JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 117, A08337, doi: /2012ja017692, 2012

Transcription

1 JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 117,, doi: /2012ja017692, 2012 On post-midnight field-aligned irregularities observed with a 30.8-MHz radar at a low latitude: Comparison with F-layer altitude near the geomagnetic equator M. Nishioka, 1 Y. Otsuka, 2 K. Shiokawa, 2 T. Tsugawa, 1 Effendy, 3 P. Supnithi, 4 T. Nagatsuma, 1 and K. T. Murata 1 Received 5 March 2012; revised 9 July 2012; accepted 11 July 2012; published 31 August [1] We investigated the relationship between post-midnight F-region field aligned irregularities (FAIs) and F-layer altitude by analyzing data of a 30.8-MHz radar installed 5at Kototabang, Indonesia (0.2 S, E; geomagnetic latitude 10.4 S) and an ionosonde installed at Chumphon, Thailand (10.7 N, 99.4 E; geomagnetic latitude 3.3 N). Chumphon is located near the geomagnetic equator on approximately the same meridian as Kototabang. Case studies show that the altitude of the F-layer rose at Chumphon a half hour before the post-midnight FAIs appeared at Kototabang. The Doppler velocity of the E-region FAIs observed simultaneously by the 30.8-MHz radar was downward, indicating that the F-layer uplift was not caused by the electric field. We also investigated seasonal variations of the post-midnight FAI occurrence and the F-layer altitude. Both the postmidnight FAIs and the uplift of the F-layer were frequently seen around midnight between May and August. The seasonal variation of the midnight F-layer uplift around the geomagnetic equator coincided with that of the post-midnight FAI occurrence at Kototabang. These results suggest that the uplift of the F-layer would play an important role in the generation of post-midnight FAIs. We evaluated the linear growth rate of the Rayleigh-Taylor instability based on the altitude of the F-layer observed at Chumphon. The result shows that the uplift of the F-layer can enhance the growth rate because gravity-driven eastward electric current increases. Therefore, we interpret that the observed FAIs were accompanied by plasma bubble, the growth rate of which was reinforced by the uplifted F-layer. Citation: Nishioka, M., Y. Otsuka, K. Shiokawa, T. Tsugawa, Effendy, P. Supnithi, T. Nagatsuma, and K. T. Murata (2012), On post-midnight field-aligned irregularities observed with a 30.8-MHz radar at a low latitude: Comparison with F-layer altitude near the geomagnetic equator, J. Geophys. Res., 117,, doi: /2012ja Introduction [2] Ionospheric irregularities in the equatorial F-region have been extensively studied for many decades. Plasma bubble, which is generated through the Rayleigh-Taylor instability, is a well-known equatorial irregularity. It is 1 National Institute of Information and Communications Technology, Tokyo, Japan. 2 Solar-Terrestrial Environmental Laboratory, Nagoya University, Nagoya, Japan. 3 National Institute for Aeronautics and Space, Bandung, Indonesia. 4 Faculty of Engineering, King Mongkut s Institute of Technology Ladkrabang, Bangkok, Thailand. Corresponding author: M. Nishioka, Applied Electromagnetic Research Institute, National Institute of Information and Communications Technology, Nukui-kita, Koganei, Tokyo , Japan. (nishioka@nict.go.jp) American Geophysical Union. All Rights Reserved /12/2012JA initiated at the time of sunset, when the eastward electric field is enhanced, which is the so-called pre-reversal enhancement. The ionospheric altitude is raised by the enhanced eastward electric field. In addition to the rise in height of the ionosphere, the density gradient in the bottom-side F-region steepens. These are favorable conditions for the Rayleigh- Taylor instability. The occurrence of the plasma bubble maximizes around the magnetic equinox, when the geomagnetic field line is parallel to the terminator [e.g., Tsunoda, 1985; Burke et al., 2004;Nishioka et al., 2008]. In the Asian sector, where the geomagnetic field line is almost parallel to the meridian, plasma bubble most frequently appears around equinoxes. [3] The scale of the density irregularities within plasma bubble varies from 10 2 mto10 5 m [e.g., Kelley et al., 1982; Tsunoda, 1980; McClure et al., 1977]. The irregularities are detected employing various observational techniques. One of the observational techniques used to measure meter-scale irregularities is VHF radar. VHF radio waves are 1of9

2 transmitted and echoes resulting from the Bragg scatter at the field-aligned irregularity (FAI) are detected. The spatial scale of the irregularity that the radar can detect is one half of the radar wavelength. VHF radars have been operated at Gadanki (13.5 N, 79.2 E; geomagnetic latitude 12.5 N) in the Indian region, at Piura (5.2 S, 80.6 W; geomagnetic latitude 13.9 N) and Jicamarca (11.9 S, 76.8 W; geomagnetic latitude 0.8 N) in the Peruvian region, and at Saõ Luís (2.6 S, 44.2 W; geomagnetic latitude 1.7 S) in the Brazilian region. In the Southeast Asia region, Equatorial Atmosphere Radar (EAR) has operated with a frequency of 47 MHz at Kototabang (0.2 S, E; geomagnetic latitude 10.4 S) in Indonesia. Several papers have reported FAIs in the nighttime F-region [e.g., Fukao et al., 2003; Otsuka et al., 2004; Yokoyama et al., 2004]. EAR is a powerful tool to observe FAIs over Indonesia. However, EAR was originally installed to conduct tropospheric and stratospheric measurements. Otsuka et al. [2009] installed a 30.8-MHz radar at Kototabang in early 2006 in order to make continuous observations of FAIs over Indonesia. The VHF radar is designed to have two observational modes; that is, FAI observation modes for the E-region and F-region. Using data from February 2006 to November 2007, Otsuka et al. [2009] reported that F-region FAIs appeared frequently before midnight around equinoxes and after midnight between May and August. Post-midnight FAIs were frequently seen between May and August in every year from 2006 to 2010 [Otsuka et al., 2012]. Although the pre-midnight FAIs would be associated with equatorial plasma bubbles, the post-midnight FAIs had a different seasonal and local time dependence from that of plasma bubble. [4] The post-midnight irregularity was observed by ionosondes in the Indian sector before solar cycle [Subbarao and Krishna Murthy, 1994; Sastri, 1999]. During the long-lasting and extremely low solar activity in this solar cycle 23 24, there were several investigations of the postmidnight irregularity. However, the generation mechanism for the post-midnight irregularity is not yet known. Yokoyama et al. [2011a] compared the F-region FAI echo observed by EAR with the in-situ plasma density observed by Planar Langmuir Probe (PLP) onboard the C/NOFS satellite. The FAI-echo regions coexisted with the in-situ density depletion regions. A statistical study was also carried out using C/NOFS-PLP data [Heelis et al., 2010; Dao et al., 2011]. It was found that during the period of the solar minimum, plasma density irregularities mostly occurred after midnight. The seasonal-longitudinal variation of the postmidnight irregularity was completely different from that of plasma bubble. The post-midnight irregularity was attributed to seeding from tropospheric sources. In the Pacific sector, the post-midnight irregularity mostly appears from November to January [Miller et al., 2010]. In the Brazillian sector, post-midnight spread-f frequently appeared around the June solstice from 2006 to 2009 [Candido et al., 2011]. It was concluded that the post-midnight irregularity was not related to equatorial processes. [5] One key to studying the generation mechanism for ionospheric irregularities is the ionospheric altitude. In this study, we investigated the relationship between the post-midnight irregularities and F-layer altitude using the 30.8-MHz radar at Kototabang and an ionosonde installed near the geomagnetic equator on almost the same meridian. 2. Data [6] A 30.8-MHz backscatter radar has been continuously operated since February 2006 at Kototabang in West Sumatra, Indonesia. Coherent echoes due to the Bragg scatter from FAIs are detected with the radar beam perpendicular to the geomagnetic field lines. For the 30.8-MHz radar, the transmitted radio wave is scattered by the FAIs with a scale of approximately 5 m. Zenith angles of the radar beams that achieve the perpendicularity are 24 35, depending on the azimuthal angle of the beams. Five beams were allocated for F-region FAI measurements on 125.8, 153.0, 180.0, 207.0, and azimuths from February 2006 to January 2008, and 153.0, 166.4, 180.0, 193.6, and since February A more detailed description of the 30.8-MHz radar is given by Otsuka et al. [2009]. [7] For Southeast Asia including the Sumatran area, an ionospheric observation network, named SouthEast Asia Low-latitude Ionospheric Network (SEALION), has been developed [e.g., Maruyama et al., 2007; Saito and Maruyama, 2006]. Frequency modulated-continuous wave (FM-CW) ionosondes were installed and continuous observations were made at unmanned stations with low electric power. Each ionosonde transmits radio waves from 2 to 30 MHz and receives echoes from the ionosphere. An ionogram is obtained every 5 min. In this study, one of the ionosondes of SEALION, installed at Chumphon, in the middle of the Malay Peninsula, is used to investigate the ionospheric altitude. Chumphon is located at (10.7 N, 99.4 E), and is on almost the same meridian as Kototabang and is close to the geomagnetic equator (3.3 N in geomagnetic latitude). The ionosonde at Chumphon has been continuously operated since [8] One ionosonde parameter, the virtual height of the bottom-side F-layer, h F, can be directly scaled from an ionogram. It reflects F-layer dynamics in the absence of sunlit, while during the day, it is mainly controlled by production and loss processes. h F is not available when the bottom-side density is less than [el/cm 3 ], which corresponds to the minimum transmitting frequency of 2 MHz. When the plasma density is extremely low, such as at night during a solar-minimum period, there are many cases that h F is not available. The peak altitude of the F-layer, hmf2, is another parameter representing the ionospheric altitude. The relationship between hmf2 and the ionospheric transmission factor, M(3000)F2 has been developed since 1950s. Shimazaki [1955] first introduced the relationship between hmf2 and M(3000)F2, assuming that the electron density is distributed parabolically. The relationship is 1490 hmf2 ¼ 176 þ Mð3000ÞF2 : ð1þ Bradley and Dudeney [1973] took the underlying E-layer into consideration to the denominator of the second term of equation (1). However, since our interest is in nighttime when the E-layer disappears, the effect of the E-layer is 2of9

3 Figure 1. (top) The Doppler velocity within F-region FAIs against time and range, as observed with five beams of the 30.8-MHz radar at Kototabang, Indonesia, between 22:00 LT on 30 August and 06:00 LT on the subsequent day. The altitude is shown on the right axis. Azimuthal angles of the five beams were 234.2, 207.0, 180.0, 153.0, and ; the results for different beam angles are shown in different panels. The right axis shows altitude at which the radar beam is perpendicular to the geomagnetic field line. Positive (negative) Doppler velocity denotes motion away from (toward) the radar. (bottom) Time series of hmf2at Chumphon from 22:00 LT on 30 August to 06:00 LT on the subsequent day. Time of spread-f is indicated by a horizontal arrow. negligible. Therefore, in this study, hmf2 is adopted as a proxy of the ionospheric altitude and derived using equation (1). 3. Results [9] Here we investigate the relationship between the occurrence of post-midnight FAIs at a low latitude and variation of the ionospheric altitude near the geomagnetic equator. First, two cases when post-midnight FAIs were observed by the VHF radar are presented. Ionograms recorded at Chumphon are also presented to show the variation of ionospheric altitude. Following the case studies, a statistical studies on the FAIs occurrence and ionospheric altitude are conducted to discuss the relationship between the occurrence of post-midnight FAIs and variation of the ionospheric altitude Case Studies [10] Figure 1 shows the Doppler velocity within the F-region FAIs against range and time observed with five beams (west to east from top to bottom) on the night of 30 August Positive (negative) Doppler velocity denotes motion away from (toward) the radar; that is, upward/ southward (downward/northward). No FAIs were observed in the post-sunset period. After midnight, FAIs first appeared for the beam directed due south (Figure 1, third panel), which means that the FAIs were generated over Kototabang. The FAIs slowly propagated to the west. The Doppler velocity of plasma within the FAIs was upward in the early stage and the velocity decreased with time. [11] Prior to the FAI appearance at the low latitude, the altitude of the ionosphere changed near the geomagnetic equator. Figure 2 shows five ionograms obtained at Chumphon at intervals of 20 min from 23:15 LT (LT = UT + 7) on the night of 30 August h F could not be scaled from these ionograms since the plasma density at the bottom-side ionosphere was less than [el/cm 3 ], which corresponds to 2 MHz. At 23:15 LT, fof2 and M(3000)F2 were 3.5 MHz and 3.1, respectively. Using equation (1), the peak altitude of the ionosphere, hmf2, was estimated to be 308 km. Figures 2a 2e show that the virtual height of the traces increased with time. At 00:15 LT, fof2 and M(3000)F2 were 2.8 MHz and 2.9, respectively, and the trace started to spread. From equation (1), hmf2 was estimated to be 326 km. We found that hmf2 increased by 20 km during 1 hour from 23:15 LT on this day. The upward velocity was approximately 5 m/s on average during the 1 hour. A time series of hmf2 is plotted in Figure 1 (bottom). A rapid increase in the F-layer altitude was seen 1 hour prior to the FAIs appearance. [12] Figure 3a shows another example of post-midnight echoes, which were observed on the night of 14 August It shows the Doppler velocity within the F-region FAIs observed with five beams; the format is the same format as that for Figure 1. No FAIs were observed in the postsunset period; however, FAIs with Doppler velocity away from the radar appeared after 22:00 LT for the easternmost beam (fifth panel). The FAIs were also consequently observed for the western beams (upper panels), which indicates that the FAI propagated to the west. Around 00:00 LT, FAIs with Doppler velocity away from the radar appeared for the westernmost beam (top panel). The FAIs were consequently observed with the eastern beams (lower panels), which indicates that these FAIs propagated eastward. The propagating direction thus differed at different times on this day. [13] Figures 4a 4e are ionograms obtained at Chumphon every 30 min from 22:30 LT on the night of 14 August of9

4 Figure 2. Ionograms obtained at Chumphon every 20 min from 23:15 LT on 30 August Traces from the F-layer appeared at virtual height of km at 22:30 LT, which were lower than the heights on the night of 30 August (Figure 2). At 22:30 LT, fof2 and M(3000)F2 were 4.8 MHz and 3.7, respectively. From equation (1), the peak altitude was estimated to be 226 km. The traces started to spread after 22:30 LT. At 23:00 LT, fof2 and M(3000)F2 were 3.7 MHz and 3.6, respectively, yielding hmf2 of 237 km. Although there was some ambiguity in the scaled values because of the spread traces, height rise of approximately 10 km was observed between 22:30 and 23:00 LT. The upward velocity was approximately 5 m/s on average during the half hour. The time series of hmf2 is plotted in Figure 3a (bottom panel). It was found that the altitude of the F-layer increased about 30 min prior to the FAI appearance. After 23:00 LT, fof2 and M(3000)F2 were not available because the traces completely spread. However, the virtual height where the spread trace appeared increased with time. This indicates that the altitude of the ionosphere increased with time. After 00:00 LT, traces from the E-layer, which is, sporadic-e also appeared. [14] FAIs embedded in the sporadic-e layer also provide information on the electro-dynamics in the F-region. Figure 3b shows the Doppler velocity within the E-region FAIs observed with five beams, in the same format as that for Figure 3a. The Doppler velocity of the E-region FAIs was downward on the beam directed due south during the whole post-midnight F-region FAI event (third panel). The FAIs appeared at altitude between 100 and 120 km. The downward Doppler velocity of the upper E-region (above 100 km) is interpreted as the line-of-sight projection of the E B drift in the F-region [Tanaka and Venkateswaran, 1982; Patra, 2002]. The geomagnetic field lines where the E-region FAIs appeared connected to the F-region over the geomagnetic equator. Therefore, the downward velocity of E-region FAIs on the beam directed due south over Kototabang was interpreted to be due to the westward electric field in F-region at the geomagnetic equator. FAI echoes with positive Doppler velocity were also observed between 23:00 and 01:00 LT at altitudes from 80 to 120 km. The positive Doppler velocity was clearly seen on western beams (top two panels). These echoes appeared simultaneously with the F-region FAIs (see Figure 3a). The echoes were not related to the E-region FAIs but were range aliasing effects of the F-region echoes Statistical Study [15] Seasonal and local time variations of the F-region FAI occurrence were investigated by analyzing backscatter signal-to-noise ratios measured by the VHF radar at Kototabang, Indonesia. A backscatter signal-to-noise ratio larger than 0 db that extends more than 50 km in range is regarded as echos from FAIs. Red and blue sections of Figure 5 show the occurrence and non-occurrence of FAI echos against local time and day of year for each year from 2006 to An absence of color in Figure 5 indicates that there is no observation time owing to instrumental problems. The figure shows that FAIs appeared most frequently post-sunset (premidnight) around equinoxes and post-midnight between May and August. Pre-midnight (post-sunset) FAIs appeared frequently around March equinox in 2006, 2009, and 2010, and around the September equinox in The yearly average of solar flux F 10.7, which is a proxy of the solar activity, was 80, 73, 68, 70, and 80 in 2006, 2007, 2008, 2009, and 2010, respectively, as given at the top of each panel in Figure 5. The occurrence rates of the pre-midnight FAIs tended to correlate with the solar activity. On the other hand, post-midnight FAIs tended to appear more frequently in 2007, 2008, and 2009 than the other two years. They sometimes appeared around the December solstice but the occurrence rate was lower than that around the June solstice; that is, between May and August. It is found that the occurrence of the post-midnight FAIs tended to be more frequent when the solar activity was lower. [16] We also investigated seasonal and local time variations of the peak altitude of the ionosphere, hmf2, at Chumphon. The peak altitude was calculated from fof2 and M(3000)F2 every 15 min for two years. Figures 6a, 6b, and 6c present hmf2 against the local time and day of year in 2006, 2008, and 2009, respectively. There are some data gaps in May-June of 2006, in January-February of 2008, and in February-March of 2009 due to instrumental problems. By compensating for the gaps in each year s data, however, we found that hmf2 was enhanced post-sunset (pre-midnight) around equinoxes and post-midnight between May and 4of9

5 Figure 3. The Doppler velocity within (a) F-region and (b) E-region FAIs against time and range, as observed with five beams of the 30.8-MHz radar at Kototabang, Indonesia, between 22:00 LT on 14 August and 05:00 LT on the subsequent day. Positive (negative) Doppler velocity denotes motion away from (toward) the radar. The altitude is shown on the right axis. Figure 3a (bottom) shows hmf2 at Chumphon between 22:00 LT on 14 August and 05:00 LT on the subsequent day. The format is the same as that for Figure 1. August. The former enhancements are due to pre-reversal enhancements. The latter enhancements, the post-midnight enhancements, have occurrence characteristics completely different from those of the pre-reversal enhancements. 4. Discussion [17] In the present work, we presented two observations of post-midnight FAIs by the 30.8-MHz radar at Kototabang, Indonesia. We found that, preceding the FAIs appearance, F- layer uplifts were observed at Chumphon, Thailand, which is located near the geomagnetic equator in almost the same meridian as Kototabang. The FAIs often appeared post-sunset (pre-midnight) around equinoxes and post-midnight between May and August. The F-layer altitude at Chumphon also frequently increased post-sunset (pre-midnight) around equinoxes and post-midnight between May and August. It is not surprising that both the FAI occurrence and the hmf2 increased post-sunset in the equinox seasons since it is wellknown that the pre-reversal enhancement largely contributes to plasma bubble occurrence through the Rayleigh-Taylor instability around equinoxes. [18] On the other hand, the coincidence of the FAI occurrence and the hmf2 enhancements at post-midnight between May and August is first reported in this paper. The uplift of the F-layer can be considered to increase occurrence rate of plasma bubbles which could be generated by the Rayleigh-Taylor instability. Here, we focus our discussion on the effect of the F-layer uplift on the Rayleigh-Taylor 5of9

6 Figure 4. Ionograms obtained at Chumphon every 30 min from 22:30 LT on 14 August instability. To investigate the effect, we evaluate the linear growth rate of the Rayleigh-Taylor instability: g ¼ E B þ g 1 n in L ; ð2þ where E, B, g, n in, and L are the eastward electric field, the geomagnetic field, gravity acceleration, the ion-neutral collision frequency, and the scale length of the vertical gradient of the plasma density in the F-region, respectively. The growth rate increases with eastward effective electric field driven by eastward electric field (E) and g B drift (g/n in ; gravity-driven current). At night, E is westward [Fejer et al., 1991], and the growth rate is thus negative or small. On the other hand, g/n in increases during night for the solar minimum condition because n in is proportional to the neutral density, n, which is smaller during the night than during the daytime and evening and decreases with decreasing solar activity. g/n in is also larger at higher altitude because n decreases with altitude exponentially. Therefore, the uplift of the F-layer increases the growth rate. [19] Figure 7a shows seasonal variation of the F-layer altitude. Each cross represents hmf2 at Chumphon at 00:00 LT, and the red curve shows the 3-month running average of hmf2. We assume that the altitude of the bottom-side of the F-layer was 40 km below hmf2 (blue curve). Chapagain et al. [2009] reported that plasma bubble is initiated at the altitude of the bottom-side F-layer. The average altitudes of the bottom-side F-layer (h F) and the peak height of the F-layer (hmf2) were 290 km and 330 km, respectively [Chapagain et al., 2009; Lee et al., 2008], at Jicamarca during a solar minimum period. Therefore, in this study, we regarded the altitude of 40 km below hmf2 as the altitude where the plasma bubble could be initiated. Figure 5. Seasonal and local time variations in the F-region FAI echo observed at Kototabang, Indonesia, from 2006 to A signal-to-noise ratio larger than 0 db was regarded as echoes from FAIs. Red and blue sections represent FAI occurrence and non-occurrence, respectively, against local time and day of year for each year from 2006 to Yearly average of the solar flux F 10.7 is given at the top of each panel. 6of9

7 Figure 6. Seasonal and local-time variations of hmf2 at Chumphon in 2006, 2008, and [20] Figure 7b shows the seasonal variation of g/n in at a fixed altitude of 220 km (black) and at an altitude of (hmf2 40 km) (blue). An enhancement of g/n in at an altitude of (hmf2 40 km) is seen between May and August, whereas the growth rate at an altitude of 220 km does not show distinct enhancement. This result indicates that the uplift of the F-layer increases the growth rate. The seasonal average of E B drift velocity (E/B) is approximately 10 m/s at night for the solar minimum condition [Fejer et al., 1991]. Since E/B is smaller than g/n in, the growth rate is positive. Consequently, around midnight in the solar minimum period, the Rayleigh-Taylor instability may operate through the enhanced gravity-driven eastward electric current and plasma bubbles may be generated. [21] The next question is what makes the F-layer uplift at the geomagnetic equator. It is well known that when the eastward electric field enhances, it shifts the F-region plasma upward by E B drift. The F-layer uplift at the evening terminator, which is due to the strengthening of the eastward electric field, is widely known as pre-reversal enhancement. During the night, however, the electric field is westward on average [e.g., Fejer et al., 1991]. Moreover, the downward Doppler velocity of the E-region FAIs as seen in Figure 3b suggests that the nighttime electric field was westward when the F-layer was uplifted over the geomagnetic equator. Therefore, the uplift of the F-layer around midnight could not be due to the electric field. [22] One possible process that can be responsible for the apparent higher F-layer height is recombination of plasma. The apparent vertical velocity of the F-layer is governed by the recombination process under some conditions; such as the F-layer heights below 300 km while it is governed by the E B drift [Bittencourt and Abdu, 1981]. Nicolls et al. Figure 7. (a) Seasonal variation of ionospheric altitude; crosses represent hmf2 derived from fof2 and M(3000)F2 at Chumphon. The running average of the daily hmf2 with a window of three months is shown by the red curve. The blue curve shows the altitude of the bottom-side ionosphere, which is defined as the altitude 40 km lower than the averaged hmf2 (red curve). (b) Estimated g B drift velocity; the blue curve shows the g B drift at the bottom-side of the ionosphere as defined in the top panel. The black curve shows the g B drift at a fixed altitude of 220 km. 7 of 9

8 [2006] conducted a model calculation of the F-layer behavior for a solar minimum period under recombination and vertical drift to explain midnight hmf2 increases that were observed at Jicamarca. They reproduced the F-layer uplift by taking the weakening westward electric field (e.g., the downward drift velocity decreases from 20 to 3 m/s between 01:00 and 03:00 LT) into account. The weakening westward electric field in conjunction with sufficient recombination would play an important role in the F-layer uplift. According to observations at Jicamacra, the westward electric field weakened clearly at night around solstices during the solar minimum period [Fejer et al., 1991]. This seasonal and solar cycle dependence of the electric field is consistent with that of the post-midnight FAIs appearance. [23] The cause of the weakening electric field might be the midnight temperature maximum (MTM) since the thermospheric wind mainly controls electric field through the F- region dynamo [Nicolls et al., 2006]. It has been reported that MTM tends to be clearly observed around solstices during solar minimum periods in the Asian sector [Niranjan et al., 2006]. This seasonal and solar cycle dependence is also consistent with that of the weakening westward electric field and the post-midnight FAIs appearance. [24] Another possible process is that transequatorial wind transports plasma along the geomagnetic field to the leeward hemisphere and affects the apparent altitude over the geomagnetic equator. Maruyama [1996] showed using model calculations that when the transequatorial wind velocity is high (e.g., 100 m/s), it appreciably raises the ionospheric altitude during the evening hours. In general, the transequatorial wind is southward around June solstices. The multipoint ionosonde observation conducted by Maruyama et al. [2008] reveals that the southward wind velocity between May and August was high and close to 100 m/s between 22:00 and 01:00 LT. The seasonal dependence of the southward wind around midnight is consistent with that of the F-layer uplift in Chumphon. Therefore, we speculate that the strong southward wind at night around June solstice triggered the F-layer uplift, which would reinforce the Rayleigh-Taylor instability. [25] It is reported that an order of 20 m/s is a necessary condition for the post-sunset plasma bubble generation during a solar minimum period [Basu et al., 1996]. On the other hand, upward velocity of approximately 5 m/s was observed before the post-midnight FAIs generation. Even if the upward velocity of approximately 5 m/s was caused by E B drift, this velocity was much smaller than g B drift estimated in our study. The contribution of the g B drift was crucial for the post-midnight FAI generation. [26] FAIs can also be generated within Traveling Ionospheric Disturbances (TIDs) [e.g., Fukao et al., 1991; Kelley and Fukao, 1991] and not only within plasma bubbles. We cannot deny the possibility that the post-midnight FAIs are generated within medium-scale TIDs (MSTIDs). Previous studies with EAR at Kototabang suggest that some postmidnight FAIs are midlatitude type FAIs that are accompanied by nighttime MSTIDs [Fukao et al., 2003; Yokoyama et al., 2011a, 2011b]. Yokoyama et al. [2011b] reported that the midlatitude type FAIs, which are characterized by negative range rates of irregularity echoes in altitude-timeintensity format, moved southeast to northwest. The early echoes which appeared on 14 August 2006 (Figure 3) had the similar feature as that of the midlatitude type FAIs. Seasonal variation of the post-midnight FAI occurrence was similar to that of MSTID occurrence in the midlatitude ionosphere in the Asian sector; MSTIDs are often observed around June solstice [e.g., Tsugawa et al., 2007; Otsuka et al., 2011]. The similarity of the occurrence characteristics between midlatitude MSTIDs and low-latitude postmidnight irregularities has also been reported for the Brazilian region [Candido et al., 2008, 2011]. There is a possibility that MSTIDs propagate to low-latitude and are observed at Kototabang. However, Shiokawa et al. [2002] reported that there would be a limit to MSTID propagation around 18 N geomagnetic latitude, according to a case study. Here, we have to note that the observation in Shiokawa et al. [2002] was conducted in 1999, when the solar activity was much higher than those of the current study. Plasma densities during the recent solar minimum were much lower than those during the higher solar activity period. The limit of the MSTID propagation might be related by the plasma density. Seamless observations from the low- to midlatitudes ionosphere should be continuously conducted to reveal the relationship between the midlatitude MSTID and the post-midnight FAIs. 5. Conclusion [27] We investigated the relationship between postmidnight F-region FAIs observed by a 30.8-MHz radar at Kototabang in Indonesia and the F-layer altitude observed by an ionosonde at Chumphon in Thailand. Chumphon is located near the geomagnetic equator on almost the same meridian as Kototabang. Through case studies, we found that the altitude of the F-layer increased near the geomagnetic equator a half hour before the post-midnight FAIs appeared at Kototabang. The Doppler velocity of the E-region FAIs observed simultaneously by the radar was downward, indicating that the F-layer uplift was not caused by the electric field. Statistical study shows that there was both frequent post-midnight FAIs occurrence and frequent midnight F-layer uplift between May and August. [28] These results suggest that the uplift of the F-layer would play an important role in the generation of postmidnight FAIs. We evaluated the linear growth rate of the Rayleigh-Taylor instability based on the altitude of the F-layer observed at Chumphon. The result shows that the uplift of the F-layer can enhance the growth rate of Rayleigh- Taylor instability because the gravity-driven eastward electric current is enhanced. Consequently, the observed FAIs would be accompanied by plasma bubble, whose growth rate is reinforced by the uplifted F-layer. The cause of the F-layer uplift would not be the eastward electric field. Recombination or the transequatorial wind would contribute to the uplift of the F-layer. [29] Acknowledgments. This work was supported by Grants-in-Aid for Scientific Research ( and ) and Strategic Funds for the Promotion of Science and Technology provided by the Ministry of Education, Culture, Sports, Science and Technology of Japan. The MHz radar is operated by STEL, Nagoya University, in collaboration with LAPAN and RISH, Kyoto University. We thank Mr. Yamazaki of NICT for his help with scaling ionograms. [30] Robert Lysak thanks the reviewers for their assistance in evaluating the paper. 8of9

9 References Basu, S., et al. (1996), Scintillations, plasma drifts, and neutral winds in the equatorial ionosphere after sunset, J. Geophys. Res., 101(A12), 26,795 26,809, doi: /96ja Bittencourt, J. A., and M. A. Abdu (1981), A theoretical comparison between apparent and real vertical ionization drift velocities in the equatorial F region, J. Geophys. Res., 86(A4), , doi: / JA086iA04p Bradley, P. A., and J. R. Dudeney (1973), A simple model of the vertical distribution of electron concentration in the ionosphere, J. Atmos. Terr. Phys., 35, Burke, W. J., L. C. Gentile, C. Y. Huang, C. E. Valladares, and S. Y. Su (2004), Longitudinal variability of equatorial plasma bubbles observed by DMSP and ROCSAT-1, J. Geophys. Res., 109, A12301, doi: /2004ja Candido, C. M. N., A. A. Pimenta, J. A. Bittencourt, and F. Becke-Guedes (2008), Statistical analysis of the occurrence of medium-scale traveling ionospheric disturbances over Brazilian low latitudes using OI nm emission all-sky images, Geophys. Res. Lett., 35, L17105, doi: /2008gl Candido, C. M. N., I. S. Batista, F. Becker-Guedes, M. A. Abdu, J. H. A. Sobral, and H. Takahashi (2011), Spread F occurrence over a southern anomaly crest location in Brazil during June solstice of solar minimum activity, J. Geophys. Res., 116, A06316, doi: /2010ja Chapagain, N. P., B. G. Fejer, and J. L. Chau (2009), Climatology of postsunset equatorial spread F over Jicamarca, J. Geophys. Res., 114, A07307, doi: /2008ja Dao, E., M. C. Kelley, P. Roddy, J. Retterer, J. O. Ballenthin, O. de La Beaujardiere, and Y.-J. Su (2011), Longitudinal and seasonal dependence of nighttime equatorial plasma density irregularities during solar minimum detected on the C/NOFS satellite, Geophys. Res. Lett., 38, L10104, doi: /2011gl Fejer, B. G., E. R. de Paula, S. A. Gonzalez, and R. F. Woodman (1991), Average vertical and zonal F region plasma drifts over Jicamarca, J. Geophys. Res., 96(A8), 13,901 13,906, doi: /91ja Fukao, S., M. C. Kelley, T. Shirakawa, T. Takami, M. Yamamoto, T. Tsuda, and S. Kato (1991), Turbulent upwelling of the mid-latitude ionosphere, 1. Observational results by the MU radar, J. Geophys. Res., 96 (A3), , doi: /90ja Fukao, S., Y. Ozawa, M. Yamamoto, and R. T. Tsunoda (2003), Altitudeextended equatorial spread F observed near sunrise terminator over Indonesia, Geophys. Res. Lett., 30(22), 2137, doi: /2003gl Heelis, R. A., R. Stoneback, G. D. Earle, R. A. Haaser, and M. A. Abdu (2010), Medium-scale equatorial plasma irregularities observed by Coupled Ion-Neutral Dynamics Investigation sensors aboard the Communication Navigation Outage Forecast System in a prolonged solar minimum, J. Geophys. Res., 115, A10321, doi: /2010ja Kelley, M. C., and S. Fukao (1991), Turbulent upwelling of the midlatitude ionosphere, 2. Theoretical framework, J. Geophys. Res., 96(A3), , doi: /90ja Kelley, M. C., R. Pfaff, K. D. Baker, J. C. Ulwick, R. Livingston, C. Rino, and R. Tsunoda (1982), Simultaneous rocket probe and radar measurements of equatorial spread Transitional and short wavelength results, J. Geophys. Res., 87(A3), , doi: /ja087ia03p Lee, C. C., B. W. Reinisch, S.-Y. Su, W. S. Chen (2008), Quiet-time variations of F2-layer parameters at Jicamarca and comparison with IRI-2001 during solar minimum, J. Atmos. Solar-Terr. Phys., 70, Maruyama, T. (1996), Modeling study of equatorial ionospheric height and spread F occurrence, J. Geophys. Res., 101(A3), , doi: /95ja Maruyama, T., M. Kawamura, S. Saito, K. Nozaki, H. Kato, N. Hemmakorn, T. Boonchuk, T. Komolmis, and C. Ha Duyen (2007), Low-latitude ionosphere-thermosphere dynamics studies with inosonde chain in Southeast Asia, Ann. Geophys., 25, , doi: /angeo Maruyama, T., S. Saito, M. Kawamura, and K. Nozaki (2008), Thermospheric meridional winds as deduced from ionosonde chain at low and equatorial latitudes and their connection with midnight temperature maximum, J. Geophys. Res., 113, A09316, doi: /2008ja McClure, J. P., W. B. Hanson, and J. H. Hoffman (1977), Plasma bubbles and irregularities in the equatorial ionosphere, J. Geophys. Res., 82(19), , doi: /ja082i019p Miller, E. S., J. J. Makela, K. M. Groves, M. C. Kelley, and R. T. Tsunoda (2010), Coordinated study of coherent radar backscatter and optical airglow depletions in the central Pacific, J. Geophys. Res., 115, A06307, doi: /2009ja Nicolls, M. J., M. C. Kelley, M. N. Vlasov, Y. Sahai, J. L. Chau, D. L. Hysell, P. R. Fagundes, F. Becker-Guedes, and W. L. C. Lima (2006), Observations and modeling of post-midnight uplifts near the magnetic equator, Ann. Geophys., 24, Niranjan, K., P. S. Brahmanandam, and B. Srivani (2006), Signatures of equatorial midnight temperature maximum as observed from in situ and ground-based ionospheric measurements in the Indian sector, J. Geophys. Res., 111, A07309, doi: /2005ja Nishioka, M., A. Saito, and T. Tsugawa (2008), Occurrence characteristics of plasma bubble derived from global groundbased GPS receiver networks, J. Geophys. Res., 113, A05301, doi: /2007ja Otsuka, Y., K. Shiokawa, T. Ogawa, T. Yokoyama, M. Yamamoto, and S. Fukao (2004), Spatial relationship of equatorial plasma bubbles and field aligned irregularities observed with an all-sky airglow imager and the Equatorial Atmosphere Radar, Geophys. Res. Lett., 31, L20802, doi: /2004gl Otsuka, Y., T. Ogawa, and Effendy (2009), VHF radar observations of nighttime F-region field-aligned irregularities over Kototabang, Indonesia, Earth Planets Space, 61, Otsuka, Y., N. Kotake, K. Shiokawa, T. Ogawa, T. Tsugawa, and A. Saito (2011), Statistical study of medium-scale traveling ionospheric disturbances observed with a GPS receiver network in Japan, in Aeronomy of the Earth s Atmosphere and Ionosphere, IAGA Spe. Sopron Book Ser., vol. 2, edited by M. A. Abdu, D. Pancheva, and A. Bhattacharyya, pp , Springer, Dordrecht, Netherlands. Otsuka, Y., K. Shiokawa, M. Nishioka, and Effendy (2012), VHF radar observations of post-midnight F-region field-aligned irregularities over Indonesia during solar minimum, Indian J. Radio Space Phys., in press. Patra, A. K. (2002), On the interpretation of the VHF radar echoes from the equatorial lower E region, Geophys. Res. Lett., 29(14), 1695, doi: /2001gl Saito, S. and T. Maruyama (2006), Ionospheric height variations observed by ionosondes along magnetic meridian and plasma bubble onsets, Ann. Geophys., 24, , doi: /angeo Sastri, J. H. (1999), Post-midnight onset of spread-f at Kodaikanal during the June solstice of solar minimum, Ann. Geophys., 17, Shimazaki, T. (1955), World-wide daily variations in the height of the maximum electron density of the ionospheric F2 layer, J. Radio Res. Labs., 2 (7), Shiokawa, K., Y. Otsuka, M. K. Ejiri, Y. Sahai, T. Kadota, C. Ihara, T. Ogawa, K. Igarashi, S. Miyazaki, and A. Saito (2002), Imaging observations of the equatorward limit of midlatitude traveling ionospheric disturbances, Earth Planets Space, 54, Subbarao, K. S. V., and B. V. Krishna Murthy (1994), Seasonal variations of equatorial spread-f, Ann. Geophys., 12, 33 39, doi: /s Tanaka, T., and S. V. Venkateswaran (1982), Characteristics of fieldaligned E-region irregularities over Iioka (36 N), Japan-II, J. Atmos. Terr. Phys., 44(5), Tsugawa, T., N. Kotake, Y. Otsuka, and A. Saito (2007), Medium-scale traveling ionospheric disturbances observed by GPS receiver network in Japan: A short review, GPS Solut., 11, , doi: /s Tsunoda, R. T. (1980), On the spatial relationship of 1-m equatorial spread F irregularities and plasma bubbles, J. Geophys. Res., 85(A1), , doi: /ja085ia01p Tsunoda, R. T. (1985), Control of the seasonal and longitudinal occurrence of equatorial scintillations by the longitudinal gradient in integrated E-region Pedersen conductivity, J. Geophys. Res., 90(A1), , doi: /ja090ia01p Yokoyama, T., S. Fukao, and M. Yamamoto (2004), Relationship of onset of equatorial F-region irregularities with sunset terminator observed with the Equatorial Atmosphere Radar, Geophys. Res. Lett., 31, L24804, doi: /2004gl Yokoyama, T., R. F. Pfaff, P. Roddy, M. Yamamoto, and Y. Otsuka (2011a), On post-midnight low-latitude ionospheric irregularities during solar minimum: 2. C/NOFS observations and comparisons with the Equatorial Atmosphere Radar, J. Geophys. Res., 116, A11326, doi: / 2011JA Yokoyama, T., R. F. Pfaff, M. Yamamoto, Y. Otsuka, M. Nishioka, and T. Tsugawa (2011b), On postmidnight low-latitude ionospheric irregularities during solar minimum: 1. Equatorial Atmosphere Radar and GPS TEC observations in Indonesia, J. Geophys. Res., 116, A11325, doi: /2011ja of9