Local time dependent response of postsunset ESF during geomagnetic storms

Transcription

1 JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 113,, doi: /2007ja012922, 2008 Local time dependent response of postsunset ESF during geomagnetic storms S. Tulasi Ram, 1 P. V. S. Rama Rao, 1 D. S. V. V. D. Prasad, 1 K. Niranjan, 1 S. Gopi Krishna, 1 R. Sridharan, 2 and Sudha Ravindran 2 Received 3 November 2007; revised 24 January 2008; accepted 25 March 2008; published 23 July [1] Development or inhibition of ESF during magnetically active periods has been an important space weather topic of interest during the recent past in view of its applications in the satellite based navigational systems. Particularly, the postsunset period exhibits significant variability for storm time development of ESF versus longitude. In this paper, we report the results of a multi-instrumental (ground based and space-borne) and multistation study on the development/inhibition of postsunset ESF during five moderate to intense geomagnetic storms occurred during the low and descending phase of the solar activity period, It has been observed that, the prompt penetration of eastward electric fields into low latitudes and subsequent development of ESF occurred in all longitudinal sectors where the local time corresponds to postsunset hours during the entire main phase of the storm. In this paper, we show the development of plasma bubble irregularities over a wide longitudinal extent of 92 owing to the dusk time penetration of eastward electric fields into low latitudes. Either the sudden increase in AE-index and/or a marked decrease in Sym-H index may be used as proxies to determine the occurrence as well as the time of penetration of electric fields into equatorial and low latitudes. However, in such cases where the AE-index does not represent any sudden increase, the dsymh/dt seems to be the better index to determine the time of penetration. In this paper, is also presented an interesting case where the prompt penetration eastward electric fields dominated the existing strong westward electric fields and subsequently caused the onset of spread-f and scintillations at both VHF (244 MHz) as well as L-band (1.5 GHz) frequencies. Citation: Tulasi Ram, S., P. V. S. Rama Rao, D. S. V. V. D. Prasad, K. Niranjan, S. Gopi Krishna, R. Sridharan, and S. Ravindran (2008), Local time dependent response of postsunset ESF during geomagnetic storms, J. Geophys. Res., 113,, doi: / 2007JA Introduction [2] The nighttime equatorial F region often contains plasma density irregularity structures with scale sizes ranging from several hundreds of kilometers to a few centimeters, which manifest themselves as spread-f on ionograms, plume like structures in HF radar maps, intensity bite-outs in airglow intensity measurements and scintillations on amplitude as well as phase of the VHF and UHF signals from satellites, which are commonly referred to as Equatorial Spread-F (ESF) irregularities [Woodman and Lahoz, 1976; Aarons et al., 1980a; Aarons, 1982; Yeh and Liu, 1982; Basu and Basu, 1985]. The generalized Rayleigh-Taylor instability process involving gravity antiparallel to the electron density gradient in the presence of electric fields and neutral winds in the equatorial ionosphere 1 Space Physics Laboratories, Department of Physics, Andhra University, Visakhapatnam, India. 2 Space Physics Laboratory, VSSC, Trivandrum, India. Copyright 2008 by the American Geophysical Union /08/2007JA is believed to be the causative mechanism for the development of ESF. Different variabilities of ESF (such as spatial, temporal and solar activity) have been extensively investigated by several researchers during the past four to five decades [Chandra and Rastogi, 1972a; Woodman and Lahoz, 1976; Sastri et al., 1979a; Fejer and Kelley, 1980; Abdu et al., 1981; Aarons, 1993; Fejer et al., 1999; Hysell and Burcham, 2002]. Although many aspects of ESF have been reasonably explained based on the past investigations, the day-to-day and storm time variability of ESF are yet to be understood comprehensively. [3] The effect of geomagnetic storms on the equatorial and low latitude ionosphere in the context of development/ inhibition of ESF is an important space weather concern and has drawn great attention during the recent past. The development/inhibition of ESF during geomagnetic storms is mainly controlled by the perturbations in the zonal electric field at the equator due to variable nature of coupling between high and low latitudes. This high latitude low latitude coupling has been thought of in terms of two basic mechanisms namely, the solar wind-magnetospheric dynamo [Senior and Blanc, 1984; Spiro et al., 1988] and the iono- 1of19

2 spheric disturbance dynamo [Blanc and Richmond, 1980]. The former is due to dynamic interactions between the solar wind and the magnetosphere that leads to changes in the polar cap potential causing prompt penetration of electric fields into low latitudes [Kelley et al., 1979; Fejer and Scherliess, 1997] giving rise to perturbations in the zonal electric fields. Whereas, the later is due to global thermospheric circulation induced by joule heating at auroral latitudes that generates long-lived (several hours) electric field disturbances at mid and low latitudes by ionospheric wind dynamo action [Scherliess and Fejer, 1997]. The local time dependence of the polarity and amplitude of electric field perturbations due to these two processes, in fact, determines the favorable or unfavorable conditions for the development of Equatorial Spread-F (ESF) irregularities at any given location. [4] Earlier Studies on Storm Time ESF. The early study by Lyon et al. [1960], based on ionosonde data taken during the International Geophysical Year (IGY), showed that the percentage occurrence of ESF was drastically reduced during the periods of geomagnetic activity. In 1970s and 1980s, advancements were made as many researchers [Chandra and Rastogi, 1972a, 1972b; Rastogi et al., 1978; Rastogi and Woodman, 1978; Chandra and Vyas, 1978] were able to distinguish between the influence of storms on postsunset and postmidnight ESF. A general consensus emerged from the studies of Rastogi and Woodman [1978], Sastri et al. [1979b], Aarons et al. [1980a], Dabas et al. [1989], and Bowman [1982] that the postmidnight ESF tend to be triggered by geomagnetic storms. Later, Kelley and Maruyama [1992] and Hysell and Burcham [1998] have shown that the geomagnetic activity in the post midnight sector can initiate ESF due to anomalous upward reversals in the vertical plasma drift under the action of eastward electric fields of Ionospheric Disturbance Dynamo (IDD) and also often due to penetration of eastward electric fields (undershielding) associated with northward turning of IMF Bz [Kelley et al., 1979]. [5] During 1990s, a pivotal advance in ESF studies during the geomagnetic storms came from the work of Aarons [1991] that explored the presence or absence of irregularities against the local time (longitude) of storm s peak excursion. Aarons [1991] and Abdu et al. [1995] have shown that, if the onset of the geomagnetic activity occurs well prior to local sunset, inhibition of ESF takes place due to the superposition of westward electric fields of Ionospheric Disturbance Dynamo (IDD) on the normal prereversal enhancement (PRE) thereby reducing the post sunset upward ExB drift. Later, Fejer et al. [1999] have shown that prolonged geomagnetic activity during daytime, in general, causes a large reduction in the occurrence of post sunset ESF at Jicamarca but only during equinoxes of solar maximum. [6] On the other hand, Abdu et al. [1995] have shown that, the rapid increase in high latitude convection around local sunset hours causes a prompt penetration of eastward electric field augmenting the normal PRE, resulting in a dramatic development of ESF including the formation of plasma bubbles. Recently, Basu Su. et al. [2001] and Basu S. et al. [2001, 2005] have reported that there is an abrupt onset of VHF and L-band scintillations due to the prompt penetration of high latitude electric fields (eastward) into low latitudes, in the longitude sector for which the early evening period corresponds to the time of rapid decrease (50 nt/h or larger) in the Sym-H/Dst index. Huang et al. [2002] have shown that the rate of change of Sym-H/Dst index is larger than 5 nt/h, for duration of more than 2 h would trigger the development of irregularities, possibly due to the penetration of high-latitude electric fields. [7] An empirical model developed by Fejer and Scherliess [1997] using the equatorial vertical plasma drifts obtained from incoherent scatter radar measurements at the Jicamarca Radio Observatory also shows disturbance upward vertical drifts during the postsunset hours due to prompt penetration effects, the pattern which is also confirmed recently by case studies [Martinis et al., 2005, and references therein]. [8] Fejer et al. [1990] have shown that the duration of prompt penetration phase can last for more than 1 h because of changes in the magnetic field configuration resulting from the changes in the polar cap potential. Recently, Basu S. et al. [2007] have shown the penetration of electric fields causing rapid uplift of equatorial F-layer and the formation of bubbles over Atlantic sector first and also over Brazilian sector after 100 min as the DMSP F14 satellite marching from Atlantic to Brazilian sector during its successive orbits. This result indicates that the prompt penetration phase persisted for more than one and half hour. Further, Huang et al. [2005, 2006] have shown the penetration of electric field into the low latitude ionosphere without shielding continued for many hours as long as the magnetic activity continues to intensify and the IMF Bz remains southward. Their results consistently show the dayside ionospheric electric field (eastward) enhancements at middle and low latitudes during the entire main phase of the geomagnetic storm under the southward orientation of IMF Bz. [9] Therefore it is plausible to expect, if the main phase of the storm lasts for many hours, then the local dusk period that coincides with the prompt penetration phase occurs over a wide longitudinal extent. Hence the equatorial ionosphere over a wide longitudinal sector is susceptible for the development of ESF irregularities. In this paper, we show the presence of plasma bubble irregularities over a longitudinal width of 92 (one fourth of the Earth s equatorial ionosphere) due to the penetration of eastward electric fields into equatorial latitudes during the entire main phase of the geomagnetic storm. [10] Further, few earlier studies [Aarons et al., 1980a; Rastogi et al., 1981] have reported that the postsunset ESF was triggered during non-esf seasons, while it was inhibited during ESF seasons due to geomagnetic storm induced perturbations in the zonal electric field. Recently, Becker-Guedes et al. [2004] have shown that the geomagnetic storm acts as an inhibitor during high spread-f season and acts as an initiator during low spread-f season over Brazilian sector, possibly due to corresponding changes in the quiet and disturbed drift patterns during different seasons. [11] The present communication reports the results of a systematic study of the response of equatorial and low latitude ionosphere to five geomagnetic storms representing both equinoxes as well as solstices (both high and low ESF seasons), in the context of development/inhibition of ESF, during moderate to low solar activity period ( ). Present investigation further deals with an interesting case, where the anomalous increase in the equatorial F-layer height and subsequent occurrence of spread-f and scintillations during the postsunset hours due to prompt penetration 2of19

3 Table 1. Details of Observations Geographic Coordinates Station Latitude Longitude Dip Latitude Parameter(s) Ionospheric Sounders Trivandrum 8.5 N 76.5 E 0.7 N h 0 F, Spread-F SHAR 13.7 N 80.2 E 6.9 N h 0 F, Spread-F Hainan Is 19.4 N 109 E 13.4 N h 0 F, Spread-F Kwajalein Is 9.4 N 167 E 4.2 N h 0 F, Spread-F Okinawa 26.4 N E N fof2 Yamagawa N E N fof2 Kokubunji N E N fof2 Wakkanai N E N fof2 VHF and L-band Scintillations Waltair 17.7 N 83.3 E 11.6 N S4 - Index Magnetometers Tirunelveli 8.7 N 77.7 N 0.9 N EEJ Strength (DH T-A ) Alibagh 18.5 N 72.9 E 12.9 N GPS Receivers Trivandrum 8.5 N 76.5 E 0.7 N TEC Bhopal N E 18.5 N TEC On board of CHAMP satellite Planar Longmuir Probe Electron density of eastward electric field even under the presence of strong ambient westward electric fields of disturbance dynamo (DD) origin. 2. Details of Observations [12] The local time dependant effects of five geomagnetic storms on equatorial and low latitude ionosphere have been studied using both ground based and space-borne instruments over Indian, Chinese, Japanese as well as Pacific regions during the moderate to low solar activity period, Data from two similar digital ionosondes, one at an Indian equatorial station Trivandrum (8.5 N, 76.5 E, dip. latitude 0.7 N) and another at a low latitude station SHAR (13.7 N, 80.2 E, dip. latitude 6.9 N) have been used to observe the storm time perturbations in the zonal electric field by monitoring the variations in the virtual height of the equatorial F-layer (h 0 F) with respect to its quiet time variations. Simultaneous observations of VHF (244 MHz) and L-band (1.5 GHz) scintillations on the amplitude of radio beacon signals from two geostationary satellites namely, FLEETSAT (73 E) and INMARSAT (65 E) have also been carried out from a low latitude station Waltair (17.7 N, 83.3 E, dip. latitude 11.6 N). Further, the data from two magnetometers situated, one at the equatorial station (within the Electrojet region), Tirunelveli (8.7 N, 77.7 N, 0.9 N dip. latitude) and the another at an offequatorial station (outside the Electrojet region), Alibagh (18.5 N, 72.9 E, 12.9 N dip. latitude) is used to observe the geomagnetic storm time variations in the Equatorial Electrojet (EEJ) strength. The difference between the DH values (H-component of the magnetic field after subtracting the nighttime base level) at Tirunelveli and Alibagh, i.e., DH T-A (nt) is taken as a measure of the Equatorial Electrojet (EEJ) strength [Rusch and Richmond, 1973; Rastogi and Klobuchar, 1990; Anderson et al., 2002]. [13] Ionogram data from ionosonde/digisonde s at Hainan Is (19.4 N, 109 E, dip. latitude 13.4 N), Okinawa (26.4 N, E, N), Yamagawa (31.12 N, E, N), Kokubunji (35.42 N, E, N), Wakkanai (45.23 N, E, N), and Kwajalen Is (9.4 N, 167 E, dip. latitude 4.2 N) has been used to study the effect of geomagnetic storms at Chinese, Japanese and Pacific sectors. [14] The in situ electron density data measured by Planar Longmuir Probe (PLP) on board the low Earth orbiting (454 km) satellite CHAMP is also used to observe the electron density depletions which are the signatures of equatorial plasma bubbles (EPBs) and the behavior of the equatorial ionization anomaly (EIA) during the geomagnetic storm periods over different longitudinal zones. [15] Further, the Total Electron Content (TEC) data obtained from two dual frequency GPS receivers operated at Trivandrum (8.5 N, 76.5 E, dip. latitude 0.7 N) and Bhopal (23.17 N, E, dip. latitude 18.5 N) under Indian ISRO-GAGAN program have also been used in this study to examine the response of EIA during geomagnetic storm periods. The details of the observations made using various instruments at different locations are summarized in Table 1 for ready reference. 3. Occurrence of ESF During Geomagnetic Storms 3.1. Storm of February 2004 [16] Two recurrent geomagnetic storms have occurred during February 2004 and 9 12 March 2004 due to the high speed solar wind stream from a recurrent coronal hole that turned into geo-effective position on 9 February 2004 (CH80) and again on 8 March 2004 (CH84). Figure 1 shows the SOHO images of coronal holes and active regions on solar disk at 0000 h UT of 9 February 2004 (Figure 1a) and 8 March 2004 (Figure 1b) obtained from Solar Terrestrial Activity Reports ( 3of19

4 Figure 1. SOHO images of a recurrent coronal hole on (a) 9 February 2004 and (b) 8 March of19

5 Figure 2. Response of equatorial and low latitude ionosphere to the geomagnetic storm of February Time histories of (a) IMF Bz, (b) AE-Index, (c) Sym-H Index, (d) dsymh/dt, (e) virtual height (h 0 F) over Hainan Is, (f) virtual height (h 0 F) over Trivandrum, and (g) Equatorial Electrojet (EEJ) Strength. The solid and dotted vertical lines indicate the times of peak negative excursions of dsymh/dt that corresponds to local dusk hours at Hainan Is and Trivandrum respectively. The shaded region indicates the respective local dusk to dawn period. solar/coronal_holes.html). The effects of the first storm (11 12 February) on equatorial and low latitude ionosphere over Indian and Chinese sectors have presented in this section, whereas the effects of the second storm (9 12 March) are discussed in the next section (section 3.2). [17] Figures 2a, 2b, and 2c depict the variation of interplanetary magnetic field (IMF Bz), AE-Index and Sym-H Index respectively during February On 11 February 2004, the interplanetary magnetic field, IMF Bz turned southward around 0600 h UT and showed a strong negative component of 12 nt at 1040 h UT as may be seen from Figure 2a. During the same time, the AE-Index exhibits a sudden increase reaching a value of 900 nt (Figure 2b). A moderate geomagnetic storm has occurred, as the Sym-H Index Figure 2c) started decreasing from 1050 h UT gradually reaching a minimum value of 107 nt around 1740 h UT. With a view to examine the rate of enhancement in the ring current, the rate of decrease in Sym-H Index (dsymh/dt) has been computed for every 10 min interval and plotted against UT in Figure 2d. It can 5of19

6 Figure 3. Series of ionograms showing the spread-f from 2016 h LT of 11 February 2004 to 0946 h LT of 12 February 2004 over a low latitude station, Hainan Is (19.4 N, 109 E, dip. latitude 13.4 N). be seen from this figure, that the dsymh/dt exhibits a negative excursion of 7 nt/10 min at 1240 h UT and the IMF Bz is also southward during the same period. [18] In Figure 2e is presented the virtual height (h 0 F) variation over a low latitude station Hainan Is (19.4 N, 109 E, dip. latitude 13.4 N), the local time of which is 7 h 16 min ahead of UT. It can be seen from this figure that the h 0 F exhibits a pronounced increase around 1245 h UT, which corresponds to local post sunset hours (2001 h LT) over that station. This is precisely the time when that dsymh/dt exhibited a sharp decrease (Figure 2d). It may also be seen from this figure (Figure 2e) that the h 0 F on the previous day (i.e., 10 February 2004) as well as the next day (i.e., 12 February 2004) was limited to an altitude less than 230 km during the post sunset hours, whereas on the storm day, (i.e., 11 February 2004) the h 0 F has increased to a high value of 275 km around 2001 h LT. [19] The near simultaneity of sharp decrease in Sym-H Index and increase in h 0 F suggests that there is a prompt penetration of eastward electric fields into low latitudes around local dusk hours that augmented the normal F- region dynamo induced prereversal enhancement (PRE), thereby lifting the F-layer to higher altitudes and creating conditions favorable for the development of ESF. As such, an intense spread-f has been observed from the next ionogram, i.e., from 1300 h UT, which has continued to occur till 0230 h UT of the next morning. In Figure 3, is presented a series of ionograms showing the presence of spread-f over Hainan Is from 1300 h UT (2016 LT) to 0230 h UT (0946 h LT) of the subsequent day. [20] Figure 2f shows the h 0 F variation over the Indian equatorial station Trivandrum (8.5 N, 76.5 E, dip. latitude 0.7 N), the local time of which is 5h8minahead of UT. The thin line with scatter bars indicates the monthly mean quiet day average pattern of h 0 F. It may also be seen from this figure, during that evening the h 0 F over Trivandrum has also increased significantly from its quiet day value and reached to a maximum value of 400 km at 1450 h UT (2000 h LT). Subsequently, intense spread-f was observed, which also continued to occur till the predawn hours of the next day. During the same night, intense scintillations were also observed at 244 MHz over an Indian low latitude station, Waltair (17.7 N, 83.3 E, dip. latitude 11.6 N) during the period from 2010 h LT to 0330 h LT as may be seen from Figure 4. [21] During geomagnetic storms, the interaction between the solar wind and the magnetosphere under the southward orientation of interplanetary magnetic field (IMF Bz) causes 6of19

7 Figure 4. VHF scintillation record at 244 MHz on 11 February 2004 over an Indian low latitude station, Waltair (17.7 N, 83.3 E, dip. latitude 11.6 N). a change in the region 1 currents leading to a sudden increase in the dawn-to-dusk polar cap potential (F pc ). The region 2 currents associated with an inner magnetospheric electric field cannot change at the same rate and an imbalance situation occurs [Kikuchi et al., 2000, 2003]. During the main phase of magnetic storm, the region 1 field-aligned current is stronger than the region 2 fieldaligned current. Thus the high-latitude electric potential, usually shielded by the region 2 currents, can penetrate into low-mid latitudes and equatorial latitudes. This situation is generally known as breakdown of shielding or undershielding [Senior and Blanc, 1984; Kikuchi et al., 1996]. The prompt penetration electric field is eastward during the daytime to the dusk sector and westward in the midnight to dawn sector [Jaggi and Wolf, 1973; Spiro et al., 1988; Fejer et al., 1990]. Further, at the dusk sector, the prompt penetration of eastward electric field may add to the normal prereversal enhancement of zonal electric field (PRE) due to the neutral wind dynamo (F-region dynamo). The resultant zonal electric field is responsible for a large uplift of the ionosphere near the terminator and sets off the Rayleigh- Taylor instability [Ossakow, 1981; Sultan, 1996] leading to the development of ESF irregularities. The present investigation further confirms the development of ESF due the penetration of eastward electric fields into low latitudes during post sunset hours. [22] In the present study, the development of ESF and its sustenance for longer duration at Hainan Island (109 E longitudinal sector) as well as Trivandrum and Waltair (77 to 83 E longitudinal sector) during this moderate geomagnetic storm has an important space weather concern. The extended sustenance of spread-f and scintillations could be due to two reasons, (1) If the F-layer is maintained at sufficiently higher altitudes where the collision frequency is less, the ESF irregularities can sustain and continue to cause spread-f/scintillations over a long period, (2) On the other hand, ESF irregularities generated at a western location may drift across the observational location as they generally drift toward east with velocities of about 100 m/s [Aarons et al., 1980b; Rama Rao et al., 2005], causing Spread-F and scintillations to persist for longer durations. However, from Figures 2e and 2f, it may be seen that the h 0 F at Hainan Is as well as Trivandrum has sharply increased during postsunset hours and decreased within 1 to 2 h, suggesting that the observed long duration spread-f and scintillations are not due to the extended maintenance of F-layer at higher altitudes. [23] Further, it may be noticed from Figure 2c that the Sym-H index started decreasing (main phase) around 1050 h UT and gradually reached to a minimum value of 107 nt at 1740 h UT. Thus the range of longitudes where the local dusk coincides with the main phase of the storm extends over a wide longitudinal sector. Hence it can be understood that the penetration of electric fields into low latitudes and subsequent development of ESF can occur over a wide longitudinal sector. Therefore it is possible that the ESF irregularities may also develop over the longitudes to the west of the (present) observational location at a later time. These irregularities subsequently move over the observational location as they generally drift toward east causing Spread-F and scintillations to persist for longer durations. The development of ESF over Hainan Island at 1300 h UT and almost 2 h later at Trivandrum (1450 h UT) further justifies this argument. [24] On the subsequent day, i.e., 12 February 2004, during the recovery phase of the storm, the IMF Bz exhibits rapid fluctuations (Figure 2a) and the dsymh/dt also show large positive and negative excursions (Figure 2d) between 3 and 9 h UT. Figure 2g shows the Equatorial Electrojet (EEJ) strength [DH T-A (nt)] variation over 75 E longitudinal region. It may be seen from this figure that there is a strong reversal in the EEJ current at 0545 h UT (1045 h LT). [25] Figure 5 shows a series of ionograms from h LT on 12 February 2004 at Trivandrum. It can be noticed from this figure that the equatorial (q-type) Sporadic E (denoted as Es-q) suddenly disappeared between the successive ionograms of 1030 and 1045 h LT, (precisely, the time when the reversal in EEJ current has been observed) and reappeared after several hours at 1715 h LT. This sudden disappearance of q-type Es in equatorial ionograms and the reversal in the EEJ current is an evidence for the contribution of large westward electric fields of ionospheric disturbance dynamo (IDD) origin associated with this geomagnetic storm. Rastogi [1972a] has explained that Es-q configuration in the ionograms at equatorial latitudes is due to scattering by irregularities generated by gradient drift instability mechanism when the electron density gradient (dn/dh) and Hall field are in the same direction. On certain occasions when the primary horizontal electric field reverses to westward direction such as during geomagnetic storm activity, the Hall field becomes downward and the Esq irregularities disappear [Rastogi, 1972b]. Further, the post sunset enhancement of h 0 F at Trivandrum is reduced compared to its quiet day average value (Figure 2f), and consequently, the occurrence of Spread-F and scintillations is inhibited on this night Storm of 9 12 March 2004 [26] A recurring moderate geomagnetic storm has occurred during 9 12 March 2004 which has similar manifestations over Indian equatorial ionosphere, and is presented in Figure 6. In Figures 6a 6g is shown the IMF Bz, AE-Index, Sym-H Index, dsymh/dt, EEJ strength, h 0 F over Trivandrum and S4-Index of VHF and L-band scintillations over Waltair, respectively, as a function of Indian local time, which corresponds to 82.5 E meridian time. 7of19

8 Figure 5. Series of ionograms from 0830 to 1800 h LT on 12 February 2004 at Trivandrum. Note the disappearance of Es-q from 1045 to 1715 h LT, during the period of reversal in Electrojet current is observed (Figure 1g). [27] Following the southward turning of IMF Bz (Figure 6a), the AE-Index (Figure 6b) started increasing from 1500 h LT reaching a value of 1200 nt around 2200 h LT. A moderate geomagnetic storm has occurred as the Sym-H index started decreasing from 1648 LT of 9 March 2004 reaching a minimum value of 75 nt at 2335 LT as seen from Figure 6c. The dsymh/dt (Figure 6d) exhibits a negative excursion of 7 nt/10 min around 8of19

9 Figure 6. Response of Indian equatorial and low latitude ionosphere to the geomagnetic storm of 9 12 March Time histories of (a) IMF Bz, (b) AE-Index, (c) Sym-H Index, (d) dsymh/dt, (e) Equatorial Electrojet (EEJ) Strength (f) virtual height (h 0 F) over Trivandrum, and (g) S4-Index of VHF (black line) and L-band scintillations. The solid vertical lines indicate the time of peak negative excursion of dsymh/dt. The shaded region indicates the local dusk to dawn period h LT. The h 0 F at the equatorial station, Trivandrum (Figure 6f) shows a marked increase from its quiet day value during the same period and intense Spread-F echoes were observed from 1945 LT, which continued up to 0530 LT of the next morning. Also, a nearly simultaneous onset of VHF and L-band (Figure 6g) scintillations were observed from 1958 LT over Waltair. The VHF scintillations continued to occur up to 0300 LT of the next morning and the L-band scintillations also continued to occur for longer duration extending to the postmidnight hours which slowly decayed by about 0130 LT. [28] Further, due to prolonged geomagnetic activity of this storm, i.e., up to the noon hours of the next day (10 March 2004), a strong reversal in the EEJ current is observed at the equator as may be seen from Figure 6e. During this evening (10 March 2004), the post sunset enhancement in the h 0 F at the equator is suppressed (Figure 6f) and the occurrence of Spread-F and scintillations is inhibited (Figure 6g). [29] Similar ionospheric response has been observed during the subsequent couple of days (11 and 12 March 2004), where the geomagnetic storm activity continued to persist. The IMF Bz turned southward at 1800 LT and the dsymh/dt showed a marked decrease of 8 nt/10 min around 2030 h LT of 11th Mar 2004, as may be seen from Figures 6a and 6c. On this evening, the h 0 F at the equator (Figure 6f) increased to an altitude of 378 km at 2100 LT and subsequently strong Spread-F echoes were observed at 9of19

10 Figure 7. Response of Indian equatorial and low latitude ionosphere to the geomagnetic storm of May Diurnal variations of (a) IMF Bz, (b) SymH-Index, and (c) virtual height (h 0 F) over an Indian low latitude station, SHAR. The shaded region indicates the local dusk to dawn period. Trivandrum from 2115 LT, which continued till 0630 LT of the next morning (12 March 2004) as seen from Figure 6f. The VHF and L-band scintillations (Figure 6g) over Waltair also show a nearly simultaneous onset at 2114 LT and persisted for about half an hour, followed by another scintillation patch around 2300 LT at both VHF and L-band frequencies. The L-band scintillations start decaying from 0100 LT, whereas the VHF scintillations continued to occur till 0600 LT of the next morning (12 March 2004) but with reduced intensities (Figure 6g). [30] Further, a strong reversal in the EEJ current is observed at the equator on the subsequent day, i.e., 12 March 2004 (Figure 6e) due to the continued geomagnetic activity during the morning to noon hours. The post sunset enhancement of h 0 F at the equator is also limited to slightly less than its quiet day value, and therefore, the subsequent occurrence of Spread-F and scintillations is inhibited as seen from Figures 6f and 6g Storm of May 2005 [31] An intense geomagnetic storm has occurred during May 2005, the SC phase of which has occurred at 0230 h UT of 15 May. The main phase started around 0620 h UT and the Sym-H reached to a minimum value of 315 nt around 0825 h UT Response Over Indian Sector [32] Over Indian longitudinal sector, the main phase (rapid decrease in SymH) of the storm has occurred around local noon hours. Unfortunately, the h 0 F data of the equatorial station Trivandrum is not available on that day. Hence the h 0 F data of a nearby low latitude station on the northern side of the equator, SHAR (13.7 N, 80.2 E, dip. latitude 6.9 N) is considered and plotted as a function of local time along with IMF Bz and SymH index and presented in Figure 7. The thin line with scatter bars indicates the monthly mean quiet day variation of the h 0 F over SHAR for the month of May It can be seen from this figure (Figure 7c) that the h 0 F did not show any significant enhancement during the post sunset hours and no spread- F was observed during that evening. The VHF and L-band scintillation observations over Waltair also did not show any postsunset scintillations during the same evening. [33] However, during the post midnight hours, the h 0 F shows quite significant increase between 0200 to 0400 h LT, possibly due to the delayed disturbance dynamo electric fields, as the polarity of these electric fields are normally eastward during the postmidnight to early dawn hours [Blanc and Richmond, 1980; Scherliess and Fejer, 1997]. Subsequently, range type spread-f echoes were observed at SHAR between 0230 h to 0515 h LT as indicated in the figure Response Over Japanese Sector [34] Over the Japanese longitudinal sector, the main phase of this storm has coincided with the local afternoon to predusk hours (1515 to 1720 h LT). In Figure 8, is presented the diurnal variation of fof2 (critical frequency of F2 layer) at Wakkanai, Kokubunji, Yamagawa and Okinawa spanning from mid latitudes to low-mid latitudes over 10 of 19

11 Figure 8. Response of ionosphere over the Japanese midlatitude region to the geomagnetic storm of May Time histories of (a) AE-index, (b) Sym-H index, (c) fof2 over Wakkanai, (d) fof2 over Kokubunji, (e) fof2 over Yamagawa, and (f) fof2 over Okinawa. The thin line with scatter bars in Figures 8c to 8f indicates the monthly mean quiet day mean variation of fof2 over the respective stations. Japanese sector. The thin line with scatter bars indicates the monthly mean quiet day variation of fof2 over the respective stations. It can be seen from this figure, that the fof2 values are much reduced compared to their quiet day mean values from 2000 h LT at all the four stations and a significant negative F2 layer storm has prevailed for the next 24 h. It is widely accepted that the marked storm time depressions in fof2 at midlatitude locations are due to increase in chemical loss brought about by the thermospheric composition changes (enhancement in molecular to atomic concentration ratio) because of modifications in the global thermospheric circulation due to storm time heating at high latitudes [Chandra and Herman, 1969; Matuura, 1972; Rishbeth, 1975; Sastri and Titheridge, 1977; Prolss, 1977]. This result further suggests that the Ionospheric Disturbance Dynamo (IDD) is active over the Japanese midlatitude sector Response Over Pacific Sector (Kwajalein Is) [35] The main phase of this storm (rapid decrease in Sym- H Index) has coincided with the local dusk hours over a pacific equatorial station Kwajalein Is (9.4 N, 167 E, dip. latitude 4.2 N), the local time of which is 11 h 9 min ahead of UT. The response of the equatorial F-region over Kwajalein Is during this geomagnetic storm is presented in Figure 9. It can be seen from Figure 9d, that the dsymh/ dt exhibits a sharp negative excursion of 94 nt/10 min at 1739 h LT. In Figures 9e and 9f are shown, the virtual height of the F-layer (h 0 F) and its rate of change (dh 0 F/dt) 11 of 19

12 Figure 9. Response of equatorial and low latitude ionosphere over Kwajalein Is to the geomagnetic storm of May Time histories of (a) IMF Bz, (b) AE-Index, (c) Sym-H Index, (d) dsymh/dt, (e) virtual height (h 0 F) over Kwajalein, and (f) rate of change of h 0 F (dh 0 F/dt). The solid vertical line indicates the time of peak negative excursion of dsymh/dt. The shaded region indicates the local dusk to dawn period. over Kwajalein Is, respectively, as a function of local time. It can be seen from these figures, that the h 0 F (Figure 9e) has increased almost instantaneously and reached to a maximum value of 495 km at 1909 h LT with a maximum rate of increase of about 100 m/s (Figure 9f) at 1809 h LT. Subsequently, intense spread-f echoes were observed from next ionogram, i.e., from 1819 h LT, which continued to occur beyond the local postmidnight hours (0309 h LT) as may be seen from Figure 10. [36] With a view to examine the longitudinal extent that is affected by the storm time electric fields through the manifestation in the development of ESF, the in situ electron density measured by the Planar Longmuir Probe (PLP) onboard the low Earth orbiting (454 km altitude) satellite CHAMP was carefully scrutinized. The top in Figure 11 shows the variation of SymH index as a function of UT, where the arrows A to F indicate the equator crossing times of six successive CHAMP orbits. Figures 11a to 11f show a sequence of latitudinal electron density profiles measured by the PLP on CHAMP in its successive orbits at various longitudinal sectors from 277 E to 162 E. It should be noted here that each profile corresponds to a satellite pass at different longitudes, but around the same local time of 2104 h. [37] The latitudinal profile of electron density over E longitude (Figures 11a) shows a well developed (asymmetric) Equatorial Ionization Anomaly (EIA) with two peaks (crests) on either side of the equator (trough). One and half hours later, the electron density profile over E longitude (Figures 11b) shows a clear signature of electron density depletion (the electron density going down to zero level) superimposed on enhanced EIA. The enhancement in EIA and the presence of depletions (which are the signatures of the equatorial plasma bubbles), suggest that the penetration of eastward electric fields into low latitudes has occurred over this longitudinal sector. However, the exact time of penetration is not known due to lack of continuous observations around these longitudinal sectors. Figures 11c 11f, also show the presence of electron density depletions (plasma bubbles) superimposed on EIA at different longitudes. Thus the total longitudinal width, where the plasma bubbles were observed extends over a longitudinal width of about 92, i.e., from E to E. [38] The penetration of eastward electric fields into equatorial latitudes and the subsequent development of plasma bubbles are first detected at E longitude (Figures 11b) around 0400 h UT (orbit B), i.e., one and half hours later from the time of sudden increase in the AE-Index. This time delay (one and half hours) and the local time (2104 h) agrees well with the Fejer and Scherliess [1997] model, for the prompt penetration of eastward electric fields into low latitudes to occur. Further, during the next orbit of the satellite (orbit C), electron density depletions were also observed at E (a western longitude with respect to the previous orbit). Similarly, depletions were also observed at E (orbit D), E (orbit E), and E (orbit F) longitudes during the successive satellite orbits. These results suggest that, the sudden increase in AE-index under the southward orientation of IMF Bz creates a favorable condition (such as breakdown of shielding) for the penetration of electric fields to take place. These penetration electric fields may add to normal PRE in the longitudinal sector where the local time corresponds the local dusk period leading to the development of ESF. For example, during the present storm, the penetration of eastward electric fields and the development of plasma bubbles are first observed over E longitude, where the local time corresponds to the post sunset hours. Later, as the solar terminator (local dusk) moves from east to west, plasma bubbles were observed during the successive orbits of satellite over E, E, 185 E, and E longitudes respectively Storm of 4 7 April 2006 [39] Figure 12 shows the response of Indian equatorial and low latitude ionosphere to another moderate geomagnetic storm that occurred during 4 7 April The Sym- H index (Figure 12c) started decreasing from 1430 h LT of 4 April 2006 and slowly reaching a minimum value of 92 nt around 1845 h LT of the next day (i.e., 5 April 2006). It can be observed from Figure 12d that the rate of change in Sym-H index (dsymh/dt) is low and varies from 3 to 2 nt/10 min during most of the time. However, 12 of 19

13 Figure 10. Series of ionograms showing spread-f from 1809 h LT of 15 May 2005 to 0319 h LT of 16 May 2005 over an Eastern equatorial station, Kwajalein Is (9 N, E, dip. latitude 4.2 N). around post sunset hours (1810 h LT) of 5 April 2006, the dsymh/dt exhibited a sharp negative excursion of 8 nt/10 min as may be seen from Figure 12d. The h 0 F over the equatorial station, Trivandrum (Figure 12e) has increased significantly from its quiet day average value and subsequently strong spread-f has been observed from 1845 to 2330 h LT. Further, intense VHF (244 MHz) scintillations were also observed over Waltair from 1910 to 2300 h LT as may be seen from Figure 12f. [40] The results presented in sections 3.1 to 3.4 are consistently showing the prompt penetration of eastward electric fields into low latitudes whenever a marked decrease in Sym-H occurs during the local postsunset hours. A total of five prompt penetration cases were presented during four moderate to intense geomagnetic storms, out of which, in four cases the AE-Index shows a sudden increase and subsequently the dsymh/dt exhibited a sharp negative excursion. Hence either AE or dsymh/dt could be considered as proxies to determine occurrence as well as the time of penetration. However, during 4 7 April 2006 storm (Figure 12), the AE-index does not represent a sudden increase and the maximum dsymh/dt occurred almost 24 h later. Owing to the gradual (as opposed to sudden) increase in AE-Index, the maximum dsymh/dt seems to be the better index to determine the time for prompt penetration Storm of 9 11 July 2005 [41] In Figure 13 is shown the response of Indian equatorial and low latitude ionosphere to another moderate geomagnetic storm that occurred during 9 11 July Following the southward turning of IMF Bz, the AE-index started increasing (Figure 13b) gradually from 1200 h LT of 9 July and the Sym-H index started decreasing (Figure 13c) almost simultaneously. It can be seen from Figures 13c and 13d, that the Sym-H index decreasing very slowly and the rate of change of Sym-H (dsymh/dt) does not show any significant negative excursions on this day. During that evening, the h 0 F is confined to less than 300 km and no spread-f has been observed on this night. [42] On the next day, i.e., on 10 July 2005, a strong reversal in EEJ current was observed (Figure 13e) during the afternoon hours. This can be inferred as due to the penetration of westward electric fields in to low latitudes associated with sudden northward turning of IMF Bz around 0900 h LT (Figure 13a). [43] In Figure 14 is shown the latitudinal electron density profiles derived from the PLP data of CHAMP over Indian 13 of 19

14 Figure 11. A series of latitudinal profiles of electron density measured by Planar Longmuir Probe (PLP) onboard CHAMP satellite during its successive orbits on 15 May of 19

15 Figure 12. Response of Indian equatorial and low latitude ionosphere to the geomagnetic storm of 4 7 April Time histories of (a) IMF Bz, (b) AE-Index, (c) Sym-H Index, (d) dsymh/dt, (e) virtual height (h 0 F) over Trivandrum, and (f) S4-Index of VHF scintillations. The vertical solid line indicates the time of peak negative excursion of dsymh/dt. The shaded region indicates the local dusk to dawn period. longitudinal sector around 1550 h LT. The thick line with solid circles shows the electron density profiles of 10 July, whereas the thin line represents the same parameter on the previous day, i.e., 9 July. It can be seen from this figure that the electron density profiles on 9 July show a well developed equatorial ionization anomaly structure with two peaks on either side of the equator. However, on 10 July, the profiles show that the ionization is confined only to one peak centered around the equator indicating a total suppression of anomaly at both longitudes (66.3 E and 89.4 E). [44] In Figure 15, is shown the diurnal variation of Total Electron Content (TEC) measured by GPS receivers located at Indian equatorial station, Trivandrum (dip. latitude 0.7 N) and at a station closer to northern crest of equatorial ionization anomaly, Bhopal (dip. latitude 18.5 N). It may be seen from this figure that the TEC at the equatorial station Trivandrum (thick line with solid circles) shows much higher values (about TEC units) than the TEC at the anomaly crest station, Bhopal (thin line) during the most of the daytime hours. This result further confirms the total suppression of EIA during the afternoon hours over the Indian longitudinal sector. [45] Strong reversal in Equatorial Electrojet current (Figure 13e) and the total suppression of EIA (Figures 14 and 15) during the afternoon hours indicates the presence strong westward electric fields over the equatorial and low latitude regions due to continued geomagnetic activity for more than 24 h from the past. [46] However, during this evening the rate of change of Sym-H index suddenly decreased to a value of 30 nt/10 min at 1910 h LT as may be seen from Figures 13c and 13d. At the same time, the virtual height of the F-layer (h 0 F) over Trivandrum showed a rapid increase (Figure 13f) indicating the prompt penetration of eastward electric fields into equatorial latitudes. Subsequently, intense spread-f over Trivandrum and scintillations at VHF (244 MHz) and L-band (1.5 GHz) frequencies over Waltair were observed as can be seen from Figures 13f and 13g. It should be emphasized here, that the prompt penetration of strong eastward electric fields corresponding to rapid decrease in Sym-H index prevailed over the strong ambient westward electric fields, 15 of 19

16 Figure 13. Response of Indian equatorial and low latitude ionosphere to the geomagnetic storm of 9 11 July Time histories of (a) IMF Bz, (b) AE-Index, (c) Sym-H Index, (d) dsymh/dt, (e) virtual height (h 0 F) over Trivandrum, (f) Equatorial Electrojet (EEJ) strength, and (g) S4-Index of VHF (black line) and L-band (gray line) scintillations. The solid vertical line indicates the time of peak negative excursion of dsymh/dt. The shaded region indicates the local dusk to dawn period. and augmented the normal PRE thereby creating favorable conditions for the development of ESF irregularities. 4. Summary [47] Using the multi-instrumental and multistation data, a detailed study has been carried out on the effects of geomagnetic storms over equatorial and low latitude ionosphere in the context of development/inhibition of ESF during five moderate to intense geomagnetic storms that occurred during the moderate to low solar activity period The results presented in sections 3.1 to 3.5 have consistently shown the prompt penetration of eastward electric fields into low latitudes leading to development of ESF in the longitudinal sectors where the time of marked decrease in Sym-H index correspond to the local post sunset hours. These results corroborate with the earlier results reported by Basu Su. et al. [2001], Basu S. et al. [2001, 2005, 2007] and Martinis et al. [2005]. Further, It is evident from the results presented in section 3.1 (11 12 February 2004 storm) and 3.3 (15 16 May 2005 storm) that the sudden increase in AE-Index and/or the marked decrease in Sym-H index under the southward orientation of IMF Bz creates favorable conditions for the prompt penetration to occur. Subsequently, penetration of eastward electric fields and consequent development of ESF may happen in all longitudinal sectors where the local time corresponds to postsunset hours during the entire period of the main phase of the storm. For example, during May storm (section 3.3), the electron density depletions were observed over a longitudinal width of 92 (one fourth of the Earth s equatorial ionosphere) i.e., from E to E during the successive orbits of CHAMP satellite within a period of 6 7 h. In the application point of view, the presence of 16 of 19

17 Figure 14. Latitudinal profiles of electron density measured by Planar Longmuir Probe (PLP) onboard CHAMP satellite during its two consecutive orbits around 1550 h LT on 9 July 2005 (thin black line) and 10 July 2005 (black line with solid circles). Figure 15. Diurnal variation of Total Electron Content (TEC) measured by dual frequency GPS receivers at the Indian equatorial station Trivandrum (black line with solid circles) and at a station nearer to northern equatorial ionization anomaly crest, Bhopal (thin black line) on 10 July of 19

18 depletions/bubbles over a wide longitudinal area could definitely lead to the degradation in the accuracy of position fixing by GPS based navigational systems as reported earlier by Bandhyopadhyay et al. [1997], Das Gupta et al. [2004], and Rama Rao et al. [2006]. [48] In the case of geomagnetic storms where the AEindex exhibits a sudden increase, either AE-index or the rate of change of SymH can be used as proxies to determine the occurrence as well as the time of prompt penetration. Whereas, in cases where the AE-index increases gradually, the dsymh/dt seems to be the better index to determine the time for prompt penetration as has been observed in the case of 4 6 April 2006 and 9 11 July 2005 storms. [49] Further, the geomagnetic storms considered in this study represent all the equinoxial (9 12 March 2004 and 4 7 April 2006) as well as the summer (15 16 May 2005 and 9 11 July 2005) and winter (11 12 February 2004) solstices. The prompt penetration of eastward electric fields leading to the development of postsunset ESF has evidenced in all seasons (both high as well as low ESF seasons) and at all longitudinal zones such as Indian (Trivandrum and Waltair), Chinese (Hainan Is), and Pacific (Kwajalein Is) as well as the Western (as depletions were observed by CHAMP over 254 to 185 E longitudes) sectors. [50] Also, it is evidenced from this study (section 3.5) that the intense eastward electric fields over the equator due to prompt penetration prevailed over the strong ambient westward electric fields leading to development of ESF even during a moderate geomagnetic storm. [51] Acknowledgments. This work has been done as a part of Indian ISRO-CAWSES WG2 project: CAWSES-06; 21 December One of the authors (S. Tulasi Ram) wishes to express his sincere thanks to CSIR, Govt. of India for providing him with SRF to carryout this research work. The authors also express their sincere thanks to J. H. Sastri, IIA for his valuable suggestions during the preparation of this manuscript. The authors further thank S. Alex, IIG for providing the EEJ data. The authors acknowledge Bodo Reinisch for UML DIDBase data, ISDC for CHAMP satellite data, ACE Science Centre for IMF Bz data, World Data Centre for Geomagnetism for AE and SymH data, World Data Centre for Ionosphere for Japanese ionogram data. [52] Amitava Bhattacharjee thanks Michael Kelley and Dr. Harish Chandra for their assistance in evaluating this paper. References Aarons, J. (1982), Global morphology of ionospheric scintillation, Proc. IEEE, 70, Aarons, J. (1991), The role of the ring current in the generation or inhibition of equatorial F-layer irregularities during magnetic storms, Radio Sci., 26, Aarons, J. (1993), The longitudinal morphology of equatorial F layer irregularities relevant to their occurrence, Space Sci. Rev., 63, Aarons, J., et al. (1980a), Seasonal and geomagnetic control of equatorial scintillations in two longitudinal sectors, J. Atmos. Sol. Terr. Phys., 42, Aarons, J., J. P. Mullen, H. E. Whitney, and E. M. Mackenzie (1980b), The dynamics of equatorial irregularity patch formation, motion and decay, J. Geophys. Res., 85, Abdu, M. A., J. A. Bittencourt, and I. S. Batista (1981), Magnetic declination control of the equatorial F region dynamo electric field development and spread F, J. Geophys. Res., 86, 11,443 11,446. Abdu, M. A., I. S. Batista, G. O. Walker, J. H. A. Sobral, N. B. Trivedi, and E. R. de Paula (1995), Equatorial ionospheric fields during magnetospheric disturbances: Local time/longitudinal dependences from recent EITS Campaigns, J. Atmos. Sol. Terr. Phys., 57, Anderson, D., A. Anghel, K. Yumoto, M. Ishitsuka, and E. Kudeki (2002), Estimating daytime vertical E B drift velocities in the equatorial F-region using ground-based magnetometer observations, Geophys. Res. Lett., 29(12), 1596, doi: /2001gl Bandhyopadhyay, T., A. Guha, A. Das Gupta, P. Banerjee, and A. Bose (1997), Degradation of navigational accuracy with Global Positioning System during periods of scintillation at equatorial latitudes, Electron. Lett., 33, Basu, Su., and S. Basu (1985), Equatorial scintillations: Advances since ISEA-6, J. Atmos. Sol. Terr. Phys., 47(8), Basu, S., Su. Basu, K. M. Groves, H. C. Yeh, S. Y. Su, F. J. Rich, P. J. Sultan, and M. J. Keskinen (2001), Response of the equatorial ionosphere in the South Atlantic region to the great magnetic storm of July 15, 2000, Geophys. Res. Lett., 28, Basu, S., Su. Basu, K. M. Groves, E. MacKenzie, M. J. Keskinen, and F. J. Rich (2005), Near-simultaneous plasma structuring in the midlatitude and equatorial ionosphere during magnetic superstorms, Geophys. Res. Lett., 32, L12S05, doi: /2004gl Basu, S., Su. Basu, F. J. Rich, K. M. Groves, E. MacKenzie, C. Coker, Y. Sahai, P. R. Fagundes, and F. Becker-Guedes (2007), Response of the equatorial ionosphere at dusk to penetration electric fields during intense geomagnetic storms, J. Geophys. Res., 112, A08308, doi: / 2006JA Basu, Su., S. Basu, and C. E. Valladares (2001), Ionospheric effects of major magnetic storms during the International Space Weather Period of September and October 1999: GPS observations, VHF/UHF scintillations, and in situ density structures at middle and equatorial latitudes, J. Geophys. Res., 106, 30,389 30,413. Becker-Guedes, F., Y. Sahai, P. R. Fagundes, W. L. C. Lima, V. G. Pillat, J. R. Abalde, and J. A. Bittencourt (2004), Geomagnetic storm and equatorial spread-f, Ann. Geophys., 22, Blanc, M., and A. D. Richmond (1980), The ionospheric disturbance dynamo, J. Geophys. Res., 85, Bowman, G. G. (1982), Spread-F occurrence in mid and low latitude regions related to various levels of geomagnetic activity, J. Atmos. Sol. Terr. Phys., 44, Chandra, H., and R. G. Rastogi (1972a), Equatorial spread F over a solar cycle, Ann. Geophys., 28, Chandra, H., and R. G. Rastogi (1972b), Equatorial spread F over a solar cycle, Ann. Geophys., 28, Chandra, S., and J. R. Herman (1969), F-region ionization and heating during magnetic storms, Planet. Space. Sci., 17, Chandra, H., and G. D. Vyas (1978), On the relationship between magnetic activity and spread-f at Kodaikanal, Indian J. Radio Space Phys., 7, Dabas, R., D. Lakshmi, and B. Reddy (1989), Effect of geomagnetic disturbances on the VHF nighttime scintillation activity at equatorial and low latitudes, Radio Sci., 24, Das Gupta, A., S. Ray, A. Paul, P. Banerjee, and A. Bose (2004), Errors in position-fixing by GPS in an environment of strong equatorial scintillations in the Indian zone, Radio Sci., 39, RS1S30, doi: / 2002RS Fejer, B. G., and M. C. Kelley (1980), Ionospheric irregularities, Rev. Geophys., 18, Fejer, B. G., and L. Scherliess (1997), Empirical models of storm time equatorial zonal electric fields, J. Geophys. Res., 102, 24,047 24,056. Fejer, B. G., R. W. Spiro, R. A. Wolf, and J. C. Foster (1990), Latitudinal variation of perturbation electric fields during magnetically disturbed periods: 1986 SUNDIAL observations and model results, Ann. Geophys., 8, Fejer, B. G., L. Scherliess, and E. R. de Paula (1999), Effects of the vertical plasma drift velocity on the generation and evolution of equatorial spread F, J. Geophys. Res., 104, 19,859 19,869. Huang, C., W. Burke, J. Machuzak, L. Gentile, and P. Sultan (2002), Equatorial plasma bubbles observed by DMSP satellites during a full solar cycle: Toward a global climatology, J. Geophys. Res., 107(12), 1434, doi: /2002ja Huang, C. S., J. C. Foster, and M. C. Kelley (2005), Long-duration penetration of the planetary electric field to the low-altitude ionosphere during the main phase of magnetic storms, J. Geophys. Res., 110, A11309, doi: /2005ja Huang, C. S., S. Sazykin, R. Spiro, J. Goldstein, G. Crowley, and J. M. Ruohoniemi (2006), Storm-time penetration electric fields and their effects, Eos Trans. AGU, 87(13), 131, doi: /2006eo Hysell, D. L., and J. D. Burcham (1998), JULIA radar studies of equatorial spread-f, J. Geophys. Res., 103, 29,155 29,167. Hysell, D. L., and J. Burcham (2002), Long term studies of equatorial spread F using the JULIA radar at Jicamarca, J. Atmos. Sol. Terr. Phys., 64, Jaggi, R. K., and R. A. Wolf (1973), Self-consistent calculation of the motion of a sheet of ions in the magnetosphere, J. Geophys. Res., 78, Kelley, M. C., and T. Maruyama (1992), A diagnostic model of equatorial Spread-F: 2. The effect of magnetic activity, J. Geophys. Res., 97, of 19