Scientific Committee on Antarctic Research

• Home
• Search:
• Members only

• About
• Science
  & Data
• News
• Events
• Antarctic
  Treaty
• Antarctic
  Information
• International
  Polar Year
• Publications
• Communication
  and Education
• Contacts

You are in:  Home » Publications » Reports » Report 23

SCAR Report 23

Overview of the research on the atmospheric impact on GPS observation in polar regions

Jan Cisak
Institute of Geodesy and Cartography
Warsaw, Poland.
jcisak@igik.edu.pl

back to AGS '03 Program

Download this document here

Abstract

Polar region is one of the best test areas on the Earth for research. The increased, in last years, number of permanent GPS stations there, provides a large amount of data and possibility to create a representative database.
Recent research developments within the framework of the project „Atmospheric impact on GPS observations in Antarctica" are presented in the paper. The effects of both ionospheric and tropospheric disturbances on GPS solutions are discussed. The GPS data, usually those provided by permanent GPS station arrays are commonly used to investigate the structure and dynamics of ionosphere. First results of the project concerned the influence of ionosphere over the Arctic and Antarctic regions on repeatability of co-ordinates of vectors of different length during the quiet and disturbed ionosphere (ionospheric storms).
The new approach of data analysis was conducted. It is based on the analysis of GPS solutions obtained from the overlapped segments of data. Time series of GPS solutions based on the processing overlapped data segments allow for investigation of atmospheric impact on GPS measurements in a new dimension. Such a series can be considered as a record of the process of variations of vector components during varying atmospheric disturbances. The experiments performed concerned the investigation of the response of the measuring system to ionospheric storms as well as the response of the measuring system to tropospheric disturbances.
The two-stage influence of ionosphere, as the fluctuation on local inhomogeneous ionosphere and the refraction on regional inhomogeneous ionosphere, makes ambiguity resolution difficult and causes the errors in vector final solutions. The deep study of the problem should lead to the rejection of bad solutions. It has also been found that the analysis of correlations gives the possibility to correct the horizontal components of the vector obtained using commercial as well as the Bernese software. Similar analysis was performed when looking for the correlation of vector solution and tropospheric data. The results obtained using the Bernese software are not correlated with the tropospheric delay. The repeatability of the vertical component of the vector obtained using commercial programmes is substantially improved after introducing the correction.

1. Introduction

A considerable progress observed in the geodynamic research is the result of a development in measuring techniques (Manning, 2001). The qualitative results on crustal movements presented in some publications (Dietrich et al., 2001, Dietrich et al., 2002) seem, however, to be at the level of their accuracy determination. A realistic estimation of the potential of the experiment is thus necessary to avoid false conclusions describing non-existent occurrences (artefacts), especially when the experiment is difficult or very expensive.
That is exactly the case of the experiments conducted in Antarctica. Seasonal changes of atmospheric conditions frequently make impossible to perform widespread continuous (yearly) GPS observations. On the other hand, those seasonal changes can cause periodic biases in data acquired, as well as periodic deformations of the Earth's crust. Besides data acquired at the growing number of Antarctic IGS permanent stations (recently about 15 stations) there is a large set of data provided by GPS Epoch Crustal Movement Campaigns (http://www.geoscience.scar.org/geodesy/giant.htm) organized for a number of years under the umbrella of SCAR GIANT (Geodetic Infrastructure of ANTarctica) program. About 50 Antarctic stations participated in those campaigns that took place during Antarctic summer only. The question arises whether the seasonality of those campaigns influences the results obtained, and if so, how that influence could be quantitatively evaluated.
The main goal of the GIANT program project on the atmospheric impact on GPS observations in Antarctica is to investigate the atmospheric impact on the quality of GPS observations in Antarctica, and possibly to develop recommendations for future Antarctic GPS campaigns, data post-processing strategies and modelling GPS solutions.
The GPS data, usually those provided by permanent GPS station arrays, are commonly used to investigate the structure and dynamics of the ionosphere (Baran et al., 2001a; Feltens and Jakowsky, 2001) as well as to investigate the troposphere (Kruczyk, 2002). During the realisation of the project it was impossible to concentrate on the atmospheric impact only and separate it from the study of the atmosphere. The results of the investigations of the ionosphere as well as of the troposphere are the output of the project too. The bibliography at the end of the paper includes the majority of publications that summarize the results obtained in the framework of the GIANT project.
2. Project background
The GIANT program project on the atmospheric impact on GPS observations in Antarctica, coordinated by Poland, has been created at the XXVI SCAR meeting in Tokyo in 2000. The Polish project, financed by Polish Scientific Committee, named “The investigation of atmospheric impact on the results of the precise geodetic measurements with GPS technique in polar conditions” was established in the Institute of Geodesy and Cartography (grant No 8T12E 045 20) in March 2001. J Cisak – the project leader, reported the first results of the international project to the projects coordinators meeting in Siena in July 2001. One week later the Third Antarctic Geodesy Symposium AGS’01 took place in St. Petersburg. One session of the Symposium was devoted to the problem of the atmospheric impact on GPS technique of measurements. The proceedings of the Symposium were published in SCAR Report, No 21, January 2002, publication of the Scientific Committee on Antarctic Research, Scott Polar Research Institute, Cambridge, UK. As the result of the Symposium, an efficient cooperation between SCAR WG on Geodesy and Geographic Information and IGS IONO WG (Feltens and Jakowsky, 2001) has been established. J. Cisak was invited to take part in the workshop of the IONO WG of IGS that was held in Darmstadt, in February 2002. Some results of the project as well as the SCAR WG GGI activity were presented and published in the special issue of the workshop. The new achievements in the project were presented at the International Workshop on "Atmospheric impact on GPS observations focused on polar regions", 15 May 2002, Warsaw, Poland. The papers and presentations of the workshop are placed on the web page of the SCAR Geoscience Standing Group: http://www.geoscience.scar.org/geodesy/warsaw/index.htm. The next interim report of the state of the project was presented to the international geodetic community at the WG GGI meeting during the XXVII SCAR, Shanghai, China, July 2002. The papers with the attempt to correct the final GPS solutions for the Antarctic vectors were presented at the AGS’02, Wellington, New Zeeland, December 2002 (Cisak et al., 2002)
http://www.geoscience.scar.org/geodesy/ags02/index.htm and at the Poland – Italy geodetic meeting, Bressanone, Italy, April 2003 (Cisak et al., 2003c). The project is still running. Everybody is very welcome to contribute to it. The final report is to be prepared and presented at the next SCAR Symposium in Bremen, 2004.
3. Ionosphere
3.1. Influence of non-homogeneity of ionosphere in polar region on the results of GPS measurements
The non-homogeneity of the atmosphere in the polar regions is an important factor when considering the influence of ionosphere for the determination of co-ordinates by use of GPS technique. It has various forms of different range.
The most spectacular form of the non-homogeneity of the atmosphere is the main ionospheric trough, which is the large-scale structure of lowering electron concentration. The concentration of the electrons in the area of ionospheric trough can be even ten times smaller than outside that area. The width of such zone, along the meridian, can reach 2-3 degrees of arc. The latitudinal gradients of the electron concentration in such area can significantly differ from those in quiet ionosphere and can lead to erroneous determination of phase ambiguities (Baran et al., 2001b; Baran et al., 2001c).
The other forms of the non-homogeneity of the ionosphere in polar regions are the bubbles with dimension of a few hundreds or even a thousand kilometres. The electron concentration inside the bubbles can exceed the concentration outside it by a factor 10 to 100. The edges of that type of non-homogeneity are characterized by large gradients.
The non-homogeneities of the atmosphere in the range of tens of kilometres can substantially affect GPS measurements and the accuracy of GPS solutions. The intensity of those non-homogeneities during the magnetic storms can grow up to several times, and cause substantial fluctuation of GPS signal phases (Epishov et al., 2002). Such fluctuations can be observed even in the latitudes below 60o.
During the occurrence of large electron concentration gradients the ionospheric refraction can strongly affect the determination of ambiguity and result in growing errors of GPS solutions. The correlation between the growth of unsolved ambiguities and variations in TEC values due to ionospheric storms is clearly visible when calculating vectors from GPS data in polar regions; it leads to erroneous determination of vector components (Cisak et al., 2002c; Cisak et al., 2003a). The GPS signal passing through the area of electron concentration changes demonstrates phase fluctuations that can result in loss of lock to satellites what further affects continuity in phase recording. Fig. 3.1.1. The number of satellites with a loss of lock on L2 against the number of satellites observed
Fig. 3.1.1 (Stewart and Langley, 1999) presents the comparison of the number of satellites for which a loss of lock on L2 occured with the number of satellites visible in Fairbanks (Alaska) during high activity of ionosphere in 27 August 1998 (upper graph) and for quiet ionosphere in 13 December 1998 (lower graph).
The influence of the electron concentration changes, resulting as the refraction of signal path and its differential lengthening with respect to different frequencies (Fig. 3.1.2) causes the second order refraction errors. Thus, the use of L3 combination does not completely remove the ionospheric effect from GPS solutions. For modelling and estimation of those effects the model errors of one layer model can be used (Zanimonskaya and Prokopov, 2001).

Fig. 3.1.2. Refraction of the signal path during the ionospheric trough. The dot-lines correspond to lower frequency signal (L2) and continuous lines - higher frequency (L1). The contour lines correspond to the electron density in units104 el/m3
Incorrectness of the ionosphere modelling using single layer approximation results in non-linearity in both TEC and pseudo-range determination (Brunner and Gu, 1991; Zanimonskaya and Prokopov, 2001). That non-linearity affects also other parameters, e.g. TZD (Krynski et al., 2002b) and vector components.
Similarly to optical systems the effect of non-homogeneity of the ionosphere is proportional to the measure of non-homogeneity itself.
Fig. 3.1.3. 3D distribution of electron density during the ionospheric storm in 13 September 1999
Fig. 3.1.4 shows the errors of the pseudo-range estimation from L3 combination, during the transition of the signal through the inhomogeneous structures of ionosphere of different size. In the compartments of inhomogeneous ionosphere of the shape of lens the largest divergence of electron contents from not violated ionosphere was considered equal to 5·105·m-3. The dimension of inhomogeneous structures of the ionosphere in the cases shown in Fig. 3.1.2 and Fig. 3.1.3 is about 1000 km and electron density change is 2·105 el/m3. Taking into account data form Fig. 3.1.4, those ionospheric disturbances affect pseudo-range by about 3 cm. It results in systematic errors in GPS solutions, mainly through the erroneous ambiguity determination.

Fig. 3.1.4. Dependence of the second order errors in pseudo-ranges on the dimension of inhomogeneous elements of ionosphere with fixed maximum of electron density changes

3.2 Dependence of variations in vector components from the electron concentration in ionosphere for the Antarctic GPS stations.
Seasonal changes of atmospheric conditions frequently disturb the continuity in tracking GPS satellites (during all seasons). Those changes can cause periodic errors in GPS solutions and suggest the periodic movements of the Earth’s crust, what naturally is an artefact. Besides an increasing number of permanent Antarctic stations, for several years the periodic GPS Epoch Crustal Movement Campaigns (Detrich, and Rülke, 2002) have been conducted. More than 50 stations take part in those campaigns. They are organized during the Antarctic summer only. It is questionable whether the results of one-season campaigns reflect the real changes in the stations co-ordinates and indicate the actual crustal movements. The first results of investigation of the influence of ionosphere on GPS solutions, obtained in the framework of this project were presented at the Antarctic Geodesy Symposium AGS’01 in St. Petersburg in July 2001 (Krankowski et al., 2001a). Data form the second part of February 1999 acquired at several stations of Northern Hemisphere and several Antarctic stations were used in the analysis (Table 3.2.1.).
Table 3.2.1. Vectors and their lengths examined in the analysis
All vectors were determined using the Bernese v.4.2. software with use of QIF strategy (Quasi Ionosphere Free) from 24h, 12h and 6h sessions. Dispersion in vector components (from 5 solutions) reaches the level of 20 cm when sub-daily sessions were processed, while for 24h sessions the dispersion do not exceed 7 cm. The maximum differences were obtained for observations collected from 12:00 – 18:00 UT, when TEC shows the large dynamics of changes. Similar large differences occured for long vectors (over 650 km) as well as for short ones, e.g. Arctowski – O’Higgins (132 km). For the stations in Northern Hemisphere (vector Onsala – Metsahovi) the differences were significantly smaller. This experiments and results obtained were encouraging to look for the confirmations of the results with use of the larger statistic material and with using new research methods.
GPS data from several permanent IGS Antarctic stations (Fig.3.2.1) were analysed, separately those from summer and from winter seasons. The Bernese v.4.2 software with QIF ambiguity resolution strategy was used to process the data in daily sessions with 23h overlap. (Cisak et al., 2003).

Fig. 3.2.1. The map of Antarctic permanent GPS stations and analysed vectors
The Ionosphere Working Group of the International GPS Service publishes the global maps and arrays of TEC values, given in function of latitude and longitude with 2h temporal resolution. Annual variations of TEC values for 2001 obtained from IONEX data by calculating daily averages over Antarctic Davis (DAV1) and European Borowa Gora (BOGO) stations as well as the Ap index representing in linear scale a measure of geomagnetic activity are presented in Fig. 3.2.2.Fig. 3.2.2. Annual variations of diurnal mean TEC over DAV1 and BOGO stations obtained from IONEX data and the Ap index
Seasonal variations of the ionosphere are clearly visible in the graph; it is especially distinct over the Antarctic station Davis.
Data of particular interest correspond to time intervals that have been marked on the graph (Fig. 3.2.2). The first data set corresponds to the vicinity of 64DOY2001 when the occurrence of ionospheric storm was detected in the analysis of time series of GPS solutions of vectors between EPN stations (Zanimonskiy et al., 2002). The third data set covers the period of extremely active ionosphere that took place in October and November 2001. The impact of the atmosphere on GPS solutions of vectors was carefully analysed for that period (Cisak, et al., 2002). The lowest electron concentration and the lowest geomagnetic activity in 2001 occurred in July. Thus the July 2001 data (second data set marked in Fig.3.2.2) was included to the analysis. Averaged solutions for lengths and vertical components of vectors between selected Antarctic permanent GPS stations for the chosen data sets are given in Table 3.2.2.
Table 3.2.2.
Differences between the average lengths and vertical components of the vectors calculated during the period of the unstable and quiet ionosphere (July 2001) are given in Table 3.2.3. Variations of the solutions shown in Table 3.2.2. for different seasons are quite substantial. In most cases they exceed their accuracy estimated by using a common error propagation procedure. The differences obtained (Table 3.2.3.) were interpreted by means of statistical analysis of correlations of GPS solutions for vector components, and by means of parameters from the processing with the Bernese software, i.e. _2, number of ambiguities resolved, number of single differences used in the solution, internal accuracy parameters, etc. with ionospheric data from IGS IONO-WG (IONEX) and also with the data received directly from IONO-WG, kindly provided by Dr. Manuel Hernandez-Pajares.
Table 3.2.3.
The problem concerns the errors, resulting from non-linearity of calculating algorithms and from second order effects of signal propagation in the ionosphere. The non-linearity of the algorithms used for the data processing of satellite observations was widely discussed in the literature (e.g. Tiberius, 1998). The weak non-linearity causes the known effect of detection, i.e. the conversion of variations of process parameters or random input signals into biases in output results. For determination of metrological properties of measurement system the qualitative and quantitative assessment of conversion of random error into systematic error is needed. The growth of random errors of GPS observations during the ionosphere’s storm (Baran et al., 2002) can be used as a signal for testing the hypothesis of the detection.
The second order effects of the ionosphere can also be considered as the source of non-linearity in the process of solving ambiguities. The existence of other sources of non-linearity cannot be excluded but the complex technology, instrumentation, software, mathematical and physical models of different sources of disturbances of GPS signal make the description of the phenomena quite difficult.
Polar regions, in particular the Antarctic - a continent being the extensive international research laboratory, are suitable test areas for investigating ionosphere’s effects on GPS solutions. During the Antarctic winter the diurnal changes of electron concentration are insignificant. It is due to a low and almost not varying altitude of the Sun over the region. The changes in electron concentration over Antarctic are caused mainly by geomagnetic activity. The Antarctic winter in 2001 was exceptionally quiet in the sense of geomagnetic activity as compared with other years. The results of GPS positioning from that winter can thus be considered as reference in studying the ionosphere’s impact on GPS solutions obtained in other years as well as in different seasons. Overlapping sessions of 24-hour were processed to smooth random errors in GPS solutions and to eliminate short-term biases. The TEC data from IONEX files was respectively averaged over 24h with 1h temporal resolution (Fig. 3.2.2).Fig. 3.2.3. Variations of vector length and uncertainty of ambiguity estimation versus TEC (a), and time series of uncertainties of ambiguity estimation for the periods investigated (b)
Fig. 3.2.3a shows the relationship between DAV1-CAS1 vector length (dD) and diurnal average of TEC (upper graph) as well as the relationship between the uncertainty of ambiguity estimation in the vector calculation and diurnal average of TEC (lower graph). Diurnal average of TEC presents stronger correlation with the uncertainty of ambiguity resolution (correlation coefficient of 0.64) than with vector length (correlation coefficient of 0.51).
Due to a regional scale of dynamics of the ionosphere in Antarctica, the ionospheric disturbances affect similarly GPS data acquired at the investigated stations. Thus the GPS vector solutions obtained with the Bernese software are practically free of TEC differences between the stations. It should also be noted that the time series of uncertainties of ambiguity resolution do not substantially differ for different vectors and do not depend on their length (Fig.3.2.3b). The same conclusions are drawn from the analysis of GPS solutions for EPN vectors.
The obtained results indicate the dependence of ambiguity resolution on the state of ionosphere. Crucial role in both performed quantitative and qualitative analysis played the use of time series of GPS solutions based on overlapped sessions (Krynski and Zanimonskiy, 2002). Correlations shown in Fig. 3.2.3 indicate a possibility of modelling the ionospheric effects on GPS solutions. For example, the solutions for vector length could get corrected by using the regression model of dD = k(TECdmv) based on TEC data. For DAV1-CAS1 vector of 1398 km, the correction equals to +3 mm/10TECU. Generally, the unstable ionosphere causes shortening of vector length obtained from GPS solution. The vector lengths corrected with the model are shown italic in Table 3.2.3. Introducing the corrections resulted in the decrease of seasonal dispersion in both October and November data and made them more similar to the July data when the ionosphere was quiet. Although in most cases the applied corrections improve the obtained results, there are exemptions when the procedure does not seem suitable (Table 3.2.3). They might happen due to relatively small amount of data processed as well as larger and more irregular disturbances of the ionosphere.
4. Troposphere
4.1 Dependence of variations in vector components from the Total Zenith Delay for the Antarctic GPS stations.
Time series of vector components obtained with the Bernese v.4.2 software for daily sessions of GPS observations from a number of permanent stations acquired in July, October and November 2001, were used to investigate the impact of varying meteorological conditions on GPS solutions. Analysed data correspond to the periods substantially distinguished in terms of dynamics of the atmosphere. Variations in monthly average of temperature, atmospheric pressure as well as vertical component of the MAW1-DAV1 vector are given in Fig. 4.1.1. Fig..4.1.1. Annual variations of monthly averages of temperature, atmospheric pressure
as well as vertical component of the MAW1-DAV1 vector
Variations in vertical components of the vectors defined by pairs of investigated GPS permanent stations in Antarctica are correlated with seasonal variations of atmospheric pressure. Similar conclusion was already drawn from the analysis of GPS solutions and meteorological data in mid-latitudes (Haefele and Kaniuth., 2001).
Tropospheric impact on GPS measurements is described in terms of tropospheric delay. To increase reliability of results obtained, tropospheric delay data from two independent sources was considered in the analysis. First, the Tropospheric Zenith Delay (TZD), available on IGS web pages, in the form of time series with 1h temporal resolution was considered. Second, the TZD data derived from radio-sounding over the majority of permanent GPS stations in Antarctica, also in the form of time series but with 12h temporal resolution. The results obtained with use of both data sources were close to each other at the acceptable level.
Time series of atmospheric pressure and TZD at the Antarctic GPS stations (Fig. 4.1.2) as well as correlation of TZD and atmospheric pressure variations with vertical components of respective vectors (Fig. 4.1.3) were analysed.
Fig. 4.1.2. Variations of atmospheric pressure and TZD for Mawson (a) and Davis (b stations

Fig. 4.1.3. Variations of differential TZD versus variations of atmospheric pressure
and GPS-derived vertical component of DAV1 – CAS1 vector
No significant effect of variations of troposphere on GPS solutions observed (Fig. 4.1.3) proves sufficient modelling of troposphere in the Bernese v.4.2 software when computing vectors of a few hundred kilometres length and longer; GPS solutions obtained are practically free of the tropospheric effect.
In case of commercial programmes used to process GPS data, the experiments conducted with EPN data indicate correlation of GPS-derived vector components with TZD. Correlation coefficients derived could efficiently be used to correct GPS solutions obtained with commercial software. It applies not only to GPS solutions for mid-latitude stations but also for those in polar regions.
5. Conclusions
The non-modelled delays of GPS signal when passing the atmosphere affect GPS solutions for station positions and vector components and result in variations in time series of such solutions. To improve GPS solutions with no better models of atmosphere, corrections to the computed vector components, calculated using correlation analysis, could be applied. Data from Antarctic GPS stations are especially suitable for modelling such correlation functions and determining their parameters due to a distinct seasonal variability of ionosphere in polar regions.
The results discussed in the paper focus on the analysis of the impact of ionospheric disturbances on variations of vector lengths obtained at high latitudes from GPS data. Variations of GPS solutions for lengths of vectors are commonly explained in terms of non-modelled variations of the ionosphere. Besides their direct effect on GPS solutions, they affect them indirectly by violating the mechanism of integer ambiguity resolution. Correlation analysis conducted using data sets from chosen Antarctic stations shows a possibility of using simple empirical models for partial eliminating the non-modelled in GPS processing software effects of ionosphere. Modelling ionospheric effects on the results of GPS data processing requires further research with use of larger data samples. GPS solutions corrected with such empirical models seem more suitable for geodynamics research.
The results of the research on the tropospheric impact on GPS solutions show seasonal dependence of height differences between Antarctic stations from changes of atmospheric pressure. Modelling the satellite signal passing through the troposphere in the Bernese v. 4.2 software seams satisfactory. No correlation between vector components obtained using the Bernese software and Total Zenith Delay was found. The analysis of time series of GPS solutions based on EPN data, obtained using commercial software shows the possibility of using empirical models to partially eliminate from GPS solutions the non-modelled in processing GPS data effects of troposphere – similarly to the ionospheric one.
Acknowledgements
The paper summarizes the results of the research conducted at the Institute of Geodesy and Cartography (IGiK), Warsaw, and at the University of Warmia and Mazury (UWM), Olsztyn. The research was partially supported by the Polish State Committee for Scientific Research (Research Project No 8 T12 E045 20). The author expresses his gratitude to all contributors to the project – particularly to Prof. W. Baran, Dr. A. Krankowski, Dr. P. Wielgosz from UWM, Dr. I. Shagimuratov from IZMIRAN, Kaliningrad, Russia, Prof. J. Kry_ski from IGiK, Dr. Y. Zanimonskiy from the Institute „Metrologia” Kharkov, Ukraine, who temporarily works at IGiK, and to Dr. Manuel Hernandez-Pajares from the Technical University of Catalonia, Spain, for processing and kind submission of high temporal resolution ionospheric data. Some fragments of the paper as well as some figures are with the approval of the authors taken from the publications cited in bibliography.
Bibliography
Baran L.W., Shagimurativ I.I., Wielgosz P., Yakimova G.A., (2001a): The structure of high-latitude ionosphere during September 1999 storm event obtained from GPS observations, EGS 2001, CD.
Baran L.W., Shagimuratov I.I., Wielgosz P., (2001b): Structure of the ionosphere during disturbance and effects GPS positioning, IAG Report 2001 Scientific Assembly, Budapest, 2-7 September 2001.
Baran L.W., Krankowski A., Shagimuratov I.I., (2001c): Influence of Ionosphere on Repeatability of Vectors Co-ordinates Determination, IAG 2001 Scientific Assembly. Budapest, 2-7 September 2001.
Baran L.W., Rotkiewicz M., Shagimuratov I.I., (2001d) Ionospheric Effects And GPS Positioning, IAG 2001 Scientific Assembly. Budapest, 2-7 September 2001.
Baran L.W., Ephishov I.I. and Shagimuratov I.I., (2001e): Ionospheric Total Electron Content Behaviour During November 1997 Storm, Phys.Chem.Earth (C), 2001, Vol. 26, No 5, pp. 341-346.
Baran L.W., Ephishov I.I., Shagimuratov I.I., (2001f): Spatial Correlation of Ionosphere During Storm Derived from GPS Measurements, Bull. Pol. Ac. Sc.: Earth Sci., Vol. 49, No 2, pp. 151-164, Warsaw, December 2001.
Baran L.W., Shagimuratov I.I., Ephishov I.I. Wielgosz P.A., Krankowski A., (2002a): The response of the ionosphere over Europe to the geomagnetic storm on March 31, 2001, EGS, Nice, April 2002, ST023; EGS02-A-01063; ST2-1WE5P-023.
Baran L.W., Rotkiewicz, M., Wielgosz, P., Shagimuratov, I.I., (2002b): Ionospheric Effects During a Geomagnetic Storm, Vistas for geodesy in the New Millennium. International Association of Geodesy Symposia, 125, Springer-Verlag Berlin Heidelberg New York 2002, pp. 291-296.
Baran L.W., Shagimuratov I.I., Aleshnikova M.V., Wielgosz P., (2002c): Latitudinal Variations of TEC over Europe Obtained from GPS Observations, 34th COSPAR Scientific Assembly, Houston, USA, 10-19 October 2002, Abstracts, CD, The New Face of Space - The World Space Congress - 2002.
Brunner F.K., Gu M., (1991): An Improved Model for Dual Frequency Ionospheric Correction of GPS Observations, Manuscripta Geodaetica, 16, pp. 205-214.
Cisak J., (2002a): Overview of the SCAR/WG-GGI/GIANT Activities as an Introduction to the Workshop, Paper presented at the International workshop on "Atmospheric impact on GPS observations focused on polar regions", 15 May 2002, Warsaw, Poland.
Cisak J., (2002b): Atmospheric Impact on GPS Observations in Antarctica - Project of GEODESY/GIANT program of SCAR WGGGI Report 2000 – 2002, Presentation on WGGGI meeting of XXVII SCAR.
Cisak J., Kry_ski J., Zanimonskiy Y.M., (2002): Variations of Ionosphere Versus Variations of Vector Components Determined from Data of IGS Antarctic GPS Stations – New Contribution to the GIANT Project “Atmospheric Impact on GPS Observations in Antarctica”, Paper presented at the AGS-02, Wellington, New Zealand, http://www.scar-ggi.org.au/geodesy/ags02/cisak_ionosphere_gps.pdf
Cisak J., Kry_ski J., Zanimonskiy Y.M., Wielgosz P., (2003a): Variations of Vector Components Determined from GPS data in Antarctic. Ionospheric Aspect in New Results Obtained within the Project “Atmospheric Impact on GPS Observations in Antarctica”, Paper presented at the scientific conference, 22-23 January 2003, Lviv, Ukraine.
Cisak J., Krynski J., Zanimonskiy Y., Wielgosz P., (2003c): Study on the Influence of Ionosphere on GPS Solutions for Antarctic Stations in the Framework of the SCAR Project „Atmospheric Impact on GPS Observations in Antarctica", Paper presented at the 7-th Bilateral Meeting Italy-Poland, Bressanone, Italy, 22-23 May 2003.
Dietrich R., Dach R., Engelhardt G., Ihde J., Korth W., Kutterer H.-J., Lindner K., Mayer M., Menge F., Miller H., Mueller C., Niemeier W., Perlt J., Pohl M., Salbach H.,
Schenke H.-W., Schoene T., Seeber G., Veit A., Voelksen C., (2001): ITRF coordinates and plate velocities from repeated GPS campaigns in Antarctica - an analysis based on different individual solutions, Journal of Geodesy, Number 74, pp. 756-766, Springer Verlag, Germany.
Dietrich R., Rülke A., (2002): The SCAR GPS Campaigns in the ITRF2000, SCAR Report, No 21, January 2002, Publication of the Scientific Committee on Antarctic Research, Scott Polar Research Institute, Cambridge, UK.
E Dongchen, Cheng Xiao, Zhou Chunxia, (2002): GPS Meteorology in Antarctica, Paper presented at the AGS-02, Wellington, New Zealand, http://www.scar-ggi.org.au/geodesy/ags02/dongchen_gps_meteorology.pdf
Ephishov I.I., Shagimuratov I.I., Wielgosz P., Cisak J., (2002): Phase Fluctuations of GPS Signals at High Latitude Ionosphere During the Storm, EMC2002, Proceedings of XVI International Wroclaw Symposium and Exhibition on Electromagnetic Compatibility, pp. 37-42, 2002.
Feltens J., Jakowsky N., (2001): The International GPS Service (IGS): Ionosphere Working Group activities, SCAR Report, Publication of the Scientific Committee on Antarctic Research, Scott Polar Research Institute, Cambridge, UK.
Góral W., (1997): Numerical estimation of differential refraction correction in GPS measurements. Reports on Geodesy, Warsaw U. of Tech., Institute of Geodesy and Geodetic Astronomy, No 5(28), pp. 181-187.
Haefele P., Kaniuth K., (2001): Analysis of Time Series of GPS Height Estimates with Regard to Atmospheric Pressure Loading, Paper presented at the IAG International Symposium on Vertical Reference Systems, 21-23 February 2001, Cartagena, Colombia.
Krankowski A., (2002): Modelling of the Ionosphere in Polar Regions, Paper presented at the International workshop on "Atmospheric impact on GPS observations focused on polar regions", 15 May 2002, Warsaw, Poland
Krankowski A., Baran L.W., Shagimuratov I.I., (2001): Influence of the Northern ionosphere in positioning precision, Abstract XXVI General Assembly EGS, Nice, France 25-30 March 2001.
Krankowski A., Baran L.W., Shagimuratov I.I., Cisak J., (2002a): Influence of Ionosphere in Antarctic Region on GPS Positioning Precision, Collection of Papers and Viewgraphs of IGS/IAACs Ionosphere Workshop, ESOC Darmstadt, Germany, 17-18 January 2002.
Krankowski A., Baran L.W., Shagimuratow I.I., Cisak J., (2002b): Influence of Ionosphere in Arctic and Antarctic Regions on GPS Positioning Precision, SCAR Report, No 21, January 2002, Publication of the Scientific Committee on Antarctic Research, Scott Polar Research Institute, Cambridge, UK.
Krankowski A., Baran L.W., Shagimuratov I.I., (2002c): Influence of the northern ionosphere on positioning precision, Physics and Chemistry of the Earth. 2002, Vol. 27, No 5, pp. 391-395.
Kruczyk M., (2002): Tropospheric Delay in GPS Permanent Networks – a Tool for Atmosphere Research, Paper presented at the International Workshop on "Atmospheric impact on GPS observations focused on polar regions", 15 May 2002, Warsaw, Poland.
Krynski J., Zanimonskiy Y., (2001): Contribution of Data from Polar Regions to the Investigation of Short Term Geodynamics. First Results and Perspectives, SCAR Report, No. 21, January 2002, Publication of the Scientific Committee on Antarctic Research, Scott Polar Research Institute, Cambridge, UK.
Krynski J., Cisak J., Zanimonskiy Y.M., Wielgosz P., (2002a): Variations of GPS Solutions for Positions of Permanent Stations – Reality or Artefact, Symposium of the IAG Subcommission for Europe (EUREF) held in Ponta Delgada, Portugal, 5-7 June 2002, EUREF Publication No 8/2, Mitteilungen des Bundesamtes für Kartographie und Geodäsie, Frankfurt am Main (in print).
Krynski J., Zanimonskiy Y.M., Cisak J., Prokopov A.V., Zanimonskaya A.E., Wielgosz P., (2002c): Global Ionospheric Impact on the Satellite Navigation System Errors (in Russian), Proceedings of the Conference in Kharkov State University, Kharkov, Ukraine, 3 October 2002.
Manning J., (2001): The SCAR Geodetic Infrastructure of Antarctica, Report from the Second SCAR Antarctic Geodesy Symposium, Warsaw, July 1999, SCAR Report No. 20, pp 22-30, SCAR, Cambridge.
Manning J., Cisak J., (2002): Overview of the SCAR/WG-GGI/GIANT activities within the project on the Atmospheric Impact on GPS Observations in Antarctica and possible fields of cooperation with IGS/Iono WG. Proceedings of IGS/IAACs Ionosphere Workshop, Darmstadt, 17-18 January 2002
Mendes V.B., Prates G., Santos L., Langley R., (2000): An Evaluation of Models for the Determination of the Weighted Mean Temperature of the Atmosphere. Proceedings of The Institute of Navigation 2000 National Technical Meeting, Anaheim, CA, U.S.A., 26-28 January 2000; pp. 433-438.
Poutanen M., Koivula H., Ollikainen M., (2001): On the Periodicity of GPS Time Series, Proceedings of IAG 2001 Scientific Assembly, 2-7 September 2001, Budapest, Hungary.
Prokopov A.V., Remayev Ye.V., (2001): Integral Ray Optics Approximation in the Electromagnetic Waves Theory and its Applications for Atmospheric Delay Modeling in GNSS Measurements, Proceedings of IAG Assembly, Budapest, 2001.
Schaer S., (1999): Mapping and Predicting the Earth’s Ionosphere Using the Global Positioning System, PhD Thesis, Astronomical Institute, University of Berne.
Shagimuratov I.I., Baran L.W., Tepenitsyna N.J., Ephishov I.I., (2002a): Variations of TEC during ionospheric storm in September 1999 Obtained from GPS Observations, Proceedings of the XX Conference on Propagation of Radio Waves. Niznij Novgorod, 2-4 July 2002, Niznij Novgorod, pp. 123-124 (in Russian).
Shagimuratov I.I., Baran L.W., Wielgosz P., Yakimova G.A., (2002b): The structure of mid- and high-latitude ionosphere during September 1999 storm event obtained from GPS observations, Ann. Geophys., 20, 1–6, 2002.
Wielgosz P., (2002): Use of Deterministic Ionospheric Component in GPS Observations Processing, Paper presented at the International workshop on "Atmospheric impact on GPS observations focused on polar regions", 15 May 2002, Warsaw, Poland.
Wielgosz P., Grejner-Brzezinska D., Kashani I., (2003): On High-Resolution TEC Derivation from Regional GPS Networks: Feasibility Study, Paper presented at 2003 IEEE AP-S Symposium and USNC/CNC/URSI National Radio Science Meeting, Columbus, Ohio.
Ulyanov Y., Zanimonska A., Zanimonskiy Y., (2000): Correlation of Surface Temperature with Average Temperature of Troposphere in Radioacoustic Sounding Data” Paper presented at the Second International Workshop on “Satellite Navigation in CEI Area” Olsztyn 3-4 July 2000.
Zablotskyj F., (2002) On Determination Precision of Tropospheric Delay at the Antarctic Coast Stations, SCAR Report, No. 21, January 2002, Publication of the Scientific Committee on Antarctic Research, Scott Polar Research Institute, Cambridge, UK.
Zanimonskaya A., Prokopov A., (2001): Estimation of the Second Order Refraction Effects in the Ionosphere Models for GPS Applications (in Ukrainian), Geodesy, Cartography and Photogrammetry, Lviv, No 61, pp. 24-29.
Zanimonskiy Y., Krynski J., Cisak J., (2002): Correlation of the Atmospheric and Geodetic time Series of GPS Results. Impact, coincidence or artefact?, Paper presented at the International Workshop on "Atmospheric impact on GPS observations focused on polar regions", 15 May 2002, Warsaw, Poland.