Plasmaspheric electron content derived from GPS TEC and FORMOSAT-3/COSMIC measurements: Solar minimum condition

Transcription

1 Available online at Advances in Space Research (2) 27 Plasmaspheric electron content derived from GPS TEC and FORMOSAT-3/COSMIC measurements: Solar minimum condition Iu.V. Cherniak a,, I.E. Zakharenkova a, A. Krankowski b, I.I. Shagimuratov a a West Department of IZMIRAN, 1 Av. Pobeda, 2361 Kaliningrad, Russia b Geodynamics Research Laboratory, University of Warmia and Mazury, Olsztyn, Poland Received December 211; received in revised form 31 March 2; accepted 2 April 2 Available online 1 April 2 Abstract The plasmaspheric electron content (PEC) was estimated by comparison of GPS TEC observations and FORMOSAT-3/COSMIC radio occultation measurements at the extended solar minimum of cycle 23/2. Results are retrieved for different seasons (equinoxes and solstices) of the year 29. COSMIC-derived electron density profiles were integrated up to the height of 7 km in order to retrieve estimates of ionospheric electron content (IEC). Global maps of monthly median values of COSMIC IEC were constructed by use of spherical harmonics expansion. The comparison between two independent measurements was performed by analysis of the global distribution of electron content estimates, as well as by selection specific points corresponded to mid-latitudes of Northern America, Europe, Asia and the Southern Hemisphere. The analysis found that both kinds of observations show rather similar diurnal behavior during all seasons, certainly with GPS TEC estimates larger than corresponded COSMIC IEC values. It was shown that during daytime both GPS TEC and COSMIC IEC values were generally lower at winter than in summer solstice practically over all specific points. The estimates of PEC (h > 7 km) were obtained as a difference between GPS TEC and COSMIC IEC values. Results of comparative study revealed that for mid-latitudinal points PEC estimates varied weakly with the time of a day and reached the value of several TECU for the condition of solar minimum. Percentage contribution of PEC to GPS TEC indicated the clear dependence from the time with maximal values (more than 6%) during night-time and lesser values (2 %) during day-time. Ó 2 COSPAR. Published by Elsevier Ltd. All rights reserved. Keywords: Plasmasphere; Ionosphere; GPS; Total electron content; FORMOSAT-3/COSMIC; Solar minimum 1. Introduction Nowadays the measurements of Global Positioning System (GPS) are widely used by the scientific community for the Earth s ionospheric studies. The dense network of GPS receivers (a few thousands all over the world) fulfils simultaneous coverage in global scale with high temporal resolution. The height of GPS/GLONASS orbits is about 2,2 km above the Earth s surface, and so most part of the propagation path of a radio signal from a GPS satellite Corresponding author. Tel./fax: addresses: tcherniak@ukr.net (I.V. Cherniak), kand@uwm.edu.pl (A. Krankowski). to ground-based GPS receiver or GPS receiver onboard a Low Earth Orbit (LEO) satellite is mainly within the plasmasphere. As the electron densities in the plasmasphere are several orders of magnitude less than in the ionosphere (e.g. Gallagher et al., 2), the plasmasphere is often ignored at analysis and estimation of GPS TEC data, however the plasmaspheric contribution to the GPS TEC can become significant under certain conditions. The total number of electrons along a ray path from GPS satellite to ground-based receiver is called total electron content (GPS TEC), this value is composed mainly by ionospheric electron content, IEC, and plasmaspheric electron content, PEC. So the contribution of PEC to the GPS TEC can be estimated from the simultaneous measurements of GPS TEC and IEC. Estimates of IEC /$36. Ó 2 COSPAR. Published by Elsevier Ltd. All rights reserved.

2 2 I.V. Cherniak et al. / Advances in Space Research (2) 27 can be retrieved as a result of integration of ionospheric electron density profiles (EDP). For this aim one can use EDPs derived from model calculations, e.g. International Reference Ionosphere (Bilitza, 21), or ground-based ionosonde measurements (Huang and Reinisch, 21). At the first case we deal with model simulation results and there are a number of papers with PEC estimation by comparison with SUPIM (Sheffield University Plasmasphere Ionosphere Model) results (e.g. Lunt et al., 1999; Balan et al., 22). At the second case, we have some limitations, first of all with number of ground-based stations. As the ionosondes provide no direct information on the profile above the maximum electron density (F2 peak), the topside part of this EDP is constructed by fitting a model to the peak value, so the complete EDP in ionosonde measurements consists of a measured bottomside and a modeled topside part. The use of ionosonde data for estimation of PEC contribution to GPS TEC over specific regions can be found in several papers (e.g. Belehaki et al., 2; Mosert et al., 27). In the given research we estimate IEC on the basis of EDPs retrieved from the FORMOSAT-3/COSMIC radio occultation (RO) measurements. GPS RO data establish the basis for a new remote sensing technique for vertical profile information on the electron density of the entire ionosphere from satellite orbit heights down to the bottomside (Kirchengast et al., 2; Jakowski et al., 2; Liou et al., 21). The new LEO mission FORMO- SAT-3/COSMIC (Taiwan s Formosa Satellite Mission #3/Constellation Observing System for Meteorology, Ionosphere and Climate) is a joint scientific mission of Taiwan and the US and was launched on April 1, 26. The mission placed six micro-satellites into six different orbits at 7 km above the Earth s surface. The orbit inclination is 72. Each microsatellite has a GPS Occultation Experiment (GOX) payload to operate the ionospheric RO measurements. Depending on the state of the constellation, COSMIC has been producing 1 2 good soundings of the ionosphere and atmosphere per day, uniformly distributed around the globe. This number of RO is much higher than ones obtained by similar missions before. The total number of ionospheric occultations for is more than 3,, (more than, profiles per month). Previous missions (e.g. CHAMP, GRACE) were able to produce only 6 RO profiles per month (only several hundred per day). Therefore, COSMIC data can make a positive impact on a global ionosphere study, providing essential information about the height electron density distribution; particularly over regions that are not accessible with ground-based measuring instruments such as ionosondes and GPS dual frequency receivers. The objective of this paper was to illustrate the scope of RO measurements application for the ionosphere/plasmasphere studies and to accentuate the importance of plasmaspheric contribution to the GPS TEC estimates, especially during periods of low solar activity. 2. Database 2.1. IGS data The permanent GPS network provides regular monitoring of the ionosphere on a global scale with high resolution of TEC measurements. IGS Global Ionospheric Maps (GIMs) of TEC in the IONEX format were used. IONEX data are accessible at the ftp server: ftp:// cddis.gsfc.nasa.gov/pub/gps/products/ionex. The GIMs are generated routinely by the IGS community with resolution of longitude and 2. latitude and temporal interval of 2 h; one TEC unit (TECU) is equal 1 electrons/ m 2. Currently, there are three types of IGS GIMs: the final, rapid and predicted respectively. There are four IGS Associate Analysis Centres (IAACs) for the final and rapid ionospheric products: CODE, ESA/ESOC, JPL and gage/upc. Detailed description of IGS GIMs computation and validation can be found in Hernandez- Pajares et al. (29). The IAACs provide ionosphere maps computed with independent methodologies that use GNSS data from different set of GPS stations. These maps are uploaded to the IGS Ionosphere Product Coordinator, who computes the official IGS combined products. Since January 2, this coordination is carried out by the GRL/UWM (Geodynamics Research Laboratory of the University of Warmia and Mazury in Olsztyn, Poland). During period of more than 1 years of continuous IGS ionosphere operation, the techniques used by the IAACs and the strategies of combination have improved in such a way that the combined IGS GIMs are now significantly more accurate and robust. In this study, the final IGS combined GIMs produced by GRL/UWM were used to calculate the global maps of monthly medians of TEC values. These median TEC maps were generated for months of 29 that corresponded to the equinoxes and solstices: March, June, September and December. So, for each month there were calculated TEC median maps (2 h resolution) COSMIC data COSMIC RO measurements and products can be available from the Taiwan Analysis Center for COSMIC (TACC, and the COSMIC Data Analysis and Archive Center (CDAAC, As it was already mentioned, COSMIC can provide 1 2 RO measurements per day, and more than 7% of the RO measurements can be successfully retrieved into EDPs, which are one of the most important products for space weather and ionospheric science. Fig. 1 illustrates the global distribution of the COSMIC RO points during one day (e.g. March 2, 29). Derivation of the ionospheric electron density from radio occultation measurements is described in more detail by Tsai et al. (21). At CDAAC, the ionospheric profiles are retrieved by use of

3 I.V. Cherniak et al. / Advances in Space Research (2) Fig. 1. Example of global distribution of the COSMIC RO points during March 2, 29 ( GPS TEC _ _ _ 1_ March 29 COSMIC IEC _ _ _ _ PEC _ _ _ _ Fig. 2. Monthly median maps of electron content derived from ground-based GPS TEC observation network (left panels), COSMIC IEC (middle panels) and calculated PEC estimates (right panels) for March 29. All maps are presented in TECU.

4 3 I.V. Cherniak et al. / Advances in Space Research (2) GPS TEC _ _ _ _ June 29 COSMIC IEC _ _ _ _ PEC _ _ _ _ Fig. 3. Same as Fig. 2 but for June 29. the Abel inversion technique from TEC along LEO GPS rays. Detailed description of CDAAC data processing and the Ne profile retrieval method can be found in Kuo et al. (2) and Syndergaard et al. (26). In paper (Krankowski et al., 211) we presented results of our study examining and estimating the accuracy of the retrieved COSMIC EDPs by comparing them with ionosonde data obtained by European network DIAS (Digital Upper Atmosphere Server). All involved ionosonde data were manually scaled. Statistical analysis for different seasons of year 2 revealed that for European mid-latitude ionospheric stations the NmF2 differences between RO and ionosonde techniques were characterized by distribution with a mean of.7% and a standard deviation of.%. In the case of hmf2 comparison, a mean of 2. km and a standard deviation of 11. km were obtained. As COSMIC RO observations have a rather dense global distribution, they are potentially able to provide twodimensional (2-D) and three-dimensional (3-D) ionosphere images that allow better understanding of the ionospheric structure and dynamics. Several recent papers demonstrate success in construction of global 3-D ionospheric maps to study the ionospheric seasonal effects (Tsai et al., 29; Lin et al., 29). For this study, we used second level data provided by CDACC ionprf files containing information about ionospheric EDPs. COSMIC RO data for different seasons corresponded to equinoxes and solstices of 29 (March, June, September and December) were analyzed. It should be mentioned that a small part of the COSMIC EDPs are affected by cycle slips in the GPS phase data. In some cases, this results in obviously distorted profiles, whereas in other cases the errors due to cycle slips are more subtle. There is a need to analyze the data quality and to fix and remove bad or questionable data electron density profiles with large and irregular spikes and data gaps. Special processing routine analyzes the shape of each profile, rate of Ne change, its comparison with confidence limit; also this

5 I.V. Cherniak et al. / Advances in Space Research (2) GPS TEC _ _ _ _ September 29 COSMIC IEC _ _ _ 1_ PEC _ _ _ 1_ Fig.. Same as Fig. 2 but for September 29. routine realizes selection of data in dependence on the length of radio occultation traces. All unsuitable RO profiles were removed from our analysis. In order to compare COSMIC RO data with GPS TEC estimates the integration of data was done. For the present study the upper limit of the ionosphere has been taken to be at 7 km (altitude of COSMIC satellites). All selected COSMIC RO EDPs were integrated up to the height of 7 km, in that way estimates of IEC were retrieved. For each month corresponded IEC values were accumulated and then they were divided into data sets by intervals with 2 h duration. For the global representation of IEC estimates a spherical harmonics expansion up to degree and order 1 was carried out, as shown in Eq. (1) IECðu; kþ ¼ X1 X n n¼ m¼ P nm ðsin uþða nm cosðmkþþb nm sinðmkþþ ð1þ where u, k are geographic latitude and longitude, P nm are the normalized associated Legendre functions of degree n and order m; and a nm and b nm are the unknown SHE coefficients which were derived using COSMIC RO observations. Fortran-based software was developed for computation of monthly averaged two-hourly maps with 2. /. of latitude/longitude resolution. The resulted maps illustrated monthly median distribution of IEC, the final outputs are at the same of IONEX format. So, for each considered month there were calculated IEC median maps (2 h resolution). 3. Results 3.1. Variability of TEC, IEC and PEC estimates on a global scale The rapid growth in the number of GPS sites provides an important data source to study the ionosphere. Joint

6 32 I.V. Cherniak et al. / Advances in Space Research (2) GPS TEC _ _ _ _ December 29 COSMIC IEC _ _ _ _ PEC _ _ _ _ Fig.. Same as Fig. 2 but for December 29. Table 1 List with coordinates of selected specific points. Geographic coordinates Corrected geomagnetic coordinates Latitude Longitude Latitude Longitude Northern Hemisphere Boulder Millstone Hill Puerto Rico Juliusruh Moscow Athens Irkutsk Wakkanai Beijing Southern Hemisphere Hermanus Canberra analysis involved GPS TEC data and LEO data is able to provide additional information about upper atmosphere structure and variability. The main aim of this research was to illustrate the scope of COSMIC RO measurements application for the ionosphere/plasmasphere studies. We consider period of low solar activity, namely the

7 I.V. Cherniak et al. / Advances in Space Research (2) Fig. 6. Geographical position of the specific points. year of 29 that corresponded to the time of the extended solar minimum of solar cycles 23/2. Different seasons of 29 were analyzed that corresponded to the equinoxes and solstices: March, June, September and December. The following figures illustrate comparison between global distribution of GPS TEC, COSMIC IEC and PEC estimates for different seasons of year 29. It should be noted that we consider the quantitative differences PEC = TEC IEC as a measure of the contribution of the PEC to GPS TEC, the similar approach was used for the ionosonde measurements by Belehaki et al. (2). These values of PEC are correspond to the part of electron content from 7 km (upper bounder of COSMIC derived Ne profile) to the height of GPS satellite orbit, 2,2 km. The resulted maps represent the monthly median values of the electron content recalculated in the grid of 2.. degree of geographical coordinates with 2 h temporal resolution. Fig. 2 presents the set of graphs with spatial and temporal (for the chosen moments of time corresponding to, 6, and 1 UT) variations of TEC, IEC and PEC (monthly median estimates) for vernal equinox. Figs. 3 illustrate the similar results obtained for summer solstice, autumnal equinox and winter solstice. The global distribution of the obtained COSMIC IEC estimates is in a rather good agreement with GPS TEC behavior for different seasons. It confirms that IEC corresponded to the ionospheric heights up to 7 km has the significant contribution to the GPS TEC. When comparing GPS TEC with IEC the similarity between them clearly observed in seasonal changes winter/summer at the Northern/Southern Hemispheres and in spatial-temporal dynamics of equatorial ionization anomaly (EIA). It is necessary to note that for global distribution of calculated PEC estimates the largest values were observed over equatorial region at local daytime Variability of TEC and IEC estimates over specific points In order to analyze seasonal behavior of PEC contribution to GPS TEC estimates in different regions we selected several specific points with coordinates, corresponded to the approximate positions of different, mainly mid-latitude, ionospheric sounding stations. Table 1 lists the names of specific points, their geographical and geomagnetic coordinates. Geomagnetic coordinates were calculated by the use of the online service for IGRF/DGRF Model Parameters and Corrected Geomagnetic Coordinates ( Fig. 6 illustrates the distribution of the given points on a global scale. For the Northern Hemisphere we selected points at the regions of Northern America (Boulder, Millstone Hill and most southern at this region Puerto Rico), Europe (Juliusruh, Moscow and Athens), Asia (Irkutsk, Wakkanai and Beijing). For the Southern Hemisphere we selected the point in South Africa (Hermanus) and Australia (Canberra). For each specific points GPS TEC, COSMIC IEC and PEC estimates were taken from the nearest grid point of the corresponding global maps. In order to illustrate diurnal variation over selected region all results are presented in Local Time. Fig. 7 shows the comparison of the diurnal variations of monthly median values of GPS TEC and COSMIC IEC (in TECU) over specific points of the Northern America for the vernal and autumnal equinoxes and for winter and summer solstices of the year 29. These three stations form triangle with mid-latitude stations Boulder, Millstone Hill and the more southern station Puerto Rico. At these graphs GPS TEC is shown by line with asterisks, IEC values are indicated by line with black dots. One can see that the shape of the diurnal variations of GPS TEC and IEC is quite similar and it is mainly regulated by solar zenith

8 3 I.V. Cherniak et al. / Advances in Space Research (2) 27 Boulder ( N, 1.3W) Millstone Hill (2.6 N, 71. W) Puerto Rico (1. N, 67.1W) Fig. 7. Diurnal variations of the monthly median values of GPS TEC and COSMIC IEC at Boulder, Millstone Hill and Puerto Rico points of Northern America region for different seasons of 29. GPS TEC is shown by line with asterisks, IEC by line with black dots. angle. The GPS TEC estimates, as expected, are larger than IEC values for all seasons. Mid-latitude stations Boulder and Millstone Hill have practically equal geomagnetic latitude, shapes of diurnal variation (both TEC and IEC estimates), as far as absolute values, are rather similar at both stations. Station Puerto Rico, located southward, demonstrates relatively sharp day-time maximum with larger absolute values of TEC and IEC during all seasons, except winter season. Figs. and 9 illustrate the similar results obtained for triangles of stations at European and Asian regions consequently. One may point to the appearance of two peaks in diurnal variations of both TEC and IEC during summer solstice for mid-latitude points Juliusruh, Moscow, Athens, Irkutsk and Wakkanai. Appearance of two peaks in diurnal variation during summer solstice is a feature known well for many years from fof2 observations by ionosondes and incoherent scatter radars at northern mid-latitude region (e.g. Evans, 196; Essex, 1977). For more southern stations Athens and Beijing the absolute values of TEC and IEC are larger in comparison with northern stations considered. Fig. 1 shows results corresponding to the Southern Hemisphere Hermanus (South Africa) and Canberra (Australia). Both kinds of observations show rather similar diurnal behavior during all seasons. Shape of variations and absolute values are quite close at the vernal and autumnal equinoxes. The maximal values were observed at summer solstice (December) and minimal ones at winter solstice (June). It should be mentioned the fact that for all stations the minimum of TEC and IEC values were observed on the winter solstice December for the Northern Hemisphere and June for the Southern Hemisphere. It is a specific fea-

9 I.V. Cherniak et al. / Advances in Space Research (2) 27 3 Juliusruh (.6 N, 13.E) Moscow (. N, 37.3 E) Athens (3. N, 23. E) Fig.. Same as Fig. 7 but for Juliusruh, Moscow and Athens points of European region. ture of the ionosphere s behavior during solar activity minimum, when so-called winter anomaly effect, which appears in the greater values of NmF2 in winter than in summer by day, is practically absent (Torr and Torr, 1973). Both TEC and IEC variations increase steadily from night minimum near LT to the day-time maximum around LT during all seasons and begin to decrease after sunset. The maximum values of TEC and IEC were observed at the Southern Hemisphere at summer solstice. At the Northern Hemisphere this seasonal maximum was less pronounced (except for Puerto Rico). These results can be explained by asymmetry of the Earth s magnetosphere/ionosphere Behavior of PEC estimates over specific points In order to estimate variability of PEC and its contribution to GPS TEC over different regions we analyze the quantitative differences PEC = TEC IEC and ratios PEC/TEC (%). Fig. 11 shows diurnal variations of the percentage contribution of PEC into GPS TEC over all considered points for different seasons of the year 29. The revealed results show that percentage PEC contribution is higher during night-time hours than at day-time. It is also important to note that the PEC over the Northern Hemisphere midlatitudes is about 3 % at day-time and 7% at night. For the most southern specific points of the Northern Hemisphere, Puerto Rico and Athens, the range between night-time and day-time PEC contribution was larger compared with other mid-latitudinal locations. The diurnal variations of PEC/TEC ratio are rather similar for all selected points. In general, the PEC contributions to GPS TEC were similar for daytime conditions over the Northern Hemisphere. At winter the PEC estimates over the Northern Hemisphere points are mostly larger (on

10 36 I.V. Cherniak et al. / Advances in Space Research (2) 27 Irkutsk (2. N, 1.3E) Wakkanai (. N, 11.7E) Beijing (39.9 N, 1. E) Fig. 9. Same as Fig. 7 but for Irkutsk, Wakkanai and Beijing points of Asian region. 1 2%) after local sunset and at night-time than noon values. The most considerable seasonal changes of PEC were observed in the equinoctial seasons. For mid-latitude locations of the Southern Hemisphere the PEC contribution consisted of about 3 % of GPS TEC estimates at day-time and 6 7% at night. The diurnal changes of PEC percentage over these locations for March, June and September months were rather similar. The variations for December (summer) revealed some differences for afternoon hours at Canberra in comparison with Hermanus. Fig. illustrates the behavior of PEC estimates in absolute values (TECU) for all considered regions. The diurnal variations of PEC are characterized by larger values during night-time (from 1 LT to LT) in comparison with the day-time values (from 9 LT to 17 LT) for low solar activity condition. In absolute values PEC over mid-latitudes of the Northern Hemisphere was about 3 6 TECU during day-time and 2. TECU during nighttime. Taken together our results indicate that PEC change is insignificant with the time of the day, i.e. they are practically independent on the solar zenith angle. It can be explained by predominating dynamical processes in the system ionosphere/plasmasphere. Table 2 illustrates main relative and quantitative characteristics of PEC variability over all specific points for different seasons of year 29. Each line contains information about average PEC values for night-time and day-time conditions expressed in TECU, as well as percentage contribution of PEC to GPS TEC also for night and daytime.. Discussion and conclusion This paper presents results of a comparison of electron content values derived from GPS and COSMIC RO

11 I.V. Cherniak et al. / Advances in Space Research (2) Hermanus (3. S, 19.2 E) March 2 June September December 2 2 Canberra (3.3 S, 19 E) March 2 June September December Fig. 1. Same as Fig. 7 but for Hermanus and Canberra points of the Southern Hemisphere. measurements. The constructed monthly median maps of COSMIC IEC data were analyzed in order to estimate the seasonal variability of IEC over different regions and calculate difference between GPS TEC and COSMIC IEC as a measure of plasmasphere contribution. Evidently, all COSMIC-retrieved IEC values were smaller than GPS TEC values derived from transionospheric measurements for distance 2,2 km. Both kinds of observations show rather similar diurnal behavior during all seasons. It was shown that during daytime both GPS TEC and COSMIC IEC values were generally lower at winter than in summer solstice practically over all individual locations. The revealed results show that the PEC contribution is higher during night-time hours than at day-time hours. It is also important to note that the PEC estimates over the Northern Hemisphere mid-latitudes are about 2 % at day-time and 7% at night. The diurnal variations of PEC/TEC ratio are rather similar for selected locations. In general, the PEC contribution to GPS TEC is rather similar for day-time conditions over the Northern Hemisphere. In winter the PEC estimates over the Northern Hemisphere are mostly larger (on 1 2%) after local sunset and night-time than noon values. The most considerable seasonal changes of PEC were observed in the equinoctial seasons. A number of studies considered the question of PEC contribution to GPS TEC, but they were mostly results of comparison with model data or with observations from limited geographical locations (mainly, over a single ground-based ionosonde). Lunt et al. (1999) used the SUPIM (Sheffield University Plasmasphere Ionosphere Model) in order to estimate the contribution of the electron content arising from the protonosphere along ray paths from GPS satellites. Results have been presented simulating effects for stations at midlatitudes in European and American sectors for solar minimum and solar maximum conditions. It is shown that for European region at solar minimum the PEC contribution is generally only a few TEC units in absolute terms. However, at night, particularly in winter, it may constitute % or more of the GPS TEC value. At solar maximum, the PEC is typically doubled in absolute magnitude but are greatly reduced in percentage terms. It was shown that during the night in winter the protonospheric contribution can exceed % at the lower latitudes and was more than 3% at the highest latitude considered at the paper. The daytime percentages were generally in range 1 3% in winter, with a slightly smaller upper limit for the lowest latitudes at other seasons. Balan et al. (22) compared GPS TEC data obtained from the GPS network (GEONET) in Japan with those calculated using the SUPIM model. It was revealed that PEC changed appreciably with season and latitude and very little with the time of the day. The percentage contribution of PEC to GPS TEC changes most significantly with the time of the day. For period of high solar activity this

12 3 I.V. Cherniak et al. / Advances in Space Research (2) 27 (a) (b) (c) (d) Boulder Juliusruh Irk utsk Hermanus M illsto ne Hill Canberra 1 Moscow Wakkanai Puerto Rico Ath ens Be ijing March June Sep tember December Fig. 11. Diurnal variations of the percentage PEC contribution into GPS TEC for different seasons of the year 29. contribution over Japan region varied from a minimum of about % during day-time at equinox to a maximum of about 6% at night in winter. In this research we analyzed PEC estimates over several locations in different regions. Results of our comparative study also revealed that for mid-latitudinal stations PEC estimates varied weakly with the time of a day and reached the value of several TECU for the condition of solar minimum. Percentage contribution of PEC to GPS TEC indicated a clear dependence on time. So the revealed estimates of PEC contribution to GPS TEC are in general agreement with previous investigations based on the GPS measurements and model simulation results. There are several studies estimated PEC contribution on the base of ionosopheric sounding data. Belehaki et al. (2) estimated PEC over Athens as a difference between GPS TEC and IEC retrieved from Athens ionosonde. These estimations of PEC corresponded to the heights from 1 km (upper limit of ionosonde profile) to 2,2 km. It was reported that for high solar activity (year 21) PEC contribution exceeded % during the night in winter with a smaller contribution (2 3%) during other seasons. The day-time percentages were generally small in winter and fall, and exceeded 1% during spring and summer months. Mosert et al. (27) reported about PEC estimates over Ebro ionosonde, which was done with the same technique as in Belehaki et al. (2). It was shown that for high solar activity (years 2 and 21) the values of PEC reached 6 TECU in winter and 6 1 TECU in summer. The

13 I.V. Cherniak et al. / Advances in Space Research (2) (a) Boulder 1 Millstone Hill 1 Puerto Rico 1 P 1 P 1 P 1 (b) 2 2 Juliusruh Moscow 1 Athens P 1 P 1 P 1 (c) Irkutsk Wakkanai 1 Beijing P 1 P 1 P 1 (d) Hermanus P Canberra P March June September December Fig.. Diurnal variations of PEC in TECU for different seasons of the year 29. largest values were observed in April 2 (1 19 TECU). During the year 21 the percentage PEC contribution had greater values in winter ( 61%) than in summer (1 36%) with intermediate values in the other two seasons (3 %). In the above-mentioned studies PEC estimates were found out for European mid-latitudes and high solar activity. In the present paper we considered period of low solar activity, so the retrieved estimates of PEC for different mid-latitudinal regions, including the European one, were, as expected, only several TECU, i.e. several times less than reported estimates for high solar activity time. The revealed percentage PEC contribution are in a rather good agreement with reported ones by Belehaki et al. (2) with maximal values (more than 6%) during night-time and lesser values (2 %) during day-time. The results obtained in this study highlight the potential of application of COSMIC RO data to be used for estimation the effect of plasmasphere on GPS TEC observations. This paper describes a phenomenon observed without detailed discussing the possible physical mechanism responsible for it. We hope to be able to explain it after having analyzed the above mentioned cases involving modeling simulation results and flux-tube geometry of the magnetosphere/ionosphere system.

14 I.V. Cherniak et al. / Advances in Space Research (2) 27 Table 2 Contribution of PEC over individual locations for different seasons of the year 29. Period March, 29 June, 29 September, 29 December, 29 Point PEC, TECU PEC/TEC, % PEC, TECU PEC/TEC, % PEC, TECU PEC/TEC, % PEC, TECU PEC/TEC, % Day Night Day Night Day Night Day Night Day Night Day Night Day Night Day Night Boulder M. Hill P. Rico Juliusruh Moscow Athens Irkutsk Wakkanai Beijing Hermanus Canberra Acknowledgments We acknowledge the Taiwan s National Space Organization (NSPO) and the University Corporation for Atmospheric Research (UCAR) for providing the COSMIC Data. We are grateful to International GNSS Service (IGS) for GPS data and products. The authors thank two anonymous reviewers for their corrections, comments and suggestions that have significantly enhanced the quality of the paper. References Balan, N., Otsuka, Y., Tsugawa, T., Miyazaki, S., Ogawa, T., Shiokawa, K. Plasmaspheric electron content in the GPS ray paths over Japan under magnetically quiet conditions at high solar activity. Earth Planets Space, 71 79, 22. Belehaki, A., Jakowski, N., Reinisch, B. Plasmaspheric electron content derived from GPS TEC and ionosonde ionograms. Adv. Space Res. 33, 33 37, 2. Bilitza, D. International Reference Ionosphere 2. Radio Sci. 36 (2), , Essex, E.A. High to low latitude variations in the evening summer total electron content and F-region electron density. J. Atmos. Terr. Phys. 39, 11 11, Evans, J.V. Cause of the mid-latitude evening increase in fof2. J. Geophys. Res. 7 (), , 196. Gallagher, D.L., Craven, P.D., Comfort, R.H. Global core plasma model. J. Geophys. Res. 1 (A), 1,19 1,33, 2. Hernandez-Pajares, M., Juan, J.M., Orus, R., Garcia-Rigo, A., Feltens, J., Komjathy, A., Schaer, S.C., Krankowski, A. The IGS VTEC maps: a reliable source of ionospheric information since 199. J. Geod. 3, , 29. Huang, X., Reinisch, X. Vertical electron content from ionograms in real time. Radio Sci. 36 (2), 33 32, 21. Jakowski, N., Tsybulya, K., Stankov, S.M., Wehrenpfennig, A. About the potential of GPS radio occultation measurements for exploring the ionosphere, in: Reigber, C., Lühr, H., Schwintzer, P., Wickert, J. (Eds.), Earth Observation with CHAMP, Results from Three Years in Orbit. Springer-Verlag, Berlin, pp. 1 6, 2. Kirchengast, G., Foelsche, U., Steiner, A. (Eds.), Occultations for probing atmosphere and climate, ISBN: , p., 2. Krankowski, A., Zakharenkova, I.E., Krypiak-Gregorczyk, A., Shagimuratov, I.I., Wielgosz, P. Ionospheric electron density observed by FORMOSAT-3/COSMIC over the European region and validated by ionosonde data. J. Geod. (), 99 96, s z, 211. Kuo, Y.-H., Wee, T.-K., Sokolovskij, S., Rocken, C., Schreiner, W., Hunt, D., Anthes, R.A. Inversion and error estimation of GPS radio occultation data. J. Meteorol. Soc. Jpn. 2 (1B), 7 31, 2. Lin, C.H., Liu, J.Y., Hsiao, C.C., Liu, C.H., Cheng, C.Z., Chang, P.Y., Tsai, H.F., Fang, T.W., Chen, C.H., Hsu, M.L. Global ionospheric structure imaged by FORMOSAT-3/COSMIC: Early results. Terr. Atmos. Ocean. Sci. 2, , TAO (F3C), 29. Liou, Y.A., Pavelyev, A.G., Matyugov, S.S., Yakovlev, O.I., Wickert, J. Radio occultation method for remote sensing of the atmosphere and ionosphere, In-Tech Press, ISBN , p. 176, 21. Lunt, N., Kersley, L., Bailey, G.J. The influence of the protonosphere on GPS observations: model simulations. Radio Sci. 3 (3), , Mosert, M., Gende, M., Brunini, C., Ezquer, R., Altadill, D. Comparisons of IRI TEC predictions with GPS and digisonde measurements at Ebro. Adv. Space Res. 39, 1 7, 27. Syndergaard, S., Schreiner, W.S., Rocken, C., Hunt, D.C., Dymond, K.F. Preparing for COSMIC: Inversion and analysis of ionospheric data products, in: Foelsche, U., Kirchengast, G., Steiner, A.K. (Eds.), Atmosphere and Climate: Studies by Occultation Methods. Springer, New York, pp. 137, 26. Torr, D.G., Torr, M.R. The seasonal behavior of the F2 layer of the ionosphere. J. Atmos. Terr. Phys. 3, , Tsai, L.C., Tsai, W.H., Schreiner, W.S., Berkey, F.T., Liu, J.Y. Comparisons of GPS/MET retrieved ionospheric electron density and ground based ionosonde data. Earth Planets Space 3, 193 2, 21. Tsai, L.C., Liu, C.H., Hsiao, T.Y. Profiling of ionospheric electron density based on FormoSat-3/COSMIC data: Results from the intense observation period. Terr. Atmos. Ocean. Sci. 2, , dx.doi.org/1.3319/tao (f3c), 29.