Session D1, Paper #1

Remote Sensing of the Earth´s Lower Atmosphere During Severe Weather Events Using GPS Technology: A Pilot Study in Victoria, Australia
S. Choy, K. Zhang, C. Wang, Y. Li, RMIT University, Australia; Y. Kuleshov, Australian Bureau of Meteorology, Australia

Predicting the development of severe weather phenomena such as cyclone and thunderstorm evolution is highly dependent on very precise estimates of water vapour contained within the lower atmosphere. The amount of water vapour contained in the troposphere also has significant implications in determining the strength and extent of the severe weather events. However, water vapour content only takes up 5% of the air and is extremely variable and is quite difficult to observe. In fact, water vapour content is one of the least well-observed parameter in the atmosphere.

In recent years, ground- and space-based GPS measurements have been well established as a powerful tool for providing precise observations of atmospheric water vapour in the troposphere and lower stratosphere. Rather than competing, these two techniques are complementary of each another. For ground-based GPS atmospheric sensing, the propagation of the microwave signals from the GPS satellites to ground-based GPS receivers is delayed by the water vapour in the lower atmosphere. This delay can be parameterised in terms of a time-varying total tropospheric Zenith Path Delay (ZPD), which is retrieved by stochastic filtering of the GPS measurements. If the surface temperature and pressure observations at the GPS receiver are known to sufficient accuracy, the retrieved ZPD can be converted into accurate estimates of the total zenith column water vapour, i.e. Precipitable Water Vapour (PWV) content, above the receiver. Ground-based GPS-PWV has become relatively mature for operational use in the recent years. On the other hand, the GPS satellite constellation also allows for radio occultation (RO) observations of the Earth´s atmosphere using one or more GPS receivers onboard Low Earth Orbiting (LEO) satellites. These space-based RO observations is a novel atmospheric sensing technique, which permit routine profiling of the atmospheric parameters, e.g. temperature, pressure and water vapour, with high vertical resolution and precision based on limb sounding geometry.

The objective of this work is to understand if the ground- and space-based GPS observations will be useful for severe weather diagnoses particularly in studying and monitoring thunderstorm occurrences. It is of interest to determine if a severe thunderstorm will leave a significant pattern in the ground-based GPS-PWV and space-based RO profiles during the storm development, maturity and dissipation stages. A severe thunderstorm event during the first week of March 2010, which brought heavy rainfall, large hails, strong winds and flash flooding to the state of Victoria, Australia, was used as a pilot case study. This study was undertaken in close collaboration with the Australian Bureau of Meteorology.

A one-week period, i.e. a few days prior to and after the storm passage, was selected during which the course of the storm extended from the west to the southeast corner of the state. Tropospheric ZPD were derived using GPS measurements obtained from a local CORS network and GPS-PWV were estimated using the temperature and pressure observations collected from the Victorian weather observation stations. These GPS-PWV estimates were then validated using radiosonde observations collected from the Australia Upper Air Network. In addition, two processing platforms for deriving tropospheric ZPD were also tested as part of this case study. They were Precise Point Positioning (PPP) and baselines processing techniques. The results from these two methods of data processing were compared to the IGS final tropospheric product. During this period, six COSMIC RO events in Victoria were also identified. Atmospheric parameters such as temperature, pressure and water vapour pressure were plotted as a function of height to examine if the storm has left a dominant signature on the retrieved atmospheric parameters profiles. The GPS RO retrieved atmospheric parameters were compared to the vertical profiles provided by radiosonde and European Centre for Medium-Range Weather Forecasts (ECMWF) data.

The preliminary results from this pilot study indicate that there is strong spatial and temporal correlation between the variations of GPS-PWV and the passage of the storm. In fact, the estimated parameter from ground-based GPS observations conformed to IGS tropospheric product, radiosonde observations as well as the amount of cumulative rainfall produced from the storm. This finding is very encouraging as ground-based GPS technique can be considered as an additional meteorological sensor in studying, monitoring, and potentially predicting severe weather events. The strength of using ground-based GPS-PWV technique is that it is capable of providing continuous observations of the storm passage with high temporal resolution; while the spatial resolution of the distribution of the water vapour is dependable on the geographical location and density of the GPS stations. Another finding worth noting is that the propagation of GPS signals is significantly disturbed (in the level of 1 - 2 dm) by the lower atmosphere during a storm passage.

The results from the space-based GPS RO limb sounding technique, on the other hand, are not as robust as the ground-based GPS-PWV in the context of studying and monitoring severe weather events. Although GPS RO could capture the dynamics of the lower Earth´s atmosphere with high vertical resolution, its limited geographical coverage in a local region and its´ occurrence rate, i.e. temporal resolution, over a short period of time raise an important question about its potential and feasibility for monitoring severe weather events, particularly local thunderstorms. This is because thunderstorms have a relatively short life-span and it usually lasts for only 1 to 2 hours. GPS RO technique will be more suited for long term climatology studies over a large area. In addition, it is discovered from this study that the quality GPS RO soundings and retrievals in the lower troposphere degrades during the storm passage as RO refractivity is extremely sensitive to moisture content in the lower atmosphere. It is anticipated that the findings from this pilot study will spur further discussion and research into using GPS meteorology technique for monitoring and forecasting severe weather events in the Australian environment.

[Return to Program]

Session D1, Paper #2

Computation of a High-Precision GPS-Based Troposphere Product by the USNO
S. Byram, C. Hackman, V. Slabinski, J. Tracey, U. S. Naval Observatory

The United States Naval Observatory (USNO) serves as one of the International GNSS Service´s (IGS) analysis centers (AC), producing solutions for the GPS constellation including the ultra rapid and rapid solutions for the GPS orbits, clocks, and the Earth orientation parameters which are submitted on a sub-daily and daily basis respectively. Recently, the USNO AC has undertaken the task of computing a high-precision troposphere product improving and expanding its troposphere product family with a goal of reproducing the quality of the current IGS final troposphere products. The USNO is one of the few (if not the only) Department of Defense sources for such products.
Unlike other IGS products, the IGS final troposphere product is not a combination of solutions from the ACs which can be affected by an individual AC changing estimation methods, the number of ACs contributing solutions, or an inconsistent set of common stations used among the contributing ACs. The current processing of the IGS final troposphere product is performed by the Jet Propulsion Laboratory (JPL). In a setup similar to the one used by JPL, the USNO computes final troposphere estimates of the zenith path delay and the East and North gradient components using a precise point position (PPP) approach to processing zero difference GPS observations directly from the RINEX files for over 300 stations world wide with the Bernese GPS software. This PPP processing of the troposphere product utilizes the IGS final orbits and clocks ensuring both consistency and the highest accuracy available for the GPS constellation inputs. The result is a set of troposphere estimates per station per day which provides internal consistency for the troposphere product users currently with a computed standard deviation of the zenith path delay in the range of 1-2 mm. This presentation will provide an overview of the final troposphere product processing with discussion of the methods and models used.

There is an increasing interest in GPS-based troposphere product by meteorological users for use in numerical weather forecast, climate research, and atmospheric studies. The large network of globally distributed ground stations yields a nearly continuous set of tropospheric estimates for users. However, to provide a consistent time series of data to the troposphere user, reprocessing of the existing IGS final troposphere products using a setup of the Bernese software at the USNO will be performed. This reprocessing gives a unique opportunity for the comparison of different estimation schemes and software packages. The comparison the USNO results to the existing IGS final troposphere products currently created by JPL using the Gipsy-Oasis software will be discussed in this presentation.

Finally, this presentation will also include a discussion of possible changes that could lead to improvements as well as what is planned for the future of the high-precision GPS-based troposphere product computation at the USNO.

[Return to Program]

Session D1, Paper #3

True Real-time Slant Tropospheric Delay Monitoring System with Site Dependent Multipath Filtering
T. Iwabuchi, C. Rocken, GPS Solutions Inc.; A. Wada, Masayuki Kanzaki Hitz, Japan

GNSS estimates of zenith tropospheric delay (ZTD) and precipitable water vapor (PWV) converted from the ZTD are now widely used in meteorology, especially to improve the water vapor distribution in the initial condition used for numerical weather model. Leading weather forecast centers all over the world such as NOAA (USA), ECMWF (EU), and JMA (Japan) have assimilate such GNSS atmospheric products in their data assimilation system, and have proven positive impact of GPS PWV on the forecast of moisture fields and severe precipitation forecasts.
The primary interest for the weather centers has changed from near real-time processing to real-time monitoring of water vapor and short-time forecast of precipitation because of frequent occurrence of severe rainfall and resulting flooding events. We have developed a true real-time processing system of ZTD (Iwabuchi et al., ION GNSS2006), where true real-time ZTD is estimated within a few seconds by using forwarding Kalman filtering processing. Currently an updated true real-time ZTD processing system (RTNet/MET V2 system based on RTNet GNSS processing software developed by GPS Solutions) is operated in Japan since May, 2010, and PWV products are provided to several weather research and operational forecast centers with short latency of a few minutes.

Every minute the RTNet/MET V2 system estimates ZTD and other parameters (station coordinates with tight constraints, receiver and satellite clocks, and float ambiguity) based on using the IGU predicted orbit and the ionospheric free phase linear combination. There is post-fit residual information at each epoch. To get more detailed information on the water vapor than PWV, we use these post-fit residuals to generate slant tropospheric delays every one minute. The period of data is from June, 2010 to February, 2011. The total length of data is enough to investigate multiple active weather conditions such as weather front passages and to get statistical information on the variability of slant delay with changing seasons.

In GNSS processing post-fit residuals consist of unmodelled signals and noise. Under active weather conditions, the inhomogenous distribution of the atmospheric refractivity (function of pressure, temperature, and vapor pressure) is the primary cause for the residual. In such disturbed weather conditions, we assume that that the slant tropospheric delay from the ground station antenna to each satellite is obtained from the sum of the estimated ZTD mapped to the observed elevation angle plus the post-fit phase residuals.

To reduce noise from site-dependent multipath and any systematic error in the assumed phase center variation (PCV) pattern of each GPS antenna with radome (including multipath from monument) the post-fit residuals are filtered with a multipath stacking map with 1 x 1 deg. resolution in azimuth and elevation for each of the 1200 processed Geonet stations . The multipath stacking map is updated every month to account for slow changes in the environment around each station caused by reflectors or changes in the reflection coefficient. The map for some stations shows oscillation patterns, suggesting effects of reflection from rod shaped poles. The multipath map worked to reduce such systematic noise. Thus we believe that most of the dominant noise affecting post-fit residuals is effectively reduced by applying the map and that the signal/noise ratio in slant tropospheric delay is significantly improved.

From the case study of a severe rainfall case in our test period, we found that the variability of slant tropospheric delay is more strongly related to the timing of passages of weather fronts rather than to the total amount of ZTD or PWV. Relatively high variability is observed when a font system is approaching a station. This suggests more active meso-scale convection of water vapor near the convergence line.

The variability of slant tropospheric delay is related to the geographical condition of the station, and depends on local time (higher in the local afternoon to night period than between midnight and morning). The long-term analysis reveals that the dominant variability occurs in the late afternoon in mountainous regions and during the passage of weather fronts.. Late afternoon is also the time of maximum diurnal variation of GPS PWV. Because such variations are considered to be induced by radiative solar heating of the slopes between mountain and plain, more active circulation of water vapor is expected near the mountains.

These facts suggest that slant tropospheric delay based on post-fit residuals represent the inhomogeneous distribution of water vapor along the ray path between satellite transmitter and receiving antenna, which is not modeled by the mapping function (tropospheric delay depending on azimuth angle) and a linear tropospheric gradient model (symmetric tropospheric delay depending on azimuth). We thus conclude that slant tropospheric delay contains meaningful information for monitoring water vapor activity during periods of disturbed weather conditions.

This further suggests that slant delays from dense GNSS networks can be used for high resolution numerical weather prediction, tomography, and tropospheric delay corrections for other ground-based sensors such as very long baseline interferometry (VLBI) or synthetic aperture radar interferometry (In-SAR) which uses microwaves passing through the atmosphere.. Additional benefits can be expected in the future from GNSS and RNSS satellites such as GLONASS, GALILEO, and Japanese QZSS (Quasi-Zenith Satellite System) as they will provide more atmospheric slant observations.

We will introduce the real-time tropospheric slant delay monitoring system, some interesting phenomena retrieved from the system, and impact of filtering of multipath stacking map on the slant delay. We will also demonstrate application of the system to meteorological application such as nowcasting of severe rainfall. The current post-fit residuals and slant tropospheric delay can be monitored with real-time in http://rtnet.info/slant.php.

[Return to Program]

Session D1, Paper #4

Integer Ambiguity Resolution to Improve Accuracy and Convergence of PPP-Inferred PWV Estimates
J. Shi, Y. Gao, University of Calgary, Canada

Precise Point Positioning (PPP) techniques have been widely investigated for precise determination of precipitable water vapor (PWV). Compared to the conventional baseline approach which involves double differencing of measurements from dual receiver stations, PPP approach has several advantages such as: 1) it eliminates the baseline requirement as PPP uses only one receiver station; 2) it eliminates calibration requirement using external PWV techniques as PPP directly estimates absolute tropospheric delay. The initialization time to estimate the PPP-inferred tropospheric delay however is very long. For example, about 4 hours are required to reach the convergence for a static GPS receiver station with an average of 7 visible satellites, a PDOP value 3 and a data sampling interval 30 seconds. In order to dramatically reduce such long initialization time, integer ambiguity resolution is introduced in the PPP-inferred PWV estimation.
PPP integer ambiguity resolution is not feasible before the fractional receiver and satellite biases could be calibrated. If not well considered, these biases would degrade the estimated parameters including coordinates, clocks, troposphere and ambiguities. Several advances have been made in recent years which involve the generation of new orbit and clock corrections to calibrate such biases so that the integer property of the ambiguity terms in the PPP model can be recovered. Previous work has demonstrated the improvement in accuracy for coordinate and clock parameters when the ambiguity terms are fixed to their integer values. No much work has been reported on the accuracy improvement of the tropospheric parameters.

The focus of this paper is to present research results to significantly reduce the initialization time of precise tropospheric delay estimation as well as to improve their accuracy through the development of an integer ambiguity resolution technique based on the PPP decoupled clock model and associated corrections. To our knowledge, no work has been reported to date except some researchers indicate the difficulty for the tropospheric delay parameters to converge to cm-level accuracy with hourly available GPS observations. A new analysis procedure will be presented in this paper which is capable of resolving integer ambiguity terms to further improve the accuracy of the tropospheric delay parameters. A new method using the integer ambiguity constraint is also proposed to significantly reduce the required initialization time for the tropospheric delay parameters.

Data analysis on the accuracy improvement of the tropospheric parameters will be first presented. GPS datasets with different observation lengths (one-hour, two-hour, 12-hour and 24-hour) at a sampling interval of 1 second from the IGS tracking network are processed using the PPP model with decoupled clocks and associated corrections. For the purpose of performance assessment, external references are established using radiosonde data or COSMIC radio occultation (RO) data for tropospheric parameters. For coordinate parameters, the known coordinates of IGS stations are employed. Comprehensive data analysis will be conducted on the float solutions. While the float solution of the 24-hour dataset is used to assess the best obtainable accuracy after all estimates are fully converged, other datasets over shorter periods are used to assess the obtainable accuracy and required initialization time for the tropospheric parameters estimated as float values. The ambiguity fixed solutions are obtained using the LAMBDA method and are analyzed to evaluate the accuracy improvement over the float solutions with different observation lengths.

Data analysis will also be conducted to investigate the reduction of the initialization time for the convergence of the tropospheric parameters through the application of integer ambiguity constraint. Since the conventional approach relies solely on the position constraint, the required initialization time based on the traditional PPP model with real-value ambiguity is about four hours for a static receiver station. For PPP ambiguity resolution with decoupled clocks, additional parameters including receiver decoupled code and phase clocks and wide-lane and N1 ambiguity parameters are also estimated along with the tropospheric delay and position parameters. As a result, a new integer ambiguity constraint is introduced to facilitate the fast convergence of the tropospheric parameters together with or even replacing the position constraint. The convergence time will be reduced to less than one hour with the proposed method.

[Return to Program]

Session D1, Paper #5

High Latitude Ionosphere Scintillations at L5 Band
S. Peng, Virginia Tech; J.Y. Morton, Miami University of Ohio; W. Pelgrum, F. van Graas, Ohio University

The newly launched PRN25 is the first GPS satellite that carries a working L5 signal at the protected ARNS band. The L5 signal is intended to provide better assurance for safety-of-life applications such as aviation. With a relatively low carrier frequency at 1.176GHz and a higher chipping rate at 10.23MHz, L5 is more susceptible to ionosphere scintillation compared to the other civilian signals at the L1 and L2 band. It is well documented that ionosphere scintillations cause increased carrier tracking error and may lead to receiver loss of lock under severe conditions. For aviation and other applications that require continuity and integrity, ionosphere scintillation poses a threat. This is especially true as we enter a new solar maximum period when scintillation activities will increase in both frequency and intensity. The objective of this work is to investigate the impact of ionosphere scintillations on L5 signals.
To achieve this objective and support ionosphere scintillation studies in general, an array of commercial GPS receivers and high end RF front ends GNSS data collection devices have been setup in Gokona, Alaska to collect ionosphere scintillation data. A detailed description of receiver array and data collection system setup is provided in [1]. Among them is a USRP2 experimental software radio receiver front end capable of recording raw GPS L5 and GLONASS L1 data at a sampling frequency of 20MHz for both in-phase and quadrature channels, as well as the L1 and L2C narrow-band GPS signals at 5MHz sampling rate. [2] describes in details the USRP2 front end configuration and its performance evaluation against an instrumentation quality GNSS receiver [3].

In July 2010, several controlled ionosphere scintillation experiments were performed at the receiver site [1]. Artificial scintillations were generated by pointing a high frequency heating beam along PRN 25 and receiver signal propagation path. Simultaneous measurements of the scintillation signals at L1, L2, and L5 were obtained. Starting in August 2010, a simple scintillation event indicator was implemented based on commercial receiver signal strength, C/N0, and carrier phase measurements [1]. This event indicator was continuously compared against a heuristically derived threshold value to trigger the software radio front ends to collect and store raw IF samples, if the event indicator surpassed the threshold value. Several naturally occurred scintillation events have been recorded as a result of this preliminary implementation.

A conventional batch-based software tracking algorithm was implemented to track the L1, L2C, and L5 signals and applied to the controlled and the naturally occurred scintillation data [2]. To ensure fair comparisons, we used the same tracking loop parameters for signals from all three bands. Our preliminary results show while the L5 signal may experience deeper fading compared to L1 and L2C, its S4 index is mostly comparable with and sometimes maybe even slightly smaller than that of L2C. For the artificially generated scintillation, carrier phase scintillation is also more prominent for naturally occurred scintillations. The paper will present implementation of the software tracking algorithms, tracking results for PRN25 L1, L2C, and L5 for both controlled and naturally occurred scintillation, and analysis of the tracking results.

[1] Pelgrum, W., Y. Morton, F. van Graas, S. Gunawardena1, M. Bakich, D. Charney, S. Peng, J. Triplett, A. Vermuru1, P. Vikram, "Measurement and analysis of artificially-generated and natural ionosphere scintillations effects on GNSS signals," Proc. ION ITM, San Diego, Jan. 2011. [2] Peng, S., Y. Morton, "A USRP2-Based Multi-Constellation and Multi-Frequency GNSS Software Receiver for Ionosphere Scintillation Studies," Proc. ION NTM, Jan. 2011. [3] Gunawardena, S., Z. Zhu, F. van Graas, "Triple Frequency RF Front-End for GNSS Instrumentation Receiver Applications," Proc. ION GNSS, Savana, GA, Sept. 2008.

[Return to Program]

Session D1, Paper #6

GPS Carrier Phase and Carrier Phase Spectrum Estimation for Ionosphere Scintillation Studies
L. Zhang, Y. Morton, Miami University

For decades, scientists used radars, rockets, and satellites to study ionosphere scintillations [1-2]. In recent years, GPS receivers have gained popularity as an alternative means for scintillation studies because of their smaller sizes, lower costs, passive nature, and ability to monitor the entire sky continuously [3-4]. To obtain measures of ionosphere scintillations, raw GPS signal measurements such as signal intensity and carrier phases are usually recorded and analyzed. The popular S4 index is the standard deviation of the received signal intensity. Carrier phase standard deviation is another indicator of the amount of signal fluctuations caused by the scintillation at the receiver plane. Additionally, the spectrums of the signal intensity and carrier phase are often computed to reveal the rate of both signal intensity and carrier phase variations [5]. These indices and measures, however, are not the true representations of the GPS signal parameters. GPS receiver signal processing has altered these parameters. This is especially true for the carrier phase estimation generated by the carrier tracking loop and the corresponding detrended carrier standard deviation spectrum. GPS carrier phase measurements are samples of local replica carrier generated by the receiver tracking loop. The carrier phase standard deviation spectrum is obtained from detrended carrier phase measurements and is often directly referred to as the received scintillation signal phase spectrum. Although the GPS receiver tracking loop makes sure that the local replica signal maintains lock with the phase of the input signal, the carrier phase of the input signal and that of the local replica are not same. This is because the GPS receiver carrier tracking loop is a feedback system. The received signal maybe regarded as the input, while the local replica signal is the output. They are related through the feedback system´s transfer function. The local replica signal can be seen as the filtered input signal. In this paper, we first present comparisons between the input carrier phase spectrum with that generated from carrier tracking loop output. Our studies show that there are significant differences between the input and output spectrums and that the differences are determined by the scintillation intensity level and receiver tracking loop parameters such as accumulation time and loop bandwidth. To obtain the true or better approximations of the scintillation signal parameters, we have to "reverse" the output measurements by applying an inverse transfer function. Simulated scintillation signals of various intensity levels based on the model created by Cornell University researchers are generated to illustrate the carrier phase spectral differences [6]. Conventional carrier phase lock loop is implemented to process the input signals. The same simulation signals are then used to demonstrate the effectiveness of applying appropriate inverse filtering to recover the input carrier spectrum. Real scintillation signals are then used in the study to show the difference between the carrier tracking loop output and the output from the inverse filter. The results suggest that it is indeed necessary to perform appropriate inverse filtering, if we want to obtain reasonable estimation of the scintillation signal parameters.

[1]. E. J. Fremouw, R. L. Leadabrand, R. C. Livingston, M. D. Cousins, C. L. Rino, B. C. Fair, and R. A. Long, "Early results from the DNA Wideband Satellite experiment - Complex-signal scintillation," Radio Science, vol. 13, pp. 167-187, Feb. 1978. [2] K. C. Yeh and C. H. Liu, "Radio wave scintillations in the ionosphere," Proceedings of the IEEE, vol. 70, no. 4, pp. 324-360, 1982. [3] A. J. Van Dierendonck, "How GPS receivers measure (or should measure) ionospheric scintillation and TEC and how GPS receivers are affected by the ionosphere," in Proc. 11th Int. Ionospheric Effects Sym., Alexandria, VA, May, 2005. [4] W. Pelgrum, Y. Morton, F. van Graas, S. Gunawardena1, M. Bakich, D. Charney, S. Peng, J. Triplett, A. Vermuru1, P. Vikram, "Measurement and analysis of artificially-generated and natural ionosphere scintillations effects on GNSS signals," Proc. ION ITM, San Diego, Jan. 2011. [5] L. Zhang, Y. Morton, Q. Zhou, F. van Graas, and T. Beach, "Characterization of GNSS signal parameters under ionosphere scintillation conditions using sequential and batch-based tracking algorithms," Proc. IEEE PLANs Conf., Palm Springs, CA, May 2010. [6] T. E. Humphreys, M. L. Psiaki, and P. M. Kintner, Jr., "Modeling the effects of ionospheric scintillation on GPS carrier phase tracking," to appear in IEEE Transactions on Aerospace and Electronic Systems.

[Return to Program]

Session D1, Paper #7

Multi-Domain Analysis of the Impact of Natural and Man-Made Ionosphere Scintillations on GNSS Signal Propagation
W. Pelgrum, Ohio University; Y. Morton, Miami University; F. van Graas, Ohio University

Ionosphere scintillations can cause significant phase and/or amplitude fluctuations of GNSS signals, thereby potentially degrading GNSS tracking performance. Improvements of GNSS receiver robustness under these circumstances require a thorough understanding of the scintillations and their impact on GNSS signal propagation. For this purpose, a high-end GNSS receiver setup has been deployed in July 2010 at the High-Frequency Active Auroral Research Program (HAARP) in Alaska. This facility can locally heat-up specific layers of the ionosphere with 3.6 MW of HF radiation controlled by a 180-element phased array antenna. Next to the ability of creating man-made scintillations, the high-latitude of the HAARP facility also provides a good platform to study naturally occurring ionosphere scintillations.
The measurement setup consists of multiple commercial dual-frequency GPS receivers, a GPS Scintillation receiver, an experimental software radio setup capable of tracking GPS L5 and Glonass, as well as an L1-L2c narrow-band (2.2 MHz) GPS RF data collection setup. The RF measurement data allows for extensive post-processing, enabling 1 kHz independent measurement update rates using open loop tracking. This, in combination with a multi-element GNSS antenna array, will help our understanding of the temporal, spectral, and spatial behavior of GNSS signal fluctuations caused by ionosphere scintillations, and support the development of robust GNSS receivers that are capable of tracking GNSS signals under scintillation conditions.

Numerous natural and man-made ionosphere scintillation events have been automatically captured since the deployment of the aforementioned setup. This paper analyzes the GNSS measurements of these events in various domains: - magnitude of the phase and amplitude fluctuations, - spectral analysis of the phase and amplitude fluctuations, - measured TEC variations during the scintillations, - tomography analysis using the measurements of the multi-element GNSS antenna array, compared with ionosphere drift velocities obtained from HAARP on-site instrumentation measurements - correlation between geomagnetic activity and natural scintillations, - and finally a comparison between natural and man-made scintillations in terms of the presence of phase vs. magnitude fluctuations, and the spectral analysis of those fluctuations.

[Return to Program]

Session D1, Alternate #1

Evaluation and Calibration of GNSS Receivers for Ionospheric Delay and Scintillation Measurements
R. Orus-Perez, R. Prieto-Cerdeira, European Space Agency, ESA/ESTEC, The Netherlands

Physical phenomena on the Earth´s ionosphere may affect GNSS measurements through two main characteristics: group delay (and its variations over time and space) and ionospheric scintillations (fluctuations of amplitude and phase of the radiowave signal due to small-scale variations of the ionospheric refractive index).

Ionospheric group delay is directly related to the integrated electron density along the satellite-receiver path (Total Electron Content) which may be estimated by dual-frequency geometry-free code and phase combinations. Those combinations suffer from unknown satellite and receiver inter-frequency biases and multipath. The later may be reduced through processing of data (such as carrier smoothing) or mitigation algorithms in the receiver (such as narrow correlator). For inter-frequency biases, ionospheric products such as those provided by the International GNSS Service (IGS), observables are processed jointly from a large number of stations iteratively solving for those unknowns. When estimating ionospheric delay for single stations, those unknowns may become a source of important errors, in which case other approaches are required such as data levelling using external global ionospheric maps from IGS or using Broadcast Group Delays in navigation message for satellite and performing calibration on the receiver and applying a model for temperature and aging variations. These various options are discussed in the paper.

Related to ionospheric amplitude and phase scintillation measurements, the quality rely in one hand on receiver characteristics (sampling frequency, loop bandwidths) but also on the method to estimate scintillation parameters: C/No, carrier phase, scintillation indices (S4 and sigma_phi), and later on detrending filter characteristics and cycle slips detection. Comparison of results from several commercial scintillation receiver monitors is analysed using synthetic time-series from different scintillation models, which basically operates in the IQ components of the raw simulated signal, and levels applied through a Spirent Radio Frequency Constellation Simulator. In this sense, the performance of the different receivers over the reference values computed with the models is evaluated jointly with the behaviour of the receiver tracking loop, counting the presence of cycle slips due to scintillation, for different scintillations events that ranges from S4=0.1 to S4=0.9. Additionally, the ability of those receivers with high sampling-frequency for narrowband channel sounding is also demonstrated.

Therefore, with the capability of simulating RF signals having a full controlled environment for the different receivers and the wide range of scintillations effect it is pretended to be able to characterize the performance capabilities and the TEC calibration parameters for the current and future scintillation monitoring receivers based on a standard procedure using the Spirent simulator.

Finally, due to increasing number of scintillation monitoring networks and the number of GNSS systems with different frequencies that can provide scintillation parameters, a format for scintillation data exchange, in line of other exchange formats like RINEX, IONEX and SINEX, is proposed to be discussed.

[Return to Program]

Session D1, Alternate #2

Ionospheric Delay Forecast Using GNSS Data
M. Cueto, E. Sardon, A. Cezon, GMV, Spain; F. Azpilicueta, C. Brunini, La Plata University, GESA, Argentina

The ionosphere affects signals broadcast by Global Navigation Satellite Systems (GNSS) by delaying the propagation of the code carried by the signal. In general ionospheric effects in mid-latitude regions are not severe, causing only gradual variations in ionospheric delays (except in the presence of magnetic storms) whose magnitude can be roughly predicted. However in the equatorial region ionospheric effects are more severe: added to the higher ionospheric delays in this area, the presence of the equatorial anomaly and other equatorial features complicate the ionospheric modeling and make more difficult the prediction of its main features and its effects on GNSS and other applications that depends on the ionospheric status.
Nowadays ionospheric algorithms are mainly focused on post-processing and now cast estimation. However, an ionospheric predicting tool could be very useful for several applications, above all for unstable conditions and for the equatorial region: an algorithm capable of predicting the ionospheric behavior in advance could be used to set up early warnings for different uses, among others civil aviation based on Ground Based Augmentation Systems (GBAS) or Satellite Based Augmentation System (SBAS), the protection of valuable communication satellites from space weather adverse conditions, as well as other technologies affected by space weather, including geophysical exploration and protection of long distance pipelines, HF radio systems, communication and surveillance systems, spacecraft operations, defence needs and alarm systems for safety applications.

Taking into account the aforementioned need, a new ionospheric forecasting tool based on the use of GNSS measurements has been developed. This tool provides predicted ionospheric delays for a set of Ionospheric Grid Point (IGP) located in the service area defined by the user. The ionospheric forecast algorithm is based on magicSBAS tool, which is an innovative tool developed by GMV which computes SBAS corrections and additional information required by a SBAS system in real-time to be broadcast to SBAS users. magicSBAS implements multi-constellation (GPS, GLONASS) state-of-the-art algorithms for precise orbit determination and time synchronization, ionospheric delay estimation, SBAS wide area correction computation and SBAS integrity determination. The tool uses as input GNSS raw data in several formats (such as Networked Transport of RTCM via Internet Protocol, NTRIP, Receiver INdependent Exchange, RINEX and European Geostationary Navigation Overlay Service, EGNOS) and provides as output SBAS information compliant with SBAS international standards (International Civil Aviation Organization Standards and Recommended Practices and Radio Technical Commission for Aeronautics DO-229D Minimum Operational Performance Standards) in two ways, as SBAS binary messages for GEO (Geostationary Earth Orbit) broadcast and as SBAS signal-in-space through Internet (SISNET).

The forecast algorithm used for ionospheric delays prediction is based on the ionospheric delay estimation from previous epochs using GNSS data and the main dependence of ionospheric delays on solar and magnetic conditions. On account of the fact that the ionospheric behaviour is highly dependent on the region of the Earth, different algorithmic modifications have been implemented in GMV´s magicSBAS ionospheric algorithms to be able to estimate and forecast ionospheric delays worldwide, adapting the ionospheric algorithms to the ionospheric characteristics at different latitudinal regions.

Simulated data provided by a GNSS End-to-end simulator developed by GMV have been used to check the performances of the ionospheric delay prediction tool. On one hand, the predicted ionospheric delays for different forecast periods have been compared with the ionospheric delay values estimated by magicSBAS for those predicted epochs. On the other hand, the aforementioned ionospheric delay values have also been compared with the simulated values for the same epochs.

This paper shows how the new ionospheric delay forecasting tool is able to provide very good forecasting results for middle latitudes, and even for those equatorial latitudes where the ionosphere is much more complicated the results obtained are quite encouraging. Forecast periods from 1/2 hour to few hours are provided, showing that the prediction periods for which adequate forecast ionospheric results are achieved at middle latitudes are longer than those for equatorial latitudes, due to the higher (and less predictable) spatial and temporal variability in this region. Several examples of comparisons between predicted, estimated and simulated ionospheric delays for different locations using Global Navigation Satellite System data for different latitudinal regions and space weather conditions are also provided in this paper.

[Return to Program]

Session D1, Alternate #3

Ionospheric Imaging Using GNSS: A New Approach for Canadian Polar Regions
O. Al-Fanek, S. Skone, University of Calgary, Canada

The study of electron density distribution is of interest to many of scientists. The solar maximum is expected to occur in 2013 with solar activity that will affect radio communication and navigation applications significantly. Some of the problems associated with solar-cycle peaks include 1) loss of aircraft communications over popular North America-Asia polar routes forcing airlines to use a different, longer, more costly routes; 2) increased radiation risk to humans in space; and 3) damage to electrical transmission equipment. During solar maximum there are higher and more variable ionospheric electron densities which offer scientists an enhanced opportunity to study the ionosphere and the complex solar-terrestrial relationship. The importance of ionosphere arises in its effect on radio waves, especially the ones transmitted by Global Navigation Satellite Systems (GNSS) such as the Global Positioning System (GPS). Ionospheric effects are considered one of the main error sources in GNSS. As the waves propagate through the ionosphere, the signals are slowed down, causing an increase in the propagation time compared to that of free space. This delay, referred to as ionospheric delay, is proportional to the line integral of the free electron distribution in a 1-m2 column along the path of the signal from the satellite to the receiver, referred to as Total Electron Content (TEC), and inversely proportional to the square of the signal frequency being transmitted. Over the past two decades, the impact on GNSS signals has been exploited to remotely sense the Earth´s ionosphere and investigate its characteristics with near global coverage. Since the ionosphere is a dispersive medium and GPS transmits signals on two frequencies: 1575.42 MHz (L1) and 1227.60 MHz (L2) with future plans for an additional broadcast at 1176.45 MHZ (L5), a dual frequency receiver can be used to derive highly accurate TEC information along each satellite-to-receiver line of sight (slant TEC). Three-dimensional ionosphere modeling using tomography techniques has been applied previously to investigate the vertical extent of electron density variations. In such methods the GNSS observations (often within ground-based networks) are manipulated to estimate line integrals of a parameter known as Slant Total Electron Content (STEC). These integrated quantities are then expressed in terms of unknown values of ionospheric electron density. For example, some approaches divide the ionosphere into voxels and assume the electron density to be constant within each cell. Inversion algorithms are used to estimate the electron density values. However, this problem is mixed-determined; that is there are regions where the data overdetermines parts of the solution but underdetermines other parts. Also, the ray paths are not available in all directions. For example there are no horizontal ray paths for GPS signals observed in a ground-based network; therefore the vertical resolution is not as good as the horizontal resolution. In this paper, a novel Computerized Ionospheric tomographic (CIT) reconstruction technique based on Empirical Orthogonal Functions (EOF) and Spherical Cap Harmonics (SCH) is developed. This model divides the ionosphere into voxels and the ionospheric parameter is assumed to be constant within a cell. To reduce the number of unknowns, a functional based model is used to represent the electron density in space. The functional based model uses EOF and SCH to describe the vertical and horizontal distribution of the electron density, respectively. The underlying model is obtained from the International Reference Ionosphere (IRI) and the necessary measurements are obtained from earth-based and satellite-based GPS databases. Based on the IRI-2007 model, a basis is formed using EOF for the required location and the time of interest. Selecting the first few basis vectors corresponding to the most significant singular values and combining them with SCH, the 3-D CIT is formulated as a weighted least squares estimation problem of the SCH coefficients. The region of interest in this paper is the polar cap, due to its importance and effect on radio propagation and communication especially during solar maximum. The polar cap region is lacking spatial resolution of TEC measurements due to the orbit limitations of spaced-based measurements and sparse networks providing such measurements. To overcome these limitations, the Canadian high Arctic Ionospheric Network (CHAIN) was designed and took advantage of Canada´s most accessible landmass in the high arctic regions. The network consists of 10 (eight within the polar cap region) high data rate Global Positioning System (GPS) ionospheric scintillation and TEC monitors and six Canadian Advanced Digital Ionosondes. Having access to such data, and applying the new tomographic technique, a clear understanding of the quality and limitations of the technique, particularly in high-latitude regions, is achieved. Spatial and temporal resolution are quantified and the benefits of three-dimensional ionospheric imaging in the polar region are investigated. Preliminary results indicate advantages for a number of practical applications. Implementation of the new tomographic method allows determination of various ionospheric parameters. For example the peak height of electron density is used in monitoring and predicting HF communication capabilities for aircraft polar routes. Imaging the three-dimensional extent of ionospheric structures allows identification of gradients associated with development of ionospheric scintillations. Velocities of such polar cap patches can be estimated and evolution of scintillation regions predicted. Such practical applications are demonstrated using the new imaging techniques.

[Return to Program]