- Email: [email protected]
Solar activity variations of nocturnal thermospheric meridional winds over Indian longitude sector
Author’s Accepted Manuscript Solar activity variations of nocturnal thermospheric meridional winds over Indian longitude sector M.K. Madhav Haridas, G. Manju, T. Arunamani
www.elsevier.com/locate/jastp
PII: DOI: Reference:
S1364-6826(16)30152-3 http://dx.doi.org/10.1016/j.jastp.2016.06.010 ATP4443
To appear in: Journal of Atmospheric and Solar-Terrestrial Physics Received date: 1 March 2016 Revised date: 21 June 2016 Accepted date: 22 June 2016 Cite this article as: M.K. Madhav Haridas, G. Manju and T. Arunamani, Solar activity variations of nocturnal thermospheric meridional winds over Indian longitude sector, Journal of Atmospheric and Solar-Terrestrial Physics, http://dx.doi.org/10.1016/j.jastp.2016.06.010 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Solar activity variations of nocturnal thermospheric meridional winds over Indian longitude sector Madhav Haridas M.K.1, 2, G. Manju1, T. Arunamani2 1 Indian Space Research Organization 2 Andhra University, Visakhapatnam Abstract The night time F-layer base height information from ionosondes located at two equatorial stations Trivandrum (TRV 8.5°N,77°E) and Sriharikota (SHAR 13.7°N, 80.2°E) spanning over two decades are used to derive the climatology of equatorial nocturnal Thermospheric Meridional Winds (TMWs) prevailing during high solar activity (HSA) and low solar activity (LSA) epochs. The important inferences from the analysis are 1) Increase in mean equatorward winds observed during LSA compared to HSA during pre midnight hours; 25 m/s for VE (Vernal Equinox) and 20 m/s for SS (Summer Solstice), AE (autumnal Equinox) and WS (Winter Solstice). 2) Mean wind response to Solar Flux Unit (SFU) is established quantitatively for all seasons for pre-midnight hours; rate of increase is 0.25 m/s/SFU for VE, 0.2 m/s/SFU for SS and WS and 0.08 m/s/SFU for AE. 3) Theoretical estimates of winds for the two epochs are performed and indicate the role of ion drag forcing as a major factor influencing TMWs. 4) Observed magnitude of winds and rate of flux dependencies are compared to thermospheric wind models 5) Equinoctial asymmetry in TMWs is observed for HSA at certain times, with more equatorward winds during AE. These observations lend a potential to parameterize the wind components and effectively model the winds, catering to solar activity variations. Keywords: meridional wind, thermosphere, equatorial ionosphere Running title: Climatology of nocturnal meridional winds 1.0 Introduction
1
Measurements of TMWs have been scarce owing to the complexities involved in the measurement process. In the recent years long-term datasets and increased number of measurements has made TMWs a subject matter of intense theoretical as well as experimental studies. Nighttime meridional wind reversal and its linkage to Midnight Temperature Maximum (MTM) were studied using optical and radar instruments (Harper 1973; Sastri et. al., 1994). Burnside and Tepley (1989) performed wind measurements using a FPI at Arecibo in the southAmerican longitude sector and partly due to measurement uncertainties at the time, did not record any significant solar activity dependences. The solar and magnetic activity dependence of meridional neutral winds at 300 km altitude have been studied from the data obtained using 15 years of measurements by the incoherent scatter facility at Saint-Santin (France) by Duboin and Lafeuille (1992). They report an increase in poleward wind magnitude with increasing solar activity in the daily mean winds obtained. The Horizontal Wind Model, an empirical model, was developed based on integration of measurements from different platforms such as FPI, Incoherent Scatter radar and satellites (Hedin et. al., 1988; Hedin et. al. 1991). A co-ordinated analysis of mid latitude data carried out by Hedin et al.(1994) however, revealed poleward shift in winds with increasing solar activity that could not be accounted by HWM. Buasanto et al., 1999 using Incoherent Scatter Radar observations of meridional winds at Millstone hill observatory report an increase in equatorward winds during 2000-2400 hr local time with decreasing solar activity. Later on, studies from the south-American sector using an extended database of three solar cycles using FPI measurements by Tepley et al. (2011) and Brum et al. (2012) report solar activity dependences. Apart from incoherent scatter technique and FPI technique, another way to determine TMWs are using ionosonde F layer heights. A technique to derive thermospheric wind from ionosonde h’F
2
measurements was developed by Krishnamurthy et. al., 1990. The winds thus obtained were validated through rocket measurements by Sekar and Sridharan, 1992. Further, the seasonal variations of thermospheric winds during high solar activity were discussed by Hari and Krishnamurthy, 1995. Nogueira et al., 2011 investigated the response of thermospheric meridional wind during geo-magnetic storm events over Brazil using a similar technique of obtaining winds. Important studies using ionosonde derived TMWs from East-Asian sector also explore responses of TMWs to seasonal/solar activity and MTM related abatement (Maruyama et. Al., 2008; Liu et al., 2003a; Liu et al., 2003b; Luan et al., 2004). Equinoctial asymmetry in winds is also reported (Maruyama et. al., 2009). It is important that a data base of the thermospheric meridional winds under different geophysical conditions is generated especially in view of the fact that there is a scarcity of wind measurements in the equatorial region. Meridional winds have an impact over the magnitude of Equatorial Ionization Anomaly (EIA) crests depending on the season and time of day (Bramley and Young, 1968; Rishbeth 1972). The upward/downward movement of ionization along the magnetic field lines due to the effects of thermospheric meridional wind causes increase/decrease of EIA strength in low latitudes as reported by Tulasiram et al. (2009). Also in the night-time F region, TMWs are having significant impact over the sustenance of Equatorial Spread F (ESF) below a critical height of F layer (Devasia et. al., 2002; Manju et. al., 2007; Madhav Haridas and Manju, 2013). The Equatorial Temperature and Wind Anomaly (ETWA) setup by TMWs due to the temperature changes caused by increased ion-drag at the EIA crests have significant impact over the vertical winds at magnetic equator (Sastri, 1990; Raghavarao et al., 1991). The present work addresses the solar activity variability aspect of TMWs, which has been hitherto unexplored with a significant database in the Indian longitude sector. In the present
3
study, using a vast database of ionosonde data from the two stations located at Trivandrum and SHAR, conclusive evidence of solar cycle dependence of the night time meridional winds are brought out. Further, mechanisms involved in the reduction of wind magnitudes are examined and the role of the ionization distribution through the ion drag locally modulating the thermospheric meridional winds is found to be a significant factor contributing to the observed solar flux differences. This study also brings out the need to improve upon the existing wind models that do not fully account for the observed solar activity variabilities in TMWs. The observations are compared with the Horizontal Wind Model (Hedin et. al. 1991) and the TIEGCM model (Roble et al., 1977; Dickinson et al., 1981; Qian et al., 2014) and their effectiveness in reproducing the observations is discussed. 2.0 Data and method of analysis The meridional wind estimation is done using ionosonde data from TRV (geographic lat/long: 8.5°N 77°E; dip 0.9° ; Dec -2.6°) and SHAR (13.5°N 80.2°E; dip 13.2° ; Dec -1.6°) following the method developed by Krishna Murthy et al.(1990). The method is based on the fact that at the magnetic equator during night-time, the F-region vertical drift is due to E × B (where E is the east-west electric field and B is the magnetic induction) while at locations away from it, the meridional component (U) of the neutral wind also has a contribution to it apart from diffusion (Rishbeth et al., 1978). An equatorward (poleward) wind pushes the ionization up (down) along the field lines. The vertical drift V at a location such as SHAR is given as V = VD cosI – U cosI sinI - WDsin2I
(1)
Where VD is the electrodynamic ExB drift, U is the meridional component of neutral wind, I is the dip angle and WD is the plasma drift due to diffusion. In view of the fact that the two stations
4
are not widely separated the assumption is that east/west electric field is constant in that limited region and so is the ExB drift at the two stations. Since the magnetic dip at SHAR is high enough the meridional neutral wind also contributes significantly to V. Simplifying equation (1), the meridional wind can be estimated from the expression (2).
2V cos I V U D WD tan I sin 2 I
(2)
The observed vertical drift velocities are initially derived from the rate of change of h’F, (d(h’F)/dt)). For Trivandrum and SHAR it is denoted by VT and VS respectively. The true vertical drift is obtained from the observed vertical drift after removing apparent drift due to recombination. VD= VT-βTHT
(3)
V = Vs-βSHS,
(4)
where the suffixes T and S denote the parameters at Trivandrum and SHAR respectively, β is the effective recombination coefficient and H is [N-1 dN/dh ]-1, N representing the electron density and h the height. Substituting for VD and V in equation 2, the meridional component of the neutral wind is obtained as,
2V cos I V 2( T H T cos I S H S ) U D WD tan I sin 2 I sin 2 I
(5)
The recombination coefficient β is given by β = K1[N2] +K2[O2], where K1 K2 are the reaction rates of [N2] and [O2] (Anderson and Rusch, 1980). [N2] and [O2] are the number densities of N2 and O2 respectively obtained from MSIS
5
model (Hedin et al., 1988). It is to be noted that the winds thus obtained are based on the estimates of contribution to magnetic meridional direction near the magnetic equator. The plasma drift due to diffusion WD is given by,
WD
1 d * Nk (Ti Te ) mi g mi in N dh 1
(6)
Here N is electron density, mi is ionic mass, k is Boltzmann’s constant and ʋin is the ion-neutral collision frequency and Te and Ti are electron and ion temperatures. TIEGCM simulations reveal that the F2 layer is isothermal considering both ion and electron temperatures in the altitude regions of interest in this study. Accordingly, the WD expression reduces to the second term in equation (6) that is, g/ʋin. The first term in equation (5) is obtained from ionogram data, while the second and third terms are derived using atmospheric models. The meridional wind U is thus estimated from equation (5). A mammoth effort was undertaken to scale each ionogram of 15 minute cadence for the years spanning 1989 to 2008 from two stations. A sum total of ~650 days of data have gone into the mean for low solar activity years (1993-1998, 2003-2008; SFU< 130) and ~300 days for high solar activity years (1989-1992; 1999-2002; SFU≥150) making it a large database of meridional wind. Periods contaminated due to spread F are eliminated from all the data. Previous references chose the level of Ap of 70 as quiet conditions as far as impact on meridional winds were concerned (Krishna Murthy et al., 1990). However, we have adopted the cut-off of Ap<18 to represent magnetically quiet days since monthly mean values of winds having more number of days with higher Ap levels are seen to bias the observations. The sources of error in wind estimation are from the ambiguity in ionosonde h’F determination (first term of equation 5), errors in model estimation of (second term of equation 5) and error due to diffusion. The error due to increases for days with low h’F while simultaneously, the error in diffusion decreases 6
and vice versa. Also, the maximum limit of error in the estimation of a single wind time series is calculated to be ±25 m/s. However, the mean is taken for a number of such points and over several years of data. As a result of the random nature of such errors, it ultimately reduces to within the standard error limits. We are depicting the standard errors in all the figures in this study as followed by previous workers (Hari and Krishnamurthy 1995; Sekhar and Sridharan 1994; Krishnamurthy et al., 1990), in the light of the fact that the random errors reduce through increased number of measurements (Bevington 1969). In order to obtain the mean seasonal electron density, the level 2 product from the Planar Langmuir Probe (PLP) onboard the CHAMP satellite has been employed (Cooke et al., 2003). The total mass density, number density and the exospheric temperatures has been obtained from the MSIS model. The observations of thermospheric winds are compared with the empirical Horizontal Wind Model (HWM) and Thermosphere Ionosphere Electrodynamics General Circulation Model (TIEGCM). The input parameters for the model are primarily the solar F10.7 cm flux, and Kp value. This is simultaneously taken into account from the date entered for simulation. Then the model solves the three-dimensional momentum, energy and continuity equations for neutral and ion species at each time step to obtain temperatures, composition and neutral winds. 3. Results Fig.1 shows the temporal average of TMWs during HSA (red) and LSA (blue) irrespective of seasonal variations obtained by averaging ~950 days of measurements. In the entire paper ‘hr’ corresponds to Indian Standard Time which is 5 hours 30 minutes ahead of Universal Time. Positive value indicates that winds are directed towards the North Pole (poleward) and negative values indicate that the winds are directed towards the equator, south of the station (equatorward).
7
The major factors that are immediately evident from the figure are i) During the post-sunset hours from 1800-2000 hr it can be seen that there is no change in TMWs with solar activity ii) Significant solar activity response of TMWs is seen for the pre-midnight hours between 2000 and 0000 hr. The equatorward winds during LSA are higher by as much as 30 m/s compared to HSA iii) Post-midnight, the magnitude of winds are again comparable, except around 0200 hr where the LSA winds are slightly more poleward than during HSA. In order to examine the seasonal variations of TMWs for different epochs, the data are segregated into mean of respective solar activity levels and plotted against local time in Fig. 2. Top left panel shows the mean TMWs obtained for HSA and LSA for the Vernal Equinox (VE) season. Deviation between HSA and LSA starts around 2100 hr and continues till 0000 hr. Magnitude of equatorward winds is higher during this period for LSA with peak magnitude of 50 m/s at 2215 hr compared to 10 m/s for HSA. Post 0000 hr, winds during both LSA and HSA are poleward directed, with no significant difference in magnitudes. The pattern of winds is the same for Autumnal Equinox (AE; left bottom panel) season also, the peak equatorward wind during HSA is 30 m/s attained around 2200 hr. The peak magnitude of equatorward winds during LSA is 50 m/s around 2200 hr, after which there is an abatement. Magnitude of winds during HSA of VE is also found to be more poleward than that for AE during the time 0200 hr – 0400 hr. During the Summer Solstice (SS) the post sunset winds are equatorward directed right from 1800 hr. Equatorward wind deviation between LSA and HSA is maximum around 2200 hr. Magnitude of maximum equatorward wind during HSA is 20 m/s and during LSA is 45m/s. MTM related abatement is evident in this season also. During Winter Solstice (WS), winds turn equatorward briefly between 2015 hr and 2145 hr for LSA in the pre-midnight sector, while it remains entirely poleward for HSA. The difference in winds between HSA and LSA during winter at
8
2100 hr is 20 m/s. The equatorward reversal during LSA is attributed to the MTM effect. Postmidnight, around 0300 hr, the winds turn equatorward again, both during LSA and HSA. It is evident from the stronger abatement seen in the equatorward winds during LSA that there are stronger influences of MTM during LSA. There is 40 m/s increase in the magnitude of abatement in equatorward winds for LSA compared to HSA. Another important aspect pertains to the asymmetry of the winds between the two equinoxes at 2200 hr. While no equinoctial asymmetry is seen between the equinoxes for LSA, there is a clear asymmetry in winds for HSA, amounting to 20 m/s at 2200 hr. The availability of data throughout the span of 2 solar cycles allows us to examine the rate of change of meridional wind magnitude with solar flux. For this, the entire dataset of 20 years during the time period 2000-0000 hr is first segregated into monthly mean values. This monthly mean meridional winds is plotted against the corresponding mean solar fluxes for all seasons (Figure 3). A linear correlation is observed for all the seasons between the winds and solar flux. The correlation coefficients obtained for the linear least squares fit is 0.76 for VE (significance>99%), 0.75 for SS (significance>98%), 0.56 (significance>95%) for AE and 0.5 for WS (significance>95%). This shows a clear trend where the equatorward winds are gradually reduced with increasing solar activity levels for all seasons. The linear relation between the winds and solar flux has been arrived at and are given as inset in each panel. The rate of change of TMWs w.r.to solar activity for VE is 0.25 m/s/SFU, 0.2m/s/SFU for SS, 0.08 m/s/SFU for AE and 0.19 m/s/SFU for WS. 4. Discussion In the thermosphere, the air may be regarded as a single fluid, subject to hydrodynamic equations of motion owing to its characteristic collisional frequencies and the fact that differential motion
9
of any of its various constituents is very much less than the overall wind speed. Solar heating around the sub solar point causes thermal expansion of the atmosphere forming the ‘diurnal bulge’ giving rise to horizontal pressure gradients that drive horizontal winds towards the antisolar point situated at nighttime longitudes. Thermospheric winds blow across the polar regions and zonally around the Earth under the dominant influence of viscous force and ion drag unlike the winds in lower altitudes where the winds roughly follow the isobars under geostrophic balance (Rishbeth 1972). During daytime the thermospheric winds diverge from the equator towards the poles and hence the direction of day time winds at a station like SHAR is poleward. As SHAR enters the night sector, convergence of winds from the dayside sub solar point (at ~100°W by midnight) takes over and the winds turn equatorward. By around midnight the phenomena of Midnight Temperature Maximum ensues as a result of the tidal forcing from lower atmosphere and adiabatic heating due to convergence and causes an outflow that is stronger than the wind magnitudes set up by the pressure gradient across earth’s great circle (Hari and Krishna Murthy 1995; Sastri et al., 1994, Niranjan et al., 2006). Earlier work related to meridional wind in the Indian sector was carried out by Hari and Krishna Murthy (1995) during high solar activity years of 1989-90 and they address the seasonal pattern of nocturnal thermospheric winds. The present study looks into the solar activity variations in TMWs by analyzing available data spread over two decades. The study conclusively shows that there is an increase in the magnitude of the thermospheric equatorward winds with decreasing solar activity in the pre-midnight hours. Maximum response to solar activity is seen in the VE months followed by SS, AE and WS. The study undertaken by Liu et. al 2003b at a station Wuhan in the East-Asian longitude sector show similar equatorward reversals in wind component along the magnetic equator with increasing
10
solar activity. Optical observations of thermospheric winds at Arecibo, in the American sector carried out by Burnside and Tepley (1989) using a limited dataset, revealed no significant solar activity differences in meridional winds. Later, Tepley et. al. (2011) undertook a detailed examination of the nocturnal TMWs using an extended dataset from the same region and hinted at the presence of a solar activity dependence in meridional winds, of very small in magnitude, during the pre midnight sector. The differences in wind magnitudes are observed only around pre-midnight hours for Indian sector in the present study also, with negligible effects post midnight. Brum et al. (2012), using the extended data set of optical measurements from the South American sector reports a mean wind response rate of 0.2-0.25 m/s/SFU for VE, 0.15-0.2 m/s/SFU for AE, 0.25m/s/SFU for SS, and 0.2m/s/SFU for WS at 0000 hr. Our observations from the Indian sector show the solar flux dependencies at nearly the same rate, of 0.23 m/s/SFU for VE, 0.09 m/s/SFU for AE and 0.18 m/s/SFU for SS and WS during 2000-0000 hr. The solar activity dependences are seen to be higher for VE and SS in our analysis. Buosanto et. al. (1999) reports a clear increase in the equatorward winds during low solar activity and attribute this to the differences in ionization and ion-drag at North American longitude sector (Millstone hill). Liu et al., 2004 have undertaken a detailed study of thermospheric winds derived from ionosondes at various locations in both northern and southern hemisphere. They report a reduction in diurnal magnitudes of meridional winds with increasing solar activity. They identify the limiting factor of winds to be the ion-drag increase during high solar activity. Luan et al., 2004 have investigated the response of equivalent wind derived from ionosonde to solar activity at various latitudes. It is interesting to note that the pattern of observations is similar to the observations made in the Indian sector. For example, they report the maximum response to solar
11
activity to be occurring in VE at 2200 hr. During AE, the high solar activity winds are more equatorward than in VE, a feature that has emerged from our observations also at 2200 hr. 4.2 Theoretical estimation of winds In order to analyze the probable causative mechanism for the observed differences in TMWs during different solar activity epochs, in the present study, the components that make up the thermospheric winds are examined. For this, the equation of motion for thermospheric winds is considered, where the gradient in the wind is balanced by the sum of the pressure gradient force, the ion-drag force, the viscous force and gravity, as follows: U 1 p 2U (7) ni (U V ) g 2 2 t x H x where, U is the horizontal wind component, x is the great-circle distance, p is pressure, ρ is the
density, υin is the ion-drag forcing, V is the ion velocity, µ is the coefficient of molecular viscosity, H is the scale height and g is acceleration due to gravity. Divergence of the wind field is the sum of pressure gradient existing across the great circle distance, the ion drag component imparted by the ions to the neutral flow, viscous force and gravity. The Coriolis force is negligible near the equator and is thus ignored. The continuity equation for the meridional wind that includes these parameters has to be satisfied locally. A sensitivity test on the changes in meridional winds due to solar activity variabilities in the individual components is performed to delineate the impacts on various forcings due to solar activity. First, the pressure gradient, which is a function of neutral density and exospheric temperature, is estimated for a typical magnetically quiet day of autumnal equinox season at 2200 hr for SHAR during HSA (2002) and LSA (2008) using the neutral densities and temperature from the MSIS model. Figure 4 shows the contour of pressure gradient estimated for the entire globe for high and low solar activity at 2200 hr local time at SHAR. There is a clear
12
enhancement in the pressure gradient force by around 38% from LSA to HSA at the location of SHAR. The ion drag forcing on the meridional wind is estimated next from the following expression,
ni
in * N e [O]
(8)
Where ʋin is the sum of ion-neutral collision frequencies of each constituent in the thermosphere, Ne is the electron density and O is the number density of atomic oxygen (Rishbeth 1969). The estimation of νni is undertaken using in-situ CHAMP electron densities available for autumnal equinox months of HSA and LSA for 2200 hr. For HSA, mean of electron densities during AE 2001-2002 is used, whereas for LSA, the mean electron densities during AE of 2007-2008 are used. Figure 5 shows the ion-drag force estimated for LSA and HSA during 2200 hr at SHAR for AE. It is very interesting to note that the pattern of ion drag is clearly reflecting the double crest pattern of EIA for HSA. This is very much in line with the expected presence of EIA well past midnight in HSA. The manifestation of EIA is found to continue well into the night even up to 0200 hr during HSA (Sastri, 1990). Thus a region of high electron density would persist in the low-latitude regions as a result of the enhanced pumping of ionized plasma from the magnetic equator at thermospheric altitudes resulting in increased ion-drag during about 1930 – 0200 hr. Its strength and duration is subject to the changes in solar activity. For LSA, the early recession of EIA leads to low electron densities and consequent low ion drag in the EIA regions. In view of this EIA induced enhanced ion drag there is an order of magnitude increase (14.3 times) seen in ion-drag forcing ʋni between LSA and HSA near Indian sector. This is at a much higher rate than in the case of the pressure gradient force (~ 1.38 times).
13
From the altitude profiles of ion-drag force, pressure gradient force, viscous force and Coriolis force given in Rishbeth (1972) it is clear that there is a region between 150 km and 400 km, where pressure gradient and ion-drag forces are predominant in the thermosphere. Below the altitude of 200 km, the Coriolis force is much greater than that of the ion drag terms and the winds follow the geostrophic approximation where the pressure gradient is balanced by the Coriolis force. Again, above the altitude of 400 km, where Coriolis force is significant, viscous forces become much stronger and hence this term has to be considered in the final solution to thermospheric wind equation. These inferences are often the boundary conditions for the numerical solutions of winds at ionospheric heights. Hence, in the altitude range of 150 km to 400 km the wind term is often expressed as the ratio of F/ ʋni and Rishbeth (1972) describes the magnitude of changes in this parameter for various times of the day as well as two solar epochs, to discuss ion-drag effects on thermospheric winds at mid-latitudes. The square of the sine component of the latitude is also to be included in the calculations to get the absolute values of the meridional winds, but that is presently ignored since that factor remains unchanged for both solar activity epochs. Liu et. al., 2004 also details the effectiveness of this parameter in capturing the variations due to ion-drag in thermospheric meridional winds. The point values in AE panel on the right column of Figure 6 shows the estimated wind for HSA and LSA at 2200 hr. For HSA, the magnitude of equatorward wind obtained is 4.6 m/s and for LSA, it is 48.3 m/s. There is an increase of 43.7 m/s in equatorward wind from HSA to LSA on account of changes in pressure gradient and ion-drag. The ion drag imparted by the increased magnitude of electron density and a stronger EIA during HSA years is having the major say in the reduction of equatorward meridional winds around equatorial and low latitudes. At equatorial and low latitude regions, there is an enhancement in the EIA soon after the Pre-Reversal
14
Enhancement (PRE) of the zonal electric field around the time after sunset, and it persists up to midnight in LSA. As the pressure gradient drives the winds across these regions of increased electron density and hence ion drag, the wind magnitudes are lower for high solar activity years. During post-midnight hours, the MTM effects become dominant and hence the ion drag effects and consequent effects on winds are not clearly visible. By around 0200-0300 hr, the MTM effects diminish, but by then the EIA is also inhibited and so the ion drag effects on winds becomes less significant. 4.2 Model simulations We compare the observational results from the Indian sector with the model outputs of the Horizontal Wind Model and TIE-GCM runs for different seasons and solar activity levels at the co-ordinates of SHAR. Fig. 6 (left column) shows the HWM derived winds. The HWM is an empirical model and has not inculcated solar flux related changes and hence is unable to reproduce the solar activity changes. However, it reproduces the seasonal trends quiet well. Equatorward nature of winds during SS and poleward nature of winds during WS is reproduced by the model. The brief equatorward reversal of winds during LSA of WS however is not seen in the model. The time of reversal of winds from equatorward to poleward direction around midnight, is reproduced well by HWM for HSA only. For LSA, HWM fails to show the MTM related changes occurring in the magnitude and time of abatement of equatorward winds. Fig 6 (right column) shows the TIEGCM obtained meridional winds for different seasons and solar activities. TIEGCM is a physical model based on the continuity equations and is the latest model available for major thermospheric constituents and parameters. It solves the threedimensional momentum, energy and continuity equations for neutral and ion species at each time
15
step to obtain temperatures, composition and neutral winds. The increased magnitudes of equatorward winds during LSA are captured well by the model. There is a brief excursion of equatorward winds during 2400 hr for LSA period of WS. The overall pattern of winds in all time sectors and for all seasons and solar activities is better captured in TIEGCM simulations than HWM simulations. However, the time of abatement of equatorward winds in response to MTM and related effects seen in the observed winds are not clearly captured by the TIEGCM model. The time of abatement (for observations) is by 2200 hr whereas in the model, the time of reversal is later by 2 hours and is only in response to changes in the direction of pressure gradient. The calculated winds for HSA is quite close to TIEGCM values. However there is a deviation w.r.to winds during HSA. These observations highlight the need for more measurements of thermospheric meridional winds especially from the equatorial and low-latitude regions as they are an important component of large scale equatorial phenomena like ESF and EIA. Future directions also include realization of a model including the viscous forces and MTM induced pressure gradient changes that better reflect the observations in the Indian longitude sector. 5. Summary A comprehensive analysis of nocturnal thermospheric meridional wind pattern encompassing two solar cycles is accomplished at the Indian longitude sector. Significant difference is seen in winds between high and low solar activity epochs, with more equatorward winds during pre midnight hours for low solar activity years. The solar flux relationship of mean winds during pre midnight hours is established. An integrated approach using the existing observational techniques, models as well as satellite based inputs are employed to understand the underlying causative mechanisms behind the observed solar activity responses of thermospheric winds. The
16
increased ion drag forcing existing during high solar activity years has been identified to be the major factor causing the observed pre midnight sector differences in the nocturnal thermospheric winds over Indian longitude sector. The study highlights the need to improve the existing models incorporating the effects of physical processes like MTM and ion-drag effects more expansively. Acknowledgements This work is funded by Indian Space Research Organization, Govt. of India. The ionosonde data used is the property of the organization. The data may be made available on specific request to the corresponding author with the permission from the concerned authorities. The CHAMP satellite data has been obtained from Information System and Data Center, hosted at http://isdc.gfz-potsdam.de/. TIEGCM simulation results have been provided by the Community Coordinated Modeling Center at Goddard Space Flight Center through their public Runs on Request system (http://ccmc.gsfc.nasa.gov). The CCMC is a multi-agency partnership between NASA, AFMC, AFOSR, AFRL, AFWA, NOAA, NSF and ONR. References Bevington P.R., (1969), Data reduction and error analysis for the physical sciences, Mc Graw hill. Bramley E.M. and Young M., (1968), Winds and electromagnetic drifts in the equatorial F2 region, J. Atmos. Terr. Phys., 80, 90. Brum, C. G. M., C. A. Tepley, J. T. Fentzke, E. Robles, P. T. Santos, and S. A. Gonzalez (2012), Long-term changes in the thermospheric neutral winds over Arecibo: Climatology based on over three decades of Fabry-Perot observations, J. Geophys. Res., 117,458, DOI: 10.1029/2011JA016. Buonsanto, M. J., and O. G. Witasse (1999), An updated climatology of thermospheric neutral winds and F region ion drifts above Millstone Hill, J. Geophys. Res., 104, 24, 675.
17
Burnside, R. G., and C. A. Tepley (1989), Optical observations of thermospheric neutral winds at Arecibo
between
1980
and
1987,
J.
Geophys.
Res.,
94,
2711–2716,
doi:10.1029/JA094iA03p02711. Cooke, D. L., W. Turnbull, C. Roth, A. Morgan, and R. Redus (2003), Ion drift-meter status and calibration, in First Champ Mission Results for Gravity, Magnetic, and Atmospheric Studies, edited by C. Reigber, H. Luhr, and P. Schwintzer, pp. 212–219, Springer, New York. Devasia, C. V., N. Jyoti, K.S. Viswanathan, K. S. V. Subbarao, D. Tiwari and R. Sridharan (2002), On the plausible linkage of thermospheric meridional winds with equatorial spread F, J. Atmos. Sol. Terr. Phys., 64, 1-12. Dickinson, R. E., E. C. Ridley and R. G. Roble (1981), A three-dimensional general circulation model of the thermosphere, J. Geophys. Res., 86, 1499-1512. Duboin, M.-L., and M. Lafeuille (1992), Thermospheric dynamics above Saint-Santin: Statistical study of the data set, J. Geophys. Res., 97, 8,661. Hari, S. S., and B. V. Krishna Murthy (1995), Seasonal variations of equatorial nighttime thermospheric meridional winds, J. Atmos. Terr. Phys., 57,1241-1246. Harper, R. M. (1973), Night time meridional neutral winds near 350 km at low to mid-latitudes, J. Atmos. Terr. Phys., 353, 2023. Hedin A.E, N. W. Spencer, and T. L. Killeen (1988), Empirical Global Model of Upper Thermosphere Winds Based on Atmosphere and Dynamics Explorer Satellite Data, J. Geophys. Res. 93, 9959- 9978. Hedin, A. E., et al. (1991), Revised Global Model of Thermosphere Winds Using Satellite and Ground-Based Observations, J. Geophys. Res., 96(A5), 7657–7688, doi: 10.1029/91JA00251. Hedin, A. E., M. J. Buonsanto, M. Codrescu, M.-L. Duboin, C. G Fesen, M. E. Hagan, K. L. Miller, and
D. P. Sipler (1994), Solar activity variations in mid latitude thermospheric
meridional winds, J. Geophys. Res., 99(A9), 17601- 17608.
18
Krishna Murthy, B. V., S. S. Hari, and V. V. Somayajulu (1990), Nighttime equatorial thermospheric meridional winds from ionospheric h0F data, J. Geophys. Res., 95(A4), 4307– 4310. Liu L., X. Luan, W. Wan, J. Lei, and B. Ning (2003a), Seasonal behavior of equivalent winds over Wuhan derived from ionospheric data in 2000-2001, Adv. Space Res., 32(9), 1765-1770. Liu L., W. Wan, X. Luan, B. Ning, and J. Lei (2003b), Solar activity dependence of effective winds derived from ionospheric data at Wuhan, Adv. Space Res., 32(9), 1719-1924. Liu, L., X. Luan, W. Wan, J. Lei, and B. Ning (2004), Solar activity variations of equivalent winds
derived
from
global
ionosonde
data,
J.
Geophys.
Res.,
109,
A12305,
doi:10.1029/2004JA010574. Luan, X., Liu, L., Wan, W., Lei, J., and Yu, T. (2004), A climatology of the F-layer equivalent winds derived from ionosonde measurements over two decades along the 120°-150°E sector, Ann. Geophys., 22, 2785-2796, doi:10.5194/angeo-22-2785-2004. Madhav Haridas, M. K., G. Manju, and T. K. Pant (2013), First observational evidence of the modulation of the threshold height h'Fc for the occurrence of equatorial spread F by neutral composition changes, J. Geophys. Res., 118, 3540–3545, doi:10.1002/jgra.50331. Manju, G. ; Devasia, C. V. ; Sridharan, R. (2007), On the seasonal variations of the threshold height for the occurrence of equatorial spread F during solar minimum and maximum years, Ann Geophys, 25 (4). pp. 855-861. ISSN 1432-0576. 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:10.1029/2008JA013031. Maruyama, T., Saito, S., Kawamura, M., Nozaki, K., Krall, J., and Huba, J. D. (2009), Equinoctial asymmetry of a low-latitude ionosphere-thermosphere system and equatorial irregularities: evidence for meridional wind control, Ann. Geophys., 27, 2027-2034, doi:10.5194/angeo-27-2027-2009.
19
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:10.1029/2005JA011386. Nogueira P.A.B., M.A. Abdu, I.S. Batista, P.M. de Siqueira (2011), Equatorial ionization anomaly and thermospheric meridional winds during two major storms over Brazilian low latitudes, J. Atmos. Sol. Terr. Phys., Volume 73, Issues 11–12, 1535-1543, ISSN 1364-6826, http://dx.doi.org/10.1016/j.jastp.2011.02.008. Raghavarao, R., L. E. Wharton, N. W. Spencer, H. G. Mayr, L. H. Brace (1991), An equatorial temperature and wind anomaly (ETWA), Geophys. Res. Lett. 18, 1193-1196. Rishbeth, H., and O. K. Garriott (1969), Introduction to Ionospheric Physics, Academic, New York. Rishbeth, H., (1972) Thermospheric winds and the F region: A review. J. Atmos. Terr. Phys. doi:10.1016/0021-9169(72)90003-7. Rishbeth, H., Ganguly, S., and Walker, J. C. G. (1978), Field-aligned and field-perpendicular velocities in the ionospheric F2 layer, J. Atmos. Terr. Phys., 40, 767–784. Roble, R. G., R. E., Dickinson and E. C. Ridley (1977), Seasonal and solar-cycle variations of zonal mean circulation in the thermosphere, J. Geophys. Res., 82, 5493-5504. Sastri,
J.
H.
(1990),
Equatorial
anomaly
in
F
region-
A
review,
Ind.
Journ.
Rad. Space Phys., 19, 225-240. Sastri, J. R, H. N. R. Rao, V. V. Somayajulu, and H. Chandra (1994), Thermospheric meridional neutral winds associated with equatorial midnight temperature maximum (MTM), Geophys. Res. Lett., 21, 825-828. Sekar R. and Sridharan R. (1992), Validity of the estimates of night-time meridional winds made from bottomside ionograms, J. atmos, terr.Phys. 54, 1197-1199. Tepley, C. A., E. Robles, R. Garcia, P. T. Santos, C. M. Brum, and R. G. Burnside (2011), Directional trends in thermospheric neutral winds observed at Arecibo during the past three solar cycles, J. Geophys. Res., 116, DOI:10.1029/2010JA016, 172.
20
Tulasi Ram, S., S.Y. Su, and C. H. Liu (2009), FORMOSAT-3/COSMIC observations of seasonal and longitudinal variations of equatorial ionization anomaly and its interhemispheric asymmetry
during
the
solar
minimum
period, J.
Geophys.
Res., 114,
A06311,
doi:10.1029/2008JA013880. Qian, L., A. G. Burns, B. A. Emery, B. Foster, G. Lu, A. Maute, A. D. Richmond, R. G. Roble, S. C. Solomon, and W. Wangm (2014), The NCAR TIE-GCM: A community model of the coupled thermosphere/ionosphere system, in Modeling the Ionosphere-Thermosphere System, AGU Geophysical Monograph Series.
21
Figure 1: Comparison of mean thermospheric wind between high and low solar activity years
22
Figure 2: Mean temporal evolution of thermospheric meridional wind during two solar activity epochs for Vernal Equinox (VE), Summer Solstice (SS), Autumnal Equinox (AE) and Winter solstice (WS)
23
Figure 3: Solar flux dependence of mean thermospheric winds during 2000 -0000 hr for Vernal Equinox (VE), Summer Solstice (SS), Autumnal Equinox (AE) and Winter solstice (WS)
24
Figure 4: pressure gradient estimated for HSA (top panel) and LSA (bottom panel) at 2200 hr for AE
25
Figure 5: Ion-drag force estimated for HSA and LSA during 2200 hr for AE
26
Figure 6: Winds obtained for different seasons/solar activity epochs using HWM (left column) and TIEGCM (Right column). The calculated winds at 2200 hr for AE is shown as stars highlights Solar activity variations of nocturnal thermospheric meridional winds established in Indian sector Equatorward wind magnitudes increase with decreasing solar activity in pre-midnight sector Ion-drag identified as a major source of variations and the observations compared with model data
27
www.elsevier.com/locate/jastp
PII: DOI: Reference:
S1364-6826(16)30152-3 http://dx.doi.org/10.1016/j.jastp.2016.06.010 ATP4443
To appear in: Journal of Atmospheric and Solar-Terrestrial Physics Received date: 1 March 2016 Revised date: 21 June 2016 Accepted date: 22 June 2016 Cite this article as: M.K. Madhav Haridas, G. Manju and T. Arunamani, Solar activity variations of nocturnal thermospheric meridional winds over Indian longitude sector, Journal of Atmospheric and Solar-Terrestrial Physics, http://dx.doi.org/10.1016/j.jastp.2016.06.010 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Solar activity variations of nocturnal thermospheric meridional winds over Indian longitude sector Madhav Haridas M.K.1, 2, G. Manju1, T. Arunamani2 1 Indian Space Research Organization 2 Andhra University, Visakhapatnam Abstract The night time F-layer base height information from ionosondes located at two equatorial stations Trivandrum (TRV 8.5°N,77°E) and Sriharikota (SHAR 13.7°N, 80.2°E) spanning over two decades are used to derive the climatology of equatorial nocturnal Thermospheric Meridional Winds (TMWs) prevailing during high solar activity (HSA) and low solar activity (LSA) epochs. The important inferences from the analysis are 1) Increase in mean equatorward winds observed during LSA compared to HSA during pre midnight hours; 25 m/s for VE (Vernal Equinox) and 20 m/s for SS (Summer Solstice), AE (autumnal Equinox) and WS (Winter Solstice). 2) Mean wind response to Solar Flux Unit (SFU) is established quantitatively for all seasons for pre-midnight hours; rate of increase is 0.25 m/s/SFU for VE, 0.2 m/s/SFU for SS and WS and 0.08 m/s/SFU for AE. 3) Theoretical estimates of winds for the two epochs are performed and indicate the role of ion drag forcing as a major factor influencing TMWs. 4) Observed magnitude of winds and rate of flux dependencies are compared to thermospheric wind models 5) Equinoctial asymmetry in TMWs is observed for HSA at certain times, with more equatorward winds during AE. These observations lend a potential to parameterize the wind components and effectively model the winds, catering to solar activity variations. Keywords: meridional wind, thermosphere, equatorial ionosphere Running title: Climatology of nocturnal meridional winds 1.0 Introduction
1
Measurements of TMWs have been scarce owing to the complexities involved in the measurement process. In the recent years long-term datasets and increased number of measurements has made TMWs a subject matter of intense theoretical as well as experimental studies. Nighttime meridional wind reversal and its linkage to Midnight Temperature Maximum (MTM) were studied using optical and radar instruments (Harper 1973; Sastri et. al., 1994). Burnside and Tepley (1989) performed wind measurements using a FPI at Arecibo in the southAmerican longitude sector and partly due to measurement uncertainties at the time, did not record any significant solar activity dependences. The solar and magnetic activity dependence of meridional neutral winds at 300 km altitude have been studied from the data obtained using 15 years of measurements by the incoherent scatter facility at Saint-Santin (France) by Duboin and Lafeuille (1992). They report an increase in poleward wind magnitude with increasing solar activity in the daily mean winds obtained. The Horizontal Wind Model, an empirical model, was developed based on integration of measurements from different platforms such as FPI, Incoherent Scatter radar and satellites (Hedin et. al., 1988; Hedin et. al. 1991). A co-ordinated analysis of mid latitude data carried out by Hedin et al.(1994) however, revealed poleward shift in winds with increasing solar activity that could not be accounted by HWM. Buasanto et al., 1999 using Incoherent Scatter Radar observations of meridional winds at Millstone hill observatory report an increase in equatorward winds during 2000-2400 hr local time with decreasing solar activity. Later on, studies from the south-American sector using an extended database of three solar cycles using FPI measurements by Tepley et al. (2011) and Brum et al. (2012) report solar activity dependences. Apart from incoherent scatter technique and FPI technique, another way to determine TMWs are using ionosonde F layer heights. A technique to derive thermospheric wind from ionosonde h’F
2
measurements was developed by Krishnamurthy et. al., 1990. The winds thus obtained were validated through rocket measurements by Sekar and Sridharan, 1992. Further, the seasonal variations of thermospheric winds during high solar activity were discussed by Hari and Krishnamurthy, 1995. Nogueira et al., 2011 investigated the response of thermospheric meridional wind during geo-magnetic storm events over Brazil using a similar technique of obtaining winds. Important studies using ionosonde derived TMWs from East-Asian sector also explore responses of TMWs to seasonal/solar activity and MTM related abatement (Maruyama et. Al., 2008; Liu et al., 2003a; Liu et al., 2003b; Luan et al., 2004). Equinoctial asymmetry in winds is also reported (Maruyama et. al., 2009). It is important that a data base of the thermospheric meridional winds under different geophysical conditions is generated especially in view of the fact that there is a scarcity of wind measurements in the equatorial region. Meridional winds have an impact over the magnitude of Equatorial Ionization Anomaly (EIA) crests depending on the season and time of day (Bramley and Young, 1968; Rishbeth 1972). The upward/downward movement of ionization along the magnetic field lines due to the effects of thermospheric meridional wind causes increase/decrease of EIA strength in low latitudes as reported by Tulasiram et al. (2009). Also in the night-time F region, TMWs are having significant impact over the sustenance of Equatorial Spread F (ESF) below a critical height of F layer (Devasia et. al., 2002; Manju et. al., 2007; Madhav Haridas and Manju, 2013). The Equatorial Temperature and Wind Anomaly (ETWA) setup by TMWs due to the temperature changes caused by increased ion-drag at the EIA crests have significant impact over the vertical winds at magnetic equator (Sastri, 1990; Raghavarao et al., 1991). The present work addresses the solar activity variability aspect of TMWs, which has been hitherto unexplored with a significant database in the Indian longitude sector. In the present
3
study, using a vast database of ionosonde data from the two stations located at Trivandrum and SHAR, conclusive evidence of solar cycle dependence of the night time meridional winds are brought out. Further, mechanisms involved in the reduction of wind magnitudes are examined and the role of the ionization distribution through the ion drag locally modulating the thermospheric meridional winds is found to be a significant factor contributing to the observed solar flux differences. This study also brings out the need to improve upon the existing wind models that do not fully account for the observed solar activity variabilities in TMWs. The observations are compared with the Horizontal Wind Model (Hedin et. al. 1991) and the TIEGCM model (Roble et al., 1977; Dickinson et al., 1981; Qian et al., 2014) and their effectiveness in reproducing the observations is discussed. 2.0 Data and method of analysis The meridional wind estimation is done using ionosonde data from TRV (geographic lat/long: 8.5°N 77°E; dip 0.9° ; Dec -2.6°) and SHAR (13.5°N 80.2°E; dip 13.2° ; Dec -1.6°) following the method developed by Krishna Murthy et al.(1990). The method is based on the fact that at the magnetic equator during night-time, the F-region vertical drift is due to E × B (where E is the east-west electric field and B is the magnetic induction) while at locations away from it, the meridional component (U) of the neutral wind also has a contribution to it apart from diffusion (Rishbeth et al., 1978). An equatorward (poleward) wind pushes the ionization up (down) along the field lines. The vertical drift V at a location such as SHAR is given as V = VD cosI – U cosI sinI - WDsin2I
(1)
Where VD is the electrodynamic ExB drift, U is the meridional component of neutral wind, I is the dip angle and WD is the plasma drift due to diffusion. In view of the fact that the two stations
4
are not widely separated the assumption is that east/west electric field is constant in that limited region and so is the ExB drift at the two stations. Since the magnetic dip at SHAR is high enough the meridional neutral wind also contributes significantly to V. Simplifying equation (1), the meridional wind can be estimated from the expression (2).
2V cos I V U D WD tan I sin 2 I
(2)
The observed vertical drift velocities are initially derived from the rate of change of h’F, (d(h’F)/dt)). For Trivandrum and SHAR it is denoted by VT and VS respectively. The true vertical drift is obtained from the observed vertical drift after removing apparent drift due to recombination. VD= VT-βTHT
(3)
V = Vs-βSHS,
(4)
where the suffixes T and S denote the parameters at Trivandrum and SHAR respectively, β is the effective recombination coefficient and H is [N-1 dN/dh ]-1, N representing the electron density and h the height. Substituting for VD and V in equation 2, the meridional component of the neutral wind is obtained as,
2V cos I V 2( T H T cos I S H S ) U D WD tan I sin 2 I sin 2 I
(5)
The recombination coefficient β is given by β = K1[N2] +K2[O2], where K1 K2 are the reaction rates of [N2] and [O2] (Anderson and Rusch, 1980). [N2] and [O2] are the number densities of N2 and O2 respectively obtained from MSIS
5
model (Hedin et al., 1988). It is to be noted that the winds thus obtained are based on the estimates of contribution to magnetic meridional direction near the magnetic equator. The plasma drift due to diffusion WD is given by,
WD
1 d * Nk (Ti Te ) mi g mi in N dh 1
(6)
Here N is electron density, mi is ionic mass, k is Boltzmann’s constant and ʋin is the ion-neutral collision frequency and Te and Ti are electron and ion temperatures. TIEGCM simulations reveal that the F2 layer is isothermal considering both ion and electron temperatures in the altitude regions of interest in this study. Accordingly, the WD expression reduces to the second term in equation (6) that is, g/ʋin. The first term in equation (5) is obtained from ionogram data, while the second and third terms are derived using atmospheric models. The meridional wind U is thus estimated from equation (5). A mammoth effort was undertaken to scale each ionogram of 15 minute cadence for the years spanning 1989 to 2008 from two stations. A sum total of ~650 days of data have gone into the mean for low solar activity years (1993-1998, 2003-2008; SFU< 130) and ~300 days for high solar activity years (1989-1992; 1999-2002; SFU≥150) making it a large database of meridional wind. Periods contaminated due to spread F are eliminated from all the data. Previous references chose the level of Ap of 70 as quiet conditions as far as impact on meridional winds were concerned (Krishna Murthy et al., 1990). However, we have adopted the cut-off of Ap<18 to represent magnetically quiet days since monthly mean values of winds having more number of days with higher Ap levels are seen to bias the observations. The sources of error in wind estimation are from the ambiguity in ionosonde h’F determination (first term of equation 5), errors in model estimation of (second term of equation 5) and error due to diffusion. The error due to increases for days with low h’F while simultaneously, the error in diffusion decreases 6
and vice versa. Also, the maximum limit of error in the estimation of a single wind time series is calculated to be ±25 m/s. However, the mean is taken for a number of such points and over several years of data. As a result of the random nature of such errors, it ultimately reduces to within the standard error limits. We are depicting the standard errors in all the figures in this study as followed by previous workers (Hari and Krishnamurthy 1995; Sekhar and Sridharan 1994; Krishnamurthy et al., 1990), in the light of the fact that the random errors reduce through increased number of measurements (Bevington 1969). In order to obtain the mean seasonal electron density, the level 2 product from the Planar Langmuir Probe (PLP) onboard the CHAMP satellite has been employed (Cooke et al., 2003). The total mass density, number density and the exospheric temperatures has been obtained from the MSIS model. The observations of thermospheric winds are compared with the empirical Horizontal Wind Model (HWM) and Thermosphere Ionosphere Electrodynamics General Circulation Model (TIEGCM). The input parameters for the model are primarily the solar F10.7 cm flux, and Kp value. This is simultaneously taken into account from the date entered for simulation. Then the model solves the three-dimensional momentum, energy and continuity equations for neutral and ion species at each time step to obtain temperatures, composition and neutral winds. 3. Results Fig.1 shows the temporal average of TMWs during HSA (red) and LSA (blue) irrespective of seasonal variations obtained by averaging ~950 days of measurements. In the entire paper ‘hr’ corresponds to Indian Standard Time which is 5 hours 30 minutes ahead of Universal Time. Positive value indicates that winds are directed towards the North Pole (poleward) and negative values indicate that the winds are directed towards the equator, south of the station (equatorward).
7
The major factors that are immediately evident from the figure are i) During the post-sunset hours from 1800-2000 hr it can be seen that there is no change in TMWs with solar activity ii) Significant solar activity response of TMWs is seen for the pre-midnight hours between 2000 and 0000 hr. The equatorward winds during LSA are higher by as much as 30 m/s compared to HSA iii) Post-midnight, the magnitude of winds are again comparable, except around 0200 hr where the LSA winds are slightly more poleward than during HSA. In order to examine the seasonal variations of TMWs for different epochs, the data are segregated into mean of respective solar activity levels and plotted against local time in Fig. 2. Top left panel shows the mean TMWs obtained for HSA and LSA for the Vernal Equinox (VE) season. Deviation between HSA and LSA starts around 2100 hr and continues till 0000 hr. Magnitude of equatorward winds is higher during this period for LSA with peak magnitude of 50 m/s at 2215 hr compared to 10 m/s for HSA. Post 0000 hr, winds during both LSA and HSA are poleward directed, with no significant difference in magnitudes. The pattern of winds is the same for Autumnal Equinox (AE; left bottom panel) season also, the peak equatorward wind during HSA is 30 m/s attained around 2200 hr. The peak magnitude of equatorward winds during LSA is 50 m/s around 2200 hr, after which there is an abatement. Magnitude of winds during HSA of VE is also found to be more poleward than that for AE during the time 0200 hr – 0400 hr. During the Summer Solstice (SS) the post sunset winds are equatorward directed right from 1800 hr. Equatorward wind deviation between LSA and HSA is maximum around 2200 hr. Magnitude of maximum equatorward wind during HSA is 20 m/s and during LSA is 45m/s. MTM related abatement is evident in this season also. During Winter Solstice (WS), winds turn equatorward briefly between 2015 hr and 2145 hr for LSA in the pre-midnight sector, while it remains entirely poleward for HSA. The difference in winds between HSA and LSA during winter at
8
2100 hr is 20 m/s. The equatorward reversal during LSA is attributed to the MTM effect. Postmidnight, around 0300 hr, the winds turn equatorward again, both during LSA and HSA. It is evident from the stronger abatement seen in the equatorward winds during LSA that there are stronger influences of MTM during LSA. There is 40 m/s increase in the magnitude of abatement in equatorward winds for LSA compared to HSA. Another important aspect pertains to the asymmetry of the winds between the two equinoxes at 2200 hr. While no equinoctial asymmetry is seen between the equinoxes for LSA, there is a clear asymmetry in winds for HSA, amounting to 20 m/s at 2200 hr. The availability of data throughout the span of 2 solar cycles allows us to examine the rate of change of meridional wind magnitude with solar flux. For this, the entire dataset of 20 years during the time period 2000-0000 hr is first segregated into monthly mean values. This monthly mean meridional winds is plotted against the corresponding mean solar fluxes for all seasons (Figure 3). A linear correlation is observed for all the seasons between the winds and solar flux. The correlation coefficients obtained for the linear least squares fit is 0.76 for VE (significance>99%), 0.75 for SS (significance>98%), 0.56 (significance>95%) for AE and 0.5 for WS (significance>95%). This shows a clear trend where the equatorward winds are gradually reduced with increasing solar activity levels for all seasons. The linear relation between the winds and solar flux has been arrived at and are given as inset in each panel. The rate of change of TMWs w.r.to solar activity for VE is 0.25 m/s/SFU, 0.2m/s/SFU for SS, 0.08 m/s/SFU for AE and 0.19 m/s/SFU for WS. 4. Discussion In the thermosphere, the air may be regarded as a single fluid, subject to hydrodynamic equations of motion owing to its characteristic collisional frequencies and the fact that differential motion
9
of any of its various constituents is very much less than the overall wind speed. Solar heating around the sub solar point causes thermal expansion of the atmosphere forming the ‘diurnal bulge’ giving rise to horizontal pressure gradients that drive horizontal winds towards the antisolar point situated at nighttime longitudes. Thermospheric winds blow across the polar regions and zonally around the Earth under the dominant influence of viscous force and ion drag unlike the winds in lower altitudes where the winds roughly follow the isobars under geostrophic balance (Rishbeth 1972). During daytime the thermospheric winds diverge from the equator towards the poles and hence the direction of day time winds at a station like SHAR is poleward. As SHAR enters the night sector, convergence of winds from the dayside sub solar point (at ~100°W by midnight) takes over and the winds turn equatorward. By around midnight the phenomena of Midnight Temperature Maximum ensues as a result of the tidal forcing from lower atmosphere and adiabatic heating due to convergence and causes an outflow that is stronger than the wind magnitudes set up by the pressure gradient across earth’s great circle (Hari and Krishna Murthy 1995; Sastri et al., 1994, Niranjan et al., 2006). Earlier work related to meridional wind in the Indian sector was carried out by Hari and Krishna Murthy (1995) during high solar activity years of 1989-90 and they address the seasonal pattern of nocturnal thermospheric winds. The present study looks into the solar activity variations in TMWs by analyzing available data spread over two decades. The study conclusively shows that there is an increase in the magnitude of the thermospheric equatorward winds with decreasing solar activity in the pre-midnight hours. Maximum response to solar activity is seen in the VE months followed by SS, AE and WS. The study undertaken by Liu et. al 2003b at a station Wuhan in the East-Asian longitude sector show similar equatorward reversals in wind component along the magnetic equator with increasing
10
solar activity. Optical observations of thermospheric winds at Arecibo, in the American sector carried out by Burnside and Tepley (1989) using a limited dataset, revealed no significant solar activity differences in meridional winds. Later, Tepley et. al. (2011) undertook a detailed examination of the nocturnal TMWs using an extended dataset from the same region and hinted at the presence of a solar activity dependence in meridional winds, of very small in magnitude, during the pre midnight sector. The differences in wind magnitudes are observed only around pre-midnight hours for Indian sector in the present study also, with negligible effects post midnight. Brum et al. (2012), using the extended data set of optical measurements from the South American sector reports a mean wind response rate of 0.2-0.25 m/s/SFU for VE, 0.15-0.2 m/s/SFU for AE, 0.25m/s/SFU for SS, and 0.2m/s/SFU for WS at 0000 hr. Our observations from the Indian sector show the solar flux dependencies at nearly the same rate, of 0.23 m/s/SFU for VE, 0.09 m/s/SFU for AE and 0.18 m/s/SFU for SS and WS during 2000-0000 hr. The solar activity dependences are seen to be higher for VE and SS in our analysis. Buosanto et. al. (1999) reports a clear increase in the equatorward winds during low solar activity and attribute this to the differences in ionization and ion-drag at North American longitude sector (Millstone hill). Liu et al., 2004 have undertaken a detailed study of thermospheric winds derived from ionosondes at various locations in both northern and southern hemisphere. They report a reduction in diurnal magnitudes of meridional winds with increasing solar activity. They identify the limiting factor of winds to be the ion-drag increase during high solar activity. Luan et al., 2004 have investigated the response of equivalent wind derived from ionosonde to solar activity at various latitudes. It is interesting to note that the pattern of observations is similar to the observations made in the Indian sector. For example, they report the maximum response to solar
11
activity to be occurring in VE at 2200 hr. During AE, the high solar activity winds are more equatorward than in VE, a feature that has emerged from our observations also at 2200 hr. 4.2 Theoretical estimation of winds In order to analyze the probable causative mechanism for the observed differences in TMWs during different solar activity epochs, in the present study, the components that make up the thermospheric winds are examined. For this, the equation of motion for thermospheric winds is considered, where the gradient in the wind is balanced by the sum of the pressure gradient force, the ion-drag force, the viscous force and gravity, as follows: U 1 p 2U (7) ni (U V ) g 2 2 t x H x where, U is the horizontal wind component, x is the great-circle distance, p is pressure, ρ is the
density, υin is the ion-drag forcing, V is the ion velocity, µ is the coefficient of molecular viscosity, H is the scale height and g is acceleration due to gravity. Divergence of the wind field is the sum of pressure gradient existing across the great circle distance, the ion drag component imparted by the ions to the neutral flow, viscous force and gravity. The Coriolis force is negligible near the equator and is thus ignored. The continuity equation for the meridional wind that includes these parameters has to be satisfied locally. A sensitivity test on the changes in meridional winds due to solar activity variabilities in the individual components is performed to delineate the impacts on various forcings due to solar activity. First, the pressure gradient, which is a function of neutral density and exospheric temperature, is estimated for a typical magnetically quiet day of autumnal equinox season at 2200 hr for SHAR during HSA (2002) and LSA (2008) using the neutral densities and temperature from the MSIS model. Figure 4 shows the contour of pressure gradient estimated for the entire globe for high and low solar activity at 2200 hr local time at SHAR. There is a clear
12
enhancement in the pressure gradient force by around 38% from LSA to HSA at the location of SHAR. The ion drag forcing on the meridional wind is estimated next from the following expression,
ni
in * N e [O]
(8)
Where ʋin is the sum of ion-neutral collision frequencies of each constituent in the thermosphere, Ne is the electron density and O is the number density of atomic oxygen (Rishbeth 1969). The estimation of νni is undertaken using in-situ CHAMP electron densities available for autumnal equinox months of HSA and LSA for 2200 hr. For HSA, mean of electron densities during AE 2001-2002 is used, whereas for LSA, the mean electron densities during AE of 2007-2008 are used. Figure 5 shows the ion-drag force estimated for LSA and HSA during 2200 hr at SHAR for AE. It is very interesting to note that the pattern of ion drag is clearly reflecting the double crest pattern of EIA for HSA. This is very much in line with the expected presence of EIA well past midnight in HSA. The manifestation of EIA is found to continue well into the night even up to 0200 hr during HSA (Sastri, 1990). Thus a region of high electron density would persist in the low-latitude regions as a result of the enhanced pumping of ionized plasma from the magnetic equator at thermospheric altitudes resulting in increased ion-drag during about 1930 – 0200 hr. Its strength and duration is subject to the changes in solar activity. For LSA, the early recession of EIA leads to low electron densities and consequent low ion drag in the EIA regions. In view of this EIA induced enhanced ion drag there is an order of magnitude increase (14.3 times) seen in ion-drag forcing ʋni between LSA and HSA near Indian sector. This is at a much higher rate than in the case of the pressure gradient force (~ 1.38 times).
13
From the altitude profiles of ion-drag force, pressure gradient force, viscous force and Coriolis force given in Rishbeth (1972) it is clear that there is a region between 150 km and 400 km, where pressure gradient and ion-drag forces are predominant in the thermosphere. Below the altitude of 200 km, the Coriolis force is much greater than that of the ion drag terms and the winds follow the geostrophic approximation where the pressure gradient is balanced by the Coriolis force. Again, above the altitude of 400 km, where Coriolis force is significant, viscous forces become much stronger and hence this term has to be considered in the final solution to thermospheric wind equation. These inferences are often the boundary conditions for the numerical solutions of winds at ionospheric heights. Hence, in the altitude range of 150 km to 400 km the wind term is often expressed as the ratio of F/ ʋni and Rishbeth (1972) describes the magnitude of changes in this parameter for various times of the day as well as two solar epochs, to discuss ion-drag effects on thermospheric winds at mid-latitudes. The square of the sine component of the latitude is also to be included in the calculations to get the absolute values of the meridional winds, but that is presently ignored since that factor remains unchanged for both solar activity epochs. Liu et. al., 2004 also details the effectiveness of this parameter in capturing the variations due to ion-drag in thermospheric meridional winds. The point values in AE panel on the right column of Figure 6 shows the estimated wind for HSA and LSA at 2200 hr. For HSA, the magnitude of equatorward wind obtained is 4.6 m/s and for LSA, it is 48.3 m/s. There is an increase of 43.7 m/s in equatorward wind from HSA to LSA on account of changes in pressure gradient and ion-drag. The ion drag imparted by the increased magnitude of electron density and a stronger EIA during HSA years is having the major say in the reduction of equatorward meridional winds around equatorial and low latitudes. At equatorial and low latitude regions, there is an enhancement in the EIA soon after the Pre-Reversal
14
Enhancement (PRE) of the zonal electric field around the time after sunset, and it persists up to midnight in LSA. As the pressure gradient drives the winds across these regions of increased electron density and hence ion drag, the wind magnitudes are lower for high solar activity years. During post-midnight hours, the MTM effects become dominant and hence the ion drag effects and consequent effects on winds are not clearly visible. By around 0200-0300 hr, the MTM effects diminish, but by then the EIA is also inhibited and so the ion drag effects on winds becomes less significant. 4.2 Model simulations We compare the observational results from the Indian sector with the model outputs of the Horizontal Wind Model and TIE-GCM runs for different seasons and solar activity levels at the co-ordinates of SHAR. Fig. 6 (left column) shows the HWM derived winds. The HWM is an empirical model and has not inculcated solar flux related changes and hence is unable to reproduce the solar activity changes. However, it reproduces the seasonal trends quiet well. Equatorward nature of winds during SS and poleward nature of winds during WS is reproduced by the model. The brief equatorward reversal of winds during LSA of WS however is not seen in the model. The time of reversal of winds from equatorward to poleward direction around midnight, is reproduced well by HWM for HSA only. For LSA, HWM fails to show the MTM related changes occurring in the magnitude and time of abatement of equatorward winds. Fig 6 (right column) shows the TIEGCM obtained meridional winds for different seasons and solar activities. TIEGCM is a physical model based on the continuity equations and is the latest model available for major thermospheric constituents and parameters. It solves the threedimensional momentum, energy and continuity equations for neutral and ion species at each time
15
step to obtain temperatures, composition and neutral winds. The increased magnitudes of equatorward winds during LSA are captured well by the model. There is a brief excursion of equatorward winds during 2400 hr for LSA period of WS. The overall pattern of winds in all time sectors and for all seasons and solar activities is better captured in TIEGCM simulations than HWM simulations. However, the time of abatement of equatorward winds in response to MTM and related effects seen in the observed winds are not clearly captured by the TIEGCM model. The time of abatement (for observations) is by 2200 hr whereas in the model, the time of reversal is later by 2 hours and is only in response to changes in the direction of pressure gradient. The calculated winds for HSA is quite close to TIEGCM values. However there is a deviation w.r.to winds during HSA. These observations highlight the need for more measurements of thermospheric meridional winds especially from the equatorial and low-latitude regions as they are an important component of large scale equatorial phenomena like ESF and EIA. Future directions also include realization of a model including the viscous forces and MTM induced pressure gradient changes that better reflect the observations in the Indian longitude sector. 5. Summary A comprehensive analysis of nocturnal thermospheric meridional wind pattern encompassing two solar cycles is accomplished at the Indian longitude sector. Significant difference is seen in winds between high and low solar activity epochs, with more equatorward winds during pre midnight hours for low solar activity years. The solar flux relationship of mean winds during pre midnight hours is established. An integrated approach using the existing observational techniques, models as well as satellite based inputs are employed to understand the underlying causative mechanisms behind the observed solar activity responses of thermospheric winds. The
16
increased ion drag forcing existing during high solar activity years has been identified to be the major factor causing the observed pre midnight sector differences in the nocturnal thermospheric winds over Indian longitude sector. The study highlights the need to improve the existing models incorporating the effects of physical processes like MTM and ion-drag effects more expansively. Acknowledgements This work is funded by Indian Space Research Organization, Govt. of India. The ionosonde data used is the property of the organization. The data may be made available on specific request to the corresponding author with the permission from the concerned authorities. The CHAMP satellite data has been obtained from Information System and Data Center, hosted at http://isdc.gfz-potsdam.de/. TIEGCM simulation results have been provided by the Community Coordinated Modeling Center at Goddard Space Flight Center through their public Runs on Request system (http://ccmc.gsfc.nasa.gov). The CCMC is a multi-agency partnership between NASA, AFMC, AFOSR, AFRL, AFWA, NOAA, NSF and ONR. References Bevington P.R., (1969), Data reduction and error analysis for the physical sciences, Mc Graw hill. Bramley E.M. and Young M., (1968), Winds and electromagnetic drifts in the equatorial F2 region, J. Atmos. Terr. Phys., 80, 90. Brum, C. G. M., C. A. Tepley, J. T. Fentzke, E. Robles, P. T. Santos, and S. A. Gonzalez (2012), Long-term changes in the thermospheric neutral winds over Arecibo: Climatology based on over three decades of Fabry-Perot observations, J. Geophys. Res., 117,458, DOI: 10.1029/2011JA016. Buonsanto, M. J., and O. G. Witasse (1999), An updated climatology of thermospheric neutral winds and F region ion drifts above Millstone Hill, J. Geophys. Res., 104, 24, 675.
17
Burnside, R. G., and C. A. Tepley (1989), Optical observations of thermospheric neutral winds at Arecibo
between
1980
and
1987,
J.
Geophys.
Res.,
94,
2711–2716,
doi:10.1029/JA094iA03p02711. Cooke, D. L., W. Turnbull, C. Roth, A. Morgan, and R. Redus (2003), Ion drift-meter status and calibration, in First Champ Mission Results for Gravity, Magnetic, and Atmospheric Studies, edited by C. Reigber, H. Luhr, and P. Schwintzer, pp. 212–219, Springer, New York. Devasia, C. V., N. Jyoti, K.S. Viswanathan, K. S. V. Subbarao, D. Tiwari and R. Sridharan (2002), On the plausible linkage of thermospheric meridional winds with equatorial spread F, J. Atmos. Sol. Terr. Phys., 64, 1-12. Dickinson, R. E., E. C. Ridley and R. G. Roble (1981), A three-dimensional general circulation model of the thermosphere, J. Geophys. Res., 86, 1499-1512. Duboin, M.-L., and M. Lafeuille (1992), Thermospheric dynamics above Saint-Santin: Statistical study of the data set, J. Geophys. Res., 97, 8,661. Hari, S. S., and B. V. Krishna Murthy (1995), Seasonal variations of equatorial nighttime thermospheric meridional winds, J. Atmos. Terr. Phys., 57,1241-1246. Harper, R. M. (1973), Night time meridional neutral winds near 350 km at low to mid-latitudes, J. Atmos. Terr. Phys., 353, 2023. Hedin A.E, N. W. Spencer, and T. L. Killeen (1988), Empirical Global Model of Upper Thermosphere Winds Based on Atmosphere and Dynamics Explorer Satellite Data, J. Geophys. Res. 93, 9959- 9978. Hedin, A. E., et al. (1991), Revised Global Model of Thermosphere Winds Using Satellite and Ground-Based Observations, J. Geophys. Res., 96(A5), 7657–7688, doi: 10.1029/91JA00251. Hedin, A. E., M. J. Buonsanto, M. Codrescu, M.-L. Duboin, C. G Fesen, M. E. Hagan, K. L. Miller, and
D. P. Sipler (1994), Solar activity variations in mid latitude thermospheric
meridional winds, J. Geophys. Res., 99(A9), 17601- 17608.
18
Krishna Murthy, B. V., S. S. Hari, and V. V. Somayajulu (1990), Nighttime equatorial thermospheric meridional winds from ionospheric h0F data, J. Geophys. Res., 95(A4), 4307– 4310. Liu L., X. Luan, W. Wan, J. Lei, and B. Ning (2003a), Seasonal behavior of equivalent winds over Wuhan derived from ionospheric data in 2000-2001, Adv. Space Res., 32(9), 1765-1770. Liu L., W. Wan, X. Luan, B. Ning, and J. Lei (2003b), Solar activity dependence of effective winds derived from ionospheric data at Wuhan, Adv. Space Res., 32(9), 1719-1924. Liu, L., X. Luan, W. Wan, J. Lei, and B. Ning (2004), Solar activity variations of equivalent winds
derived
from
global
ionosonde
data,
J.
Geophys.
Res.,
109,
A12305,
doi:10.1029/2004JA010574. Luan, X., Liu, L., Wan, W., Lei, J., and Yu, T. (2004), A climatology of the F-layer equivalent winds derived from ionosonde measurements over two decades along the 120°-150°E sector, Ann. Geophys., 22, 2785-2796, doi:10.5194/angeo-22-2785-2004. Madhav Haridas, M. K., G. Manju, and T. K. Pant (2013), First observational evidence of the modulation of the threshold height h'Fc for the occurrence of equatorial spread F by neutral composition changes, J. Geophys. Res., 118, 3540–3545, doi:10.1002/jgra.50331. Manju, G. ; Devasia, C. V. ; Sridharan, R. (2007), On the seasonal variations of the threshold height for the occurrence of equatorial spread F during solar minimum and maximum years, Ann Geophys, 25 (4). pp. 855-861. ISSN 1432-0576. 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:10.1029/2008JA013031. Maruyama, T., Saito, S., Kawamura, M., Nozaki, K., Krall, J., and Huba, J. D. (2009), Equinoctial asymmetry of a low-latitude ionosphere-thermosphere system and equatorial irregularities: evidence for meridional wind control, Ann. Geophys., 27, 2027-2034, doi:10.5194/angeo-27-2027-2009.
19
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:10.1029/2005JA011386. Nogueira P.A.B., M.A. Abdu, I.S. Batista, P.M. de Siqueira (2011), Equatorial ionization anomaly and thermospheric meridional winds during two major storms over Brazilian low latitudes, J. Atmos. Sol. Terr. Phys., Volume 73, Issues 11–12, 1535-1543, ISSN 1364-6826, http://dx.doi.org/10.1016/j.jastp.2011.02.008. Raghavarao, R., L. E. Wharton, N. W. Spencer, H. G. Mayr, L. H. Brace (1991), An equatorial temperature and wind anomaly (ETWA), Geophys. Res. Lett. 18, 1193-1196. Rishbeth, H., and O. K. Garriott (1969), Introduction to Ionospheric Physics, Academic, New York. Rishbeth, H., (1972) Thermospheric winds and the F region: A review. J. Atmos. Terr. Phys. doi:10.1016/0021-9169(72)90003-7. Rishbeth, H., Ganguly, S., and Walker, J. C. G. (1978), Field-aligned and field-perpendicular velocities in the ionospheric F2 layer, J. Atmos. Terr. Phys., 40, 767–784. Roble, R. G., R. E., Dickinson and E. C. Ridley (1977), Seasonal and solar-cycle variations of zonal mean circulation in the thermosphere, J. Geophys. Res., 82, 5493-5504. Sastri,
J.
H.
(1990),
Equatorial
anomaly
in
F
region-
A
review,
Ind.
Journ.
Rad. Space Phys., 19, 225-240. Sastri, J. R, H. N. R. Rao, V. V. Somayajulu, and H. Chandra (1994), Thermospheric meridional neutral winds associated with equatorial midnight temperature maximum (MTM), Geophys. Res. Lett., 21, 825-828. Sekar R. and Sridharan R. (1992), Validity of the estimates of night-time meridional winds made from bottomside ionograms, J. atmos, terr.Phys. 54, 1197-1199. Tepley, C. A., E. Robles, R. Garcia, P. T. Santos, C. M. Brum, and R. G. Burnside (2011), Directional trends in thermospheric neutral winds observed at Arecibo during the past three solar cycles, J. Geophys. Res., 116, DOI:10.1029/2010JA016, 172.
20
Tulasi Ram, S., S.Y. Su, and C. H. Liu (2009), FORMOSAT-3/COSMIC observations of seasonal and longitudinal variations of equatorial ionization anomaly and its interhemispheric asymmetry
during
the
solar
minimum
period, J.
Geophys.
Res., 114,
A06311,
doi:10.1029/2008JA013880. Qian, L., A. G. Burns, B. A. Emery, B. Foster, G. Lu, A. Maute, A. D. Richmond, R. G. Roble, S. C. Solomon, and W. Wangm (2014), The NCAR TIE-GCM: A community model of the coupled thermosphere/ionosphere system, in Modeling the Ionosphere-Thermosphere System, AGU Geophysical Monograph Series.
21
Figure 1: Comparison of mean thermospheric wind between high and low solar activity years
22
Figure 2: Mean temporal evolution of thermospheric meridional wind during two solar activity epochs for Vernal Equinox (VE), Summer Solstice (SS), Autumnal Equinox (AE) and Winter solstice (WS)
23
Figure 3: Solar flux dependence of mean thermospheric winds during 2000 -0000 hr for Vernal Equinox (VE), Summer Solstice (SS), Autumnal Equinox (AE) and Winter solstice (WS)
24
Figure 4: pressure gradient estimated for HSA (top panel) and LSA (bottom panel) at 2200 hr for AE
25
Figure 5: Ion-drag force estimated for HSA and LSA during 2200 hr for AE
26
Figure 6: Winds obtained for different seasons/solar activity epochs using HWM (left column) and TIEGCM (Right column). The calculated winds at 2200 hr for AE is shown as stars highlights Solar activity variations of nocturnal thermospheric meridional winds established in Indian sector Equatorward wind magnitudes increase with decreasing solar activity in pre-midnight sector Ion-drag identified as a major source of variations and the observations compared with model data
27