Newcastle University eprints

Transcription

1 Newcastle University eprints Tiwari R, Ghafoori F, Fenek O-Al, Hadad O, Skone S. Investigation of highlatitude ionospheric scintillation observed over Canadian region. In: 23rd International Technical Meeting of The Satellite Division of the Institute of Navigation (ION GNSS). 2, Portland, Oregon, USA: The Institute of Navigation. Copyright: 24 The Institute of Navigation OR The definitive version of this article, published by The Institute of Navigation in Proceedings of the 23rd International Technical Meeting of The Satellite Division of the Institute of Navigation (ION GNSS 2). September 2-24, 2, Oregon Convention Center, Portland, Oregon, Portland, OR., 2, is available at: Always use the definitive version when citing. Date deposited: August 24 This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3. Unported License eprints Newcastle University eprints

2 Investigation of High Latitude Ionospheric Scintillation Observed in the Canadian Region R. Tiwari #, F. Ghafoori, O. Al-Fanek, O. Haddad and S. Skone Department of Geomatics Engineering, Schulich School of Engineering, University of Calgary, Canada # (at present) School of Electrical, Electronics & Computer Engineering, Newcastle University, United Kingdom BIOGRAPHY Rajesh Tiwari, Ph.D. (Physics), is research associate at School of Electrical, Electronics and Computer Engineering, Newcastle University, UK. He is currently doing an M.Sc. in Geomatics Engineering at the University of Calgary. He received his M.Sc. (24) in Physics at Barkatullah University, Bhopal, India. He has represented scientific team member in the XXVI Indian Scientific Expedition to Antarctica under the project Space Weather Program at Antarctica. Fatemeh Ghafoori is a Ph.D. candidate in Geomatics Engineering at the University of Calgary. She received her B.Sc. (22) in Electrical and Communication Engineering at the Ferdowsi University, Iran, and M.Sc. (27) in Digital Communication Systems and Technology at the Chalmers University of Technology, Sweden. Ossama Al-Fanek is a Ph.D. candidate in the Department of Geomatics Engineering at the University of Calgary. He received a B.S. (24) in Mechanical Engineering at the Jordan University of Science and Technology. His research interests focus on Ionospheric Tomography using Global Navigation Satellite Systems. Oday Haddad received a B.Sc. in 24 and an M.Sc. in 27 in Electrical Engineering at the Jordan University of Science and Technology. He joined the University of Calgary in 28, and is currently working on an M.Sc. degree in Geomatics Engineering. His research interests focus on Global Navigation Satellite Systems and the Wide Area Augmentation System. 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. ABSTRACT GPS is a satellite-based radio navigation system, with diverse applications such as surveying, enroute navigation and precision approach of aircraft, deformation monitoring, and land/marine navigation. Many such applications require very high levels of accuracy, reliability and availability. A challenge to such applications exists, however, during high levels of ionospheric activity. The GPS signals are refracted by an amount dependent on the given signal frequency and total electron content along the signal path. In regions of smallscale irregularities in electron density the GPS signals may experience scintillation effects random rapid variations in signal phase and amplitude. Such effects can cause errors in receiver signal tracking loops and, in some cases, loss of signal lock. Ionospheric scintillations occur primarily in the equatorial, auroral and polar regions. In each region there are different physical processes leading to formation of electron density irregularities. For example, the polar cap is a region of magnetic field lines open to the solar wind; polar patches of ions and electrons can form and convect over the pole in the anti-sunward direction. In steep gradients at the edges of these patches, gradient-drift instabilities can develop with associated small-scale irregularities and scintillation effects. GPS signals propagating through the polar cap can experience both phase and amplitude scintillation effects. In the auroral oval, strong phase scintillation effects are observed when energetic electrons precipitate into the night side ionosphere along closed magnetic field lines during substorm events; amplitude scintillations are minimal in this region. Since 23 the University of Calgary has operated a number of specialized GPS receivers in Canada for ION GNSS 2, Session D, September 2-24, Portland, OR

3 ionosphere monitoring as part of the CANGIM (CANadian GPS Network for Ionosphere Monitoring) project. These receivers are modified dual-frequency survey-grade NovAtel Euro4 receivers, with specialized firmware capable of deriving phase and amplitude scintillation information per minute from detrended 5 Hz observations. Observations also include raw GPS pseudorange and carrier phase observations, total electron content (TEC), 5 Hz phase (ADR) and amplitude observations, phase scintillation indices, and amplitude (S4) scintillation indices. Such data have been collected at three sites in western Canada over the past seven years. Sites are located at similar longitudes and in a range of latitudes spanning the sub-auroral region into the polar cap. Observations from these stations allow latitude profiling of the spatial extent of scintillation effects and investigation of auroral scintillation and polar scintillation effects. In this paper we exploit the CANGIM data set, augmented with other ground- and satellite-based instruments, to study the processes and characteristics of high-latitude scintillation effects. Scintillation events are identified and classified as auroral and polar. Phase and amplitude scintillation characteristics (magnitude, frequency, and spatial extent) are then investigated for each event, in order to gain insight into physical drivers and processes. Ground-based fluxgate magnetometer observations are obtained from CARISMA (Canadian Array for Real-time Investigations of Magnetic Activity) to determine the auroral/polar cap boundary through inferred latitude range of the auroral electrojets. COSMIC (Constellation Observing System for Meteorology, Ionosphere, and Climate) mission satellites are used to find co-located GPS radio occultation events, and derive vertical profiles of electron density to study the altitude distribution of irregularities. Interplanetary magnetic field (IMF) data are obtained from the Advance Composition Explorer (ACE) spacecraft to investigate the IMF orientation during scintillation events. This study includes twenty six events from the years Analysis of scintillation events as a function of geomagnetic activity is also conducted. This study is intended to provide insight into the physical drivers of ionospheric scintillations, with application to physicsbased simulation models used in development of more robust receiver designs for modernized GPS or Galileo in high latitude regions. INTRODUCTION The Global Positioning System (GPS) is a satellite-based navigation tool developed by the U.S. for their military use, which is made available for civilian. The GPS users in various fields rely on a GPS-based location, and some of the applications need a very high level of precision, reliability, and availability. But the ionosphere is the major source of error in GPS, and this error can be very large (3 m or more) with receiver performance degraded during intense ionospheric storms. Generally, there exist small-to-medium scale electron density fluctuations in the ionosphere, and these are termed ionospheric irregularities. The formation of these irregularities results from solar-terrestrial interactions for example solar flares causing enhanced geomagnetic activity. When the GPS signals propagate through such irregularities, the received signal (at the user s receiver) can exhibit rapid random fluctuations in its characteristics such as phase, amplitude or strength of the signal. This phenomenon is known as ionospheric scintillation [Yeh and Liu, 982]. The random fluctuation in phase is termed phase scintillation and amplitude scintillation occurs for random fluctuations in amplitude of the received signal. Ionospheric scintillation degrades the tracking performance of a receiver [Skone and Knudsen, 2; Skone, 2], which can affect GNSS navigation accuracy. Under strong phase scintillation, loss of signal lock causes cycle slips [Datta-Barua et al., 23]. Scintillation occurs primarily in high and low latitude regions. The high latitude region is mainly characterized by auroral substorms and polar energetic precipitation, which is directly influenced by solar activity. The frequency of scintillation events is expected to increase during the upcoming solar maximum period, with a peak expected approximately 23. During the severe scintillation periods, accuracy and availability of GNSS applications may be compromised for aviation, marine navigation, and other safety-of-life applications. In order to investigate expected impact on GNSS, a physics-based model can be developed to simulate ionospheric scintillation. However, an understanding of physical driving mechanisms responsible for ionospheric scintillation must be developed, with appropriate spatial distribution of irregularities. This research explores physical characteristics and mechanisms responsible for auroral and polar influenced scintillation, based on analysis of GPS data (23-27) covering sub-auroral to polar regions in western Canada. SOLAR-TERRESTRIAL INTERACTION The upper layer of the Sun, the corona, continuously ejects a tremendous amount of energetic charged particles into space also termed as solar wind. These energetic particles collide with the Earth s magnetosphere with ~ 37 W/m 2 of power density [Hunsucker and Hargreaves, 23]. Furthermore, the solar wind transports the interplanetary magnetic field (IMF), which interacts with the Earth s dayside magnetic field lines as shown in Figure. Interaction of the IMF and terrestrial magnetic lines can result in some energetic particles entering the Earth s magnetosphere - known as leak in ION GNSS 2, Session D, September 2-24, Portland, OR 2

4 Altitude (km) [Gordon, 998]. The leak in energy depends on the orientation of IMF; if the orientation is parallel (southnorth orientation) to the Earth s magnetic line, then a very small fraction of solar energetic particles will enter; this is a closed magnetosphere configuration. If the IMF is antiparallel (north-south orientation), the Earth s magnetic lines are broken through reconnection and therefore open to solar energetic particles; this is an open magnetosphere configuration. However, the mechanism of insertion of solar particles occurs in two different ways; (i) through open magnetic field lines at the poles, the energetic particles directly enter the Earth s atmosphere, (ii) through energetic particles entering the nightside magnetotail at neutral points. Field lines reconnected at the magnetotail neutral point trap particles which are accelerated towards the Earth and eventually diverted along the terrestrial magnetic field lines to the high-latitude nightside ionosphere - causing auroral substorms. The aurora in visible form appear as dancing green, yellow, or purple lights with global distribution in an oval pattern (referred to as auroral oval). The auroral oval is threaded by closed magnetic lines mapped to the magnetotail. This oval is centered approximately over the geomagnetic pole, and expands equatorward at geomagnetic midnight [Feldstein, 963]. Figure illustrates the overall mechanism of solarterrestrial interaction; the gray solid arrows indicate the flow of energetic particles at open and closed magnetic lines. The consequences of open magnetic lines are polar patches at locations indicated with the blue circle red outline. Aurora occurring on closed magnetic field lines are shown in orange for the night sector. The mechanisms of polar scintillations and auroral substorms are discussed in subsequent sections. Hamza, 2]. An image of the auroral oval is shown in Figure 2; the polar region of open magnetic lines is indicated. The auroral oval is located approximately 6-7 geomagnetic latitude near local midnight. The auroral oval expands equatorward at midnight during enhanced ionospheric conditions. The poleward oval boundary is observed in the noon sector approaching 77 geomagnetic latitude [Feldstein, 963]. E-region electron precipitation (resulting in auroral emissions) is one characteristic of auroral substorms, which produce irregularities. Figure 3 shows a vertical profile of electron density and irregularities obtained from a high pass filter; this figure is consistent with theory showing irregularities observed primarily at altitudes - 5 km. GNSS signals propagate through this region and experience phase scintillation. Phase scintillation of GPS signals can arise from electron density variations in E- region of scale sizes -2 km [Skone et al., 29], sometimes extending to F-region. The correct estimation of auroral boundaries is very important in selecting auroral ionospheric scintillation databases and ultimately developing physics-based scintillation models. Auroral Region Polar Region Figure 2. Image of auroral oval [courtesy of NASA], and polar and auroral regions. Vert. Prof. of Electron Density 4 Irregularities Figure. Solar-terrestrial interaction with earth s atmosphere. AURORAL SUBSTORMS CLOSED FIELD LINES As discussed above, the auroral oval is a region of closed magnetic field lines, in which occur visible patterns due to precipitation of electrons into the neutral atmosphere at an altitude of approximately km [St.-Maurice and el/cm 3 x 5 el/cm 3 x 5 Figure 3. Vertical profile of electron density and irregularities observed in auroral oval 6:45 UT October 3, 25 from CHAMP radio occultation observations. ION GNSS 2, Session D, September 2-24, Portland, OR 3

5 H (nt) Auroral Oval Estimation The auroral oval contains electric currents (auroral electrojets) that flow primarily westward at km altitude. Therefore a ground-based magnetometer can be used to detect magnetic fluctuations due to growth of these currents and identify enhanced ionospheric activity. Auroral electrojets are observed in the south-north H component (positive northward) of the magnetic field. Figure 4 indicates the presence of aurora at 8: UT with a large negative perturbation in the H-component magnetic field H-Component Fort Churchill -5 : 4: 8: 2: 6: 2: 24: Time (UT) Figure 4. Variation of H component at ground-based magnetometer at Fort Churchill on September 28, 27. Based on similar theory, a global index, Kp, is obtained from a chain of ground-based magnetometers located at sub-auroral latitudes. When Kp is greater than 5, it is said that a strong geomagnetic storm is occurring, while Kp less than 4 is considered geomagnetically quiet [SWPC, 2]. Furthermore, the distribution of auroral events is not based on the visible luminosity of a given formation, but is instead based on the measurement of particle energy, since not all energetic particles will be visible to the naked eye [Hardy et al., 985]. Hardy studied the zones of electrons and ion flux in terms of Magnetic Local Time (MLT) at all levels of geomagnetic activity quantified by Kp. Chubb and Hicks in 97 also used Kp to study the mechanism of aurora associated with various points on the Kp scale, observing that the lower boundary of the oval expands about.7 deg equatorward per unit of Kp on the dayside of the Earth and.3 deg on the nightside while it also moves by -3 deg of latitude during a substorm. Gussenhoven et al. [983] express the equatorward boundary (L) of the oval in terms of the Kp index as L L a Kp () o where, L o and a are coefficients dependent on MLT (for details see Gussenhoven et al. [983]). While in general the auroral oval characteristics may be estimated based on Kp value alone, the Kp index is unfortunately a very coarse measurement, as values are provided for three-hourly periods and may not be appropriate for all local regions due to the global nature. The use of Kp is unsuitable for precise local estimation of auroral activity and oval boundaries. For this reason, other researchers, including those at Athabasca University, have established a network of magnetometer stations covering Canadian high latitude regions including those along the Churchill meridian. The Churchill meridian network is a set of ground-based fluxgate magnetometers installed in central Canada covering mid- to high-latitudes as shown in Figure N 35 W Figure 5. Fluxgate magnetometers for Churchill meridian, part of the CARISMA network. Each magnetometer station on the Churchill meridian records Earth s magnetic field in local X, Y and Z components, which are further converted into the Horizontal (H; H X 2 Y 2 ), vertical (Z), and Dip angle (D) components. The extrema of the H component of latitudinal variation approximates the boundaries of the auroral electrojet. The difference between the positive Z peak and negative Z peak represent the width of the auroral oval. Once the two data points (equatorward and poleward boundaries) are estimated at any local time, it is then straightforward to compute colatitude of the oval locus at magnetic local midnight using Equation (2). According to the Feldstein Oval, the average difference in auroral oval latitudes between noon and midnight is 6.5 deg ([ 2 ] 6. 5 ). Substituting this value into Equation (2) defines the oval (one-hour averages) at any desired local time. The regression fit of 24 consecutive one-hour points of equatorward and poleward data produces the auroral oval [Rostoker and Skone, 99]. An example is shown in Figure 6. cos 2 (2) 2 [Kamide and Fukushima, 97] where 6 N 2 W 75 N Churchill Meridian 5 W TALO RANK ESKI FCHU GILL ISLL PINA 9 W 75 W = co-latitude of oval locus at local time 6 W ION GNSS 2, Session D, September 2-24, Portland, OR 4

6 Altitude (km) 2 = co-latitude of oval locus at midnight = co-latitude of oval locus at noon = azimuth angle as function of local time 2 complementary magnetometer dataset from the Fort Churchill meridian is used derived auroral oval boundaries. To determine the IMF orientation, ACE spacecraft data sets are used. The observations and measurement techniques are described as follows below. Vert. Prof. of Electron Density 4 Irregularities Figure 6. The auroral oval estimated using CARISMA ground-based magnetometer data. POLAR REGION OPEN FIELD LINES Previous research shows at high latitude, gradient drift instability [Tsunoda, 988], precipitation of soft particles and shear electric filed are some mechanisms responsible to generate irregularities [Basu et al., 983; Kersley et al. 988]. At the polar latitudes, F-region structures (distributed across all altitudes as shown in Figure 7) of - km scale size occur due to the gradient drift instability mechanism [Tsunoda, 98; Mitchell et al., 25; de Francheschi et al., 28]. This instability occurs during periods of southward (negative) IMF, which leads to open terrestrial magnetic field lines along which energetic particles diffuse into the Earth s high latitude ionosphere causing polar patches. Generally, auroral and polar regions are defined in the geomagnetic latitude domain (auroral region 65-75º geomagnetic latitude and polar region poleward of 75º geomagnetic latitude [Pi et al., 22]), but more generally the differentiating factor is magnetic field line characteristics - open field lines in the polar region and closed field lines in the auroral region (Figure 2). The corresponding scintillation effects are referred to as polar and auroral scintillation. DATA SET The data set includes GPS observations obtained from ground- and space-based receivers. The ground-based GPS data are obtained from CANGIM to estimate ionospheric scintillation; space-based GPS observations from LEO satellites (COSMIC) are used to derive vertical profiles of electron density. Furthermore, a el/cm 3 x 5 el/cm 3 x 5 Figure 7. Vertical profile of electron density, and irregularities obtained using high pass filter in polar region 5:5 UT August 26, 25. CANGIM Observations To study the characteristics of high-latitude ionospheric scintillation, the Canadian GPS Network for Ionosphere Monitoring (CANGIM) was established in 23. The network includes three stations: Yellowknife ([Yell] 62.48º N, 4.48º W), Athabasca ([Athb] 54.72º N, 3.3º W), and Calgary ([Calg] 5.8º, N 4.3º W). Sites are equipped with NovAtel Modulated Precision Clock (MPC) receivers (see Figure 8 for location of sites and Table for coordinates) and NovAtel 6 antennas. The receivers are capable of tracking L/L2 carrier phase with 2 channel capacity, based on NovAtel OEM4 L/L2 GPS technology with Euro4 card. The receiver contains Patented Pulsed Aperture Correlator (PAC) technology with powerful microprocessor; this technology makes the receiver multipath resistant, and has very good timing for acquisition and re-acquisition. Receivers have specialized firmware capable of deriving phase () and amplitude (S4) scintillation indices from 5 Hz detrended GPS L observations. The CANGIM network does not currently operate in realtime. However, data are accessed remotely from a central unit at University of Calgary. The CANGIM data used in this research work are phase (one-second average) and amplitude scintillation S4 at one-minute intervals for all satellites above 2 deg elevation from 23 to 27. The data are categorized for two regions, depending on the ionospheric pierce point (IPP) latitudes and estimated auroral oval boundaries: (i) polar scintillation, and (ii) auroral scintillation. The threshold for scintillation detection is a phase scintillation index of.2 radian. For ION GNSS 2, Session D, September 2-24, Portland, OR 5

7 No. of Obs. SSN example, IPPs of satellites with phase scintillation above the threshold at latitudes in the auroral oval are considered to be auroral scintillation, while observations with IPPs poleward of the auroral oval are considered to be polar scintillation (as shown in Figure 9). The solar activity for year 23 was very high and sunspot numbers were in the range 3-2. Year 27 represents relatively low solar activity and sunspot numbers were below 5. The dataset thus covers transition from solar maximum to solar minimum periods. Phase scintillation in the auroral and polar regions has good correlation with solar cycle as shown in Figure. We analyzed 3,63,24 raw GPS scintillation observations from CANGIM sites for years From these observations, the auroral and polar phase scintillations above the threshold value of.2 rad (and satellite above 2 deg elevation) are selected. These observations are identified as auroral or polar by comparing IPP locations with estimated auroral boundaries. Overall,288 auroral observations and 36 polar observations were selected. Polar Scintillation Auroral Scintillation Figure 9. Identification of auroral and polar scintillation. The green tracks indicate IPPs for all satellites in view at the three CANGIM sites over a 24- hour period. Red points indicate occurrence of scintillation. Approximate location of the auroral oval is shown. 2 Daily SSN variation Jan 23 Dec 23 Dec 24 Dec 25 Dec 26 Dec 27 Number of Phase and Amplitude Scintillation >=.2 (Elv Cutoff = 2 o ) 4 Aurora Phase Obs. Polar Phase Obs. 2 Figure 8. CANGIM stations: Yellowknife (Yell), Athabasca (Athb) and Calgary (Calg) for high latitude ionospheric studies. Table. Geographic and geomagnetic coordinates of CANGIM stations. Station code Geographic Geomagnetic Lat. Long. Lat. Long. Yell 62.48º -4.48º 68.8º º Athb 54.72º -3.3º 6.52º 35.67º Calg 5.8º -4.3º 57.86º 36.43º Year Figure. The upper panel represents daily sunspot number and the lower panel represents the number of phase scintillation observations above.2 radian with 2 deg elevation cutoff. Radio Occultation Observations Hajj et al. [994] proposed a space-based GPS monitoring system in order to cover the unseen region of the ionosphere above the oceans. In this technique, a GPS receiver is placed on a low earth orbiter (LEO), with an altitude between 5 and 8 km. Since the GPS satellite is at 2,2 km altitude, the LEO satellite is able to receive dual-frequency GPS signals along lines-of-sight passing through successive horizontal ionosphere layers; this is the concept of radio occultation (RO). Figure illustrates the link between the LEO and GPS satellite. Many missions have been conducted to exploit this method: e.g. CHAMP [Wickert et al., 2], and COSMIC [Sokolovskiy et al., 27]. This technique is used to estimate an ionospheric vertical electron density profile [Hajj and Romans, 998]; this method exploits ION GNSS 2, Session D, September 2-24, Portland, OR 6

8 H Index dhvar/dt Sigma (rad) bending angle dependence on refractive index of the ionosphere (which depends on electron density) [Schreiner et al. 999; Ghafoori, 29]. Generally, the RO events have durations of 2-3 minutes; approximately 2 such events are recorded daily globally. Therefore, only selected events are considered here which are close to the IPPs of GPS satellites visible in the CANGIM network which observe scintillation. The availability of RO data enhances the scintillation study in terms of considering altitude distribution of electron density irregularities: e.g., Skone et al. [28] and Skone et al. [29] studied auroral scintillation due to E-layer irregularities over the Canadian region. LEO GPS Figure. Radio occultation geometry. The solid yellow circles represents ionospheric irregularities observed along the GPS-LEO link. CARISMA Ground-based Magnetometer Observations CARISMA ground-based fluxgate magnetometer data are publicly available ( The CARISMA data are used in two different ways; (i) estimation of auroral oval boundaries; and (ii) estimation of a local magnetic index, referred as H index. The estimated H index is used to establish the presence of aurora independently from the phase scintillation indices; a one-to-one correlation is expected. The raw H- component magnetic field data are obtained from the Yellowknife (YKC) magnetometer station (62.48 N, E) at 5 s intervals. The rate of change of H dh var component variation ( ) is then estimated to dt detrend the data. The estimated standard deviation for one minute of samples (e.g., 2 samples for 5 s sample interval) gives one-minute H index ( H In ) as derived in Equation (3). H In n n i dh dt var i dh dt var 2 i where, n 6 f and f is the sampling rate of raw H- component data. Correlation of the phase scintillation index ( ) for auroral scintillation events at CANGIM site YELL with estimated H index is shown in Figure 2. The estimated H index has one-to-one correlation with phase scintillation. Thus this further validates the selected auroral scintillation dataset through independently establishing the presence of local auroral substorm activity with H index values. Interplanetary Magnetic Field Observations The solar IMF is a three-directional vector quantity (Bx, By, and Bz). The Bz component is perpendicular to the elliptical plane. The IMF orientation can be parallel or antiparallel to Earth s magnetic lines. The antiparallel (N- S or negative Bz) IMF reconnects with the terrestrial field lines leading to an open magnetosphere configuration, and resulting in solar energetic particles entering directly into the polar ionosphere causing formation of polar patches. Auroral substorms may also occur nightside due to closed field lines. The IMF data are publicly available at as obtained from the Advance Composition Explorer (ACE) spacecraft. The IMF orientation is a good indicator to understand and validate the polar scintillation dataset..5 - GPS at Yellowknife(62.48 o N -4.5 o E) 238 Magnetometer at Yellowknife (62.48 o N -4.5 o E) 238 Magnetometer at Yellowknife (62.48 o N -4.5 o E) Hours (UT) Figure 2. The upper panel represents the phase scintillation obtained from CANGIM site YELL. The corresponding rate of change of H-component and H index are shown in the lower two panels as derived from local ground-based magnetometer observations. 2 (3) ION GNSS 2, Session D, September 2-24, Portland, OR 7

9 H Index Sigma (rad) RESULTS AND DISCUSSION Researchers have studied the morphology of scintillation at GPS frequencies at geomagnetic low, middle and high latitudes [Kelley 989; Fejer et al., 999; Aarons 982, 995, 997; Meggs et al., 28; de Francheschi et al., 26, 28]. The characteristics and physical driving mechanism for scintillation are significantly different in all three regions due to the different physical processes leading to formation of electron density irregularities that produce scintillations. An objective here is to investigate scintillation at high latitudes. threshold) and corresponding H indices in Figure 5. Overall the phase scintillation shows a highly positive correlation with the H index, validating the presence of auroral substorm activity during periods of higher auroral phase scintillation. This further validates selection of the auroral scintillation dataset. Auroral Oval : 45 UT 2. < Sigma (rad).2.2 < Sigma (rad) Thus two different mechanisms are studied here: one related to polar regions (open magnetic field lines), and the second related to auroral regions (closed magnetic field lines). Phase scintillations are considered in terms of IMF orientation as well as vertical electron density profiles obtained from COSMIC and CHAMP GPS radio occultation observations. Ground-based fluxgate magnetometer observations are obtained from CARISMA to determine the auroral/polar boundary through inferred latitude range of the auroral electrojets. 8 Yell Athb Calg 6 Auroral Scintillation In Figure 3, the red points represent auroral phase scintillation observations at GPS IPPs (where the satellitereceiver lines-of-sight pierce an ionospheric shell at 35 km altitude) for satellites with 2 deg elevation cutoff observed from CANGIM 6:3-6:45 UT October 3, 25. The higher phase scintillation values are located within the auroral oval boundaries. In the high latitude ionosphere, the electron precipitation carries most of the particle energy, which is the major driver of auroral emissions. In the auroral region energetic electron precipitation is the dominant factor in producing both the ionospheric layers and the irregularities. To confirm this, the left panel in Figure 3 shows the electron density profile derived from corresponding co-located CHAMP auroral observations and the right panel shows the higher order filtered electron density at 6:45 UT. It was observed that the maximum scintillation is well-correlated with the peak in electron density at 5 km altitude. This E-region enhancement is due to the precipitating electrons. This result is consistent with the findings of Skone et al. [28, 29]. As discussed previously, the H index should have one-toone correlation with auroral scintillation occurrence. To study this example event, the H index is correlated with phase scintillation indices of all visible satellites (observations exceeding scintillation detection threshold.2 rad) at YELL as shown in Figure 4. Overall a positive correlation is observed. Further analysis was conducted for the 26 selected auroral scintillation events (observations exceeding.2 rad Figure 3. Auroral oval and auroral scintillation recorded on October Figure 4. Auroral phase scintillation and H index. Figure 5. Auroral phase scintillation and H index correlation. GPS Yell(62.48 o N -4.5 o E) 253 Yell(62.48 o N -4.5 o E) UT (Hours) ION GNSS 2, Session D, September 2-24, Portland, OR 8

10 Sigma (rad) Commulative Distribution The magnitudes of phase and amplitude scintillation derived for the auroral region, based on our selected dataset, are shown in Figure 6. A range of phase scintillation values were observed, with some as high as -2 rad (severe). The corresponding amplitude scintillation (S4) was lower than.3 (low) for the entire dataset. Auroral Scintillation; Elevation Cutoff = 2 o.8.6 Phase Amplitude polar dataset. The results show strong influence of IMF orientation on the occurrence of polar phase and amplitude scintillation and thus strong influence of the IMF on polar patch formation and associated irregularities In Figures 2 and 2 the highest scintillation values occur during periods of negative (southward) Bz for open magnetosphere configurations. These results are consistent with theory. Figure 22 shows that large amplitude and large phase scintillation occur in the polar region. This is in contrast to the distribution shown in Figure 6 for the auroral region where only minimal amplitude scintillation is observed in the data set..4.2 Auroral Oval : 5 UT Magnitude of S4 and phase scintillation index Figure 6. Distribution of auroral scintillation for the auroral dataset. Polar Scintillation Figure 7 shows an example of polar scintillation observed 5:5 UT August 26, 25, where satellite IPPs for detected scintillations are poleward of the estimated auroral oval boundaries. The time series of polar phase scintillation at YELL is shown for the entire day August 26 in Figure 8. The gaps in the plot are due to selection of only those IPPs which are poleward of the auroral oval. The vertical distribution of co-located electron density irregularities for this event is shown in Figure 7. The irregularities extend through F-region altitudes as expected for polar events. Soft electron precipitation into the F region is known to result in formation of these types of irregularities which produce phase scintillation [Basu et al., 983; Kerslay et al., 988]; however, amplitude scintillation is also significant as shown in Figure 9. In the polar cap scintillations are produced by ionospheric irregularities created in the F-region density gradients associated with polar cap patches [Coker et al., 24]. Polar cap patches appear to enter the polar cap on the dayside and drift toward midnight [Buchau et al., 984]. Mitchell et al. [25] reported scintillation associated with steep total electron content (TEC) gradients observed southwest of Svalbard (Norway) during the Halloween storm of October 23. In their analysis, it was found that the high TEC values originated from convection of plasma from the North American sector over the polar cap. While investigating the dynamics of ionospheric plasma during severe storm events, de Francheschi et al. [25] established that the formation and movement of polar patches was strongly influence by IMF orientation. 8. < Sigma (rad).2.2 < Sigma (rad) Yell Calg 6 Figure 7. Polar scintillation observed on open field lines poleward of the auroral oval on August 26, Universal Time (hours) Figure 8. Polar phase scintillation indices for CANGIM site YELL August 26, 25. Polar Phase Scintillation; 26 August 25 Local Time (hours) 6:3 2:3 24:3 4:3 8:3 2:3 6:3 We therefore investigate the relationship of IMF orientation and phase and amplitude scintillation in our ION GNSS 2, Session D, September 2-24, Portland, OR 9

11 Commulative Distribution Bz (nt) Bz (nt) S4 S4 and Phase Scintillation Correlation on 26 August Phi Sec Figure 9. Amplitude and phase scintillation recorded for polar region on August Sigma (rad) Figure 2. IMF correlation with polar phase scintillation for polar dataset Polar Phi Vs IMF Bz; Elevation Cutoff = 2 o S4.6.8 Figure 2. IMF correlation with all polar amplitude scintillation for polar dataset Polar S4 Vs IMF Bz; Elevation Cutoff = 2 o Polar Scintillation, Elevation Cutoff = 2 o Phase Amplitude Magnitude of S4 and Sigma (rad) Figure 22. Distribution of amplitude and phase scintillation for polar dataset. CONCLUSION High-latitude ionospheric scintillation is investigated in Canada using GPS receivers (ionospheric scintillation monitors) covering sub-auroral, auroral and polar regions. A method has been developed to clearly differentiate between auroral and polar scintillation events, by dynamically estimating auroral boundaries, and generating full auroral and polar scintillation databases. Auroral scintillation occurs in a region of closed magnetic field lines. Phase scintillation is prominent in this region and highly positively correlated with a local magnetic index - the H index which reflects the presence of auroral substorm activity. Polar scintillation occurs on open magnetic field lines, and it is observed that both amplitude and phase scintillation effects are significant. In the polar region, sources of ionospheric structure are polar patches and sun-aligned arcs. Thus, polar phase scintillation shows a high level of correlation with southward IMF orientation. There are less occurrences of phase scintillation with northward IMF. Direct coupling of solar wind and ionosphere occurs due to open field lines, thus creating polar electron density irregularities at altitudes 2-4 km, causing both phase and amplitude scintillation of comparable severity in the polar region. Results overall validate the distinctions between auroral and polar scintillations, and the different physical mechanisms for each region. Results indicate that a single phase screen model for high latitude scintillation would not be sufficient, and high-latitude physics-based simulations must account for both polar and auroral regions distinctly. The WBMOD is one model of wellknown scintillation climatology; future work includes plans to validate WBMOD scintillation predictions using our selected database. ACKNOWLEDGEMENT Authors wish to acknowledge I. Mann and team of CARISMA for the use of magnetic data for Canadian region. The CARISMA network is operated by University of Alberta, and is funded by Canadian Space Agency (CSA). Author wish to acknowledge COSMIC and CHAMP group for providing radio occultation data, and extend acknowledge to National Space Science Data Center for providing IMF data under web access Author, Rajesh Tiwari wishes to acknowledge Newcastle University, UK for arranging partial travel support to attend ION meeting. ION GNSS 2, Session D, September 2-24, Portland, OR

12 REFERENCES Aarons, J., Global morphology of ionospheric scintillations, Proceeding in IEEE, 7(4), pp , 982. Aarons, J., L. Kersley and A. S. Rodger, The sunspot cycle and auroral F layer irregularities, Radio Science, 3 (3), pp , 995. Aarons, J., M. Mendillo, and R. Yantosea, GPS phase fluctuations in the equatorial region during sunspot minimum, Radio Science, 32, pp , 997. Basu, S., S. Basu, J. P. McClure, W. B. Hanson, and H. E. Whitney, High resolution in-situ data of electron densities and VHF/GHz scintillations in the equatorial region, Journal of Geophysics Research, 88, pp , 983. Buchau, J., E. J. Weber, D. N. Anderson, Jr., H. C. Carlson, J. G. Moore, B. W. Reinisch, and R. C. Livingston, Ionospheric structures in the polar cap: Their origin and relation to 25MHz scintillation, Radio Science, 2(3), pp , 985. Carlson, H. C., Characterizing the polar-cap ionosphere, in: Characterising the ionosphere, edited by: Wyman, G.; Technical Report RTO-TR-IST-5, Chapter, pp Neuilly-sur-Seine, France: RTO, Chubb, T. A and G. T. Hicks, Observation of the aurora in the far ultraviolet from OGO 4, Journal of Geophysics Research 75, pp , 97. Coker, C., G. S. Bust, R. A. Doe, and T. L. Gaussiran, High-latitude plasma structure and scintillation, Radio Science, 39, 24. 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 23, Portland, OR, September 9-2, pp , 23. De Franceschi G, L. Alfonsi, and V. Romano, Isacco: an Italian project to monitor the high latitude ionosphere by means of GPS receivers, GPS Solutions,, pp , 26. De-Franceschi, G., L. Alfonsi, V. Romano, M. Aquino, A. Dodson, C. N. Mitchell, P. Spencer, and A. W. Wernik, Dynamics of high-latitude patches and associated smallscale irregularities during the October and November 23 storms, Journal of Atmospheric and Solar- Terrestrial Physics 7, pp , 28. Fejer B. G., L. Scherlies, and E. R. de Paula, Effects of the vertical plasma drift velocity on the generation and evolution of equatorial spread F, Journal of Geophysics Research, 4, pp , 999. Feldstein, Y. I., Some problems concerning the morphology of auroras and magnetic disturbances at high latitudes, Geomagnetic Aeron., English Translation, 3, pp , 963. Ghafoori, F., Correcting the negative values of the retrieved ionospheric electron density profiles using the NNLS algorithm, Proceedings of the 2st International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS 29), Savannah, GA, September 29, pp Gussenhoven, J. M., D. A. Hardy, and N. Heinemann, Systematic of the equatorward diffuse auroral boundary, Journal of Geophysics Research, 88, pp , 983. Hajj, G. A., R. Ibanez-Meir, E. R. Kursiniski and L. J. Romans, Imaging the ionosphere with Global Positioning System, International Journal Imaging System Technology, 5, pp , 994. Hajj, G. A. and L. J. Romans, Ionospheric electron density obtained with Global Positioning System: results from the GPS/MET experiment, Radio Science, 33, pp. 75-9, 998. Hardy, D. A., M. S. Gussenhoven and D. Brautigan, A statistical model of auroral ion precipitation, Journal of Geophysics Research, 9, pp , 985. Hunsucker, R. D. and J. K. Hargreaves, The high latitude ionosphere and its effects on radio propagation, st edition, Cambridge University Press, UK, 23. Kamide, Y., and N. Fukushima, Spatial extent of the return current of the auroral zone electrojet, Part II, Rep. Ionos. Space Res. Jpn., 24, pp. 5-25, 97. Kersley L, S. Pryse, and N. Wheadon, Amplitude and phase scintillation at high latitudes over northern Europe, Radio Science, 23(3), pp 32 33, 988. Kelley, M. C., The Earth's Ionosphere, The Academic Press, San Diego, 989. Kintner P. M., B. M. Ledvina, and E. R. de Paula, GPS and ionospheric scintillations, Space Weather, 5, 27. Meggs, R. W., C. N. Mitchell, and F. Honary, GPS scintillation over the European Arctic during the November 24 storms, GPS Solution, 2, pp , 28. Mitchell, C. N., L. Alfonsi, G. D. Franceshi, M. Lester, V. Romano and A. W. Wernik, GPS TEC and scintillation measurements from the polar ionospheric during the October 23 storm, Geophysical Research Letters, 32 (2), 25. Pi, X., Boulat, B. M., Mannucci, A. J. and D. A. Stowers, Latitudinal characteristics of L-band ionospheric scintillation, Proceedings of the ION GPS 22, Portland, OR, September 24-27, 22. ION GNSS 2, Session D, September 2-24, Portland, OR

13 Rostoker, G., and S. Skone, Magnetic flux considerations in the auroral oval and the earth s magnetotail, Journal of Geophysics Research, 98, pp , 993. Rostoker G., Nowcasting of space weather using the CANOPUS magnetometer array, La Physique Au Canada, pp , 998. 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 Science, 34(4), pp , 999. Skone, S., and K. Knudsen, Impact of ionospheric scintillations on SBAS performance, Proceedings of the ION GPS 2, Salt Lake City, UT, September, 2. Skone S., The impact of magnetic storms on GPS receiver performance, Journal of Geodesy, 75(6), pp , 2. Skone, S., M. Feng, F. Ghafoori and R. Tiwari, Investigation of scintillation characteristics for high latitude phenomena, Proceedings of the 2st International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS 28), Savannah, GA, September 28, pp Skone, S., M. Feng, R. Tiwari and A. Coster, Characterizing ionospheric irregularities for auroral scintillations, Proceedings of the 22nd International Technical Meeting of The Satellite Division of the Institute of Navigation, Savannah, GA, September 29, pp Sokolovskiy, S.V., C. Rocken, D.H. Lenschow, Y.H. Kuo, R.A. Anthes, W.S. Schreiner, and D.C. Hunt, Observing the moist troposphere with radio occultation signals from COSMIC, Geophysical Research letters, Vol. 34, September 8, 27. St.-Maurice, J. P. and A. M. Hamza, A new non-linear approach to the theory of E region irregularities. Journal of Geophyics Research, 6, pp , 2. SWPC, access on October 3, 2. Tsunoda, R. T., Magnetic field-aligned characteristics of plasma bubbles in the nighttime equatorial ionosphere, Journal of Atmospheric and Terrestrial Physics, 42, pp , 98. Tsunoda R. T., High-Latitude F region irregularities: a review and synthesis, Rev Geophys, 26(4), pp.79 76, 988. Wickert, J., C. Reigber, G. Beyerle, R. Konig, C. Marquardt, T. Schmidt, L. Grunwaldt, R. Galas T. K. Meehan, W. G. Melboutne, and K. Hocke, Atmosphere sounding by GPS radio occultation: first results from CHAMP, Geophysics Research Letter, 28, pp , 2. Yeh, K. C., and Chao-Han Liu, Radio wave scintillations in the ionosphere, Proceeding IEEE, 7(4), pp , 982. ION GNSS 2, Session D, September 2-24, Portland, OR 2