 1AU), and an intense geomagnetic storm on Earth in subsequent days (Dst-index of about −147 nT) due to magnetic cloud (MC) embedded in the interplanetary CME.The ionospheric and thermospheric response to the initial, and main phase of this geomagnetic storm on 14-15 December has been investigated using both observational and model data sets (Lei et al., 2008;Pedatella et al., 2009).During the initial phase of the storm, a positive ionospheric storm effect occurred in the American longitude sector to the shock associated with the CME.Long lasting positive ionospheric storm in the EIA (Equatorial Ionization anomaly) crest region was also observed over the Pacific Ocean starting from the main phase of the storm for more than 12 hr.It was estimated that the combination of PPEF, equatorward neutral wind, and thermospheric compositions were the source of these positive storms, and the equatorward neutral wind may have played an important role in maintaining it.
In this paper, the thermosphere-ionosphere (T-I) perturbations over the Indian longitude sector due to various space weather events which occurred during 3-20 December 2006 have been investigated.Both ground and space-based observations of key ionospheric and magnetic parameters (magnetometers, ionosonde, GPS-TEC, COSMIC, CHAMP, TIMED/GUVI) have been employed in this study.The data sources and data processing used in this study are presented in Section 2. The solar disturbances and the general geophysical context including the solar wind characteristics and the geomagnetic indices are presented in Section 3. In Section 4, the understanding based on the multi-instrument data of key ionospheric parameters over the Indian longitude region due to solar forcing is presented.The new results obtained by using the recent advances such as wavelet analysis of magnetic field fluctuations due to ionospheric electric fields are presented in Section 5 and subsequently the conclusions are outlined in Section 6.

Data Sets and Data Processing
In this study, the response of geomagnetic field, and ionospheric parameters to the continuously occurring multiple space weather events (flares, high intensity long duration continuous auroral activities (HILDCAAs), CMEs  2 shows the starting, maximum, and end UT time of all the four flares with their classes.The effect of a powerful (X9) solar flare on 5th of December (10:35 UT), in the ionospheric D and E-region is investigated by using 10 magnetometer stations over the globe (Table 1, Figure 1; Black shaded region represents the nightside at 10:30 UT on 5th of December 2006 and black scatter plot represent the magnetic equator).The first 10 stations of Table 1 are also generally used in the estimation of SYM-H, and ASY-H indices.One minute cadence magnetic field data of these stations with their base values for everyday are obtained from International Real-time Magnetic Observatory Network (2006, https://www.intermagnet.org/).The TIR magnetometer data set is obtained on request from Indian Institute of Geomagnetism (IIG, https://www.iigm.res.in/).The base value for TIR magnetometer station is obtained from taking the mean of magnetic field at about midnight during  the four International Quiet Days (IQDs; 2, 4, 27, and 29 December) (World Data Center for Geomagnetism, Kyoto (2006), https://wdc.kugi.kyoto-u.ac.jp/qddays/).The base value of each station is subtracted from the measured magnetic field for every day to obtain the variation ΔH (nT).
To characterize the solar, interplanetary, and geomagnetic conditions in this study, solar wind and geomagnetic parameters such as interplanetary magnetic field (IMF) B 0 , IMF B z , solar wind velocity (v), SYM-H, ASY-H, AU, AL, and AE indices are obtained from OMNIWeb interface (Goddard Space Flight Center, OMNIWeb Plus (2006), https://omniweb.gsfc.nasa.gov/).Ap, and Kp indices available at International Service of Geomagnetic Indices (2006, https://isgi.unistra.fr/)are also used to estimate the intensity of the geomagnetic storms.
The O/N 2 ratio calculated using the intensity ratio of oxygen 135.6 nm and N 2 Lyman-Birge-Hopfield emissions from TIMED/GUVI are utilized to show the compositional changes over Indian longitude region throughout the considered period of our study (TIMED/GUVI (2006), https://spdf.gsfc.nasa.gov/pub/data/timed/guvi/).The O/ N 2 ratio provided by TIMED/GUVI was referenced to N 2 column density of 10 17 cm −2 to minimize its uncertainties (Strickland et al., 2004;Zhang et al., 2004).

Processing of Magnetic Field Data
The fluctuations in Earth's equatorial magnetic field (ΔH) in the dayside during geomagnetic storms can be expressed by: where S q is the average variance of the Earth's magnetic field on designated magnetic quiet days (ΔH quiet , EEJ emphasis on dayside magnetometers); D mag denotes the magnetic disruption caused by the magnetospheric currents (ring currents, and tail current, etc.) (Fukushima & Kamide, 1973;Le et al., 2004).D mag can be approximated on the dayside at each station by calculating SYM-H cosΦ (magnetic dip latitude Φ is used to standardize the ring current impact at associated station), since it is the largest contributor to dayside magnetospheric currents during extreme space weather conditions.
D iono is the magnetic disturbance on the dayside due to disturbed ionospheric currents.At low latitudes D iono is related to 2 main physical processes: (a) prompt penetration of the magnetospheric convection electric field PPEF (Vasyliunas, 1970) at the origin of the magnetic disturbance DP 2 (Nishida, 1968), and (b) the ionospheric disturbance dynamo (Blanc & Richmond, 1980), DDEF, at the origin of the magnetic disturbance D dyn (Le Huy & Amory-Mazaudier, 2005).
To analyze the distinct periods of waves and their existence, a spectrum analysis on D iono is performed with continuous wavelet transform using Morlet wavelet for the duration of our investigation (Nava et al., 2016).To extract the D dyn magnetic disturbance we use the method of wavelet + semblance (Younas et al., 2022).

vTEC Processing
We used a standard procedure for processing GPS measurements (Hofmann-Wellenhof et al., 1992;Schaer, 1999) and for the conversion of slant TEC to vertical TEC, we use the single layer mapping function (Schaer, 1999), see Azzouzi et al. (2015) for more details.We use the daily measurements of the GPS stations (Table 1) of the International GNSS network (IGS) which are in RINEX 30s format.
The slant Total Electron Content (sTEC) is calculated by combining the two pseudo-range measurements at frequencies f1 and f2 (Hofmann-Wellenhof et al., 2001).The satellite differential biases (DCB) are obtained from the values published on the website of the University of Bern (ftp.aiub.unibe.ch/CODE/).The receiver bias is estimated by the daily difference between these STEC values on those recalculated from the GIM/CODG model for all the GPS paths and for elevation angles greater than 30° to avoid multi-paths.The use of the single-layer ionospheric model mapping makes it possible to calculate the vTEC (v for Vertical) from the calibrated sTEC considering the altitude of the IPP point (Ionospheric Positioning Point) at 420 km.

Global Sun-Earth Contexts
To specify the Sun-Earth context we characterize the Solar disturbances (solar flare, coronal hole, and CME) in Section 3.1 and the solar wind variations and geomagnetic indices in Section 3.2.

The Solar Events: Solar Flare, Coronal Hole, and CME
All the four X-class solar flares that occurred during our period of investigation (5, 6, 13, and 14 December 2006, respectively) are shown in the first row of Figure 2 with a sudden variation in brightening captured by GOES-12 SXI.The origin of all these flares was active sunspot region 10,930.The abrupt increase in electromagnetic emissions during a flare at different wavelengths (from radio waves to gamma rays) can be responsible for the rapid ionospheric disruptions and subsequent perturbation of the Earth's magnetic field known as magnetic crochet.Solar flares are facets of energy released from closed field lines on the Sun, while coronal holes comprise open field lines to space where solar plasma can radially escape as high-speed streams (HSS).Soft X-rays, and EUV pictures by GOES-13 SXI, and SOHO/EIT (second row of Figure 2) show solar coronal holes as dark patches, because of their colder and less dense plasma characteristics than surroundings.As the fast HSS associated with a coronal hole interact with background solar wind a co-rotating interaction region (CIR) forms with fluctuating IMF-B z polarity, which can result in low to moderate space weather activity after the interaction with Earth's magnetosphere (Tsurutani et al., 2006).
CMEs are the most dramatic eruptions in the solar corona that have been identified as main contributors of interplanetary disruptions.SOHO/LASCO C2, and SOHO/EIT both observed a strong halo CME accompanied by X3.4 flare on December 13 (bottom row of Figure 2, as it was also observed by GOES-12, first row, 3rd figure from the left).It was the largest halo CME since the Halloween storms (Liu et al., 2008).After interacting with Earth's magnetosphere, this CME produced an intense geomagnetic storm on 14th-15th of December.This paper goes into further detail on the impact of solar flares, coronal holes, and the CME on the Earth's ionosphere and magnetosphere.
Figure 3 depicts the solar X-ray fluxes (Watts.m−2 ) observed by the GOES-11 Solar X-ray Imager during 3-20 December that included four X-class flares.The blue, and red curves in the figure denote the short wavelength (0.05-0.3 nm), and long wavelength (0.1-0.8 nm) X-ray irradiances, respectively.The vertically green dotted lines represent the maximum of these X-class flares, in which the X9 flare on December 5 is the most powerful of the four in terms of X-ray output.The effects of the SRBs associated with these flares on the ionosphere, and GPS functioning has been studied in the past (Afraimovich et al., 2009;Carrano et al., 2009;Cerruti et al., 2008).The starting, maximum, and the end time of these flares are also provided in Table 2.The sudden brightening at the Sun about the maximum of these flares was also captured by the GOES-12 SXI, which is also shown in the first row of Figure 2 (from left to right, X9: 5 December (10:37 UT), X6.5: 6 December (18:45 UT), X3.4: 13 December (2:40 UT), X1.5: 14 December (22:15 UT)).2).It can be seen in Figure 4 that the geomagnetic conditions on the Earth are very quiet before this flare, which could be due to the particular location of the flare (solar eastern limb, Figure 2).However, the effects of solar energetic particles (SEPs) associated with this flare were observed at Venus, and Mars (Futaana et al., 2008).The effect of this flare on Earth's magnetic field around the globe is discussed in the next section.Enhanced geomagnetic activity, as observed in geomagnetic indices (Figures 4c-4f) just after the sudden storm commencement (SSC) at 23:59 UT on 5th of December could be due to the coronal holes shown in Figure 2 (5 December).The CIR associated with southward IMF-B z (Figure 4a) caused by the coronal hole could have played an important role in creating a low to moderate geomagnetic activity on 6th of December.This geomagnetic activity was then followed by High Intensity Long Duration Continuous Auroral Activities (HILDCAAs) on 7th and 8th of December due to continuous fluctuation in the polarity of IMF-B z associated with the CIR.Before the geomagnetic conditions  3.

Results
This section is organized taking into account the different solar disturbances and their effects on Earth's ionosphere.In Section 4.1 we analyze the impact of the X9 Solar flare on the Earth's magnetic field, the F2 region of the ionosphere and the vTEC.In Section 4.2, we analyze the impact of CIR and CME on the ionosphere, for this we study the F2 region of the ionosphere, the vTEC and the O/N 2 ratio.In Section 4.3 we also investigate the impact of CIR and CME on the Earth's magnetic field and we use signal processing techniques to separate the signatures of different physical processes (PPEF and DDEF).Figure 5 shows the magnetic field variations on December 5 observed in 10 magnetic stations (Table 1, Figure 1) located on the dayside and nightside, at the bottom of the figure the magnetic indices SYM-H and ASY-H are also plotted.Based on the local time of each considered magnetometer station (Figures 5a-5j) at the time of the flare (10:35 UT, represented in green), the stations in the sunlit hemisphere (AMS, CLF, HER, SJG, TIR, ABG) detected solar flare effects (Sfe) as a rapid drop or rise (crochet-like shape) in the horizontal component of the Earth's magnetic field compare to the     During the solar flare events, the enhanced conductivity of D-region or boundary between D and E-region (below the quiet time dynamo region) due to additional atmospheric ionization in sunlit hemisphere plays an important role in Sfe variations presented above (Curto, 2020;Curto et al., 1994;Gaya-Piqué et al., 2008;Rastogi et al., 1999;Sripathi et al., 2013).It was   10.1029/2023SW003595 9 of 21 calculated that the Sfe are not just intensified diurnal variation (Sq) as the location of the Sfe current system can be a bit higher in latitude (3°-4°N) and 1 hr of local time eastward in comparison to Sq.The relocation of the ionospheric current system and its center may sometimes generate magnetic changes in the opposite sense of Sq (reversed Sfe) in both hemispheres.The ionospheric effects of this flare are discussed further ahead.

Impact of the X9-Flare on vTEC
The total electron content (TEC) is an essential property of the total ionization level in Earth's atmosphere.
In the GPS era, global measurements of vertical TEC (vTEC) can assist us in estimating the effects of space weather disturbances on the ionosphere (Tsurutani et al., 2009).Figure 6a illustrates the vTEC measurements by the GPS satellites with different PRN (Pseudo random noise) number at different GPS receivers (TLSE-prn06, HRAO-prn09, for middle latitudes; IISC-prn24, DGAR-prn02, for low latitudes with positions provided in Table 1) on 5th of December during the occurrence of X9 flare.The three blue vertical lines at each receiver represent the beginning, maximum, and end UT hour, respectively, of the X9 flare (Table 2).As all the receivers shown in the figure are in the dayside at the time of the flare, it can be seen that after about 7-9 min from the beginning of the flare, TEC starts an abrupt increase at each station, and reaches a total increment of 1.2-1.4tecu (1 tecu = 10 16 el/m 2 ) for middle latitudes and less than 1.0 tecu at low latitudes.These values are significantly lower than the increases of 25 tecu preserved during the Halloween events of 2003 (Tsurutani et al., 2005).This increase in TEC at each station is due to the enhanced/additional ionization caused by the extreme ultraviolet and X-rays bands that radiated during the solar flare.This enhanced ionization of iono- spheric D and E-layers during large flares is important but the region is limited in altitudes so the increase of TEC is of few TEC units.Normally, the maximum is near the sub-solar point but in our case, the increase is maximum in middle latitudes.The increase in TEC lasts about 50 min and therefore is maintained well beyond the end of the solar flare.
In the D layer, the frequency of collisions with the neutral particles is also important, an increase in the number of electrons strongly increases the absorption of the radio waves which cross it and in particular the HF waves (3-30 MHz).For the sounding of the ionosphere by ionosondes, the consequence is an absence of echoes from the first HF frequencies which can lead to a total blackout on all frequencies for several tens of minutes during strong flare eruptions (Chakraborty et al., 2018;Tao et al., 2020).Figures 6b and 6c illustrate this phenomenon.On both the stations of Roquetes (Spain) and TIR (India), echoes from regions E and F are present at 10:15 UT.At 10:30 UT (16:00 IST), there is a total radio fadeout with even the disappearance of jammers.In general, the foF2 are above the median at the beginning of the month (from 03 to 13) and below at the end of the period.We see two diurnal maximums on foF2 (e.g., days 04, 06, 08, 12, 13, 16, 20), one in the morning and the other in the afternoon with a decrease around local noon ("bide-out") (Chen et al., 2021;Kane, 2006).This is typical of low latitudes.
On the top panel, the variations of the B z component of the solar wind magnetic field represents the signature of the different solar events.The oscillatory variations reveal the presence of CIR.A first CIR arrived on December 6 and peaked at a speed ∼700 km/s on December 7 (see Figure 4 panel b), a second CIR arrived on December 10 which peaked on December 12 at a speed of ∼700 km/s (see Figure 4 panel b).Then on December 14 and 15, the impact of a CME and negative B z component of the solar wind magnetic field was observed (see Figure 4 panel a).At the end of December 18, a new third CIR was also observed.The two CIRs produced an increase in the foF2 component at TIR, on December 7 and 8 for the first CIR, and on December 12 and 13 for the second CIR.The increase of foF2 was observed to be 2 MHz on December 7 and about 1 MHz on December 8, 12 and 13.The foF2 reached up to about 10.5 MHz over Tirunelveli (Mag.Lat.= −0.18°N)near local noon on 7th.The impact of the CME was very different.On December 15th, a decrease of foF2 by 2 MHz as well as the presence of an oscillation of the foF2 were observed, and on December 16th a small increase in foF2 was also observed.
The first CIR produced a positive ionospheric storm and the CME a negative storm followed by a positive ionospheric storm.The CME caused an intense magnetic storm with strong negative SYM-H values.Geomagnetic storms generally produce a negative ionospheric storm in summer hemisphere and positive ionospheric storm in winter hemisphere (Fuller-Rowell et al., 1996) except for strong geomagnetic storms where there is a negative phase in the winter hemisphere too.This scenario seems to fit the observations.Concerning the decrease of foF2 (negative ionospheric storm) observed on December 15th, it is also well explained by UT time of the beginning of the storm (Fuller-Rowell et al., 1994).
The last panel of Figure 7A represents the variation in NmF2 (computed from foF2) measured by TIR ionosonde (lime color).The magenta triangle and black asterisk scatter points indicates the NmF2 measured by COSMIC-1 and electron density measured by CHAMP (at about 360 km), respectively, in the vicinity of TIR ionosonde (7-12°N, and 65-80°E) for the same time period.It can be seen in the Figure that the COSMIC-1 and CHAMP observations agree with the TIR ionosonde to a large extent throughout the considered period of the study.For example, the positive ionospheric storm on 7th of December is evident in all three instruments, along with the negative ionospheric storm observed on 15th of December.
Another interesting feature of the figure is the presence of Traveling Ionospheric Disturbances (TIDs) in the foF2 just after the main phase of an intense geomagnetic storm on 15th of December.A zoom part of Figure 7A on 15th of December is also shown in Figure 7B. Figure 7B depicts the TIDs present in the early UT hours indicated by the arrows.The presence of both the peaks are at about 2.25 UT, and 5.5 UT, respectively, which is just before the UT time when the same TIDs are observed near IISC in measured vTEC (Figure 8b).There are also sudden enhancements in h'F2 during the earlier mentioned main phase of moderate storm on 6th of December, and the main phase of intense storm near the midnight of 14th of December.

Impact of CIR and CME on vTEC
To comment on the temporal variations of vTEC during this period, we use the daily solar flux over 10.7 cm measured in Ottawa as a proxy (figure not shown) for ionizing solar radiation.It presents a maximum of 100 sfu around 05-06/12 and decreases steadily toward a minimum of 70 sfu (−30%) on 20/12.For the 3 stations (JMSM, IISC, and HYDE) the maximum of vTEC diurnal trend is observed around 08-09/12 and it decreases to reach a minimum on 29/12 (figure for x-axis limit 21-31 December not shown, Figure S1).The date of the maximum occurs a few days after that of the solar flux indicating a reaction time of the ionosphere.We find in the daily evolution of the vTEC that of the 27-day rotation of the Sun.On the other hand, for the DGAR station in the southern hemisphere, the evolution is different with a maximum on 03-04/12, a minimum level on 15/12 and a second diurnal maximum on 26/12.The level of ionization at this station strongly depends on the presence of a southern equatorial crest and its geographical position.
Figure 8a is divided into five panels; the first (from the top) panel shows the temporal variation of SYM-H index (nT) throughout 3-20 December 2006 (red color).The last four panels show the temporal variation in vTEC (tecu) at JMSM, HYDE, IISC, and DGAR GNSS stations, respectively (Table 1).The vTEC levels increase as the day begins, peaking before sunset and then decreasing as the day concludes.The five vertical black lines represent the UT occurrence of SSCs for the same period (Table 3).It is noticeable in the figure that there is a significant decrease in vTEC at DGAR (Table 1; Mag.Lat.= −15.48N, near southern equatorial ionization anomaly (EIA) crest) at about local noon on 7th of December.As it will be discussed later that, there is a westward electrojet (WEJ) present at the same time (near local noon) in the Indian longitude region, which can cease the fountain effect, and hence the less vTEC observed at the same time compared to previous quiet day (4th of December).The JMSM station (Mag.Lat.= 19.24N) is also near the northern EIA crest region, however, there is no depletion in vTEC there.The enhanced O/N 2 ratio near the JMSM station (Figure 9e) may compensate for the effect of WEJ, which results in negligible perturbation of vTEC compared to quiet days at JMSM near local noon on 7th of December.Another evident change in vTEC occurred on 15th of December, when DGAR station showed a depleted vTEC, which could be due to large depletion of O/N 2 ratio at the same location (Figure 9m).An interesting feature of Figure 8a is the presence of traveling ionospheric disturbances (TIDs) in the vTEC measured at all the mentioned GNSS stations on 15th of December.Figure 8b shows the zoom on vTEC profiles at each of the four stations (color coded) on 15th of December from 0 to 8 UT.The zoom on the two maximum of 15 December on the fof2 (Figure 7B) goes in the same direction as the vTEC (Figure 8b) and we do not see any notable movement on h′F (evolution in slight descent).We can hypothesize that these undulations are only at the level of the F2 region and do not concern the upper ionosphere or little above 400 km.

Impact of CIR and CME on the Ratio O/N 2
The thermospheric-ionospheric response to the geomagnetic storm time joule heating and accompanying thermospheric dynamics changes is also controlled by the local time at the start of the storm and the season (Fuller-Rowell  , 1994, 1996).The compositional changes (mostly in O/N 2 column density) in the thermosphere caused by storm-induced dynamics play a major role in determining the positive/negative phases of ionospheric storms.9f).The TIMED/GUVI data set isn't available for the whole globe at the time, however it is evident in Figure 9m that, there was a depletion in O/N 2 ratio again in equatorial region which even penetrated toward the winter hemisphere just after the CME induced intense geomagnetic storm on December 15th.After that, the O/ N 2 ratio starts coming back to its pre-storm values (Figures 9p-9r).

Impacts of CIR and CME on the Earth's Magnetic Field
During magnetic storms, two large scale disturbances processes of ionospheric electric currents act at low latitudes.(1) The penetration of the magnetospheric convection electric field transmitted from the auroral zone, PPEF (Vasyliunas, 1970) (2) The disturbed ionospheric dynamo (Blanc & Richmond, 1980) generated by the dissipation of Joule energy from auroral electrojets, DDEF.These two physical processes affect the magnetic field at low latitudes and generate magnetic field disturbances DP2 for PPEF (Nishida, 1968) and Ddyn for DDEF (Le Huy & Amory-Mazaudier, 2005).
In this section we will separate the two disturbances DP2 and Ddyn using wavelet and semblance techniques.
Figures 10a and 10b show the local variations in dayside horizontal component of Earth's magnetic field (ΔH, blue) along with the quiet time (S q , green), and ionospheric disturbances (D iono , red) over Indian longitude regions (Tirunelveli (TIR), and Alibag (ABG)) from 3 to 20 December 2006.Some of the flares and SSC (according to Tables 2 and 3 of flares and SSC) are visible at these stations.Figure 10 panel (c) shows the temporal variations in SYM-H, and ASY-H indices corresponding to the same UT.During the geomagnetic storms (SYM-H/ASY-H variations), almost similar patterns in the ΔH component of the geomagnetic field (black line) are seen at both the stations at the same UT.A substantial fall in ΔH can be seen due to enhanced westward ring current during the main phases of both the CIR and CME induced geomagnetic storms on 6th and 15th of December, respectively.However, it can also be seen in the figure that a small difference in latitude, and local time for both the stations can still play a crucial role in determining the magnitude of ΔH decrement.The magnitude difference of ΔH variations is also present in D iono (red) variations (Equations 1 and 2), where its pattern seems to be the same at both the stations with difference in magnitude.As the storm starts to recover due to continuously reducing ring current effects (both SYM-H, and ASY-H starts coming back to its controlled values), ΔH variations also start coming back to their pre-storm values.D iono is the magnetic signature of disturbed ionospheric electric currents circulating on the dayside.At low latitudes D iono is the sum of two magnetic disturbances DP 2 and D dyn (Equations 1-3 in Section 2).The magnetic perturbation DP 2 is composed of short fluctuations (1-3 hr).The signature of the magnetic perturbation D dyn is clearly identified as an anti-Sq circulation on the dayside.At the TIR station we can clearly see the fluctuations of the DP 2 signal every day as well as the anti-Sq signature of D dyn on December 7, 13, 14, 15, and 19, which is discussed further ahead.These days correspond to the impact of CIRs and CMEs.The ABG Station presents a bit different result concerning the anti-Sq disturbance of D dyn observed on the same days.Figure 11 represents the power spectrum of continuous wavelet transformation for D iono during 3-20 December 2006 at both the stations ABG (top panel), and TIR (bottom panel), respectively.Presence of about 12-16 hr period oscillations can clearly be seen at TIR station on 6th-7th of December with high confidence (black contour lines) during the first CIR induced moderate storm and HILDCAA event associated with it.These waves are also present for 14-17 December during and after the CME induced intense geomagnetic storm at both TIR, and ABG.As we are only considering the local dayside for the calculation of D iono , these periodicities could represent the diurnal component associated with the ionospheric DDEF (Blanc & Richmond, 1980;Le Huy & Amory-Mazaudier, 2005;Nava et al., 2016).Short term oscillations of about 1-3 hr due to DP2 fluctuations during PPEF (Nishida, 1968;Vasyliunas, 1970) are also present throughout the storms, which may be associated with continuous IMF-B z fluctuations during the considered period of our study (Figure 7A, first panel).As can be seen in both the figures, waves of even larger periodicities are also present especially during and after the CME induced storm.
Figure 12 shows the Ddyn disturbance (Equation 3) computed by the wavelet + semblance method (Younas et al., 2022), from 3 to 20 December 2006.The top panel is for the ABG magnetic station and the bottom panel for the TIR one.As theorized by Blanc and Richmond (1980), the ionospheric disturbance dynamo (DDEF) causes a decrease in the day-time Sq currents near the equator.Therefore, an anti-Sq variation observed in the D iono can precisely be linked to Ddyn.In Figure 12, we see an increase in the period of 20-28 hr at ABG on 12, 13, and 17 December 2006, during the second CIR and after the CME.On the other hand, at TIR the signature of DDEF is observed on 6-7 and 12-17 December, during the two CIRs and the CME.It is interesting to notice that the computation of D dyn is in good agreement with the observations of D iono (Figure 10).Indeed, we observe in Figure 10 an anti-Sq variation simultaneous to the D dyn for the station of TIR.For the station ABG there is no anti-Sq circulation for the first CIR (December 6-7) and no D dyn .D dyn disturbance predicted by Blanc and Richmond (1980) varies with latitude.This disturbance is composed of two equivalent current cells (see Figure 4 of Blanc & Richmond, 1980).The fact that we do not observe equivalent disturbances at ABG and TIR means that these two stations are not on the same portion of the equivalent current cell.

Discussion
In this work we analyzed the impact of various solar events (Sfe, CME, CIR, HILDCAA etc.) on the ionosphere over Indian longitudes from December 3 to 20, 2006.Earlier studies have reported few of these variations separately, but no detailed study has been done on the ionosphere over Indian longitudes.We studied the day-to-day variability of ionospheric and magnetic parameters due to succession of the solar disturbances in the light of new scientific findings.We studied the appearances of crochets in the horizontal component of the Earth's magnetic field observed in the dayside magnetometers at about 10:35 UT on 5th of December due to the intense X-class (X9) solar flare.The impact of this solar flare is also observed in SYM-H and ASY-H indices (Figure 5).The impact in terms of magnitude of enhancement is larger on ASY-H index, compared to SYM-H index.Recently, Liu et al. (2021), using numerical simulations demonstrated "that solar flare effects extend throughout the geospace via electrodynamic coupling, and are not limited-as previously believed-to the atmospheric region where radiation energy is absorbed." The SYM-H and ASY-H indices allow us to know the symmetrical and asymmetrical components of the magnetospheric ring current.The solar flare signatures on SYM-H and ASY-H show that the solar flare impacts on the magnetosphere, this is in agreement with Liu et al. (2021).It is important to note that the magnetometers located on the night side (Figure 5, right side) do not show any impact from the X-class (X9) Sfe.Gaya-Piqué et al. (2008) studied the X-class (X9) Sfe.They made the equivalent current system due to this Sfe.The equivalent currents are calculated using a two-dimensional model of the ionospheric electric currents closing on the hemisphere which is not the reality.The numerical simulation is based on a three-dimensional approach with a connection between the atmospheric region (where the radiation energy is absorbed) and the magnetosphere.Previous studies (Anad et al., 2016;Mazaudier & Venkateswaran, 1985;Van Sabben, 1961) have already raised the problem of the impossibility of closing regular ionospheric currents only in the ionosphere of one hemisphere.With the work of Liu et al. (2021), the subject is open.
The impact of the X9 solar flare is also seen in ionosphere as a sudden radio fade-out in ionograms (ROQUETES and TIR), and enhanced vTEC in both low to mid-latitude regions due to enhanced ionization of D and E-ionospheric region.Interestingly, the increase in vTEC is larger in mid-latitude (TLSE, and HRAO) compared to low latitudes (IISC and DGAR) (Figure 6a).This could be due to larger increase in conductivity near local noon (TLSE, and HRAO) by the solar flare in comparison to the local late afternoon (IISC, and DGAR) (Figure 1).As can also be seen in Figure 5, the magnetic crochets near TLSE, and HRAO at CLF and HER, respectively, are also stronger in comparison to TIR, and ABG.
Large-scale traveling ionospheric disturbances (LSTIDs) was also observed just after the CME induced intense storm near early UT hours on 15th of December by TIR ionosonde and GPS TEC data sets (Figures 7B and 8b).The TEC perturbations are smaller at DGAR as compared to other stations, which could be due to smaller background TEC in summer (Southern) hemisphere compared to winter (Northern) hemisphere on the same day.Hayashi et al. (2010) also reported multiple LSTIDs near the same UT hours propagating both southward, and northward.However, as can be seen in Figure 8b, based on GPS TEC observation, it's a bit hard to interpret the direction of the LSTID, as the presence of this TID is occurring in late UT hours at both DGAR (southern) and JMSM (northern) from IISC/HYDE.After a multi-instruments analysis using both ground (ionosonde) and space-based (CHAMP/COSMIC satellites) observations over TIR (magnetic equator), it is found that, there is a positive ionospheric storm over TIR during 7th of December (Figure 7A).At the same time, enhanced Joule heating due to HILDCAA (Figure 4) in high latitude region, and westward electric currents at the equator (Figure 10) are also present.The TIMED/GUVI observed (at about 10 LT) O/N 2 depletion in mid to low-latitude of summer (southern) hemisphere along with O/N 2 enhancement in equatorial to low-latitude of winter (northern) hemisphere in Indian longitude region (over TIR, magnetic equator) (Figure 9e).Which could be due to the enhanced equatorward wind with background summer to winter wind caused by continuous auroral Joule heating (HSS/CIR + ongoing HILDCAA event) present on 7th of December.The presence of WEJ at the same time (7th of December) can suppress the fountain effect over magnetic equator (TIR), and with the enhanced O/N 2 over TIR, it can result in higher electron density over magnetic equator (TIR) than quiet times.The second CIR induced positive ionospheric storm over TIR is also observed during 12-13 December 2006.
The equatorial to low-latitude ionospheric response to CME induced intense geomagnetic storm in Indian longitude region is also presented by investigating the variation of vTEC (Figure 8), and TIR ionosonde observed h′F2 & foF2 (Figure 7A).It is found that there is a negative ionospheric storm on 15th of December with TID occurrence in electron density profile observed by TIR ionosonde, and in the vTEC profile over DGAR station.Which could be due to the compositional (O/N 2 ) changes (Figure 9m) caused by the UT time of the beginning of the storm.This negative ionospheric storm is also followed by a positive one on 16th of December.The response of the low latitudes to the CME are well explained by the numerical simulations of Fuller-Rowell et al. (1994, 1996).
We have analyzed the variations of the H component in two Indian stations TIR and ABG (Figure 10).We observe the signature of two known physical processes: (1) The DP 2 (Nishida, 1968), magnetic signature of the penetration of the magnetospheric convection electric field PPEF (Vasyliunas, 1970).PPEF is the transmission of the magnetospheric convection electric field from auroral zone to low latitudes.(2) D dyn (Le Huy & Amory-Mazaudier, 2005), magnetic signature of the Ionospheric disturbance Dynamo DDEF (Blanc & Richmond, 1980).In the auroral zone the Joule heating disturbs the thermospheric circulation of winds and generate storm induced meridional winds from the pole to the equator (Amory- Mazaudier et al., 1985;Mazaudier & Bernard, 1985;Mazaudier & Venkateswaran, 1990).These disturbed winds produce electric fields and electric current by dynamo action.Using Wavelet and Wavelet analysis + semblance analysis we separate these 2 magnetic disturbances (Figures 11  and 12).
Forty years ago, the Saint-Santin incoherent scatter sounder located at middle latitudes allowed observation of the penetration of the magnetospheric convection electric field PPEF (Mazaudier, 1985), by in situ measurements.PPEF is a phenomenon in universal time whose magnetic signature DP 2 is clearly identifiable.In Figures 10 and 11, this phenomenon causes fluctuations of the magnetic field of small periods (<3hr) simultaneous to ABG and TIR.Many articles (more than a thousand) have been published on PPEF.The DP2 magnetic disturbance (Nishida, 1968) due to PPEF, was found experimentally before the theory made by Vasyliunas (1970).
The existence of the disturbed circulation of thermospheric winds during magnetic storms has also been observed with the incoherent scatter sounder (Mazaudier & Bernard, 1985).The amplitude of the wind disturbances observed was in perfect agreement with the simulation of Blanc and Richmond (1980) magnetic disturbance.The wind disturbance by dynamo effect produced disturbed electric field DDEF.The DDEF disturbance was measured by the Jicamarca incoherent scattering sounder located at the magnetic equator (Fejer et al., 1983).But it was only 25 years later that the D dyn magnetic disturbance associated with this phenomenon was isolated for simple cases (Le Huy & Amory-Mazaudier, 2005).The D dyn magnetic disturbance due to perturbed thermospheric winds is a local weather phenomenon that causes an anti-Sq circulation at equatorial latitudes.The extraction of magnetic disturbance associated with perturbed thermospheric winds has been developed during the last decade and different techniques have been used, moving averages (Fathy et al., 2014), Wavelet analysis (Nava et al., 2016), and Wavelet + semblance analysis (Younas et al., 2022).All these studies have highlighted the anti-Sq circulation at low latitudes.However, Blanc and Richmond's model predicts the DDEF disturbance and its associated magnetic signature at the different latitudes of a hemisphere and this is not always anti Sq (see Figure 4 of Blanc and Richmond (1980)).The difference observed between the perturbation D dyn at ABG and TIR (Figure 12) can be explained by Blanc and Richmond's model.

Conclusions
In this study we have presented a detailed multi-instrument study of the ionospheric and magnetic variations over India during 3-20 December 2006.This period included an intense solar X9 flare, strongest CME after Halloween storm, HSS with CIR.Many studies have reported the effects of the individual events on the Earth's ionosphere.Our study brings a new understanding of day -to -day variability of the Indian longitude ionosphere under the influence of these space weather events (Solar Flare, CME, and CIR).The ionospheric impact of CIR occurred during December 2006 was not reported earlier.It is only after the IHY project in the 2010s that the study of CIR induced ionospheric perturbation began.The IHY project has changed the geophysical approach to a heliophysical approach and leads to the analysis of the ionospheric variability as a function of the variability of the Sun.This remains as the main motivation of this study.
The current studies are generally focused on understanding the effects of individual events (Solar Flare CME and CIR) and statistics on all phenomena of the same nature.With space weather, we must follow the impact of all solar events on a daily basis and also aim to develop the prediction of their impact on the ionosphere as meteorology does for the lower atmosphere.This is one of the challenges of Space Weather.In our study we have also addressed the impact of solar and geomagnetic events on the variations of the Earth's magnetic field directly related to the circulation of ionospheric electric currents.We have mainly highlighted the magnetic disturbance D dyn related to the ionospheric dynamo with the new technique of wavelet + semblance.Recall that this disturbance was extracted from the data in 2005 and again it is during the 2010s that the study of this disturbance has developed.
Solar flare induced enhanced conductivity in D and E region ionosphere resulted in small but sudden increase in dayside vTEC, appearance of magnetic crochets, and a total radio blackout at the time of the flare (10:35-10:40 UT, 5 December).Presence of reversed or weakened EEJ due to D dyn over Tirunelveli on 7 and 8 December along with the O/N 2 enhancement at the same time resulted in a positive ionospheric storm over the Indian magnetic equator.Enhanced Joule heating induced dynamics in the thermosphere on 15 December resulted in D dyn over TIR, and depletion in O/N 2 at the same time could have played an important role in the observed negative ionospheric storm over Indian longitude regions.
We have been able to fully explain the day-to-day variation of the ionosphere over India with physical phenomena (PPEF, DDEF, change in composition, TID etc.) observed for a long time.The challenge of Space Weather is to integrate all these phenomena in Heliophysical science and also to define their impacts on our widely used technologies such as GNSS (Global Navigation satellite system).
This study also showed the signature of the X9 solar flare of December 5 on the SYM-H and ASY-H indices.A question arises, is this a real impact as a recent numerical simulation shows or is it related to the calculation of SYM-H and ASY-H indices?We sincerely thank NASA/SPDF OMNI-Web database, TIMED/GUVI, World Data Center (WDC) for Geomagnetism, Kyoto-Japan, INTERMAGNET database, and IGSI database for their openly available data sets used in this study.We also thank technical team of IIG for operating/ maintaining the Ionosonde and magnetometer network.We are grateful for the SOHO/EIT, SOHO/LASCO, GOES-11/ GOES-12/GOES-13/SXI, COSMIC-1, and CHAMP observed data sets utilized in this study.We also acknowledge the use of CDDIS/UNAVCO provided data sets used in this study.One of the author AKR thanks Ministry of Education, New Delhi for the financial support in the form of graduate assistantship.
during 3-20 December 2006 are investigated.The data sets, as well as the techniques used in data processing and analysis for this study, are detailed as follows.GOES-12 solar X-ray imager (SXI) is used to capture the pictures associated with sudden brightness for all the 4 solar flares (NOAA National Centers for Environmental Information (2022), https://www.ncei.noaa.gov/data/).GOES-13 SXI along with SOHO/EIT (Extreme ultraviolet Imaging Telescope, https://www.ias.u-psud.fr/eit/movies/)are used to indicate the dark regions in the solar corona also known as coronal holes.The SOHO satellite images by EIT, and LASCO C2 (C2 camera of the Large Angle and Spectrometric Coronagraph) are also used to capture the strong X3.4 class solar flare along with a coronal mass ejection (CME) that appeared to be headed toward Earth on 13th of December 2006 (https://soho.nascom.nasa.gov/pickoftheweek/old/13dec2006/).Solar X-ray fluxes measured by GOES 11 are used to observe and identify the occurrence of solar flares with their classes throughout the considered period (NOAA National Centers for Environmental Information (2015), https://www.ngdc.noaa.gov/stp/satellite/goes/dataaccess.html).Table

Figure 1 .
Figure 1.Map with the location of the magnetometers (blue stars), ionosondes (red and brown stars) and GPS receivers (pink stars).

Figure 4
Figure 4 illustrates the characteristics of the solar wind and the geomagnetic parameters.Panels from the top to bottom are plotted as (a) the solar wind magnetic field B 0 and the B z component in nT, (b) the speed of the solar
of the X9-Flare on the Earth's Magnetic Field

Figure 3 .
Figure 3. GOES-11 observational data on the solar flares from 3 to 20 December 2006.
variations.The stations chosen in Figures5a-5jare the stations over the globe (except for TIR), which are generally used in the estimation of SYM-H/ASY-H indices.It can be seen in Figures5k and 5lthat the presence of this sudden disturbance in the magnetic field is also present in ASY-H, and SYM-H indices as a sudden increase.The effect on another M1.8-class at about 8 UT (lime vertical dotted line) on the same day can also be seen at TIR station as a sudden enhancement in earth's magnetic field.The effects of other three solar flares mentioned in

Figure 4 .
Figure 4. Solar wind and magnetic parameters from 3 to 20 December 2006, (a) solar wind magnetic field B 0 and B z in nT, (b) solar wind speed in km/s, (c) magnetic index Ap in nT, (d) Kp index, (e) SYM-H and ASY-H in nT, (f) AE, AU and AL indices in nT.

Figure 5 .
Figure 5. Impacts of the solar flare on the Earth's magnetic field on 5 December 2006.

Figure 6 .
Figure 6.(a) Impact of the solar flare on GPS receivers TLSE, IISC, HRAO, and DGAR.(b) Impact of the solar flare on the ionosonde of Roquetes/Spain.(c) Impact of the solar flare on the ionosonde at TIR/India.

Figure 8 .
Figure 8. (a): On the top sub-figure the variations of SYM-H, and vTEC measured in various GPS stations in Asia: JMSM, HYDE, IISC, and DGAR from December 3 to 20.(b) Zoom on the vTEC observations on December 15.

Figures
Figures 9a-9r shows the TIMED/GUVI observed O/N 2 variations in Indian latitude-longitude regions from about 9:30 to 12:30 LT throughout 3 to 20 December 2006.It can be seen in the Figure that after the first CIR induced moderate storm on December 6th, a continuous auroral activity (HILDCAA) after the main phase of the storm (Figures4e and 4f) resulted in O/N 2 enhancement in equatorial-low latitude regions of winter (northern) hemisphere with depletion in mid-low latitude summer (southern) hemisphere on 7th and 8th December (Figures9e and 9f).The TIMED/GUVI data set isn't available for the whole globe at the time, however it is evident in Figure9mthat, there was a depletion in O/N 2 ratio again in equatorial region which even penetrated toward the winter hemisphere just after the CME induced intense geomagnetic storm on December 15th.After that, the O/ N 2 ratio starts coming back to its pre-storm values (Figures9p-9r).

Figure 10 .
Figure 10.From 3 to 20 December 2006, panel (a) Variations in dayside Earth's magnetic field at TIR (black curves) with the Sq variations (green curve) and D iono (blue curves), panel (b) same curves for ABG, panel (c) SYM-H and ASY-H indices.

Table 1
Geographic and Geomagnetic Coordinates of the Stations

Table 2 Table of X
Solar Flares

Table 3
Table of SSCs (Sudden Storm Commencements)