Investigation of height gradient in vertical plasma drift at equatorial ionosphere using multifrequency HF Doppler radar

Transcription

1 JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 109,, doi: /2004ja010641, 2004 Investigation of height gradient in vertical plasma drift at equatorial ionosphere using multifrequency HF Doppler radar S. R. Prabhakaran Nayar and C. V. Sreehari Department of Physics, University of Kerala, Trivandrum, India Received 22 June 2004; revised 21 September 2004; accepted 25 October 2004; published 28 December [1] A multifrequency HF Doppler radar installed at the magnetic equatorial station Trivandrum provides an opportunity to study the height gradient in vertical plasma drift at the bottomside of equatorial F region during evening time. The multifrequency radar gives near-simultaneous observation of vertical plasma drift at three close by F region heights above the sounding station. The height gradient of the vertical drift shows a negative value during the prereversal enhancement (PRE) period and turns to positive value after the prereversal enhancement. The average height gradient in vertical plasma drift remains negative around PRE and its magnitude decreases with altitude, below F peak. This could be a clear-cut manifestation of the curl-free nature of the low-latitude electric field, and it could also indicate a partial signature of the postsunset velocity vortex at the equatorial F region. The magnitude of the mean height gradient around PRE exhibits a seasonal variation. INDEX TERMS: 2415 Ionosphere: Equatorial ionosphere; 2411 Ionosphere: Electric fields (2712); 2437 Ionosphere: Ionospheric dynamics; 2407 Ionosphere: Auroral ionosphere (2704); 0694 Electromagnetics: Instrumentation and techniques; KEYWORDS: height gradient, velocity vortex, vertical plasma drift Citation: Prabhakaran Nayar, S. R., and C. V. Sreehari (2004), Investigation of height gradient in vertical plasma drift at equatorial ionosphere using multifrequency HF Doppler radar, J. Geophys. Res., 109,, doi: /2004ja Introduction [2] The low-latitude ionosphere has been studied for over four decades using extensive ground based observations, space-borne instruments and theoretical and numerical models. The general properties of equatorial F region plasma drifts and electric fields were determined mostly from incoherent scatter radar measurements at Jicamarca (11.9 S, 76.8 W, dip 2 N) [Woodman, 1970]. The vertical plasma drift at the equatorial F region during the evening period is primarily driven by a zonal electric field formed as a part of the vertical polarization field developed due to enhanced thermospheric winds and decay of E region conductivity [Rishbeth, 1971; Farley et al., 1986; Haerendel and Eccles, 1992; Eccles, 1998a]. Single frequency HF Doppler radars at Trivandrum (8.33 N, 77.0 E, dip 0.4 N) [Namboothiri et al., 1989; Balan et al., 1992] and Kodaikanal (10.2 N, 77.5 E, dip 3 N) [Ramesh and Sastri, 1995] were used to study the characteristics of F region vertical drifts at the equatorial ionosphere at the Indian sector. Jicamarca observations correspond to the height interval of km, whereas the HF Doppler data corresponds to reflection height of the F layer, which varies with local time. The general behavior of the vertical drift patterns measured with HF Doppler radar technique is consistent with those of the Jicamarca drifts [Fejer, 1997]. The Copyright 2004 by the American Geophysical Union /04/2004JA equatorial vertical plasma drift and the corresponding zonal electric field exhibits a sharp enhancement after local sunset, known as the prereversal enhancement (PRE). The PRE in the vertical plasma drift at the equatorial F region is due to the combined effect of the dynamo effect of the F region neutral wind blowing eastward near the dusk and the changes in the E region conductivity near the sunset terminator at conjugate magnetic latitudes [Rishbeth, 1971; Farley et al., 1986; Eccles, 1998b]. A simple model based on eastward F region wind during evening time can account for the prereversal enhancement of electric field quite adequately [Rishbeth, 1971]. The variations in the PRE are explained using variations in F region wind, changes in E/F region conductivity ratio, E region dynamo contributions, strength of equatorial electrojet, effect of lack of symmetry between the conjugate E regions of the northern and southern hemispheres and the separation between geographic and dip equators [Batista et al., 1986; Abdu et al., 1993]. [3] Murphy and Heelis [1986] estimated the existence of height gradient in vertical drift using Jicamarca vertical and zonal drifts, to be consistent with the curl free nature of the low-latitude electric field. Pingree and Fejer [1987] observed height gradient in the evening equatorial F region vertical drifts using Jicamarca data. Further, negative height gradient in vertical plasma drift was reported from ionosonde derived vertical drifts on two locations in Brazil [Abdu et al., 1993] and in Trivandrum [Subbarao and Krishna Murthy, 1994]. Sastri et al. [1995] reported a 1of7

2 small positive height gradient in evening F region vertical drifts using simultaneous data from Trivandrum and Kodaikanal. Kudeki et al. [1981] reported a height gradient in zonal plasma drift in the postsunset F region. Kudeki and Bhattacharyya [1999] reported very interesting profiles of a postsunset velocity vortex from vertical and zonal drift estimates of Jicamarca radar using a new experimental and processing scheme. Eccles et al. [1999] used San Marco-D electric field measurements to investigate the evening plasma drift vortex and compared with model and theory of low latitude electric fields. Maruyama et al. [2002] have also found the signature of a vortex-like flow pattern in the evening F region from simultaneous ionosonde measurements over two stations in Philippines. The velocity vortex has an important role in the electrodynamics of the post sunset equatorial ionosphere. The observations of height gradient in vertical plasma motion at the bottomside of F layer provide an indirect evidence of velocity vortex during evening hours. In this background, this paper presents first results of the height gradient in vertical plasma drift and its variations below F peak in the evening hours using a multifrequency HF Doppler radar. 2. Experimental Method and Observation Scheme [4] The multifrequency HF Doppler radar installed at the University of Kerala is a coherent monostatic pulsed system capable of quadrature detection of ionospheric echoes. It uses a single broadband transmitter made to work at three fixed frequencies: 2.5, 3.5, and 4.5 MHz. Pulsed RF signal of desired frequency (f 0 ) is generated by a frequency synthesizer unit which employs direct digital synthesis (DDS) technique and is amplified and transmitted by the broadband transmitter through a three element folded dipole antenna. The Doppler shifted reflected signal (f 0 ± f D ) from the ionosphere is received by a dipole antenna, the same kind of one used for transmission. The received signals are fed to the phase coherent receivers to extract Doppler signals. The quadrature components of these signals are digitized using the data acquisition system and this Doppler data is Fourier analyzed to evaluate the Doppler frequency (f D ) at regular one minute intervals. The magnitude and direction (up/down) of the vertical plasma drift is evaluated from f D as V = f D l/2 where l is the operating wavelength. The receiver, data acquisition hardware and software are accurately standardized after transmitting and recording standard very low-frequency signals. The uncertainties in the evaluation of the magnitude of vertical drifts are 0.25, 0.18, and 0.14 m s 1 for 2.5, 3.5, and 4.5 MHz respectively. [5] In the case of single frequency operation, for every one minute, one vertical drift data point is evaluated. For multifrequency operation, a sequential switching scheme is followed such that the radar is operated in one frequency for one minute and after a delay of one minute the next frequency is operated for one minute and then the third frequency and so on. The reflection height for each frequency is continuously changing in the post sunset period. The reflection height is measured from the time delay between the transmitted pulse and the echo for each operating frequency and has a resolution of 1.5 km. The height gradient between a pair of frequencies is evaluated from the difference in vertical drifts and difference in the corresponding reflection heights of the pair of probing frequencies. The one-minute delay in between data collections for two frequencies is required for changing the frequency settings of the radar. Thus we are getting nearly simultaneous vertical drift data at three different heights. One set of drift measurements at three heights is obtained within six minutes. With these measurements, we are able to calculate height gradient at three altitudes between the reflection heights of (1) 2.5 and 3.5 MHz, (2) 3.5 and 4.5 MHz, and (3) 2.5 and 4.5 MHz. The third set will behave like the the mean of the first two. [6] The plasma drift observed by HF Doppler radar is apparent as it includes the contribution due to chemical loss induced by recombination reactions. This is to be considered because on some of the days, the reflection height is below 300 km, where layer decay is prominent in the evening time [Bittencourt and Abdu, 1981]. To account for the vertical velocity introduced due to chemical loss, a correction scheme is applied to the measured vertical drift. The loss coefficient b during evening time is given by, b = k 1 n[o 2 ]+k 2 n[n 2 ], [Rishbeth, 1986]. The rate coefficients k 1 and k 2 were calculated using the relations, k 1 = (T n /300) 1/2 and k 2 = [Anderson and Rusch, 1980]. The number densities of O 2 and N 2 and the neutral temperature T n are calculated using MSIS-86 model [Hedin, 1987] at each height and local time corresponding to the vertical drift data. The velocity due to chemical loss bh is then estimated assuming a reasonable value for the electron density scale height (H) as 10 km. Hari et al. [1996] have evaluated the electron density scale height at the F region during evening hours using ionosonde at Trivandrum between 2.5 and 3 MHz as 10 km. Similarly Ramesh and Sastri [1995] also have assumed a value of scale height 10 km at 4 MHz for explaining the seasonal variation in F region vertical drifts. The correction in the vertical drift due to the chemical loss is of the order of 1 2ms 1 and is subtracted from the observed drift velocity to get the true electrodynamic vertical drift (V z ). The scale height is expected to vary with time and with solar conditions. Since the probing altitudes for the selected operational frequencies are close by, evaluation of the height gradient in vertical drift may not be significantly affected by the variation in H. During daytime, due to the presence of equatorial electrojet, the reflected signal from the F region received by the HF Doppler radar is weak especially around noon time. So, for the present study, the HF Doppler radar operation is restricted to local evening to morning hours. The general behavior of PRE in vertical drift observed by HF Doppler radar on 3 February 2004 is consistent with global equatorial vertical drift model [Scherliess and Fejer, 1999]. The magnitude of the peak value of V z observed by the HF Doppler radar at Trivandrum is higher by 10 m s 1 compared to the model value as noticed in earlier comparisons by Scherliess and Fejer [1999]. To investigate the altitude dependence of vertical drift, the HF Doppler radar system was operated both in single frequency and multifrequency modes during August 2003 to April For the present study, from 80 days of single frequency and 40 days of multifrequency observa- 2of7

3 Figure 1. Multifrequency sounding results on 3 February Variations in (a) vertical drifts, (b) reflection heights corresponding to three operating frequencies, and (c) height gradients in three altitude ranges. tions made during evening hours, 60 quiet days of single frequency and 20 quiet days of multifrequency observations are selected after avoiding the days with spread F and considerable geomagnetic activity. 3. Height Gradient in Vertical Plasma Drift and Its Variation [7] Observations of vertical plasma drift (V z ) at different reflection heights reveal that the plasma motion is height dependent in the range 260 to 420 km. In general, the plasma drift observed decreases with sounding frequency while the reflecting height increases with increase in sounding frequency. On an average, the height gradient of the vertical plasma drift (dv z /dh) has a negative value around the evening prereversal enhancement (PRE) period. Figure 1 presents the time variation in V z, reflection heights and the height gradients in vertical drift corresponding to three probing frequencies observed on 3 February Figure 1a depicts V z observed at three different height regions presented in Figure 1b. It is noticed that up to the PRE period, V z increases and is found to be larger for lower height regions and decreases with height (frequencies). Beyond PRE period, V z goes on decreasing and the height of observation increases up to the reversal point around 19:30 LT. The time variation of (dv z /dh) for three height regions, i.e., for the three pairs of frequencies (1) 2.5 and 3.5 MHz (open circles), (2) 3.5 and 4.5 MHz (solid circles), and (3) between 2.5 and 4.5 MHz (squares) are presented in Figure 1c. It is observed that the height gradients have negative values during the prereversal period. Beyond PRE, the height gradient in V z fluctuates between positive and negative values with a mean positive value. The height gradient obtained between probing frequencies 3.5 and 4.5 MHz is at a higher height compared to that between 2.5 and 3.5 MHz. From Figure 1c it is clear that the magnitude of the negative height gradient of vertical plasma drift prior to the prereversal period is consistently less at higher heights in the height region below the F layer peak. The height gradient evaluated between 2.5 and 4.5 MHz exhibits a median value between the other two. So, around PRE period, the height gradients in V z have negative value at the equatorial ionosphere and its magnitude decreases with altitude. [8] To investigate the nature of the mean height gradient, multifrequency measurements of vertical drift in the evening F region for the 20 quiet days with A p 25 are selected. Figure 2 depicts the time variation of the mean (dv z /dh) at 3of7

4 Figure 2. Average of the observed height gradients between a pair of operating frequencies: (a) 2.5 and 3.5 MHz, (b) 3.5 and 4.5 MHz, and (c) 2.5 and 4.5 MHz. (d) Mean height gradient for the three categories. heights corresponding to operating frequencies (1) between 2.5 and 3.5 MHz in Figure 2a, (2) between 3.5 and 4.5 MHz in Figure 2b, (3) between 2.5 and 4.5 MHz in Figure 2c, and (4) average of the above three data sets in Figure 2d. From a comparison of Figures 2a and 2b it is noticed that (dv z /dh) is more prominent at lower altitudes (Figure 2a). The magnitude of standard deviations are represented by vertical bars in Figure 2. The value of the standard deviation is caused by the distribution of height gradient data over different altitudes and seasons. In general, the pattern of the height gradient is negative till the PRE time and then sharply changes to positive values when the plasma drift tends to decrease after PRE. After the reversal, (dv z /dh) oscillate between positive and negative values. The mean value of all height gradients between 17:00 LT and 21:00 LT for these 20 days is m s 1 km 1 or s 1. The mean height gradient between 18:00 LT and 19:00 LT is m s 1 km 1 and the mean gradient beyond 19:00 LT is m s 1 km 1. [9] In addition to the multifrequency sounding, the HF Doppler radar was operated in single-frequency mode (either at 2.5 MHz or 3.5 MHz or 4.5 MHz) for a number of days. The peak value of vertical drift during the PRE (V zp ) for each day and the corresponding reflection height (h zp ) are noted. The seasonal variation of V zp for operating frequency 3.5 MHz is presented in Figure 3a. The observed values of V zp (circles) are presented in Figure 3a. The seasonal trend in V zp is obtained by fitting an eighth-order polynomial to the observed values and presented as curve in Figure 3a. Similar polynomial fits are obtained for V zp and for the corresponding heights h zp for each operating frequency separately using the multifrequency observation data and the fitted curves are depicted in Figures 3b and 3c. In Figures 3b and 3c, to avoid cluttering of data points, fitted curves alone are shown. The peak values of PRE (V zp ) depicted in Figures 3a and 3b and reflection height h zp in Figure 3c exhibit maximum around the September equinox. The height gradient between 18:00 LT and 19:00 LT is averaged separately for the frequency ranges and MHz for each day of multifrequency operation and are depicted in Figure 3d. The circles in Figure 3d indicate the gradient values observed between 3.5 and 4.5 MHz and the dashed line indicates the sixthorder polynomial fit of the corresponding data. Similarly, the asterisks in Figure 3d indicate the height gradient observations made between frequencies 2.5 and 3.5 MHz, and the solid line indicates the polynomial fit. The magnitude of the vertical velocity height gradient is larger for lower probing frequency (lower altitudes) compared to higher probing frequencies (higher altitudes). The magnitude of the average 4of7

5 Figure 3. Seasonal variation of (a) V zp corresponding to 3.5 MHz observation (curve indicates the polynomial fit) and (b) polynomial fits to V zp for each observation frequency. (c) Polynomial fits to h zp for three frequencies and (d) the mean height gradients around PRE (1) between 2.5 and 3.5 MHz (asterisks and solid line) and (2) between 3.5 and 4.5 MHz (circles and dashed line). negative height gradient around PRE exhibits a seasonal trend with larger magnitude around December solstice. 4. Discussion [10] At equatorial latitudes, both the geomagnetic field and the F region zonal electric field are almost perpendicular to each other. Then, from Maxwell s equations, r E = (@B/@t) and if one assumes local time invariance of geomagnetic field during quiet conditions, re = 0. i.e., The electric field should be curl-free or irrotational. On the basis of this fact, Murphy and Heelis [1986] developed a model connecting the drift velocity components. General behavior of height gradients presented in Figure 1 are in agreement with their model prediction, which signifies the curl-free condition of low-latitude electric field during the post sunset period. Eccles [1998b] compared various theoretical models in finding the actual driving mechanism for the prereversal enhancement and concluded that the curl-free nature itself is the real motivation behind the zonal electric field enhancement. The curl-free nature was in fact theoretically suggested by Rishbeth [1971] indicating that the vertical polarization electric field lines are curved [Rishbeth, 1971, Figure 5] during sunset. Then the vertical component of plasma drift will have to be decreased with increase in altitude below the F peak, of course, satisfying the curl-free nature at every height. Observations presented in Figure 1 correspond to altitude region km range, below the F region peak. In this altitude range the vertical plasma drift exhibits a clear altitude dependence with V z decreasing with altitude. [11] Farley et al. [1986] explained the observed PRE on the basis of conjugate E region physics. By their modeling study, the evening F region vertical polarization field maps to conjugate E regions via nearly equipotential geomagnetic field lines. However, the weak conductivity of nightside hemisphere stimulates the generation of an east-west electric field at the conjugate E regions which when maps back to the equatorial F region producing the prereversal enhancement of zonal electric field. This is the case with a perfectly aligned magnetic meridian and sunset terminator, which is hardly realized in geomagnetic environment. Abdu et al. [1993] discuss possible consequences of the noncoincidence of magnetic meridian and sunset terminator. If there is such a noncoincidence, the separation between magnetic meridian and sunset terminator increases as one moves away from magnetic equator. This will cause a decrease of charge accumulation with latitude negative latitudinal gradient in zonal electric field will be developed. 5of7

6 This negative latitudinal gradient when mapped back to equatorial F region manifest itself as a negative height gradient in zonal electric field as seen from Figure 1. [12] Figure 2 depicts the characteristics of the variation of height gradients at two altitude regions in the equatorial F region. At the lower altitude (the region probed between frequencies 2.5 and 3.5 MHz) presented in Figure 2a, the magnitude of the height gradient is more prominent compared to that in the higher-altitude region (probed between frequencies 3.5 and 4.5 MHz) presented in Figure 2b. As the altitude of the F region is higher, the magnetic flux tubes maps to the conjugate E region at higher latitudes where the separation between the magnetic meridian and the sunset terminator is larger. This results in reduced vertical drift height gradient at higher altitudes in agreement with present observation. Figure 3d indicates the seasonal trend in the vertical velocity gradient around PRE for the two ranges of probing frequencies. The negative height gradient in the probing frequency range MHz is larger than that probed using MHz frequency range throughout the observation period. [13] The seasonal variation in PRE can be explained in terms of the seasonal variation of the alignment of the solar terminator with respect to the magnetic field lines. During periods other than those close to the equinoxes, sunset times are different at the E regions of the two hemispheres which are magnetically linked to equatorial F region and the prereversal enhancement is less pronounced. From Figure 3 it is noticed that both the peak values of the prereversal enhancement (V zp ) depicted in Figure 3a and height gradients in vertical plasma drift depicted in Figure 3d exhibit seasonal variations. While the magnitude of V zp exhibits larger values around equinoxes the height gradient in V z around PRE is negative in all seasons and its magnitude has smaller values around equinoxes and larger values around the December solstice. This result indicates the effect of seasonal variation in the asymmetry between conjugate E region points in deciding the vertical plasma drift gradient and the prereversal enhancement. From Figures 3b and 3c it is noticed that for all the probing frequencies (i.e., at all the altitudes) both the vertical drift peak V zp and corresponding reflection altitude h zp depict a seasonal variation. However, a close look at Figures 3b and 3c indicates that around equinox, the altitude variation at higher frequencies is sharper than that at lower frequencies and V zp variation is sharper at lower frequencies compared to that corresponding to higher frequencies. This observation indicates an inverse relation between the h zp variation and the V zp variation resulting in a less negative gradient around the equinox. [14] The prediction of a postsunset two dimensional velocity vortex in the zonal plane was made by Tsunoda et al. [1981] using ALTAIR radar maps. The existence of the vortex is illustrated by Kudeki and Bhattacharyya [1999] using renewed Jicamarca radar measurements. The negative height gradient in vertical drifts around PRE and its reversal to positive gradient beyond PRE observed by HF Doppler radar can be a signature of the vortex-like flow pattern present at the equatorial F region ionosphere during the post sunset period. Present observations are limited to vertical sounding. At present no zonal drift measurement is made along with the multifrequency vertical drift measurements, but the augmentation of the system to observe zonal plasma drift in the near future will provide more details of the pattern of the plasma drift vortex. 5. Conclusions [15] The observation of the height gradient in vertical plasma drift at equatorial F region using multifrequency HF Doppler radar provide the following results: (1) The height gradient of vertical plasma drift exhibits a time variation with negative values around the period of the prereversal enhancement and then turning to positive values after the reversal. (2) The magnitude of the negative height gradient of vertical plasma drift decreases as the probing altitude range increases below the F peak. (3) The mean vertical plasma drift gradient around the prereversal enhancement exhibits a seasonal variation. The magnitude of the negative gradient is larger around the December solstice. [16] Acknowledgments. The HF-Doppler Radar system is built and operated with financial support from Indian Space Research Organization(ISRO) under its RESPOND program. [17] Arthur Richmond thanks Jorge Chau and S.P. Namboothiri for their assistance in evaluating this article. References Abdu, M. A., B. G. Fejer, I. S. Batista, J. H. A. Sobral, and E. P. Szuszczewicz (1993), Equatorial ionosphere sunset electrodynamics in the American sector from SUNDIAL December 1988 campaign results, Geomagn. Aeron., 33(1), Anderson, D. N., and D. W. Rusch (1980), Composition of the nighttime ionospheric F 1 region near the magnetic equator, J. Geophys. Res., 85(A2), Balan, N., B. Jayachandran, R. B. Nair, S. P. Namboothiri, G. J. Bailey, and P. B. Rao (1992), HF doppler observations of vector plasma drifts in the evening F-region at the magnetic equator, J. Atmos. Terr. Phys., 54, Batista, I. S., M. A. Abdu, and J. A. Bittencourt (1986), Equatorial F-region vertical plasma drifts: Seasonal and longitudinal asymmetries in the American sector, J. Geophys. Res., 91(A11), 12,055 12,064. Bittencourt, J. A., and M. A. Abdu (1981), A theoretical comparison between apparent vertical velocities and real vertical E B plasma drift velocities in the equatorial F-region, J. Geophys. Res., 86(A4), Eccles, J. V. (1998a), A simple model of low-latitude electric fields, J. Geophys. Res., 103(A11), 26,699 26,708. Eccles, J. V. (1998b), Modeling investigation of the evening prereversal enhancement of the zonal electric field in the equatorial ionosphere, J. Geophys. Res., 103(A11), 26,709 26,720. Eccles, J. V., N. C. Maynard, and G. Wilson (1999), Study of the evening plasma drift vortex in the low latitude ionosphere using San Marco electric field measurements, J. Geophys. Res., 104(A12), 28,133 28,143. Farley, D. T., E. Bonelli, B. G. Fejer, and M. F. Larsen (1986), The prereversal enhancement of the zonal electric field in the equatorial ionosphere, J. Geophys. Res., 91(A12), 13,723 13,728. Fejer, B. G. (1997), The electrodynamics of the low-latitude ionosphere: Recent results and future challenges, J. Atmos. Sol. Terr. Phys., 59, Haerendel, G., and J. V. Eccles (1992), The role of the equatorial electrojet in the evening ionosphere, J. Geophys. Res., 97, Hari, S. S., K. S. Viswanathan, K. S. V. Subbarao, and B. V. K. Murthy (1996), Equatorial E and F region zonal electric fields in the postsunset period, J. Geophys. Res., 101(A4), Hedin, A. E. (1987), MSIS-86 thermospheric model, J. Geophys. Res., 92(A5), Kudeki, E., and S. Bhattacharyya (1999), Postsunset vortex in equatorial F-region plasma drifts and implications for bottomside spread F, J. Geophys. Res., 104(A12), 28,163 28,170. Kudeki, E., B. G. Fejer, D. T. Farley, and H. M. Ierkic (1981), Interferometer studies of equatorial F region irregularities and drifts, Geophys. Res. Lett., 8(1), Maruyama, T., K. Nozaki, M. Yamamoto, and S. Fukao (2002), Ionospheric height changes at two closely separated equatorial stations and implications in spread F onsets, J. Atmos. Sol. Terr. Phys., 64, Murphy, J. A., and R. A. Heelis (1986), Implications of the relationship between electromagnetic drift components at mid and low latitudes, Planet. Space Sci., 34, of7

7 Namboothiri, S. P., N. Balan, and P. B. Rao (1989), Vertical plasma drifts in the F-region at the magnetic equator, J. Geophys. Res., 94(A9), 12,055 12,060. Pingree, J. E., and B. G. Fejer (1987), On the height variation of the equatorial F region vertical plasma drifts, J. Geophys. Res., 92(A5), Ramesh, K. B., and J. H. Sastri (1995), Solar cycle and seasonal variations in F-region vertical drifts over Kodaikanal, India, Ann. Geophys., 13, Rishbeth, H. (1971), Polarization fields produced by winds in the equatorial F-region, Planet. Space Sci., 19, Rishbeth, H. (1986), On the F2-layer continuity equation, J. Atmos. Terr. Phys., 48, Sastri, J. H., V. K. M. Varma, and S. R. P. Nayar (1995), Height gradient of F region vertical drift in the evening equatorial ionosphere, Geophys. Res. Lett., 22(19), Scherliess, L., and B. G. Fejer (1999), Radar and satellite global equatorial F region vertical drift model, J. Geophys. Res., 104(A4), Subbarao, K. S. V., and B. V. Krishna Murthy (1994), Post-sunset F region vertical velocity variations at magnetic equator, J. Atmos. Terr. Phys., 56(1), Tsunoda, R. T., R. C. Livingston, and C. L. Rino (1981), Evidence of a velocity shear in bulk plasma motion associated with the post-sunset rise of the equatorial F-layer, Geophys. Res. Lett., 8(7), Woodman, R. F. (1970), Vertical drift velocities and east-west electric fields at the magnetic equator, J. Geophys. Res., 75(31), S. R. Prabhakaran Nayar and C. V. Sreehari, Department of Physics, University of Kerala, Kariavattom, Trivandrum , India. (srp@md2.vsnl.net.in) 7of7