Ionosphere Variability in Low and Mid-Latitudes of India Using GPS-TEC Estimates from 2002 to 2016 Ionosphere Variability in Low and Mid-Latitudes of India Using GPS-TEC Estimates from 2002 to 2016

Continuous Global Positioning System (cGPS) observations spanning 14 years at 24 cGPS sites located in low and mid-latitudes (5 – 35 (cid:1) N) in the Indian subcontinent are analyzed to extract the ionosphere delay from one-way residuals for each satellite. Absolute integrated electron content (IEC) is the integral of electrons per m 2 along the line of sight between the satellite and receiver. Total electron content (TEC) is determined from IEC using elevation mapping function to normalize the variation of the ray path length through the ionosphere based on the GPS satellite elevation angle. In this study, GPS TEC estimates temporally cover two solar cycles (23 and 24) and spatially cover equatorial ionization anomaly (EIA) region and beyond, thus depicting the ionosphere variability in space, time and geographi- cal location. Results capture different phases of solar cycle, EIA, annual, daily, diurnal and seasonal variability of ionosphere in northern hemisphere. This chapter gives significant insights in to the high and random variability of TEC associated with the changes in solar activity, intensity of the sun radiation, zenith angle at which they impinge the earth ’ s atmosphere, equatorial electrojet (EEJ) and plasma flow. Intensity of EIA and its latitude of crest development vary with the strength of EEJ, season and solar activity. Our study indicates that the northern crest of EIA region extends up to about 17 – 18 (cid:1) N geomagnetic latitude in Indian region. Diurnal variability of ionosphere depends on the intensity of solar activity, season and strength of geomagnetic field with high TEC values recorded in 2004 and 2011 and low values in 2009. Day-to-day variability is more pronounced for the high solar activity years when compared to low solar activity years. Maxima occurs during midday (7 – 13 h UT) with longer duration for geomagnetic latitudes between 0 and 9 (cid:1) N and pronounced peaks for greater than 9 (cid:1) N. Minima occurs after midnight (20 – 24 h UT) between 0 and 9 (cid:1) N geomagnetic latitude whereas it is for longer duration (15 – 24 h UT) for northern crest of EIA region and beyond. Day-to-day variability of maxima is more pronounced in the crest of EIA regions (9 – 18 (cid:1) N geomagnetic latitude). Day-to-day variability of diurnal TEC is high and random during December 2004 and September 2011 due to seismo-ionosphere disturbance caused by 2004 Sumatra and 2011 Sikkim earthquake. Also anomalous day-to-day TEC variability is observed for Gujarat stations (MABU, KHAV, BELP) in 2011 which needs further detailed study. Diurnal maxima and minima vary significantly during the equinox and solstice of summer and winter seasons with lower values during summer solstice in EIA region and higher values during equinox and winter solstice. Beyond EIA (>18 (cid:1) N), maxima with pronounced peak occurs in the equinox


Introduction
Ionosphere consists of layers of earth's atmosphere containing free electrons as a result of ionization of the atoms in this region by high energy from sun and cosmic rays. These layers of free electrons surrounding the earth from 60 to 1100 km altitude influence the GPS signal propagation, causing errors in positioning. Total electron content (TEC) is estimated from the dual frequency GPS receiver signals by extracting the phase advances and code delays caused by ionosphere. Precise TEC estimates give significant insights into the variability of ionosphere in space, time, geographical location and solar and cosmic activity. GPS based ionosphere research was initiated globally [1][2][3][4][5][6] for large scale studies, local earthquakes, mine blasts, and so on. Spatial and temporal variability of ionosphere based on GPS-TEC estimates was studied by several researchers [7][8][9][10][11][12][13][14][15][16][17][18] using GPS data in different regions of the world giving insights into the response of ionosphere due to the variations in solar activity, geomagnetic storms, and so on. Ionosphere maps for few regions were prepared from the GPS-TEC estimates from a network of stations.
In India, GPS based ionosphere studies were initiated after the establishment of dual frequency GPS stations in 2003 by Indian Space Research Organization and Airport Authority of India as a part of the GAGAN (Geo And GPS Augmented Navigation) program. For the first time in India, spatial and temporal variability [19] of equatorial ionosphere is studied using GPS-TEC estimates for a 16-month period (March 2004-June 2005) with low sunspot activity (LSSA) using 18 GPS station data covering a geomagnetic range of 1 S to 24 N. Using the same GAGAN network GPS data [20], estimates of GPS TEC were compared with the International Reference Ionosphere (IRI) predicted TEC values. They have also investigated diurnal, seasonal and annual variability of ionosphere over Indian subcontinent during the 16-month LSSA period. For a low latitude station Rajkot located near the equatorial ionization anomaly crest region in India [21], ionosphere variability during LSSA period (2005)(2006)(2007) was investigated to give insights into solar activity dependence and effects of geomagnetic storm on GPS-TEC. Variability of GPS-TEC at a single station Udaipur in Rajasthan for a period of 2005-2010 was studied [22] and the result of seasonal variations are compared with IRI-2007 Model. Similarly diurnal and seasonal variation of GPS-TEC at a single station Agra, for the LSSA period (2006)(2007)(2008)(2009) was studied [23]. GPS-TEC estimates for Surat GPS station [24,25] were compared with model predictions from IRI-2007 and IRI-2012 and the ionosphere variability was investigated. GPS-TEC derived [26,27] from a chain of Indian stations for a 1 year period (2011-2012) was used to study the diurnal, seasonal and latitude variability and its relation to geomagnetic storms, solar eclipse, and so on. They gave comparison of GPS-TEC with IRI-2012, Standard Plasmasphere-Ionosphere Model (SPIM) and Global Ionospheric Maps (GIM).
All the above studies so far reported in the Indian subcontinent were for a period of 1-2 years over single and network of GPS stations. For the first time we report the GPS-TEC estimates for a period spanning 14 years (2002-2016) covering solar cycle 23 (1996-2008) and 24 (2008-2019) from a network of about 24 cGPS stations ( Figure 1; Table 1) with geodetic latitude ranging from 5 to 35 N and longitude ranging from 70 to 95 E in Indian subcontinent. New set of cGPS station data is used for the present study compared to majority of earlier ionosphere studies which use GAGAN network data and hence give an independent estimate of ionospheric TEC in this region. The geomagnetic latitude and longitude of these GPS stations ( Table 1) is 0-26 N and 145-168 E which is very important for the study of ionosphere variability as equatorial region has the high ionosphere activity compared to the rest of the regions in the world. In addition, for the first time TEC estimates are reported for region beyond the EIA region in the Indian subcontinent using cGPS data. Annual, spatial, seasonal, diurnal variability of ionosphere is presented using these TEC estimates and its relation to solar activity, EEJ, EIA is investigated.   Figure 1) located in the Indian subcontinent with data span ranging from 2 to 13 years with sampling interval of 30 s has been used. The dataset span is more than any previous study till date and spatially covers the length and breadth of the Indian subcontinent. Details of the cGPS sites and the data used are listed in Table 1 and the data is analyzed using GAMIT software [28]. Data sampling interval of 30 s and elevation cut-off angle of 20 is used for the analysis. The quality check of GPS data at each station was done using TEQC software [29] to remove data with several cycle slips, multipath and span of less than 18 h. The daily data of all the stations after quality check is analyzed using GAMIT to extract the ionospheric delays suffered by GPS signals in L-band with frequency f 1 (1.57542GHz) and f 2 (1.2276GHz) from the one-way residuals of each satellite at 30 s interval. These one-way residuals for each satellite are output after cleaning the observables for cycle slips, multipath, outliers during the analysis. The IEC along the line of sight (LOS) between the satellite and receiver are calculated from the carrier phase delays L 1 and L 2 and the group delays P 1 and P 2 (code pseudo-ranges) in range units obtained from GAMIT analysis as detailed earlier. The absolute IEC [30] is given by where, ambiguity constant B is given by where, n is number of phase measurements in a given arc [1]. L G and P G are linear combinations as given above whereas L G is precise and smooth with unknown phase ambiguity constant and P G is noisy and less precise and not ambiguous [2,31]. To estimate the absolute and precise estimate of IEC, we fit L G on P G using ambiguity constant B. λ 1 and λ 2 are the wavelengths of the L-band GPS signals. IEC is given in TEC units where 1 TECU = 10 16 electrons/m 2 . To calculate vertical equivalent TEC along the satellite receiver path, elevation mapping function (emfθ) is used to account for the variation in the ray path length (L θ ) based on GPS satellite elevation angle θ varies with the orbital pass of each satellite as given below: where, H ion is the mean ionosphere thickness, R is earth's radius and H t and H b are the top and bottom altitudes of the ionosphere layer. Vertical equivalent TEC along the satellite receiver path in TEC units is given by TEC is computed at 30 s interval during the orbital pass of the each satellite at the each GPS station for all the 24 cGPS stations during 2002-2016. Two-sigma iterated average of TEC at 30 s interval is computed from the TEC of all the visible satellites at that epoch. GPS-TEC thus estimated is used to discuss the ionosphere variability over Indian subcontinent.

Ionosphere variability
Ionosphere is highly variable in space (geographical location) and time (solar cycle, seasonal, diurnal) and with solar-related ionospheric disturbances and earthquakes. About 15 cGPS sites are located in the equatorial ionization anomaly (EIA) region from geomagnetic equator to northern crest of EIA region (17 N) where the low latitude ionosphere exhibits annual, spatial, seasonal and diurnal variability. Nine cGPS sites are located in mid-latitude region beyond the EIA region in northern India and Himalaya. Using the GPS-TEC estimates, variability of ionosphere is discussed in the subsequent sections.

Inter-annual
There are about 12 sites with data span covering the two solar cycles. Daily mean value of GPS-TEC is plotted for these stations from 2002 to 2016 in Figure 2 to study the annual variability of ionosphere over the located beyond the EIA region, the TEC variability is low and depends upon the geographic location of these sites which is discussed in detail in the subsequent sections. Daily mean value of TEC plotted in Figure 2 indicates significant semi-annual and annual cycles. Sensitivity of TEC to solar activity is stronger at low latitudes when compared to mid-latitudes in the northern hemisphere.

Spatial variability
For the Indian subcontinent, geomagnetic equator passes through the southern bottom tip and the northern crest of EIA (15 N geomagnetic latitude) lies in the middle ( Figure 1) providing a unique opportunity for studying the ionosphere variability. Since we have several stations and a larger spread of data, we give detailed in-depth study of spatial variation of ionosphere from 5 to 35 N latitude (0-26 N geomagnetic latitude) and 70 to 95 E longitude. Also since the data covers different phases of two solar cycles, the results are given separately for each solar cycle.

Solar cycle 24
The

Diurnal variability
Diurnal variability of TEC depends on the Sun's orbit, changes in solar activity and intensity of radiance, earth's magnetic field and dynamics of neutral winds (diffusion of transported electrons from the equator). Plasma flow associated with the EIA effects the day-to-day variability of diurnal TEC for the stations located in EIA region. Geomagnetic and seismoionosphere disturbances also effect the day-to-day variability of diurnal TEC. Results of diurnal variability for two solar cycles is given below.    [32] which affected the cGPS stations in southern India.     high and random variability related to seismo-ionospheric disturbance due to Mw 6.9, 18 September 2011 Sikkim earthquake. It can be observed that the diurnal variability depends on the solar activity, solar radiance, geomagnetic field, latitude, longitude and plasma flow related to EIA effects.

Solar cycle 23
Monthly diurnal mean values of TEC are plotted from November 2004 to December 2005 for all the stations between geomagnetic equator and northern crest of EIA (0-17 N geomagnetic latitude) in Figure 9 and beyond EIA region in Figure 10.

Solar cycle 24
Monthly diurnal mean TEC values are plotted from January to December for low solar activity period of 2009 ( Figure 11) and ascending solar activity period of 2011 ( Figure 12). Monthly and seasonal cycle is not very pronounced during the low solar activity period of 2009 with marginally higher peak values recorded during October for IISC (35 TECU) and March for GHTU (40 TECU) and lower peaks during January (20)(21)(22)(23)(24)(25)(26)(27)(28)(29)(30) in the EIA region. For the ascending solar activity period of 2011, the monthly and seasonal variation is distinct with the highest (80 TECU) in EIA region during October and November and the lowest during January (20-30 TECU) consistent with the winter anomaly observed in the northern hemisphere. GPS-TEC values increase from geomagnetic equator to the crest of EIA region (17 N geomagnetic latitude) after which they gradually decrease toward mid-latitudes in the northern hemisphere. Latitude variability of ionosphere is more pronounced during the high solar activity years (2002)(2003)(2004)) when compared to low solar activity years (2008-2010). Diurnal peak TEC value has longer duration between 0 and 9 N geomagnetic latitude. Diurnal maxima have pronounced peaks and diurnal minima is observed for longer duration in the northern crest of EIA region and beyond. Ionosphere variability with longitude is observed for longitude difference of 19 E and above during the ascending phase of current solar cycle 24. Normally, solar radiation strikes the atmosphere more obliquely with increasing latitude decreasing its intensity and production of free electrons, whereas near the geomagnetic equator its strikes horizontally with eastward electric field during day and westward during night. This causes plasma diffusion along magnetic field lines at approximately AE15 geomagnetic latitudes forming crests on both the hemispheres (EIA region). Hence, TEC increases gradually from geomagnetic equator to the EIA crest, beyond which it decreases toward the mid-latitude regions. Intensity of EIA and its latitude of crest development vary with the strength of EEJ, season and solar activity. Our study indicates that the northern crest of EIA region extends up to about 17-18 N geomagnetic latitude in Indian region.

Summary
Diurnal variability of ionosphere depends on the intensity of solar activity, season and strength of geomagnetic field with high TEC values recorded in 2004 and 2011 and low values in 2009. Day-to-day variability is more pronounced for the high solar activity years when compared to low solar activity years. Maxima occurs during midday (7-13 h UT) with longer duration for geomagnetic latitudes between 0 and 9 N and pronounced peaks for greater than 9 N. Minima occurs after midnight (20-24 h UT) between 0 and 9 N geomagnetic latitude whereas it is for longer duration (15-24 h UT) for northern crest of EIA region and beyond. Day-to-day variability of maxima is more pronounced in the crest of EIA regions (9- Monthly diurnal mean TEC values are the highest in November and the lowest in the months June to August for solar cycle 23 and increase with latitude in the EIA region. This is due to the winter anomaly observed in the EIA region of northern hemisphere and is consistent with previous studies. Beyond EIA region, the high values are observed in the summer equinox months and November and minimum values occur during January. For the current solar cycle 24, the monthly and seasonal variability is marginal for the low solar activity year (2009) when compared to 2011. In the EIA region, the highest values are recorded during October-November and the lowest during January for ascending phase (2011) of current solar cycle 24. The seasonal and monthly variation is random depending upon the intensity of solar cycle and seasons in each year.
In summary, the temporal and spatial variability of equatorial, low and mid-latitude ionosphere reported using the GPS-TEC estimated from new GPS data during 2002-2016 are broadly consistent with previous studies globally and specific to the Indian subcontinent. When compared to previous studies, present study with longer data span and spatial spread gives significant insights into the randomness of day-to-day variability of ionosphere as detailed above. This high and random variability of TEC is due to the changes associated with solar activity, intensity of the sun radiation and zenith angle at which they impinge the earth's atmosphere. TEC variability on quiet days depends on the changes in Earth's magnetic field and EEJ strength. In equatorial and low-latitude region of Indian subcontinent there is intense east-west electric current (EEJ) due to neutral winds and the plasma flow associated with the EIA plays a significant role in the day-to-day variability of diurnal TEC. Ionosphere is also affected by solar and geomagnetic storms, solar eclipse, seismic disturbances, volcanic eruptions, tsunamis, and so on. Indian Space Research Organisation in collaboration with Airports Authority of India developed a model to predict TEC in the Indian region which can be used to provide TEC maps. They have used GAGAN (GPS Aided Geo Augmented Navigation) ground network of 18 stations for this model and predict TEC between 8 and 30 N latitude and 60-100 E longitude. Since the present study uses a new set of cGPS data for a 14 year period, benchmarking ISRO ionosphere model with the current data and combining with the current TEC estimates would give an opportunity to develop precise ionosphere models and maps for this region. In addition these GPS-TEC estimates can be used to model the spatial and temporal variability of the low and mid latitude ionosphere specific to Indian subcontinent. GPS TEC study has several applications in varied fields such as precise positioning, navigation, seismo-ionosphere coupling, propagation of radio waves and solar-terrestrial events.