PUBLICATIONS. Radio Science. Low-latitude ionospheric effects on SBAS RESEARCH ARTICLE /2015RS005863

Transcription

1 PUBLICATIONS RESEARCH ARTICLE Special Section: URSI AT-RASC (Atlantic Radio Science Conference) Key Points: Main low-latitude ionospheric effect for SBAS are TEC gradients, scintillations, and depletions EGNOS preserves integrity in all cases, normally penalizing accuracy and availability Correspondence to: J. Arenas, Citation: Arenas, J., E. Sardón, A. Sainz, B. Ochoa, and S. Magdaleno (2016), Low-latitude ionospheric effects on SBAS, Radio Sci., 51, , doi: / 2015RS Received 18 NOV 2015 Accepted 12 MAY 2016 Accepted article online 13 MAY 2016 Published online 7 JUN American Geophysical Union. All Rights Reserved. Low-latitude ionospheric effects on SBAS J. Arenas 1, E. Sardón 1, A. Sainz 1, B. Ochoa 1, and S. Magdaleno 1,2 1 GMV, Tres Cantos, Spain, 2 Now at ESSP, Madrid, Spain Abstract Satellite-based augmentation systems (SBAS) provide augmentation to Global Navigation Satellite Systems (GNSS) users in three areas: (1) broadcasting accurate corrections to GNSS satellite ephemeris, (2) providing a real-time empirical ionospheric model in the service area, and (3) providing integrity information in the form of estimates of the confidence of the ephemeris corrections and ionospheric delays. Ionospheric effects on SBAS are twofold: (a) the input data used by the SBAS will be affected by ionospheric effects, and (b) the more perturbed the ionosphere is, the more difficult it will be to provide accurate and reliable ionospheric information to the users. The ionosphere at low latitudes presents larger variability and more intense phenomena than at midlatitudes. Therefore, SBAS providing service to low-latitude regions will be more affected than those at other latitudes. From the different low-latitude ionospheric effects, this paper will focus on those having the largest impact on SBAS, which are total electron content temporal and spatial gradients, ionospheric scintillations, and depletions. This paper will present the impact of these effects on EGNOS (European Global Navigation Overlay System), the European SBAS. Although EGNOS can be considered as a midlatitude SBAS, it has to provide coverage down to rather low latitudes, so sometimes low-latitude ionospheric effects are observed in the EGNOS data. It will be shown how EGNOS performs under nominal conditions and how its performance is degraded when low-latitude ionospheric phenomena occur. Real EGNOS data affected by low-latitude ionospheric phenomena will be used. 1. Introduction One of the main sources of concern in satellite-based augmentation is the effect of the ionosphere on the radio signals. This effect can range from a few meters to many tens of meters depending on several factors. The main problem of the ionospheric effect is the difficulty of predicting and modeling its contribution due to the high variability of the ionosphere. Radio signals crossing the ionosphere suffer several ionospheric effects like phase advance, group delay, additional contribution to the geometric Doppler shift, Faraday rotation, ray bending, and amplitude and phase scintillation. Most of those effects depend on the ionospheric total electron content (TEC). The total electron content is the number of electrons in a vertical column of 1 m 2 cross section. Values for the TEC are between e/m 2 and e/m 2 along the propagation path, depending on local time, position, season, solar, geomagnetic activities, etc. There is a long-term variation of the TEC with the solar cycle, in such a way that the largest values of TEC occur during the solar maximum; the last one has just finished. There is also a diurnal variation of the TEC, and there can be more than 1 order of magnitude of difference between diurnal and nocturnal TEC values. There are also regional changes of the ionosphere due to geographicgeomagnetic relationships. In this case, the most relevant phenomenon is the equatorial crest, a region of enhanced electron density on both sides of the geomagnetic equator. This is the region with the largest horizontal TEC gradients, where rather high values of TEC can be observed even during low solar activity. Apart from those relatively well-known dependencies of the TEC, there are other irregular effects that affect the behavior of the TEC like ionospheric storms, ionospheric irregularities, TEC depletions, and scintillations. For some of those irregular effects the affected area can be rather small, which makes their observability by the reference stations in the satellite-based augmentation systems (SBAS) very difficult. This can provoke events of loss of integrity at the user level or degradation in the availability of the system if the SBAS ionospheric bounds (grid ionospheric vertical error, GIVE) are increased to cope with those localized effects. The ionospheric impact will be different for single-frequency SBAS (as those currently deployed) than for dual-frequency SBAS (future generation of SBAS). The analysis performed and presented in this paper will consider only single-frequency SBAS. ARENAS ET AL. LOW-LATITUDE IONOSPHERIC EFFECTS ON SBAS 603

2 SBAS provide ionospheric information to single-frequency Global Navigation Satellite Systems (GNSS) users as a grid of vertical TEC in their service area using as input data GNSS observations; therefore, those systems are directly impacted by ionospheric effects. Most operational SBAS in the world are designed for midlatitude regions, and SBAS standards assume a smooth ionosphere behavior. However, even midlatitude SBAS cover some regions with low-latitude ionospheric effects, where the hypothesis of a smooth Figure 1. The European Geostationary Navigation Overlay Service, EGNOS. ionosphere does not hold, making it more difficult to provide accurate and reliable information. European Global Navigation Overlay System (EGNOS), the European SBAS [Ventura-Traveset and Flament, 2006], has to provide service in an area from 20 N to 70 N in latitude, very close to the equatorial ionospheric region, and is therefore subject to the ionospheric effects described previously. Here we will concentrate on TEC temporal and spatial gradients, the equatorial depletions, and the ionospheric scintillations as the ionospheric effects with the largest impact on SBAS covering low latitudes. 2. SBAS Satellite-Based Augmentation Systems (SBAS) broadcast differential corrections to ephemeris satellite position and clock and ionospheric delays at a set of ionospheric grid points. Additionally, those systems provide boundings for the errors associated with those corrections, known as integrity information. From this information, the users can determine more accurately their position, and most importantly for safety-of-life application they can determine the confidence in that position solution. SBAS consist of a ground network of reference stations collecting dual frequency GNSS data that are sent to a set of processing centers where the corrections and integrity information are computed. SBAS use geostationary satellites to broadcast GNSS integrity and corrections data to their users and to provide ranging signals. Several SBAS are deployed and provide currently operational signals to their users: Wide Area Augmentation System (WAAS, U.S.), EGNOS (Europe), GPS Aided Geo Augmented Navigation (GAGAN, India), and Multifunctional Satellite Augmentation System (MSAS, Japan). Additionally, there are other SBAS that are currently under development (as, for example, System for Differential Corrections and Monitoring (SDCM) in Russia) or under study as it is the case of Sistema de Aumentación para el Caribe, Centro y Sudamérica (Augmentation System for the Caribbean, Central and South America) SACCSA (for Central and South America). EGNOS, the European SBAS, augments the GNSS satellite system making it suitable for civil aviation safety critical approaches. EGNOS was conceived as a multimodal GNSS service and has been built as a system interoperable with other SBAS. EGNOS is composed of three different segments, as presented in Figure 1. The EGNOS Ground Segment consists the following: (1) Ranging Integrity Monitoring Stations (RIMS): A dense network of reference stations for GPS, Geostationary Satellite (GEO), and Global Navigation Satellite System (Russia) (GLONASS) data reception; (2) Central Processing Facilities (CPF) and Central Control Facilities (CCF): Processing and Control Centers, both collocated in the Master Control Centers (MCC); (3) Navigation Land Earth Stations (NLES): uplink stations; and (4) EGNOS Wide Area Network (EWAN): Network for communications capabilities between the different ground-segment elements. The EGNOS Space Segment consists of a navigation payload in geostationary satellites. The EGNOS User Segment consists of user reception and data processing. EGNOS users are singlefrequency users. ARENAS ET AL. LOW-LATITUDE IONOSPHERIC EFFECTS ON SBAS 604

3 Figure 2. User ionospheric interpolation. The purpose of the SBAS ionospheric corrections is to provide single-frequency navigation users with information necessary to correct their measurements for ionospheric delay effects. These ionospheric corrections are computed by the Central Processing Facility-Processing Set (CPF-PS) and take the form of vertical ionospheric delays (Grid Ionospheric Vertical Delay, GIVD), in meters at L1 frequency, at a set of ionospheric grid points (IGPs), located at an altitude of 350 km over WGS 84 ellipsoid, covering the service area (European Civil Aviation Conference, ECAC). The estimates of the confidence of the ionospheric corrections are provided as grid ionospheric vertical error bounds (GIVE). The information is refreshed at least every 300 s. The user obtains the delay on his measurements, through interpolation of the GIVD values in the four IGPs surrounding his IPP (ionospheric pierce point or point in which the user line of sight crosses the ionosphere, assumed to be a thin shell at 350 km height) and applying an elevation mapping, to convert from vertical to slant, see Figure 2. EGNOS provides ionospheric information for a region which extends from 20 N to 70 N in latitude and from 40 E to 40 W in longitude, see Figure 3. The final set of grid points for which ionospheric information is provided each time will depend on the observability of the grid point (IGP) and conditions. IGPs are flagged as follows: Not Monitored, if there is no data from enough reference stations around the IGP; Do Not Use: in case integrity cannot be guaranteed for the IGP; and Use, otherwise. No ionospheric delay is provided for Not Monitored or Do Not Use IGPs. The following fundamental assumptions are considered in the estimation of the SBAS ionospheric vertical delays (GIVD) and integrity information (GIVE): (1) Ionospheric vertical total electron content can be Figure 3. EGNOS ionospheric grid point mask representation. ARENAS ET AL. LOW-LATITUDE IONOSPHERIC EFFECTS ON SBAS 605

4 accurately modeled as if the entire ionosphere were concentrated on a thin layer at 350 km. (2) Vertical ionospheric delay at points which are not part of the grid can be satisfactorily interpolated from the values at the grid points. (3) The mapping function from the vertical to the line of sight can be accurately modeled through a simple formula elevation dependent. SBAS allows the user to estimate protection levels, horizontal protection level (HPL), and vertical protection level (VPL). HPL and VPL are computed from Grid Ionospheric Vertical Error (GIVE) and the User Differential Range Error (UDRE, the integrity bound for ephemeris corrections) values. SBAS operations are only allowed while the HPL and VPL are lower than certain alert limits, HAL and VAL, respectively, defined for each civil aviation service level (e.g., en route and precision approach). Usually SBAS performances are measured in terms of the following four concepts: 1. Availability. Horizontal/vertical protection levels (HPL and VPL) not exceeding alarm limits (HAL and VAL) for the corresponding service level. Currently, EGNOS provide APV-I (approach with vertical guidance) service, what corresponds to a HAL of 40 m and a VAL of 50 m. 2. Accuracy. It is given as the difference between estimated and real user position. 3. Continuity. Service level is declared available for the whole operation. 4. Integrity. Navigation error is not exceeding the alarm limits. It has to be noted that SBAS give priority to integrity (confidence in the provided information), and in case of conflict, availability and accuracy of SBAS services may be degraded. The goal is to broadcast GIVE values the lower the better so that their contribution to protection levels makes the user solution available while integrity is always preserved. 3. Low-Latitude Ionospheric Effects on SBAS As mentioned above, there is a large dependency on ionospheric effects with latitude. In this sense, three main regions can be considered: (1) the midlatitude region; (2) the low-latitude region, including the equatorial and equatorial crest regions; and (3) the auroral and polar caps. The main ionospheric features vary depending on the region of the Earth, and therefore, the impact of the ionosphere on GNSS (and in particular on SBAS) consequently varies depending on the location considered. At low latitudes, the ionospheric activity can become a significant problem on GNSS. The main ionospheric problems associated to the low-latitude region are the presence of the equatorial crest or anomaly, the occurrence of scintillations, and the existence of ionospheric bubbles provoking depletions. Moreover, quite frequently these effects occur simultaneously. 1. The presence of the equatorial crest complicates the ionospheric modeling for systems like SBAS, due to the existence of great horizontal and temporal gradients [Cueto et al., 2013]. TEC gradients also result in large mapping function errors, which have been reported to reach up to 10 m in the vertical domain [Hoque and Jakowski [2013]. The main effects on the SBAS of those gradients are as follows: (1) Increment of the internal modeling errors, (2) rejection of some data due to internal checks, (3) increment in the error of the interpolations performed at system and user level, and (4) nonvalidity of fundamental assumption. From a high level point of view and in terms of impact on SBAS, the equatorial anomaly becomes a limitation affecting mainly availability and integrity. 2. The presence of small scale electron density irregularities in the ionosphere F region may cause scintillations on the received signals. Ionospheric scintillations produce rapid variations in signal amplitude and phase. The amplitude of fading can be so severe that the receiver has to continuously reacquire the signal, which compromises the tracking capabilities of GNSS receivers [Hlubek et al., 2014]. The region of strongest effects is around 15 either side of geomagnetic equator, typically in the post sunset period. Significant effects can also be observed in the auroral and polar region. Scintillation phenomenon has mainly a local effect on GNSS data. This implies that ionospheric information provided by SBAS does not include specific information on scintillation. The main effects on SBAS of ionospheric scintillations are as follows: (1) Increment in the input data noise, at system, and user level, which could degrade the accuracy of the corrections broadcast, (2) loss of input data to the SBAS processing facilities, and (3) loss of SBAS corrections, due to loss of GEO signal to the users ARENAS ET AL. LOW-LATITUDE IONOSPHERIC EFFECTS ON SBAS 606

5 From a high level point of view and in terms of impact on SBAS, the presence of scintillations becomes a limitation affecting mainly accuracy and availability/continuity. It is important to note that this phenomenon will also affect future dual-frequency SBAS. 3. The presence of ionospheric bubbles or depletions, which produces severe radio signal disruptions when crossing them [Cueto et al., 2013]. Plasma depletions (or bubbles) are strong reductions in the Figure 4. Analysis tools. ionospheric F region plasma density due to the appearance of Rayleigh-Taylor instability in the postsunset, producing severe decreases in the ionospheric delay of radio signals crossing them. Most of the plasma depletions are confined to the equatorial region. Depending on the size of the depletions, they can be local phenomena, not observed by the SBAS reference stations but affecting SBAS users or vice versa. The main effects of those depletions on the SBAS are as follows: (1) Rejection of affected data, (2) increment of all modeling and estimation errors, (3) degraded user position accuracy, and (4) increment of integrity risk. From a high level point of view and in terms of impact on SBAS, the depletions become a threat affecting mainly availability and integrity. In general, the main low-latitude ionospheric effect is the nonvalidity of the main ionospheric assumptions used in SBAS, which increment all modeling errors and estimations and in the worst case could cause an increment of integrity risk. 4. EGNOS Data To analyze the degradation of the EGNOS performances when low-latitude ionospheric phenomena are present, one EGNOS real scenario from year 2014 for each one of the main three ionospheric problems associated with the low-latitude region has been chosen: (1) Signal in Space (SIS) from 25 February 2014, affected by great horizontal and temporal gradients; (2) SIS February 2014, impacted by the presence of scintillations; and (3) SIS 06 November 2014, affected by the existence of ionospheric bubbles provoking depletions. Two types of EGNOS data have been used to perform the analyses of the selected days: (1) Raw GPS and navigation data as collected by the EGNOS monitoring reference network (RIMS-A and NLES), coming from the EGNOS Data Access Service (EDAS) service, which provides the data collected and generated by the EGNOS infrastructure, and (2) EGNOS messages, i.e., the augmentation messages broadcast by EGNOS, coming from the EGNOS Message Server (EMS). 5. Analysis tools Several tools have been used to perform the different analyses, as depicted in Figure 4: 1. magicsbas. It is a state-of-the-art, multiconstellation, SBAS test bed developed by GMV to offer SBAS regional differential corrections and nonsafety critical integrity augmentation to any interested region. The magicsbas product suite also includes a postprocessing version, which is the ideal tool to support SBAS engineering and feasibility studies where the user expects fast executions. This version can process prestored scenarios to perform trade-offs, architecture optimization, and fine tuning of the algorithms. In this work, the postprocessing version of magicsbas has been used to assess the impact of low-latitude ionospheric effects on EGNOS processing algorithms (in particular, measurement preprocessing and corrections computation). 2. ECLAYR. It is a specialized engineering tool for detailed performance assessment of SBAS and for verifying the level of compliance with respect to predefined requirements (e.g., GNSS-Standards and Recommended Practices (SARPs, ICAO)). The tool automatically collects and processes SBAS and reference real data and generates comprehensive performance assessment reports. This tool is currently being used by EGNOS Service ARENAS ET AL. LOW-LATITUDE IONOSPHERIC EFFECTS ON SBAS 607

6 Figure 5. VTEC estimated with EGNOS algorithms at 13:30 (quiet) and 19:30 (perturbed) on 25 February Provider to generate daily and monthly performance reports. In this case, this tool has been used to analyze EGNOS performances on the whole service area. 3. magicgemini. It is a state-of-the-art, operational GNSS performance analysis and monitoring tool specifically designed to meet the needs of air navigation service providers and airspace users who need to evaluate the performance of GNSS and their augmentations. Here we have taken advantage of magicgemini capabilities to assess EGNOS user performances at the locations of some IGS receivers [Dow et al., 2009] within the service area. Further information about these tools can be found in [Caro et al., 2013]. ARENAS ET AL. LOW-LATITUDE IONOSPHERIC EFFECTS ON SBAS 608

7 Radio Science Figure 6. EGNOS APV-1 Availability Map on 25 February 2014 for GEO PRN 120. For the accuracy and integrity evaluation a true reference is needed. In this case, Ionosphere Map Exchange Format (IONEX) files (TEC values) obtained from IGS are used as a reference to compare the estimated GIVD values against them. 6. Ionospheric Gradients: Spatial and Temporal Notable TEC temporal and spatial gradients are observed on 25 February 2014 between 17:00 h and 20:00 h. Figure 5 presents the vertical ionospheric delay estimated by the EGNOS CPF-PS at 13:30 (quiet period) and at 19:30 h (perturbed period), showing in the later hour high TEC gradients in the area of the Canary Islands and the northwest of Africa. This great horizontal and temporal TEC variability at low latitudes affects the southernmost EGNOS performances due to the nonvalidity of fundamental ARENAS ET AL. LOW-LATITUDE IONOSPHERIC EFFECTS ON SBAS 609

8 Figure 7. EGNOS Integrity Map on 25 February 2014 for GEO PRN 120. assumptions for the estimation of the ionospheric vertical delays (GIVD) and integrity information (GIVE). The increase of the error in the interpolations performed at system level causes higher errors in the ionospheric delay estimation (GIVD) at IGPs and finally degradation of the accuracy performances in the southern ECAC. As a consequence, the ionospheric bounds (GIVE) are increased to cope with these higher errors in the ionosphere delay estimation (GIVD) that together with the rejection of some data due to some internal checks finally causes an important degradation of the APV-1 availability performance in the south and center of the ECAC (see Figure 6). While during the quiet period (13:00 14:00) the system is available for APV-I (>99.9%, red area); for the perturbed period (19:00 20:00), there is availability (red area) only for the northern part of the ECAC. In spite of the APV-1 service degradation, the integrity performance of the EGNOS is ensured for all monitored locations, as can be seen in Figure 7, where the area in which HPE < HPL and VPE < VPL 100% of the time is presented. ARENAS ET AL. LOW-LATITUDE IONOSPHERIC EFFECTS ON SBAS 610

9 Figure 8. S4 and σϕ for mste01 station on 27 February 2014 (courtesy from the Ionosphere Monitoring and Prediction Center (IMPC) of DLR [Berdermann et al., 2014]). 7. Scintillations The EGNOS service area covers a wide range of latitudes, from 20 N to 70 N. In periods of high ionospheric activity, the presence of ionospheric scintillations is observed for the northernmost and southernmost regions of this area. At high latitudes, this phenomenon is often correlated with the occurrence of aurora borealis. On the other hand, at low latitudes, scintillation effects are commonly seen up to 20 N and sometimes up to 30 N. In particular, the Canary Islands are located between 25 N and 30 N; therefore, they are sometimes impacted by this effect. In particular, for this area, the effect of scintillations is observed from 21 h to 3 h (UTC). From the SBAS point of view, we can separate the effects on the system itself (through the dualfrequency receivers at the reference stations) from the ones on a particularsbasuser(single-frequency receiver). Ionospheric scintillation can be characterized by two signal properties: S4 (amplitude scintillation index) and σϕ (phase scintillation index). Both magnitudes are higher when scintillation is stronger. Figure 8 shows the values of S4 and σϕ for an external receiver mste01 (located at N, E, in the Canary Islands) on 27 February Theincreaseofσϕ and, mainly, S4 indices during the first and last hours of the day is evident; the latter reaching values close to 1.0 between 21 h and 2 h, approximately. Figure 9. Number of GPS satellites tracked by NOUA station minus satellites geometrically in view on 26 February 2014 and 27 February Intense ionospheric scintillation results in several losses of lock for GPS receivers, including SBAS ground stations. As an example, Figure 9 depicts the difference between the number of tracked GPS satellites minus the number of satellites which should be geometrically in view for the EGNOS NOUA station (located at N, E, in Mauritania). This difference represents the number of satellites for which the station has lost signal lock. ARENAS ET AL. LOW-LATITUDE IONOSPHERIC EFFECTS ON SBAS 611

10 Figure 10. Smoothed vtec measurements from NOUA station on 26 February 2014 and 27 February Apart from the loss of raw measurements, the effect of scintillation is also evident on smoothed TEC measurements, visible as an increase of measurement noise and instability. Figure 10 shows the values of vtec computed by the EGNOS algorithms for the NOUA station during this period. The loss of raw measurements has an immediate effect on the SBAS algorithms that compute ionospheric corrections. In particular, having fewer observables to compute the GIVD results in less accurate corrections. On the other hand, the confidence interval of these corrections also gets larger, which implies a higher GIVE value. In the extreme case, the loss of measurements could be so important that some points of the ionospheric grid (IGPs) are declared Not Monitored by the SBAS, and no corrections are provided for them. Both events, high GIVE values and not monitored IGPs, result in losses of availability and continuity of the SBAS service. To illustrate this effect, Figure 11 shows the availability map for APV-I service in EGNOS on 27 February 2014 for two periods: from 0 h to 2 h (UTC), affected by strong scintillations in the southwest, and for 6 h to 10 h, a nominal period with low ionospheric activity. As observed, there is a clear reduction of the availability of APV-I service (red area) over the Canary Islands and North Africa for the period with scintillations (0:00 2:00 h). The availability maps shown above are computed assuming that the user obtains pseudorange measurements from all the GPS satellites that are monitored by the SBAS. However, at the user level, the same impact as on reference stations can be expected: loss of tracking and increased noise levels. This means that the accuracy achieved by the user when computing its position will be worse than the one predicted assuming that it tracks all monitored satellites. Moreover, this will also imply higher protection levels, meaning that the availability of SBAS services will also be penalized for a user who is losing measurements. Finally, it must be noted that ionospheric scintillation does not only affect GPS signals, but also the ones sent by the SBAS GEO satellites. The loss of GEO signals may eventually result in a lack of SBAS corrections, stopping completely the usage of SBAS services by the user. To mitigate this effect, SBAS implement GEO satellite redundancy to broadcast the information to the users. 8. Depletions In this section, the impact of TEC depletions, caused by ionospheric plasma bubbles, in the SBAS is assessed. Ionospheric depletions are local effects, and therefore, the main problem may occur when the depletion is seen by the SBAS and not by the SBAS user or vice versa. If the system sees the depletion, the broadcast ionospheric information will reproduce this phenomenon. If the user data are also affected by the same depletion, the ionospheric information both broadcast and observed by the user will be coherent. If the user is not affected by the depletion, then the broadcast and user ionospheric information will be different causing degradation of accuracy, and in the worst case, an integrity failure will occur if the computed GIVE is not large enough to bound the difference. If the system does not see the depletion, this could be due to two reasons: either the system input data are not affected (the reference station lines of sight do not cross the ionospheric bubble) or the raw input ARENAS ET AL. LOW-LATITUDE IONOSPHERIC EFFECTS ON SBAS 612

11 Figure 11. EGNOS APV-I Availability Map on 26 February 2014 and 27 February 2014 for GEO PRN 120. data are rejected inside the system because some of the internal barriers consider this as not reliable data. In this case, the broadcast ionospheric information will not reproduce the depletion. If the user is also not affected by the depletion, their information will be coherent. But if the user data are affected by the depletion, then again there will be a degradation of accuracy and, in the worst case, an integrity failure if the computed GIVE is not large enough to bound thedifferencebetweenthebroadcastanduser ionospheric information. It is noted that as a general barrier, the ionospheric vertical error bounds (GIVE) are computed in such a way that they depend on the number of valid input measurements and their geometry. Therefore, GIVE values are increased to protect users from hidden structures in case input data are rejected or not available. As an example the effect of an ionospheric depletion detected using the IBS application [Magdaleno et al., 2012] is presented here below. ARENAS ET AL. LOW-LATITUDE IONOSPHERIC EFFECTS ON SBAS 613

12 Figure 12. Depletion found at MAS1 on 06 November 2014 using IBS (local time). Figure 12 shows the depletion found by IBS on 6 November 2014 at the MAS1 IGS station (Maspalomas, Canary Islands). The estimated depth of the depletion is near 30 TECU (total electron content unit, 1 TECU = el m 2, more than 4.5 m at L1). The analysis of lines of sight between the EGNOS RIMS CNR, very close to Maspalomas, and PRN5 and PRN26 (illustrated as L2-L1 raw pseudoranges in Figures 13 (top) and 13 (bottom), respectively) shows that both lines of sight present TEC depletions. So it is confirmed that the system input data are affected by the depletion. Figure 14 shows the position versus time of ionospheric pierce points for both satellites used in the ionospheric correction estimation. Additionally, the Slant TEC (STEC) values of both lines of sight are provided in color scale. The period when the depletion appears is marked by a red circle in both lines of sight. It is noted that in the case of line of sight CNR-PRN26, most of the data affected by the depletion were rejected by the internal barriers and were not used by the CPF-PS ionospheric algorithm. In the case of line of sight CNR-PRN5, some data are used for ionospheric corrections estimation (noted that Figure 13 shows data gaps during the first and last epochs of the depletion). Finally, Figure 14 also provides the location of the four closest IGPs. The ionospheric slant delay (STEC) computed for the user receiver at Maspalomas (MAS1) using the ionospheric information broadcast by the SBAS for these four IGPs is presented in Figure 15 together with the ionospheric slant delay computed using dual-frequency data from MAS1 IGS receiver. As it can be seen, the SBAS does not reproduce the ionospheric depletion due to the rejection of the affected input data. Therefore, there is a large degradation on accuracy. In order to check the impact on integrity, Figure 16 shows the horizontal and vertical errors against the corresponding protection level. In both cases, the protection levels are quite above the errors, preserving integrity with a large margin. ARENAS ET AL. LOW-LATITUDE IONOSPHERIC EFFECTS ON SBAS 614

13 Figure 13. TEC depletions found in PRN5 and PRN26 in RIMS CNR. Figure 14. TEC depletions (red circle) and the closest IGPs (black dots). ARENAS ET AL. LOW-LATITUDE IONOSPHERIC EFFECTS ON SBAS 615

14 Figure 15. STEC computed from SBAS ionospheric information (green) and STEC computed from dual frequency data (red) at Maspalomas receiver (MAS1). It is observed that the horizontal range error increases from 20.6 to 21.6 h, which corresponds with the depletion time. During this period, the HPL covers well the error. However, during some epochs the protection level is too high (>40 m) which implies that APV-I service is not available in this area. The vertical range error also increases from 20.9 to 21.2 h. As for horizontal deviation, the vertical error is well covered by the protection level. However, it is also too high (>50 m) during some periods, doing APV-I service not available. Therefore, integrity is always kept ( position error < protection level for % of cases) although accuracy and availability are penalized. Figure 16. User (top) horizontal and (bottom) vertical position errors (red) and protection levels (green). 9. Conclusions This paper has presented the main low-latitude ionospheric effects impacting SBAS: temporal and spatial gradients, scintillations, and depletions. As a case study, the impact on EGNOS, the European SBAS, has been shown. The effect of low-latitude ionospheric effects is clear in EGNOS, the affected area being the Canary Islands and North Africa and, in general, the EGNOS service area below 30 N. In extreme cases, the impact is propagated to the Mediterranean region. EGNOS, and the SBAS in general, are designed to preserve integrity in all cases, normally penalizing accuracy and availability. Therefore, the main consequences of the low-latitude ionospheric effects, as expected, have been loss of SBAS service availability and accuracy degradation. However, integrity of SBAS information is preserved even in case of high ionospheric perturbations. ARENAS ET AL. LOW-LATITUDE IONOSPHERIC EFFECTS ON SBAS 616

15 Acronyms and Abbreviations Acknowledgments The authors want to express their gratitude to the IGS service, the EGNOS Service Provider, and the European Space Agency (ESA) for making all data and products used for these analyses available. IGS data and products (IONEX and RINEX files) are publicly available at compindex.html. EDAS data are available upon registration at Finally, EGNOS Message Server can be accessed at ftp://ems.estec.esa.int/pub APV approach with vertical guidance CCF Central Control Facility CPF Central Processing Facility CPF-PS Central Processing Facility Processing Set DLR Deutsches Zentrum für Luft- und Raumfahrt (German Aerospace Center) ECAC European Civil Aviation Conference EDAS EGNOS Data Access Service EGNOS European Geostationary Navigation Overlay Service EMS EGNOS Message Server ESA European Space Agency EWAN EGNOS Wide Area Network GAGAN GPS Aided Geo Augmented Navigation GEO Geostationary Satellite GIVD Grid Ionospheric Vertical Delay GIVE Grid Ionospheric Vertical Error GLONASS Global Navigation Satellite System (Russia) GNSS Global Navigation Satellite System GPS Global Positioning System HAL horizontal alert limit HPE horizontal position error HPL horizontal protection level IBS Ionospheric Bubble Seeker IFB interfrequency bias IGP ionospheric grid point IGS International GNSS Service IMPC Ionosphere Monitoring and Prediction Center IONEX Ionosphere Map Exchange Format IPP ionospheric pierce point MSAS Multifunctional Satellite Augmentation System NLES Navigation Land Earth Stations RIMS Ranging Integrity Monitoring Stations SACCSA Sistema de Aumentación para el Caribe, Centro y Sudamérica (Augmentation System for the Caribbean, Central and South America) SARPs Standards and Recommended Practices (ICAO) SBAS Satellite-Based Augmentation System SDCM System for Differential Corrections and Monitoring SIS signal in space STEC slant TEC TEC total electron content TECU TEC unit (= el/m 2 ) UDRE User Differential Range Error UTC coordinated universal time VAL vertical alert limit VPE vertical position error VPL vertical protection level VTEC vertical TEC WAAS Wide Area Augmentation System References Berdermann, J., N. Jakowski, M. M. Hoque, K. D. Missling, M. Kriegel, C. Borries, V. Wilken, H. Barkmann, and M. Tegler (2014), Ionospheric Monitoring and Prediction Center (IMPC), Proceedings ION GNSS+ 2014, Tampa, Florida. ARENAS ET AL. LOW-LATITUDE IONOSPHERIC EFFECTS ON SBAS 617

16 Caro, J., J. Barrios, V. M. Esteban, V. Izquierdo, A. Madrazo, M. Odriozola, J. Ostolaza, and J. Simon (2013), Enhanced SBAS with satellite dynamic mask and precise orbit and clock corrections, Proceedings ION GNSS+ 2013, Nashville, Tenn. Cueto, M., S. Magdaleno, A. Cezon, and E. Sardon (2013), Characterization of equatorial ionospheric features on the verge of the next solar cycle maximum, Proceedings ION GNSS+ 2013, Nashville, Tenn. Dow, J. M., R. E. Neilan, and C. Rizos (2009), The International GNSS Service in a changing landscape of Global Navigation Satellite Systems, J. Geod., 83, , doi: /s Hlubek, N., J. Berdermann, V. Wilken, S. Gewies, N. Jakowski, M. Wassaie, and B. Damtie (2014), Scintillations of the GPS, GLONASS, and Galileo signals at equatorial latitude, J. Space Weather Space Clim., 4, A22, doi: /swsc/ Hoque, M. M., and N. Jakowski (2013), Mitigation of ionospheric mapping function error, Proceedings ION GNSS+ 2013, Nashville, Tenn. Magdaleno, S., M. Herraiz, and S. Radicella (2012), Ionospheric Bubble Seeker: A Java application to detect and characterize ionospheric plasma depletion from GPS data, IEEE Trans. Geosci. Rem. Sens., 50(5), , doi: /tgrs Ventura-Traveset, J., and D. Flament (Eds.) (2006), EGNOS. The European Geostationary Navigation Overlay System A cornerstone to Galileo (ESA SP-1303), ESA Publications Division, Noordwijk, Netherlands. ARENAS ET AL. LOW-LATITUDE IONOSPHERIC EFFECTS ON SBAS 618