By continuing to browse this site you agree to us using cookies as described in About Cookies
Notice: Wiley Online Library will be unavailable on Saturday 7th Oct from 03.00 EDT / 08:00 BST / 12:30 IST / 15.00 SGT to 08.00 EDT / 13.00 BST / 17:30 IST / 20.00 SGT and Sunday 8th Oct from 03.00 EDT / 08:00 BST / 12:30 IST / 15.00 SGT to 06.00 EDT / 11.00 BST / 15:30 IST / 18.00 SGT for essential maintenance. Apologies for the inconvenience.
 Ionospheric vertical sounding observations are being carried out at Sao Jose dos Campos (23.2°S, 45.9°W; dip latitude 17.6°S), Brazil, under the southern crest of the equatorial ionization anomaly (EIA) since August 2000. In this paper, we present and discuss the observations of daytime F2-layer stratification near the crest of EIA, for the first time, under magnetically quiet high solar activity conditions. Three examples and a year of statistics are presented. The F2-layer stratification and F3-layer were observed between 10:40 and 11:45 UT on 31 December 2000, between 13:30 and 14:30 UT on 1 January 2001, and between 13:15 and 15:15 UT on 11 February 2001. The statistics during September 2000 to August 2001 shows that the F3-layer occurs only for 66 days (18% occurrence), and it occurs only during September–February (spring–summer), with maximum occurrence in September–October and longest duration in February. The F2-layer stratification seems to be associated with gravity waves (GWs), which have periods of about 30–60 min, downward phase velocities of about 60–140 m/s, and vertical wavelengths of about 200–500 km. The presence of powerful gravity waves in a vertically extended F-layer seems to stratify the F2-layer and produce the F3-layer. Because the stratifications are observed during geomagnetically quiet periods, the source of the gravity waves are most likely to be associated with local tropospheric disturbances and not with high-latitude disturbances.
 During the last decade, several investigators have studied both theoretical aspects and observational results related to the additional F3-layer and considerable progress has been achieved [Balan and Bailey, 1995; Balan et al., 2000; Batista et al., 2002, 2003; Depuev and Pulinets, 2001; Jenkins et al., 1997; Pulinets et al., 2002]. The large morning upward drift of the F-layer around the magnetic equator raises the plasma to higher altitudes, and the strong meridional wind blowing from the summer hemisphere to the winter hemisphere acts to raise the plasma in the summer hemisphere. This uplifting of the plasma and the influence of transequatorial meridional wind produces an additional F3-layer, which has a peak value greater than the peak value of the F2-layer just before local moon. However, the meridional wind at the magnetic equator has a smaller vertical component than at a few degrees of latitude away from the magnetic equator, and consequently, the F3-layer is weaker at the magnetic equator and stronger a few degrees away [Jenkins et al., 1997]. On the other hand, On the other hand Rama Rao et al.  have reported from observations at Ahmedabad (23°N, under the northern equatorial anomaly crest: Indian sector) complete absence of the additional F3-layer. This supports the idea that the formation of F3-layer is strongly dependent upon upward plasma motion and could be seen only at locations closer to the magnetic equator influenced by vertical plasma dynamics associated with electrojet current system. Lynn et al.  presented a study, which agrees with Jenkins et al. , related to the latitudinal dependence of F2-layer stratification in the Southeast Asia, and suggested that the region of maximum F2-layer stratification lies between the magnetic equator and the peak of the southern equatorial anomaly. However, using satellite topside radio sounding, Depuev and Pulinets  and Pulinets et al.  have reported a F3-layer, with foF3 less than foF2, between the equatorial anomaly crest and magnetic equator. They have reported maximum in occurrence just above the equator and that the occurrence decreases poleward within ±10° dip. They have pointed out that the F3-layer occurs during both daytime and nighttime. It should be mentioned that topside sounders observe the F3-layer as topside ledges, which are observed even after sunset when the layer disappears from bottomside ionosphere.
Abdu et al.  showed a few special cases where traveling ionospheric disturbances (TIDs), possibly associated with geomagnetic disturbances, cause bifurcations of the F-layer trace during pre-sunrise periods. In one of the cases studied, the F-layer profile during dawn showed two reasonably well-defined layers, reassembling the F1 and F2 bifurcations of the regular daytime ionosphere. In addition, they have pointed out that the occurrence of F-layer satellite traces in the ionograms is a usual characteristic observed near the southern crest of the equatorial ionization anomaly (EIA) region especially at sunset hours, followed by range spread-F echoes. They have also pointed out that multiple F-layer traces are observed during the daytime as well as the nighttime following geomagnetic storms or TIDs (for example, multiple F-layer traces in ionograms; see, e.g., Abdu et al. [1982, Figure 1b], 27 June 1978 at 04:00 LT).
 This paper presents and discusses observations of the F3-layer formation detected in the region of the southern crest of the equatorial ionization anomaly, a region where the presence of F3-layer was not predicted by model simulations [Balan and Bailey, 1995; Balan et al., 1998, 2000; Jenkins et al., 1997]. Also, we show that the F3-layer observed at Sao Jose dos Campos (SJC; 23.2°S, 45.9°W), Brazil, is strongly related to wave-like disturbance structures propagating through the F-layer. It appears that these observations reporting the formation of the ionospheric F3-layer outside the equatorial region are associated with a different mechanism than suggested by previous workers [Balan and Bailey, 1995; Balan et al., 1998, 2000; Jenkins et al., 1997] or a combination of the two mechanisms (i.e., vertical upward movement of the F2-layer peak and gravity waves). In addition, the observations reported here are quite different from those reported by Abdu et al.  from a location very close to the present site. The present study shows the effects of gravity waves in the F-layer during the daytime and geomagnetic quiet conditions, whereas Abdu et al.  studied its effects during the nighttime with passage of TIDs. It should be mentioned that the electrodynamics of the ionosphere at low latitude during the daytime is strongly affected by the photoionization process and vertical upward movement caused by the E × B drift. However, the present investigations indicate that sometimes the gravity waves produce a strong stratification of the F-layer and are very closely related with the generation of F3-layers in the EIA region.
 A Canadian Advanced Digital Ionosonde (CADI) is in operation at SJC since August 2000. The CADI is simultaneously operational in two different modes:
 (1) It sweeps 180 frequencies from 1 to 20 MHz sampling with a temporal resolution of 300 s (5 min), and these measurements provide usual ionograms.
 (2) It operates only at six preselected frequencies (3.1, 4.1, 5.1, 6.3, 7.1, and 8.1 MHz) with sampling at a high temporal resolution (100 s), and these measurements make available isofrequency plots.
 Using the virtual height values extracted directly from the second mode of operation, with high rate of sampling (100 s), it is possible to investigate the daily virtual height variations for each of the six frequencies (isofrequency plots). Also, the rate of sampling of 100 s allows us to study GW propagations with periods larger than about 5 min (300 s). Since the observed GW periods in this study are about 30–60 min, the rate of sampling of 100 s appears to be appropriate. The bottomside vertical electron density profile (about 400–500 km in height range) of the ionospheric F-region allows us to study gravity wave vertical wavelength with a good confidence ranging from about 100 to 700 km. The periods and vertical wavelengths mentioned above are the limitations of the instrumentation for studying GW using isofrequency plots. Becker-Guedes et al.  and Lima et al.  have used similar isofrequency plots to study the response of the ionospheric F-layer during geomagnetically disturbed periods and TIDs propagating to low latitude. The present investigation demonstrates that it is possible to study GW propagations during the daytime and their effects on the low-latitude ionosphere using digital ionosonde observations with higher sampling rates.
3. Results and Discussion
Figures 1a–1c show the daily virtual height variations for 3.1, 4.1, 5.1, 6.3, 7.1, and 8.1 MHz (isofrequency plots) for 31 December 2000, 1 January 2001, and 11 February 2001, respectively. On these days, it is possible to identify the presence of two well-known features of the low-latitude ionosphere. The first one takes place during nighttime from about 01:00 to 08:00 UT (22:00–05:00 LT), and multiple height reflections at 3.1, 4.1, 5.1, and 6.3 MHz frequencies are easy to see; these are the signatures of well-known range type equatorial spread-F at this low-latitude station, a common feature observed during the December solstice period (November to February) in the Brazilian sector. The second feature starts around 21:00 UT (18:00 LT) and is the low-latitude extension of the well-known equatorial prereversal enhancement peak caused by the action of the eastward electric field during the nighttime that induces a F-layer E × B upward vertical drift starting around 21:00 UT (18:00 LT). Notice in Figures 1a–1c (enlarged right-hand-side panels) that around 21:00 UT (18:00 LT), the F-layer starts drifting upward, due to the E × B vertical drift, and after reaching its maximum height around 22:30 UT (19:30 LT), it drifts downward. The upward and downward F-region drift presents synchronized changing of height simultaneously for all the sounding frequencies without any time lag; in other words, a zonal electric field displaces up or down the entire F-Layer. The geomagnetic indices (Dst and Kp) for the 3 days studied are presented in Table 1. It is noted that the days investigated are quiet and that perturbations with origin at high latitude (auroral region) are not expected in the low-latitude F-region. It should be mentioned that the AE indices on these 3 days were less than 500 nT, which also indicates low magnetic activity in the high-latitude region.
Table 1. Dst and Daily ∑Kp Geomagnetic Indices for 31 December 2000, 1 January 2001, and 11 February 2001
31 December 2000
1 January 2001
11 February 2001
 It is noted in Figures 1a–1c that between 09:00 and 20:00 UT (06:00–17:00 LT), the height variations for the frequencies 6.3 (orange) and 7.1 MHz (green) show wave-like oscillations. It is important to mention that virtual height echoes at the frequencies 6.3 and 7.1 MHz shown in Figures 1a–1c (enlarged left-hand panel) present sometimes two traces for each frequency. The traces are fairly separated in the case of 7.1-MHz echoes (Figure 1, green traces), and these two virtual height echoes refer to the ordinary and extraordinary traces. Also, Figures 1a–1c show that the maximum and minimum virtual height variations occur first at 7.1 MHz and then after a few minutes later at 6.3 MHz, representing downward phase propagation velocities of about 60 to 140 m/s with periods of about 40 to 60 min and inferred vertical wavelengths of about 150 to 500 km. This agrees with the principal characteristics of propagation of gravity waves in the ionospheric F-region [Hines, 1960; Nicolls and Kelly, 2005]. The time lag features are indicated in Figures 1a–1c (enlarged left-hand panels), with solid lines. Therefore Therefore, using the isofrequency plots (Figures 1a–1c, left and right panels), we are able to distinguish between the electric field (simultaneous influence in all the sounding frequencies) and the gravity wave (downward phase propagation) effect at F-layer height.
 The present investigation demonstrates that it is possible to study GW propagations during the daytime and their effects on the low-latitude ionosphere, using digital ionosonde observations with higher sampling rates. Walker et al. , using ionospheric sounding observations with typical sampling rates of 5 to 10 min, have reported that ionospheric wave-modulated structures may be difficult to observe during the daytime because of the presence of large latitudinal ionization gradients associated with the equatorial anomaly and the variability of the location of equatorial anomaly crest. However, the high temporal mode of operation opens the possibility of monitoring ionospheric gravity wave activity during the daytime. One of the important points to show is related to the direct effect of the gravity waves when it crosses the F-region heights.
 In this investigation, we have shown three case studies when the F-region was strongly modulated by intense gravity wave activity and its (gravity waves) strength were sufficiently intense to modify the vertical ionospheric profiles during several hours. The presence of F-layer stratification was observed only when the signatures of gravity waves at F-region heights were present. In order to show how the gravity waves modified the F-layer vertical profile, Figures 2, 3, and 4present a sequence of ionograms for the 3 days considered in this study. However, we should point out that the transformation/perturbation due to gravity waves in the neutral atmosphere (thermosphere) in the ionospheric F-region is not direct and simple. In all the cases studied (Figures 1a–1c), the manifestation of gravity wave oscillations in the ionosphere (during daytime) shows the virtual height changing for fixed frequencies. This possibly indicates that the gravity waves induced perturbations in the meridional wind and such perturbations resulted in vertical redistribution of ionization.
 Regarding the role of viscosity on the wavelength observed, it should be pointed out that Vadas and Fritts  have recently investigated the thermospheric responses to gravity waves related to increasing viscosity and thermal diffusivity. However, they mention that gravity wave propagation and effects at higher altitudes are poorly understood at present and damping from molecular viscosity and thermal diffusivity become increasingly important because of the decreasing background density. These processes are highly dependent on GW parameters; GWs with larger vertical wavelengths (in the present investigations, the vertical wavelengths are large and of the order of 200–500 km) and vertical group velocities propagate higher into the thermosphere before dissipating than GWs with smaller vertical wavelengths and vertical group velocities.
 The first case studied is on 31 December 2000 (Figure 2); notice that around 10:00–10:30 UT, the F1-layer was displaced downwards and the F2-layer was displaced upward. However, between 10:45 and 11:45 UT, the F1-layer height remains almost unchanged, but the F2-layer is completely distorted including a stratification of the F2-layer that appears as F3-layer. After 12:00 UT, a cusp is observed between F1- and F2-layers and propagates downward. In order to draw attention to these points, pairs of ionograms are superimposed to compare such features in Figure 5a. Also, it is important to mention that the F3-layer appears when the F-layer virtual height profiles seen on ionograms extended from about 200 to 800 km, i.e., having virtual height profile extending for about 600 km.
 In the second example, for 1 January 2001 (Figure 3), notice that between 12:45 and 13:15 UT, the stratification of the F-layer begins with a cusp between F1- and F2-layers, and at around 13:30 UT, there appears a F3-layer that remains for about an hour. Again, the structuring of the F-layer resulting in F3-layer appears when the F-layer virtual height profiles on ionograms extend for about 500 km. Figure 5b shows more details of the development of this structure when pairs of ionograms are superimposed.
 In the third example on 11 February 2001 (Figure 4), the stratification of the F-layer starts with a F3-layer around 13:15–14:15 UT, when the F-layer virtual height profiles extend for about 500 km. Also, around 14:30 UT, the F-layer structuring appears at lower heights and remains until 15:15 UT when the virtual height extension is about 400 km. See Figure 5c for more details.
 According to the work of Balan et al. , the physical mechanism leading to the formation of the F3-layer in regions close to the magnetic equator is due to the combined effects of upward E × B drifts and neutral winds, which possibly provides vertical upward plasma drift velocities at altitudes near and above the F2 peak, resulting in the formation of the F3-layer. They used several patterns of vertical plasma drifts in order to investigate the day-to-day variations in the F3-layer occurrence and concluded that the vertical plasma drift (Vz from equations (1) and (2) of Balan et al.) should raise the F2-layer peak by over 210 km for the appearance of the F3-layer around the equator. Balan et al. have also mentioned that the F2-layer could be raised either gradually or rapidly, above the threshold displacement (210 km), since the time-accumulative velocity was sufficient to lift the F-layer for the F3-layer formation. However, in the present study, near the EIA crest, the vertical lift calculated using the height displacements for the frequencies 6.3 and 7.1 MHz showed that the F-layer was raised to lower heights (see Table 2 for details) than the threshold level (210 km) reported by Balan et al. for equatorial locations. It should be pointed out that on 1 January 2001, the F-layer uplift was only by 125 km during the daytime (Figure 2). It should be noted that the ionospheric peak height decreases with latitude, and hence the altitude of the additional F2-layer stratification at the EIA crest is lower than that around the equator.
Table 2. Main Characteristics of F-Layer Vertical Upward Movement Calculated Using the Height Displacements for the Frequencies 6.3 and 7.1 MHz, Duration of F3-Layer, and Phase of Gravity Waves When the F3-Layer Occurred
Daytime F-Layer Upward Displacement, km
Duration of F3-Layer UT (LT)
31 December 2000
Descending and ascending
1 January 2001
11 February 2001
 All the cases studied in this investigation show the presence of oscillations in the F-layer lasting several hours (see Figure 1). However, the stratification appears only during the daytime for about 1 or 2 hours in this period, when the F-layer has unusual vertical extension. Therefore, it appears we need a combination of both the F-layer extension and gravity waves to create favorable conditions for F-layer stratification and formation of F3-layer near EIA crest.
 The combined effect of upward E × B drift and equatorward neutral wind is known to vertically extend the low-latitude ionosphere. Large daytime upward E × B drift and/or large equatorward wind can cause unusual vertical extension of the F2-layer near the EIA crest. The upward propagating gravity waves in such an unusually extended F2-layer seem to be the possible reason for the F2-layer stratification and F3-layer observed near the EIA crest.
 In order to illustrate the interaction between the F-layer profiles with a perturbation wave, we generated two hypothetical F-layer profiles interacting with perturbation wave having a wavelength of 350 km (which is consistent with the wavelength calculated in this work of 200–500 km). The first F-layer profile extends from 200 to 700 km, and the second one extends from 200 to 500 km. This schematic illustration presented in Figure 6 shows that the occurrence of F-layer stratification is more probable during some special relationship between the F-layer extension and the wave vertical wavelength. This schematic illustration indicates that the occurrence of F-layer stratification will have favorable conditions of occurrence when the F-layer extension is larger than the gravity wave wavelengths. Possibly the viscosity should increase the vertical wavelength significantly. However, the gravity wave propagations at higher altitudes and its interaction with the ionosphere (F-layer) are still poorly understood, especially the role of the viscosity and thermal diffusivity. These are the limitations of the proposed idea. Also, the present study is important in studying the coupling between ionosphere and neutral atmosphere. Again, it should be pointed out that during geomagnetic quiet conditions, the troposphere is a strong candidate for gravity waves source.
 A recent work carried out in the Indian sector by Rama Rao et al.  presented a study of F2-layer stratification during the daytime for a half solar cycle, and their analysis shows that the F3-layer appeared more frequently during summer time of low solar activity and that it does not have dependence on geomagnetic activity. Also, they reported that the occurrence of F3-layer is rare over Trivandrum (8.4°N, Mag. Lat. 0.47°N), an equatorial station, that F3-layer has a weak presence over SHAR (14°N, Mag. Lat. 6.8°N), that there are more frequent observations of F3-layer over Waltair (17.7°N, Mag. Lat. 8.2°N), a subtropical station, and that F3-layer is completely absent over Ahmedabad (23°N, Mag. Lat. 14.4°N), located under the northern crest of the equatorial ionization anomaly. A comparison between the present investigations (South American sector) and that carried out in the Indian sector indicates that the occurrence of F3-layer under the crest of EIA presents a remarkable asymmetry. Such asymmetry may be due to the longitudinal differences in possibly the vertical extension of the F-layer in the two sectors combined with the presence of strong oscillation source causing GW perturbations in the F-region in the South American sector. Table 3 shows the statistics of the seasonal variations of F3-layer during 1 year (September 2000 to August 2001) of observations at high solar activity. It is noted that the occurrence of the F3-layer under the equatorial anomaly southern crest is limited to only 66 days during a full year of observations (18%) with no occurrence during the months from March 2001 to August 2001 (autumn and winter). Also, it should be pointed out that in terms of days, the maximum occurrence takes place during the months of September and October 2000, while in terms of hours, the maximum occurrence takes place during the month of February 2001. The statistics (Table 3) shows that stratification occurs only during September–February and does not occur during March–August. The observed seasonal dependence of the stratification agrees with the seasonal variation of effective upward plasma velocity, mainly due to E × B drift and partly due to neutral wind. The daytime E × B drift, although upward in all seasons, is largest in spring [Fejer et al., 1991]. The neutral wind near the EIA crest, in general, is equatorward during spring–summer and poleward during autumn–winter [Hedin et al., 1994]. Hence the effective upward plasma velocity, needed for the vertical extension of the F2-layer, can be larger during spring–summer than during autumn–winter. Further statistics is needed to confirm the seasonal variation of the stratification observed near the EIA crest.
Table 3. Seasonal Variations of the Daytime F3-Layer Occurrence During the Period From September 2000 to August 2001 (High Solar Activity)
Number of Days
Number of Hours
 In this work, we present ground-based ionospheric observations carried out at Sao Jose dos Campos, Brazil, in the region of the southern crest of the equatorial ionization anomaly (EIA), showing stratification of the F-layer with occurrence of the F3-layer and its possible association with gravity wave events during geomagnetic quiet conditions.
 The major findings in this paper are summarized as follows:
 (1) Ionospheric sounding with sampling at a high temporal rate of about 100 s allows us to study in more detail the perturbation of the F-region caused by gravity waves, especially during daytime.
 (2) Ground-based observations showed the presence of the F3-layers during the daytime in the region of southern crest of the EIA.
 (3) The statistics during September 2000 to August 2001 shows that the F3-layer occurs only for 66 days (18% occurrence), and it occurs only during September–February (spring–summer), with maximum occurrence in September–October and longest duration in February.
 (4) It appears that gravity waves play an important role in the F-layer stratification in the region of the EIA.
 (5) A schematic illustration shows that the F-layer vertical extension must be larger than the gravity wave wavelength to create favorable conditions for stratification and formation of the F3-layer.
 Thanks are due to the Brazilian funding agencies CNPq and FAPESP for the partial financial support through grants: 301222/2003/7, 300843/2003-8, and 305625/2003-9 (CNPq); 2004/10104-9 (FAPESP).
 Wolfgang Baumjohann thanks Michael Nicolls and another reviewer for their assistance in evaluating this paper.