SEARCH

SEARCH BY CITATION

Keywords:

  • scintillation;
  • maps

Abstract

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Database
  5. 3. Data Analysis
  6. 4. Modeling-Measurement Comparison
  7. 5. Undersampling and IGS Data
  8. 6. Generation of Scintillation Maps
  9. 7. Conclusion
  10. Acknowledgments
  11. References
  12. Supporting Information

[1] This paper presents the results of a scintillation measurement campaign. The receivers used during this campaign were located in five regions of the globe which allowed us to investigate geographic similarities and differences. Most of the receivers were located in the equatorial region, and one of them was located at high latitudes. Significant results were achieved even though the measurement period was around solar minimum. An intercomparison with the Wide-Band Scintillation Model and the Global Ionosphere Scintillation Model has been carried out, and the construction of scintillations maps has been initiated.

1. Introduction

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Database
  5. 3. Data Analysis
  6. 4. Modeling-Measurement Comparison
  7. 5. Undersampling and IGS Data
  8. 6. Generation of Scintillation Maps
  9. 7. Conclusion
  10. Acknowledgments
  11. References
  12. Supporting Information

[2] As a result of propagation through ionosphere electron density irregularities, transionospheric radio signals may experience amplitude and phase fluctuations. These signal fluctuations may occur specially during equinoxes, after sunset, and last a few hours. They are more intense in periods of high solar activity. These fluctuations result in signal degradation from VHF up to C band. They may affect several applications as navigation systems, communications, remote sensing and Earth observation systems.

[3] This paper is focused on propagation through small-size ionosphere irregularities (below 1 km). These inhomogeneities develop under several deionization instability processes, which induce random fluctuations of the medium's refractive index. Propagation through this medium induces signal scintillation. Received intensity normalized standard deviation S4 is used to classify scintillation events: weak scintillations for S4 < 0.3, moderate scintillations for 0.3 < S4 < 0.7 and strong scintillations for S4 > 0.7. The occurrence of strong scintillation events increases with solar activity.

[4] Two globe regions are impacted by such events: equatorial regions (−20° to +20° magnetic latitude) and polar regions (>60°). Scintillations are more intense at the equator, and characteristics are different between these two regions. This was first presented by Aarons [1982]. For a propagation study, main parameters are the field statistical moments, up to the fourth order [Yeh and Liu, 1982].

[5] Many scintillation measurement campaigns have been conducted, in particular in South America [Doherty et al., 2000; De Paula et al., 2001], India [Chandra et al., 2007] and Japan [El-Arini et al., 2003; Otsuka et al., 2006]. Several review papers were published on this specific topic [Basu and Basu, 1985; Basu et al., 1996]. Over the last years, the number of scintillation receivers expanded significantly worldwide. However, very few data were measured in Africa. Also, very little work was carried out regarding scintillation prediction and scintillation maps construction. These topics are addressed in this paper.

[6] In the framework of an ESA/ESTEC contract, we conducted such a measurement campaign with worldwide coverage, including Africa and high latitudes (Sweden). This study is presented in this paper. The frequency of interest was the L band and we used GPS receivers. Some of these receivers were regrouped in order to derive the medium's correlation properties, especially in Vietnam and in Africa. Data from the different locations were stored under a standard format in order to build a database. This database was analyzed, and results obtained were compared with predictions given by GISM and WBMOD models. In addition, algorithms were developed to build forecasting scintillations maps.

2. Database

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Database
  5. 3. Data Analysis
  6. 4. Modeling-Measurement Comparison
  7. 5. Undersampling and IGS Data
  8. 6. Generation of Scintillation Maps
  9. 7. Conclusion
  10. Acknowledgments
  11. References
  12. Supporting Information

[7] Data were collected over 2006 and 2007, near solar cycle minimum. Two monitor's types were deployed at low latitudes in South America, Africa, Asia and at Europe high latitudes, as shown on Figure 1. The first monitor model is a Legacy (from Javad or Topcon, both identical in construction and operation mode). The second monitor type is a GSV4004 based on a Novatel receiver. Both types are operating at 50 Hz and provide additional channels to track geostationary satellites (when available). A database was built. Owing to limited local support, storing raw data files was not always possible. In some locations the data were processed online, stored, and sent back to Europe on a monthly basis. All data were to derive scintillation characteristics and improve scintillation models.

image

Figure 1. Receivers deployment.

Download figure to PowerPoint

[8] Several receivers detected many multipaths, difficult to cancel out. One way to remove these multipaths is indicated in the GSV4004 user manual and based on a phase code divergence criterion. This procedure was systematically used, but was found to be insufficient in a number of cases. Receivers in Hué, Ho Chi Minh, Cayenne and Douala provided good quality data. There were practically no multipaths. In addition, the Cayenne receiver provided raw data used for the spectrum analysis. The receiver in Hanoï, inside the city, suffers from multipaths but their level is not significant and they could be removed. The receiver in N'Djamena was badly located over the first three months and the level of multipaths was too high. It was then moved to another location. Strong multipaths effects mitigated the analysis of Javad receivers data, especially the Bandung's one, whereas phase measurements were satisfactory.

[9] We excluded data which did not correspond to ionosphere scintillations. Therefore, we discarded all measures recorded with elevation angles lower than 20° and S4 level lower than 0.2. Assuming this selected data only shows scintillations ionosphere originated, we obtained a reduced data set for subsequent analysis. Table 1 presents, for each station, the number of selected samples according to previous criteria.

Table 1. Percentage of Selected Measurements
StationLatitude Geographic (Magnetic)Longitude Geographic (Magnetic)Recorded SamplesSelected SamplesSelected Samples (%)
Bandung (Indonesia)6°53′39″S (16°8S)107″35′11″E (178°44E)2,261,51313,5430.6
Kiruna (Sweden)67°50′26″N (64°4N)20°24′37″E (104°08E)2,614,466130.00
Sancto de la Cruz (Canary Islands)28° 25′4.5″N (19°86N)16° 33′15″W (61°15E)2,096,3333,3510.16
N'Djamena (Chad)12° 6′N (0°88S)15° E (86°5E)4,074,0219,9780.24
Ho Chi Minh (Vietnam)10°50′54″N (2°34N)106°33′35″E (177°42E)6,538,1681,0300.02
Hué (Vietnam)16°27′47″N (8°49N)107°35′05″E (178°38E)4,127,4641,5460.04
Hanoï (Vietnam)21°28′26″N (13°75N)105°47′59″E (176°70E)5,445,4972,3960.04
Cayenne (French Guiana)4°49′27″N (15°23N)307°38′12″E (21°99E)2,137,02912,5120.59
Douala (Cameroon)4°N (9°37S)9°E (80°16E)3,257,67119,7500.61
Total  32,552,16264,1190.2

[10] This study is based on the small part of the database that was selected. The ionospheric scintillation was a rare event during the measurement campaign: only 0.2% of the measured satellite/Earth links were affected by unambiguous scintillations. However, although measured near solar minimum, high scintillation values were recorded. This enabled us to derive the scintillation statistical characteristics.

[11] However, some additional parameters modify scintillation characteristics at receiver level, namely, the respective satellite drift velocities to the ionosphere and the magnetic field. The apparent drift velocity at receiver level, is a combination of ionosphere drift velocity and motion of the satellite link ionosphere pierce point (IPP). The IPP motion modifies the fades duration. It can be an increase or a decrease depending on the geometry. This ratio of the two velocities depends on the elevation angle and on the magnetic field, as a result of elongated bubbles in that direction. The increase in fade duration modifies the recorded signals statistical properties. It may also increase the number of losses of lock. This point was addressed by Kintner et al. [2001] using several collocated receivers, and by DasGupta et al. [2006]. These two parameters, drift velocities and magnetic field, were not considered in this study. As a first assumption, GPS trajectories are equally distributed versus the ground receivers. Therefore, recorded values used in this study may be seen as first-order approximations.

3. Data Analysis

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Database
  5. 3. Data Analysis
  6. 4. Modeling-Measurement Comparison
  7. 5. Undersampling and IGS Data
  8. 6. Generation of Scintillation Maps
  9. 7. Conclusion
  10. Acknowledgments
  11. References
  12. Supporting Information

3.1. Preprocessed Files

[12] Preprocessed files contain scintillation indices values for intensity (S4) and phase (sigma phi). Raw data were processed in situ. Indices are calculated from the 1-min successive samples. To perform this analysis, the software is included in the GSV receiver. Figure 2 details standard measurements observed. Figure 2 corresponds to data recorded in Vietnam and presents intensity (S4) and phase (sigma phi) scintillations standard deviations on a monthly basis. In that case, maximum scintillation activity occurs during equinoxes, on both sides of the equatorial anomaly.

image

Figure 2. Histogram of intensity and phase scintillations recorded in Vietnam. Horizontal axis: (left) the value of S4 to be multiplied by 0.1; (right) the month number. Vertical axis: normalized number of events.

Download figure to PowerPoint

[13] This analysis was performed for all deployed receivers, with same results achieved for all of them. However, although in all cases no scintillations were recorded in July and August, the activity peak was not always recorded at equinoxes. This was especially the case in Cayenne (French Guiana), where maximum scintillation activity occurred in December. The time dependency, plotted in Figure 3, exhibits the classical behavior, starting after sunset and lasting a few hours.

image

Figure 3. Distribution of 1 year of scintillation events from June 2006 to July 2007 (6-min intervals).

Download figure to PowerPoint

3.2. Raw Data Files

[14] Scintillation spectra are calculated through raw data files. To simplify the analysis, raw data from all receivers was translated into Rinex format. Figure 4 presents the power spectra of GPS signal amplitude measured in Bandung in 2006 during ionospheric scintillation effects (Figure 4, left), in comparison with a spectrum of normal signal without ionospheric scintillations (Figure 4, right). The system noise high value is apparent in Figure 4.

image

Figure 4. Power spectra of GPS signal amplitude in Bandung station (right) in a regular case and (left) during a scintillation event.

Download figure to PowerPoint

[15] In order to analyze all downloaded raw data, the Power Spectral Density (PSD) must be synthesized into a small number of parameters. The PSD is usually characterized by its strength at 1 Hz (T) and by its power law decay index (p). On a log-log representation, these parameters are obtained by fitting a line to data points, using least squares. The frequency range for this line fit begins at 0.1 Hz, which is the high-pass filter cutoff frequency. This filter removes low-frequency parasitic fluctuations, due in particular to the satellite motion on its orbit. In order to prevent this fit from expanding into system noise, which biases p toward lower values, the considered frequency range ends at 1 Hz.

[16] This processing is applied to every 1-min sample of the downloaded RINEX files. Most of the parameters used for multipaths rejection and filter convergence checking are available in other files. However, for software simplification purposes, these parameters were not used. Instead, simplified criteria were used: (1) no missing measurements (3000 points in a 1-min sample); (2) S4 > 0.2 to avoid multipaths; (3) σphi < 2.0 to check the filter convergence.

[17] The following values present the p index histogram for power and detrended phase PSD (Figure 5). These histograms are directly related to the probability density distributions of p. It can be observed that both distributions are centered around 2.8. Usually, p is considered to be in the 1–4 range. In addition, 2.5 is a commonly chosen value for p at equatorial latitudes. The observed distributions are compatible with these statements.

image

Figure 5. Intensity and phase spectrum slope.

Download figure to PowerPoint

[18] The phase PSD can be approximated by Tf−p. As a consequence, the phase standard deviation σphi is expressed as follows, where fc is the cutoff frequency:

  • equation image

Since S4 is not a standard deviation, such a simple relation between S4 and the power PSD is not expected.

[19] Figure 6 presents σphi computed from the previous equation, as a function of the σphi computed as the time series standard deviation, for every 1-min sample. The correlation between the frequency domain deducted value and the time domain deducted value is remarkable. This observation and the previous realistic p distribution seem to prove that the spectral parameters extraction is reliable.

image

Figure 6. Sigma phi calculated from the spectral analysis versus sigma phi calculated in the time domain.

Download figure to PowerPoint

4. Modeling-Measurement Comparison

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Database
  5. 3. Data Analysis
  6. 4. Modeling-Measurement Comparison
  7. 5. Undersampling and IGS Data
  8. 6. Generation of Scintillation Maps
  9. 7. Conclusion
  10. Acknowledgments
  11. References
  12. Supporting Information

[20] The measurements, selected according to our criteria, were used for comparison with scintillation models. For each one of these selected data, we calculated the corresponding values through models: GISM [Béniguel, 2002] developed under contract with ESA/ESTEC and WBMOD [Secan et al., 1995]. For this calculation, GISM input used is the Yuma file at corresponding date. GISM has an orbit generator, and data were provided simultaneously to WBMOD. As a result, both models are executed with the same input data, corresponding to the valid measurement. This was performed for the 2 years 2006–2007.

[21] Figure 7 shows a comparison of data measured in Vietnam (Hanoï and Hué) and estimated using GISM and WBMOD models in year 2006. As indicated before, the 0.2 noise measurement threshold was due to multipaths.

image

Figure 7. Comparison of measurements and modeling in Hanoï and Hué depending on day of year and local time.

Download figure to PowerPoint

[22] To analyze latitudinal and longitudinal (or temporal) extent of scintillations at ionosphere level, we have approximated ionosphere by a thin shell at 350 km of altitude. The Ionosphere Pierce Point (IPP) is defined as the intersection point of the satellite/Earth link and this thin shell. For each valid sample, the recorded azimuth and elevation angle are used to compute the IPP latitude and longitude. The longitude is used to determine local time at the IPP.

[23] With a 20° elevation mask angle, the latitudinal extent of observed zone is about 14° geographic. Temporal extent is set to 24 H, centered around midnight, since scintillations occur during nighttime. The observation zone is divided into a grid of 20 latitude points and 70 temporal points. For each station, all IPPs observed during measurement campaign are classified into grid cells.

[24] For each grid cell, samples are counted and the mean value of measured S4 and σphi is computed. These three parameters are good indicators of scintillation activity (only unambiguous scintillations measurements are taken into account). This kind of map is the first step toward a cartography of ionosphere irregularities.

[25] Figure 8 shows the results obtained in Cayenne using measured data. The line in the middle of each plot corresponds to the geostationary satellite track. The phase is usually wrong for the geostationary satellites [Van Dierendonck and Arbesser-Rastburg, 2004]. Contrary to GPS satellites, this explains their higher phase level compared to intensity level. A few points are outside the −5 to +5 theoretical range area, where they are expected to occur. They probably correspond to nonremoved multipaths.

image

Figure 8. Number of samples and scintillation indexes in Cayenne.

Download figure to PowerPoint

[26] The four stations located in Vietnam and Indonesia are almost located at the same longitude. Measurements recorded have been positioned on the same plot which depicts global pattern. Horizontal stripes between −25° and −10° and between 4° and 10° correspond to multipaths in Hanoï and Bandung. Modeling results are given for comparison purposes in Figure 9 and show a reasonable agreement.

image

Figure 9. Comparison of measurements and modeling in Vietnam and Indonesia.

Download figure to PowerPoint

[27] WBMOD exhibits a peak on both sides of the equatorial anomaly, identical to what could be observed on a TEC map. On the contrary, this is neither apparent with GISM, nor with measurements. GISM uses the NeQuick model [Leitinger et al., 2002] as a background to get the electronic density mean value. NeQuick model [Radicella and Zhang, 1995] gives an analytical representation of the electron density vertical profile, with continuous first and second derivatives. This quick-run model is particularly tailored for transionospheric applications and enables the calculation of electron concentration at any given location in the ionosphere. The NeQuick TEC map reproduces satisfactorily the equatorial anomaly with peaks on both sides of the magnetic equator. As a conclusion, it is not obvious that a scintillation map should reproduce the same trends as a TEC map, and this example shows the contrary. The same issue has been addressed by Cervera and Thomas [2006] with similar observations.

5. Undersampling and IGS Data

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Database
  5. 3. Data Analysis
  6. 4. Modeling-Measurement Comparison
  7. 5. Undersampling and IGS Data
  8. 6. Generation of Scintillation Maps
  9. 7. Conclusion
  10. Acknowledgments
  11. References
  12. Supporting Information

[28] The aim of this section is to analyze the ability to get, through the use of IGS network data to get some information on the scintillation activity for forecasting purposes. The loop bandwidth of the receiver was 10 Hz (default value for the GSV4004 receiver). Although this value could be optimized to increase the receiver performance in a scintillation environment [Humphreys et al., 2005; Guichon et al., 2008], this is not the topic of this section.

[29] Figure 10 presents the effect of an undersampling for one typical example (satellite PRN 2, day 315, year 2006 at Cayenne, French Guiana). The detrended 50 Hz raw data were sampled down to 1 Hz. As a consequence, the explored frequency range drops from 25 Hz to 0.5 Hz. In this example, it appears that the effect of the 1 Hz sampling is moderate.

image

Figure 10. Undersampling at 1 Hz: (left) time domain and (right) frequency domain.

Download figure to PowerPoint

[30] In order to investigate the effect of the smaller frequency band, the integral used to compute σphi from the PSD is decomposed as follows:

  • equation image

Setting p to 2.8, as observed in the measurements, we get:

  • equation image

The part of the spectrum above 0.5 Hz is not material. In other words, a sampling rate of 1 Hz seems to be sufficient to evaluate σphi in that case. In order to check this statement, every available 1-min sample was downsampled to 1 Hz. The σphi computed over these 60 points samples was compared with the σphi computed over the full 3000 points samples. Figure 11 presents this comparison. It is observed that the 1 Hz σphi is a good estimation of the 50 Hz σphi.

image

Figure 11. Sigma phi calculated from 1 Hz data versus sigma phi calculated from 50 Hz data during 4 days at Cayenne.

Download figure to PowerPoint

[31] This undersampling study has been complemented by a 1 Hz IGS data analysis from Kourou, located at a short distance from Cayenne. The chosen index for IGS data is the ROTI index [Valette et al., 2007]. This phase index is deducted from the change rate in geometry after removing mean variation using L1 and L2 frequencies. A day of intense scintillations at Cayenne was selected for this comparison. Figure 12 shows the comparison between the 1-min S4 index recorded by the GSV receiver in Cayenne, and the empirical ROTI scintillation index derived from the IGS 1 s data in Kourou. This enables, through IGS network use, to get some information on scintillation activity. However, a more systematic analysis should be performed at higher solar activity to confirm these observations.

image

Figure 12. Comparison of scintillation indices from (top) PRIS (50 Hz)/Cayenne and (bottom) IGS (1 Hz)/Kourou for 1 December 2006 (the black lines are the projection of the GPS satellite pierce points at an altitude of 400 km) with the following S4 color scale: orange, 0.25–0.4; red, 0.4, 0.55; purple, >0.55.

Download figure to PowerPoint

6. Generation of Scintillation Maps

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Database
  5. 3. Data Analysis
  6. 4. Modeling-Measurement Comparison
  7. 5. Undersampling and IGS Data
  8. 6. Generation of Scintillation Maps
  9. 7. Conclusion
  10. Acknowledgments
  11. References
  12. Supporting Information

[32] The generation of scintillation maps is a challenging task for future GNSS applications in high- and low-latitude regions with enhanced scintillation activity. Predicted scintillation maps enable a decision at user level in advance, thus avoiding/reducing signal degradation, i.e., positioning errors and safety risks.

[33] Owing to the small-scale structure of ionospheric irregularities causing radio scintillations, the maps require a high spatial resolution. Considering the present data situation worldwide, it can be stated that existing high-rate GNSS networks are not dense enough to fulfill this requirement. On the other hand, small-scale irregularities with lengths from 1 m up to 1 km do not usually occur as isolated perturbations. Geophysical conditions enhancing the probability of generating plasma instabilities are not restricted to the meter level. Therefore, there is a good chance to describe the scintillation activity horizontal distribution by a grid spacing up to a few 100 km. The correlation distance is a key parameter for creating scintillation maps. First attempts to construct scintillation maps were performed by applying the same technique as the one developed for generating TEC maps in DLR [Jakowski, 1998]. To generate S4 maps, the deducted S4 values are geo-located and then assimilated into the GISM background model. Thus, the model is updated by available measurements. The characteristic width s of the distance, depending on weight functions constructed around measurement points, can be matched to available data density. High data density allows reducing s to the mean distance of measurement points with enhanced spatial resolution. Figure 13 demonstrates the construction of a scintillation map by using the GISM background model and scintillation observations carried out in Brazil on 11 January 2002 [Jakowski et al., 2007]. The s parameter is fixed at 8 deg.

image

Figure 13. Scintillation maps over Brazil on 11 January 2002 at 0030:00 UT. (left) GISM model and (right) map reconstruction using s = 8.

Download figure to PowerPoint

[34] Generally, one has to take into account the fact that the data are unevenly distributed. Nevertheless, owing to the construction technique, grid points at the margin area are also influenced by the overall measurements depending on the choice of the s parameter.

7. Conclusion

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Database
  5. 3. Data Analysis
  6. 4. Modeling-Measurement Comparison
  7. 5. Undersampling and IGS Data
  8. 6. Generation of Scintillation Maps
  9. 7. Conclusion
  10. Acknowledgments
  11. References
  12. Supporting Information

[35] The scintillation characteristics were derived from a data set including measurements from South America, Africa and Asia in years 2006 and 2007, corresponding to years close to solar minimum. All receivers provided data for more than 1 year. Although being observed at solar minimum, high values of scintillation may occur, allowing deriving statistical characteristics of scintillations. The expected behavior of scintillation occurrences was determined, with some slight discrepancies between the different locations. In particular, the maximum of scintillation occurrences occurs in winter in South America and at equinoxes in Vietnam. No scintillations were recorded at high latitudes and very few with the receiver located in the Canary Islands. However, at this location, the produced IPP maps show a few points at very low latitudes, corresponding to the crest of equatorial anomaly.

[36] An extensive comparison was made between the measurements and the two models, GISM and WBMOD. For this analysis, in order to avoid misinterpretation and wrong conclusions, a severe criterion was chosen to decide on valid measurements. Only 0.2% of the measurements recorded during the campaign were considered.

[37] The comparisons show that the two models have a good correlation in most cases, and that the correlation of the models with the experiments is not so good on a particular event. This is not surprising, as we try to estimate one realization of an event with a model. On a larger point of view, one week or one month, the comparison is greatly improved.

[38] The plotted latitude versus IPP maps allow a global view of the scintillation activity. It is quite easy to separate the different contributions: multipaths, geostationary satellite contribution and scintillations. The comparison between measurements and models, although perfectible, is satisfactory and the maps analysis is of great interest.

[39] First attempts of generating scintillation maps were made through combination of the GISM model and S4 data deducted from 50 Hz scintillation measurements. Owing to low data density, spatial resolution is scarce. Considering the growing number of GPS stations capable of measuring scintillations, the situation permanently improves. Another way to feed the model with a larger number of data could be using 1 Hz data from GNSS networks such as IGS. We showed in one particular case that undersampling may produce acceptable results. This promising option will be analyzed in more details in the future.

Acknowledgments

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Database
  5. 3. Data Analysis
  6. 4. Modeling-Measurement Comparison
  7. 5. Undersampling and IGS Data
  8. 6. Generation of Scintillation Maps
  9. 7. Conclusion
  10. Acknowledgments
  11. References
  12. Supporting Information

[40] The authors want to express their acknowledgment to the different people who have contributed to the success of this measurement campaign: M. A. Alimi and M. Noubatessem (ASECNA company, N'Djamena), M. Le Huy (Hanoï Institute of Geophysics, Vietnam), M. H. Thai (Vietnam), M. Fleury (ENSTB, Brest, France), and LAPAN institute in Indonesia.

References

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Database
  5. 3. Data Analysis
  6. 4. Modeling-Measurement Comparison
  7. 5. Undersampling and IGS Data
  8. 6. Generation of Scintillation Maps
  9. 7. Conclusion
  10. Acknowledgments
  11. References
  12. Supporting Information
  • Aarons, J. (1982), Global morphology of ionospheric scintillations, Proc. IEEE, 70(4), 360378, doi:10.1109/PROC.1982.12314.
  • Basu, S., and S. Basu (1985), Equatorial scintillations: Advances since ISEA-6, J. Atmos. Terr. Phys., 47, 753768, doi:10.1016/0021-9169(85)90052-2.
  • Basu, S., et al. (1996), Scintillations, plasma drifts, and neutral winds in the equatorial ionosphere after sunset, J. Geophys. Res., 101(A12), 26,79526,809.
  • Béniguel, Y. (2002), GIM: A global ionospheric propagation model for scintillations of transmitted signals, Radio Sci., 37(3), 1032, doi:10.1029/2000RS002393.
  • Cervera, M., and R. M. Thomas (2006), Latitudinal and temporal variations of equatorial ionospheric irregularities determined from GPS scintillation observations, Ann. Geophys., 24, 33293341.
  • Chandra, H., P. V. S. Ramarao, P. N. Vijay Kumar, B. M. Pathan, K. N. Iyer, A. K. Gwal, R. P. Singh, B. Singh, and A. Dasgupta (2007), VHF scintillations at low latitudes in India, paper presented at Beacon Satellite Symposium, Int. Union of Radio Sci., Boston, Mass.
  • DasGupta, A., A. Paul, S. Ray, A. Das, and S. Ananthakrisnan (2006), Equatorial bubbles as observed with GPS measurements over Pune, India, Radio Sci., 41, RS5S28, doi:10.1029/2005RS003359.
  • De Paula, E., P. H. Doherty, and T. Dehel (2001), Scintillation activity and recent measurements from the Brazilian region, paper presented at International Satellite Based Augmentation Systems (SBAS–IONO) Meeting, Eur. Space Agency, Maastricht, Netherlands, May .
  • Doherty, P. H., S. H. Delay, C. E. Valladares, and J. Klobuchar (2000), Ionospheric scintillation effects in the equatorial and auroral regions, paper presented at ION GPS 2000, Inst. of Navig., Salt Lake City, Utah, 19 – 22 Sept.
  • El-Arini, M. B., R. S. Conker, S. D. Ericson, K. W. Bean, F. Niles, K. Matsunaga, and K. Hoshinoo (2003), Analysis of the effects of ionospheric scintillation on GPS L2 in Japan, paper presented at ION GPS 2003, Inst. of Navig., Portland, Oreg., Sept.
  • Guichon, H., N. Martin, S. Lannelongue, and M. Crisci (2008), Impact of ionospheric scintillations on ground stations and user receivers, paper presented at Navitec Symposium, Noordwijk, Netherlands.
  • Humphreys, T., M. Psiaki, B. Ledvina, and P. Kintner (2005), Performance of GPS carrier tracking loops during ionospheric scintillations, in Proceedings of Ionospheric Effects Symposium, JMG Assoc., Alexandria, Va.
  • Jakowski, N. (1998), Generation of TEC maps over the COST area based on GPS measurements, paper presented at 2nd COST251 Workshop, 30 – 31 March , Side, Turkey.
  • Jakowski, N., E. de Paula, V. Wilken, and Y. Béniguel (2007), Assimilation of Brazilian data into the GISM model, paper presented at International Satellite Based Augmentation Systems (SBAS–IONO) Meeting, Eur. Space Agency, Boston, Mass.
  • Kintner, P., H. Kil, T. Beach, and E. de Paula (2001), Fading timescales associated with GPS signals and potential consequences, Radio Sci., 36(4), 731743.
  • Leitinger, R., S. Radicella, and B. Nava (2002), Electron density models for assessment studies—New developments, Acta Geod. Geophys. Hung., 37, 183193, doi:10.1556/AGeod.37.2002.2-3.7.
  • Otsuka, Y., K. Shiokawa, and T. Ogawa (2006), Equatorial ionospheric scintillations and zonal irregularity drifts observed with closely spaced GPS receivers in Indonesia, J. Meteorol. Soc. Jpn., 84A, 343351, doi:10.2151/jmsj.84A.343.
  • Radicella, S. M., and M. L. Zhang (1995), The improved DGR analytical model of electron density height profile and total electron content in the ionosphere, Ann. Geofis., 38, 3541.
  • Secan, J. A., R. M. Bussey, and E. J. Fremouw (1995), An improved model of equatorial scintillation, Radio Sci., 30, 607617, doi:10.1029/94RS03172.
  • Valette, J. J., et al. (2007), Observations of ionospheric perturbations on GPS signals at 50 Hz, 1 Hz and 0.03 Hz in South America and Indonesia, Eur. Space Weather Week, Brussels, Nov.
  • Van Dierendonck, A. J., and B. Arbesser-Rastburg (2004), Measuring ionospheric scintillation in the equatorial region over Africa, including measurements from SBAS Geostationary Satellite signals, paper presented at Beacon Satellite Symposium, Int. Union of Radio Sci., Trieste, Italy, Oct.
  • Yeh, C. K., and C. H. Liu (1982), Radio wave scintillations in the ionosphere, Proc. IEEE, 70(4), 325378.

Supporting Information

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Database
  5. 3. Data Analysis
  6. 4. Modeling-Measurement Comparison
  7. 5. Undersampling and IGS Data
  8. 6. Generation of Scintillation Maps
  9. 7. Conclusion
  10. Acknowledgments
  11. References
  12. Supporting Information
FilenameFormatSizeDescription
rds5669-sup-0001-t01.txtplain text document1KTab-delimited Table 1.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.