Investigation of Scintillation Characteristics for High Latitude Phenomena

Transcription

1 Investigation of Scintillation Characteristics for High Latitude Phenomena S. Skone, F. Man, F. Ghafoori and R. Tiwari Department of Geomatics Engineering, Schulich School of Engineering, University of Calgary, Canada; BIOGRAPHY Susan Skone, Ph.D., is an Associate Professor in Geomatics Engineering, Schulich School of Engineering, at the University of Calgary. She has a background in space physics and conducts research in ionospheric and tropospheric effects on GPS. She has developed software for mitigation of atmospheric effects and is currently chair of the Canadian Navigation Society. Man Feng is an MSc Student in Geomatics Engineering at the University of Calgary. She received her first Masters degree (2007) in Space Physics at the Peking University, China. She received her B.S. (2004) in Computer Science and Technology at the Tianjin University, China. Fatemeh Ghafoori is a Ph.D. candidate in Geomatics Engineering at the University of Calgary. She received her B.Sc. (2002) in Electrical and Communication Engineering at the Ferdowsi University, Iran, and M.Sc. (2007) in Digital Communication Systems and Technology at the Chalmers University of Technology, Sweden. Rajesh Tiwari is a Ph.D. Candidate in Geomatics Engineering at the University of Calgary. He received his M.Sc (2004) in Physics at Barkatullah University, Bhopal, India. He has been a Research Fellow in the XXVI Indian Scientific Expedition to Antarctica under the project "Space Weather Program at Antarctica" during summer ABSTRACT High latitude irregularities in electron density have dimensions of meters to kilometers and can cause severe scintillation effects for GNSS signals traveling through this region of the ionosphere. Extreme scintillation effects, and associated degradation or loss of navigation capabilities, have been observed in North America for severe events. Auroral scintillation has been studied for decades, with typically strong phase scintillation and weak amplitude scintillation observed at high latitudes due to the precipitating electrons. Auroral scintillation is most commonly observed at nightside local times. Since 2003 the University of Calgary has operated a number of specialized GPS receivers in Canada for ionosphere monitoring - as part of the CANGIM (CANadian GPS Network for Ionosphere Monitoring). These receivers are modified dual-frequency survey-grade NovAtel Euro4 receivers, with specialized firmware capable of deriving phase and amplitude scintillation information. These data have been collected at three sites in western Canada over the past five years. These sites are located at similar longitudes and over a range of latitudes spanning the sub-auroral region into the polar cap. Observations from the three stations allow latitude profiling of the spatial extent of scintillation effects. In this paper, we investigate nightside auroral scintillation using the extensive CANGIM GPS data set. Scintillation events throughout the years are identified. In order to better understand the nature of these effects, the CANGIM data are also augmented with observations from spaceborne GPS receivers. GPS observations from the CHAMP satellite are used to infer the presence of ionospheric irregularities at various altitudes. For an occulting GPS satellite, the line-of-sight to the low-earth orbiter (LEO) successively passes though horizontal layers at different altitudes. Times series of such observations can therefore reflect the presence of smallscale electron density variations in a given height range if rapid random phase variations are observed. When considered in conjunction with the ground-based scintillation measurements in Canada, local LEO observations are therefore used to determine the vertical extent of irregularities. The nature of physical processes leading to formation of these irregularities (and associated scintillation effects) is then determined. Results are applicable to the development of physics-based scintillation simulations - which are used to assess GNSS receiver tracking performance.

2 INTRODUCTION The ionosphere is a dispersive medium, in which RF signals are refracted by an amount dependent on the given signal frequency and ionospheric electron density. In regions of small-scale irregularities in electron density, rapid random phase variations can be produced by phase irregularities in the emerging wavefront (Figure 1). These are referred to as phase scintillations, and such effects can lead to cycle slips and corrupt phase observations [Datta- Barua et al, 2003]. Phase scintillation is quantified by σ ϕ which is the standard deviation of high-frequency detrended phase observations over an interval of up to 60 seconds. Diffraction of the signal (interference across the wavefront see Figure 1) also leads to variations in signal amplitude referred to as amplitude scintillation (or amplitude fading, for degradations in signal strength). Strong amplitude fades can result in loss of signal tracking and, in extreme cases, loss of navigation capabilities entirely [Skone and Knudsen, 2000]. Scintillation effects are strongest in the equatorial (±10 geomagnetic latitude), auroral (65-75 geomagnetic latitude), and polar cap (> 75 geomagnetic latitude) regions [Pi et al., 2002]. Wave propagation Wavefront In order to assess the impact of scintillations on receiver tracking performance for existing and future GNSS signals and services, several researchers have developed simulation capabilities. Hegarty et al. [2001] and Conker et al. [2003] have simulated the signal phase and amplitude variations associated with scintillation using statistical models and then formulated equations for predicted phase errors based on amplitude and phase scintillation parameters. Similar models were used in Skone et al. [2005] to simulate signal characteristics including the effects of scintillation, with signals then processed in a software receiver to test tracking loop designs. Humphreys et al. [2005] evaluated GPS receiver tracking performance in the presence of scintillation using single frequency wideband data collected in the field and then post-processed using a software receiver. Such approaches are limited. For example, the random errors generated from statistical models do not allow for correlation of effects on different frequencies. The ionospheric index of refraction (which is frequency dependent) varies slightly for GPS L1 versus L2 frequencies. Propagation paths of the L1 and L2 signals are therefore separated by tens of meters in the E-and F- region ionosphere. In the case of small-scale irregularities (meter-level scale size) the signals may experience different phase and amplitude variations. In realistic simulations the level of correlation on multiple frequencies must be represented in the simulation method. This is especially important for simulations developed to assess multi-frequency GNSS receiver capabilities (e.g. modernized GPS and Galileo). f 1 f 2 Emerging wavefront Figure 1. GNSS signal propagating through region of ionospheric electron density irregularities. Auroral irregularities result from an acceleration of energetic electrons along terrestrial magnetic field lines into the high latitude ionosphere. These electrons precipitate at approximately 110 km altitude resulting in enhanced ionization, optical and UV emissions (aurora borealis/australis) and formation of irregularities and sporadic E. This phenomenon characterizes the auroral substorm, where associated irregularities in electron density (at altitudes of km) lead to scintillation effects [cf. Aarons, 1982]. The auroral oval can expand several degrees equatorward during such events (i.e. over Northern Europe, Northern United States). rx Figure 2. GNSS signal propagation for two different frequencies through one-dimensional phase screen. An alternate approach has been developed by Dyrud et al. [2005] in which the physical ionospheric irregularity distribution is simulated. The model can then allow for characterization of the signal phase and amplitude fluctuations by ray racing the signal propagation through

3 the phase screen (Figure 2). This approach can be applied for multiple frequencies simultaneously, allowing the correct error correlation between signals. In this approach, however, the electron density distribution is generally limited to one or two dimensions with the irregularities concentrated at a single altitude. In reality, the vertical distribution of electron density irregularities can span several hundred kilometers. As the signals are refracted and diffracted at each layer, the additive effects must be considered for a more realistic simulation model. In this paper we evaluate the vertical electron density distributions at auroral locations during scintillation events. In this manner we assess the validity of a single layer phase screen assumption and determine the necessity for more complex models. DATA SETS In this paper two data sets are investigated to determine the vertical distribution of electron density irregularities in the high latitude auroral region: the CANGIM groundbased data and the CHAMP satellite radio occultation database. Both are described here. (TEC), absolute TEC values, and WAAS messages. The firmware version (scintw) was developed by A.J. Systems [Van Dierendonck et al., 1996]. The CANGIM network does not currently operate in realtime. Data are downloaded at periodic intervals and archived at a central processing facility at University of Calgary. After several hardware upgrades, the three installed stations have been operating reliably since May The relationship between high-latitude phase and amplitude scintillations is shown in Figure 4, where scintillation parameters are plotted for all satellites in view from the Calgary CANGIM site for two severe ionospheric storm events combined. Figure 4 demonstrates little correlation between high-latitude phase and amplitude scintillations. For high phase scintillations, S 4 values vary from less than 0.05 (low) to 0.3 (moderate). CANGIM Observations The Canadian GPS Network for Ionosphere Monitoring (CANGIM) was developed for the purpose of ionosphere monitoring for GPS applications. The CANGIM currently includes three stations in Western Canada: Calgary (51.08 N, W), Athabasca (54.72 N, W) and Yellowknife (62.48 N, W) see Figure 3. These stations allow latitude profiling of both the auroral and sub-auroral regions. Figure 4. The relationship between amplitude (S 4 ) and phase scintillations observed at CANGIM Site Calgary. CHAMP Observations Figure 3. Locations of CANGIM sites for high-latitude scintillation monitoring. The CANGIM sites are equipped with NovAtel Modulated Precision Clock (MPC) receivers and NovAtel 600 antennas. These units contain dual-frequency Euro4 cards with an internal integrated PC and precise oscillator. A user-interface command structure allows direct access via modem or internet connection. The CANGIM receivers have specialized firmware which provides scintillation parameters extracted from 50 Hz L1 phase observations, in addition to raw GPS code and phase observations, rates of change of total electron content The CHAMP micro-satellite was launched July 15, 2000 with a planned mission lifetime of five years. One purpose of this mission was to obtain space-based GPS observations from a low-earth orbiter (LEO) for atmospheric and ionospheric soundings using radio occultation methods [Wickert et al., 2001]. In cases of an occulting GPS satellite where the LEO-GPS line-ofsight passes through the Earth s atmosphere it is possible to observe variations in the ionospheric range delay for successive horizontal layers. Using dual frequency observations to compute the TEC at each epoch, inversion techniques are then applied to compute values of the ionospheric index of refraction and electron density profiles. Methods for ionospheric electron density retrieval are described in Hajj and Romans [1998] and Schreiner et al. [1999]. Figure 5 shows a sample electron density profile for quiet ionospheric conditions near the

4 CANGIM sites. Note that the peak electron density is at 250 km altitude. Figure 5. Electron density profile derived from CHAMP observations. Data for the CHAMP mission are publicly available through the Information System and Data Center at GFZ Potsdam. Various levels of data products can be selected depending on access permission; these range from raw GPS observations to the final electron density profiles. For the purposes of this study raw observations were obtained (1 Hz data rate) for processing using University of Calgary in-house software. The University of Calgary electron density retrievals were compared with CHAMP final products to assess validity of the processing methods. Agreements were consistently within a few percent under a variety of ionospheric conditions. The CHAMP orbit is almost circular and near polar. The initial satellite altitude was approximately 454 km, with several hundred occultation events observed each day. Each event has a duration of 2-3 minutes. Given the global CHAMP coverage, only a small subset of the occultation profiles were located near the CANGIM sites in Canada. Only those specific occultation events located within the CANGIM coverage area are considered in this study. RESULTS AND ANALYSIS An extensive CANGIM database ( ) was analysed to identify cases in which significant scintillation effects were present near at least one of the three GPS reference sites. Only those events occurring during nightside hours local time were considered, in order to ensure that scintillation effects could primarily be attributed to auroral substorm activity. Scintillations were assumed to be present when the phase scintillation index exceeded a predefined threshold (a threshold at which degraded GPS receiver tracking performance is typically observed). For such textbook events, it is expected that increased values of electron density will be observed at E- region altitudes with electron density irregularities extending to higher altitudes. Over 30 events were identified in this manner. A search was then conducted of the CHAMP database to find corresponding radio occultation events. Approximately ten events were selected in which the CHAMP electron density profiles mapped to the CANGIM region. The CHAMP raw GPS data were then processed to derive the electron density profiles with a spatial resolution of a few kilometers. Note that the spatial resolution is limited by the data rate of 1 Hz. Higher frequency data were only available at altitudes below 150 km. The raw data time series were then filtered to isolate only the high frequency components associated with smaller scale sizes. These filtered data were then mapped to corresponding altitudes using in-house software. In this manner the electron density variations with scale sizes of a few kilometers are represented as a function of altitude. From such analysis the vertical distribution of electron density irregularities can then be determined. Results for three representative events are given here. October 1, 2006 During this event larger phase scintillation values were observed at all three CANGIM sites at approximately 0600 UT. This corresponds to 2200 local time and the scintillation effects are attributed to nightside auroral substorm activity. Figure 6 shows the temporal variation of phase scintillation indices at Athabasca. Athabasca is generally a sub-auroral site but for strong events such as this the auroral oval expands southward. Figure 7 shows the ionospheric pierce points (where the satellite-receiver line-of-sights pierce an ionospheric shell at 350 km altitude) for satellite observations which exceed the phase scintillation detection threshold. The tangent point of the radio occultation is plotted in yellow for the duration of the CHAMP observations. In this case the auroral region has expanded considerably, with strong scintillations observed near Calgary. The CHAMP tangent points are well within the region of scintillation effects. Figure 8 shows the electron density profile derived from corresponding CHAMP observations. Compared to Figure 5 (quiet conditions) a prominent feature is the peak in electron density at 110 km altitude. This is consistent with expectations, given that aurora are characterized by E- region precipitation of energetic electrons resulting in ionization and increased electrons at 110 km altitude. It is also noted that there are fluctuations in the electron density profile extending to higher altitudes indicating the presence of smaller scale structures or irregularities in electron density.

5 Figure 9 shows the filtered electron density values, where only scale sizes of several kilometers or less are represented. It must be noted that such scale sizes are at the larger limit of dimensions contributing to scintillation effects. However, there are electron density irregularities of 15 percent at E-region altitudes. At higher altitudes (up to 200 km) the relative electron density variations are only 5 percent. Figure 8. Electron density profile at 0600 UT on 1 October 2006 derived from CHAMP observations. Figure 6. Phase scintillation indices for all satellites in view at Athabasca on 1 October 2006 (elevation cutoff angle of 30 degrees). Figure 9. Electron density variations for scale sizes less than 3 km at 0600 UT on 1 October 2006 derived from CHAMP observations. June 23, 2005 Figure 7. Ionospheric pierce points for CANGIM satellite-receiver lines-of-sight for high phase scintillation values (green, blue and magenta) and radio occultation tangent points (yellow). CANGIM site locations are shown as red stars. During this event larger phase scintillation values were observed at Calgary and Yellowknife at 1210 UT which corresponds to 0410 local time. Figure 10 shows the temporal variation in phase scintillation indices measured at Calgary. The time period studied coincides with the end of several hours of enhanced ionospheric activity. Note that Athabasca data were not available for this event. Figure 11 shows the spatial distribution of scintillation effects and the radio occultation tangent points. Similar to the previous event, the occultation observations are within the identified area of strong phase scintillation.

6 Figure 10. Phase scintillation indices for all satellites in view at Calgary on 23 June 2005 (elevation cutoff angle of 30 degrees). Figure 12. Electron density profile at 1210 UT on 23 June 2005 derived from CHAMP observations. Figure 11. Ionospheric pierce points for CANGIM satellite-receiver lines-of-sight for high phase scintillation values (green, blue and magenta) and radio occultation tangent points (yellow). CANGIM site locations are shown as red stars. Figure 12 shows the electron density profile derived from corresponding CHAMP observations. A peak in electron density is observed at 110 km altitude. Similar to the October 1, 2006 event this E-region peak is consistent with auroral substorm activity. A second peak is observed at 220 km altitude with high-frequency fluctuations extending to the higher altitudes. Figure 13 shows the filtered electron density values (scale sizes of several km or less). In this case the relative E-region (110 km altitude) electron density variations are 12 percent. At higher altitudes the relative electron density variations are 5-7 percent. During auroral substorm activity the irregularities extend up to altitudes of 250 km - with magnitudes lower than those in the E-region but which cannot be considered negligible. Figure 13. Electron density variations for scale sizes less than 3 km at 1210 UT on 23 June 2005 derived from CHAMP observations. January 7, 2005 High values of phase scintillation were observed at Yellowknife over a period of several hours for this event. These values are shown in Figure 14. At 1530 UT (0730 local time), phase indices are as high as 0.25 rad for some of the satellites in view from Yellowknife. This is classified as intense scintillation. Figure 15 shows the spatial distribution of the intense activity, with the CHAMP radio occultation tangent points in the vicinity of Yellowknife observations.

7 Figure 14. Phase scintillation indices for all satellites in view at Yellowknife on 7 January 2005 (elevation cutoff angle of 30 degrees). Figure 16. Electron density profile at 1530 UT on 7 January 2005 derived from CHAMP observations. Figure 15. Ionospheric pierce points for CANGIM satellite-receiver lines-of-sight for high phase scintillation values (green, blue and magenta) and radio occultation tangent points (yellow). CANGIM site locations are shown as red stars. Figure 16 shows the electron density profile derived from CHAMP observations. As for the previous two events, a strong E-region enhancement of electron density is observed (at 100 km altitude). Very large fluctuations in electron density extend up to altitudes of 300 km. This is verified in Figure 17 where the electron density irregularities for scale sizes of several meters are plotted. In this case, the E-region relative variations are percent and the higher altitude variations (altitudes of km) are percent. For this vertical profile, the full three-dimensional electron density distribution will contribute to scintillation of satellite signals as they propagate downward. A simple phase screen model for one fixed altitude would not realistically simulate the scintillation effects for this event. Figure 17. Electron density variations for scale sizes less than 3 km at 1530 UT on 7 January 2005 derived from CHAMP observations. This event took place early morning and cannot be considered a textbook nightside auroral event. There is a possibility in this case that the effects observed are not simply due to precipitating electrons in the E region but may also be due to polar cap effects. The polar cap is a region north of the auroral oval in which magnetic field lines are directly open to the solar wind. Charged particles originating from the Sun can enter this region and create electron density structures and patches associated with scintillation. In order to further investigate the nature of this event, three-dimensional electron density values were generated by G. Bust using the IDA3D model. This model assimilates total electron content observations from a global network of GPS ground reference stations and other observations, such as electron density profiles from radio occultations, as available. For this event, the model did not include the CHAMP electron density profile. Figure 18 shows a northern polar projection of the electron density at an altitude of 100 km. The enhanced

8 electron density approximately circling the pole is the auroral oval. The CHAMP radio occultation tangent points are plotted in magenta. It is observed that the tangent points track directly through the auroral region. Additionally, Figure 19 shows a slice of the electron density at 1530 UT (coinciding with the scintillation event) approximately along the CANGIM meridian. The largest electron density values occur near 62 deg geographic latitude and 110 km altitude. This region corresponds to the auroral oval. A latitudinally localized enhancement of electron density also extends upwards from the aurora at this location, to an altitude of approximately 300 km. While the model smoothes smallscale features, such that irregularities are not apparent in this plot, the region of enhanced electron density corresponds to the extent of electron density fluctuations in Figure 17. Figure 19. Electron densities at 1530 UT and -114 deg geographic longitude (G. Bust). CONCLUSION Three case study auroral scintillation events have been presented. In all cases the electron density irregularities extended to altitudes well above the E-region, spanning the range km. For moderate events the higher altitude relative variations were as low as five percent. For an intense event, however, relative irregularities of percent extended over several hundred kilometers of altitude. In developing scintillation models for simulation testing, a physics-based approach is advantageous in that correlations between signal ray paths can be included. In order to implement such an approach, however, it may be necessary to model the irregularities not as one simple phase screen but as multiple phase screens at various altitudes. REFERENCES Figure 18. Northern polar projection at 1530 UT of electron density values at 100 km altitude (G. Bust). Top (bottom) of the figure corresponds to local noon (midnight). Aarons, J., Global morphology of ionospheric scintillation, Proc. IEEE, 70, pp , Conker, R. S., M. B. El-Arini, C. J. Hegarty and T. Hsiao, Modeling the effects of ionospheric scintillation on GPS satellite-based augmentation system availability, Radio Science, Vol. 38, No. 1, Datta-Barua, S., P. H. Doherty and S. H. Delay, Ionospheric scintillation effects on single and dual frequency GPS Positioning, Proceedings of the ION GPS 2003, Portland, OR, September 9-12, pp , Dyrud, L., N. Bhatia, S. Ganguly and A. Jovancevic, Performance analysis of software based GPS receiver using a generic ionospheric scintillation model, Proceedings of the ION GNSS 2005, Fort Worth, TX, Sept , pp , 2005.

9 Hajj, G.A. and L.J. Romans, Ionosphere Electron Density Profiles Obtained With The GPS System: Results from the GPS/MET experiment, Radio Science, Vol. 33, No. 1, pp , January-February Hegarty, C., M.B. El-Arini, T. Kim, and S. Ericson, Scintillation modeling for GPS/Wide Area Augmentation System receivers, in Radio Science, Vol.36, No.5, pp , Humphreys, E. D., M. L. Psiaki, and P. K. Kinter, Jr., GPS Carrier Tracking Loop Performance in the Presence of Ionospheric Scintillation, in Proceedings of ION GPS/GNSS, Sept., Long beach CA, pp , Institute of Navigation, Pi, X., Boulat, B.M., Mannucci, A.J. and D.A. Stowers, Latitudinal characteristics of L-band ionospheric scintillation, Proceedings of the ION GPS 2002, Portland, OR, September 24-27, Schreiner, W. S., S.V. Sokolovskiy, C. Rocken, and D. C. Hunt, Analysis and validation of GPS/MET radio occultation data in the ionosphere, Radio Sci., 34(4), , Skone, S., and K. Knudsen, Impact of ionospheric scintillations on SBAS performance, Proceedings of the ION GPS 2000, Salt Lake City, UT, September, Skone, S., G. Lachapelle, D. Yao, W. Yu, and R. Watson, Investigating the Impact of Ionospheric Scintillation using a Software Receiver, in Proceedings of ION GPS/GNSS, Sept, Long Beach CA, pp , U.S. Institute of Navigation, Fairfax VA, Van Dierendonck, A. J., Hua, Q., Fenton, P., and J. Klobuchar, Commercial ionospheric scintillation monitoring receiver development and test results, Proceedings of the ION 52th Annual Meeting, Cambridge, MA, June, Wickert, J., C. Reigber, G. Beyerle, R. Konig, C. Marquardt, T. Schmidt, L. Grunwaldt, R. Galas, T. K. Meehan, W. G. Melbourne and K. Hocke, Atmosphere sounding by GPS ratio occultation: First results from CHAMP, Geophys. Res. Lett., 28, , 2001.