Quantcast


CURATOR
A pinboard by
Anna Riddell

An interest for science and the changing size and shape of the Earth piqued my curiosity and led me to do a degree in Surveying and Spatial Sciences at The University of Tasmania (UTas), Australia. Upon completion of my degree, I was successful in gaining a graduate placement at Geoscience Australia (GA). I am now completing a PhD at UTas.

I am currently undertaking a PhD in the field of Geodesy at the University of Tasmania, working with Prof. Matt King and Dr. Christopher Watson.

PINBOARD SUMMARY

Processing GNSS data requires complex processing software and models to account for uncertainty.

Processing GNSS data from continuously operating sites requires complex processing software and multi-layered geophysical models as correction inputs. Models applied during processing include: satellite orbit and clock models; solid earth, ocean tide and pole tide models; antenna models; ionosphere and troposphere models. All of these models contain complexity that cannot perfectly represent reality, but are the current best practice. Each model contains uncertainty and errors which accumulate during the processing and result as 'noise' in the coordinate time series and velocity product.

6 ITEMS PINNED

MIDAS robust trend estimator for accurate GPS station velocities without step detection.

Abstract: Automatic estimation of velocities from GPS coordinate time series is becoming required to cope with the exponentially increasing flood of available data, but problems detectable to the human eye are often overlooked. This motivates us to find an automatic and accurate estimator of trend that is resistant to common problems such as step discontinuities, outliers, seasonality, skewness, and heteroscedasticity. Developed here, Median Interannual Difference Adjusted for Skewness (MIDAS) is a variant of the Theil-Sen median trend estimator, for which the ordinary version is the median of slopes vij  = (xj-xi )/(tj-ti ) computed between all data pairs i > j. For normally distributed data, Theil-Sen and least squares trend estimates are statistically identical, but unlike least squares, Theil-Sen is resistant to undetected data problems. To mitigate both seasonality and step discontinuities, MIDAS selects data pairs separated by 1 year. This condition is relaxed for time series with gaps so that all data are used. Slopes from data pairs spanning a step function produce one-sided outliers that can bias the median. To reduce bias, MIDAS removes outliers and recomputes the median. MIDAS also computes a robust and realistic estimate of trend uncertainty. Statistical tests using GPS data in the rigid North American plate interior show ±0.23 mm/yr root-mean-square (RMS) accuracy in horizontal velocity. In blind tests using synthetic data, MIDAS velocities have an RMS accuracy of ±0.33 mm/yr horizontal, ±1.1 mm/yr up, with a 5th percentile range smaller than all 20 automatic estimators tested. Considering its general nature, MIDAS has the potential for broader application in the geosciences.

Pub.: 27 Sep '16, Pinned: 07 Aug '17

Assessment of second- and third-order ionospheric effects on regional networks: case study in China with longer CMONOC GPS coordinate time series

Abstract: Abstract Higher-order ionospheric (HOI) delays are one of the principal technique-specific error sources in precise global positioning system analysis and have been proposed to become a standard part of precise GPS data processing. In this research, we apply HOI delay corrections to the Crustal Movement Observation Network of China’s (CMONOC) data processing (from January 2000 to December 2013) and furnish quantitative results for the effects of HOI on CMONOC coordinate time series. The results for both a regional reference frame and global reference frame are analyzed and compared to clarify the HOI effects on the CMONOC network. We find that HOI corrections can effectively reduce the semi-annual signals in the northern and vertical components. For sites with lower semi-annual amplitudes, the average decrease in magnitude can reach 30 and 10 % for the northern and vertical components, respectively. The noise amplitudes with HOI corrections and those without HOI corrections are not significantly different. Generally, the HOI effects on CMONOC networks in a global reference frame are less obvious than the results in the regional reference frame, probably because the HOI-induced errors are smaller in comparison to the higher noise levels seen when using a global reference frame. Furthermore, we investigate the combined contributions of environmental loading and HOI effects on the CMONOC stations. The largest loading effects on the vertical displacement are found in the mid- to high-latitude areas. The weighted root mean square differences between the corrected and original weekly GPS height time series of the loading model indicate that the mass loading adequately reduced the scatter on the CMONOC height time series, whereas the results in the global reference frame showed better agreements between the GPS coordinate time series and the environmental loading. When combining the effects of environmental loading and HOI corrections, the results with the HOI corrections reduced the scatter on the observed GPS height coordinates better than the height when estimated without HOI corrections, and the combined solutions in the regional reference frame indicate more preferred improvements. Therefore, regional reference frames are recommended to investigate the HOI effects on regional networks.AbstractHigher-order ionospheric (HOI) delays are one of the principal technique-specific error sources in precise global positioning system analysis and have been proposed to become a standard part of precise GPS data processing. In this research, we apply HOI delay corrections to the Crustal Movement Observation Network of China’s (CMONOC) data processing (from January 2000 to December 2013) and furnish quantitative results for the effects of HOI on CMONOC coordinate time series. The results for both a regional reference frame and global reference frame are analyzed and compared to clarify the HOI effects on the CMONOC network. We find that HOI corrections can effectively reduce the semi-annual signals in the northern and vertical components. For sites with lower semi-annual amplitudes, the average decrease in magnitude can reach 30 and 10 % for the northern and vertical components, respectively. The noise amplitudes with HOI corrections and those without HOI corrections are not significantly different. Generally, the HOI effects on CMONOC networks in a global reference frame are less obvious than the results in the regional reference frame, probably because the HOI-induced errors are smaller in comparison to the higher noise levels seen when using a global reference frame. Furthermore, we investigate the combined contributions of environmental loading and HOI effects on the CMONOC stations. The largest loading effects on the vertical displacement are found in the mid- to high-latitude areas. The weighted root mean square differences between the corrected and original weekly GPS height time series of the loading model indicate that the mass loading adequately reduced the scatter on the CMONOC height time series, whereas the results in the global reference frame showed better agreements between the GPS coordinate time series and the environmental loading. When combining the effects of environmental loading and HOI corrections, the results with the HOI corrections reduced the scatter on the observed GPS height coordinates better than the height when estimated without HOI corrections, and the combined solutions in the regional reference frame indicate more preferred improvements. Therefore, regional reference frames are recommended to investigate the HOI effects on regional networks.

Pub.: 26 Sep '16, Pinned: 07 Aug '17

Evaluating mass loading products by comparison to GPS array daily solutions

Abstract: We evaluate the benefit of different global geophysical loading products on the internal scatter of GPS position time-series from 240 globally distributed sites. We focus on the non-tidal atmospheric pressure loading predicted from NASA's Modern-Era Retrospective Analysis for Research and Applications (MERRA-NATML) and the European Center for Medium-Range Weather Forecasts operational model (ECMWF-NATML), non-tidal ocean loading predicted from the Ocean Model for Circulation and Tides model (OMCT-NTOL), and the continental water storage loading predicted from the MERRA model (MERRA-CWSL) and the GFZ's Land Surface Discharge Model (LSDM-CWSL), respectively. The result shows that the root mean square (RMS) discrepancy of different CWSL models is larger than that of NATML models in the vertical component due to the varying model parameters and input data choices. We discuss the performance of different loading models and their combination to reduce the weighted RMS of GPS up-coordinates. MERRA-NATML & OMCT-NTOL & MERRA-CWSL reduced the weighted RMS (WRMS) in 96 per cent (JPL solutions) and 86 per cent (SOPAC solutions) of the cases, and ECMWF-NATML & OMCT-NTOL & LSDM-CWSL reduced the WRMS in 95 per cent (JPL solutions) and 88 per cent (SOPAC solutions) of the cases. The result shows that local effects and technical uncertainties in GPS time-series hamper the meaningful comparison between GPS observations and mass loading models. Hence, simply using the RMS of the time-series as the assessment criteria may lead to biased comparison results. Nonetheless, we give a detailed comparison (differences in phase and amplitude at seasonal timescales) for eight representative stations located adjacent to great rivers, lakes and reservoirs. We find that LSDM can provide a complementary model to study the small-scale hydrological loading like loading extremes along river channels. However, such small-scale hydrological loading effects are still instable to be modelled in some regions with its current accuracy. Finally, we discuss the impacts of mass loading corrections on the velocity and noise estimates. The noise reductions have the consistent performance as WRMS reductions for most sites, whereas some sites have their noise increased even though GPS signal WRMS is decreased there, suggesting that our posterior correction is potentially feasible, but not sufficient.

Pub.: 08 Nov '16, Pinned: 07 Aug '17

Ionospheric error contribution to GNSS single-frequency navigation at the 2014 solar maximum

Abstract: Abstract For single-frequency users of the global satellite navigation system (GNSS), one of the main error contributors is the ionospheric delay, which impacts the received signals. As is well-known, GPS and Galileo transmit global models to correct the ionospheric delay, while the international GNSS service (IGS) computes precise post-process global ionospheric maps (GIM) that are considered reference ionospheres. Moreover, accurate ionospheric maps have been recently introduced, which allow for the fast convergence of the real-time precise point position (PPP) globally. Therefore, testing of the ionospheric models is a key issue for code-based single-frequency users, which constitute the main user segment. Therefore, the testing proposed in this paper is straightforward and uses the PPP modeling applied to single- and dual-frequency code observations worldwide for 2014. The usage of PPP modeling allows us to quantify—for dual-frequency users—the degradation of the navigation solutions caused by noise and multipath with respect to the different ionospheric modeling solutions, and allows us, in turn, to obtain an independent assessment of the ionospheric models. Compared to the dual-frequency solutions, the GPS and Galileo ionospheric models present worse global performance, with horizontal root mean square (RMS) differences of 1.04 and 0.49 m and vertical RMS differences of 0.83 and 0.40 m, respectively. While very precise global ionospheric models can improve the dual-frequency solution globally, resulting in a horizontal RMS difference of 0.60 m and a vertical RMS difference of 0.74 m, they exhibit a strong dependence on the geographical location and ionospheric activity.AbstractFor single-frequency users of the global satellite navigation system (GNSS), one of the main error contributors is the ionospheric delay, which impacts the received signals. As is well-known, GPS and Galileo transmit global models to correct the ionospheric delay, while the international GNSS service (IGS) computes precise post-process global ionospheric maps (GIM) that are considered reference ionospheres. Moreover, accurate ionospheric maps have been recently introduced, which allow for the fast convergence of the real-time precise point position (PPP) globally. Therefore, testing of the ionospheric models is a key issue for code-based single-frequency users, which constitute the main user segment. Therefore, the testing proposed in this paper is straightforward and uses the PPP modeling applied to single- and dual-frequency code observations worldwide for 2014. The usage of PPP modeling allows us to quantify—for dual-frequency users—the degradation of the navigation solutions caused by noise and multipath with respect to the different ionospheric modeling solutions, and allows us, in turn, to obtain an independent assessment of the ionospheric models. Compared to the dual-frequency solutions, the GPS and Galileo ionospheric models present worse global performance, with horizontal root mean square (RMS) differences of 1.04 and 0.49 m and vertical RMS differences of 0.83 and 0.40 m, respectively. While very precise global ionospheric models can improve the dual-frequency solution globally, resulting in a horizontal RMS difference of 0.60 m and a vertical RMS difference of 0.74 m, they exhibit a strong dependence on the geographical location and ionospheric activity.

Pub.: 24 Nov '16, Pinned: 07 Aug '17

CODE’s five-system orbit and clock solution—the challenges of multi-GNSS data analysis

Abstract: Abstract This article describes the processing strategy and the validation results of CODE’s MGEX (COM) orbit and satellite clock solution, including the satellite systems GPS, GLONASS, Galileo, BeiDou, and QZSS. The validation with orbit misclosures and SLR residuals shows that the orbits of the new systems Galileo, BeiDou, and QZSS are affected by modelling deficiencies with impact on the orbit scale (e.g., antenna calibration, Earth albedo, and transmitter antenna thrust). Another weakness is the attitude and solar radiation pressure (SRP) modelling of satellites moving in the orbit normal mode—which is not yet correctly considered in the COM solution. Due to these issues, we consider the current state COM solution as preliminary. We, however, use the long-time series of COM products for identifying the challenges and for the assessment of model-improvements. The latter is demonstrated on the example of the solar radiation pressure (SRP) model, which has been replaced by a more generalized model. The SLR validation shows that the new SRP model significantly improves the orbit determination of Galileo and QZSS satellites at times when the satellite’s attitude is maintained by yaw-steering. The impact of this orbit improvement is also visible in the estimated satellite clocks—demonstrating the potential use of the new generation satellite clocks for orbit validation. Finally, we point out further challenges and open issues affecting multi-GNSS data processing that deserves dedicated studies.AbstractThis article describes the processing strategy and the validation results of CODE’s MGEX (COM) orbit and satellite clock solution, including the satellite systems GPS, GLONASS, Galileo, BeiDou, and QZSS. The validation with orbit misclosures and SLR residuals shows that the orbits of the new systems Galileo, BeiDou, and QZSS are affected by modelling deficiencies with impact on the orbit scale (e.g., antenna calibration, Earth albedo, and transmitter antenna thrust). Another weakness is the attitude and solar radiation pressure (SRP) modelling of satellites moving in the orbit normal mode—which is not yet correctly considered in the COM solution. Due to these issues, we consider the current state COM solution as preliminary. We, however, use the long-time series of COM products for identifying the challenges and for the assessment of model-improvements. The latter is demonstrated on the example of the solar radiation pressure (SRP) model, which has been replaced by a more generalized model. The SLR validation shows that the new SRP model significantly improves the orbit determination of Galileo and QZSS satellites at times when the satellite’s attitude is maintained by yaw-steering. The impact of this orbit improvement is also visible in the estimated satellite clocks—demonstrating the potential use of the new generation satellite clocks for orbit validation. Finally, we point out further challenges and open issues affecting multi-GNSS data processing that deserves dedicated studies.

Pub.: 24 Nov '16, Pinned: 07 Aug '17

GPS code phase variations (CPV) for GNSS receiver antennas and their effect on geodetic parameters and ambiguity resolution

Abstract: Abstract Precise navigation and geodetic coordinate determination rely on accurate GNSS signal reception. Thus, the receiver antenna properties play a crucial role in the GNSS error budget. For carrier phase observations, a spherical radiation pattern represents an ideal receiver antenna behaviour. Deviations are known as phase centre corrections. Due to synergy of carrier and code phase, similar effects on the code exist named code phase variations (CPV). They are mainly attributed to electromagnetic interactions of several active and passive elements of the receiver antenna. Consequently, a calibration and estimation strategy is necessary to determine the shape and magnitudes of the CPV. Such a concept was proposed, implemented and tested at the Institut für Erdmessung. The applied methodology and the obtained results are reported and discussed in this paper. We show that the azimuthal and elevation-dependent CPV can reach maximum magnitudes of 0.2–0.3 m for geodetic antennas and up to maximum values of 1.8 m for small navigation antennas. The obtained values are validated by dedicated tests in the observation and coordinate domain. As a result, CPV are identified to be antenna- related properties that are independent from location and time of calibration. Even for geodetic antennas when forming linear combinations the CPV effect can be amplified to values of 0.4–0.6 m. Thus, a significant fractional of the Melbourne–Wübbena linear combination. A case study highlights that incorrect ambiguity resolution can occur due to neglecting CPV corrections. The impact on the coordinates which may reach up to the dm level is illustrated.AbstractPrecise navigation and geodetic coordinate determination rely on accurate GNSS signal reception. Thus, the receiver antenna properties play a crucial role in the GNSS error budget. For carrier phase observations, a spherical radiation pattern represents an ideal receiver antenna behaviour. Deviations are known as phase centre corrections. Due to synergy of carrier and code phase, similar effects on the code exist named code phase variations (CPV). They are mainly attributed to electromagnetic interactions of several active and passive elements of the receiver antenna. Consequently, a calibration and estimation strategy is necessary to determine the shape and magnitudes of the CPV. Such a concept was proposed, implemented and tested at the Institut für Erdmessung. The applied methodology and the obtained results are reported and discussed in this paper. We show that the azimuthal and elevation-dependent CPV can reach maximum magnitudes of 0.2–0.3 m for geodetic antennas and up to maximum values of 1.8 m for small navigation antennas. The obtained values are validated by dedicated tests in the observation and coordinate domain. As a result, CPV are identified to be antenna- related properties that are independent from location and time of calibration. Even for geodetic antennas when forming linear combinations the CPV effect can be amplified to values of 0.4–0.6 m. Thus, a significant fractional of the Melbourne–Wübbena linear combination. A case study highlights that incorrect ambiguity resolution can occur due to neglecting CPV corrections. The impact on the coordinates which may reach up to the dm level is illustrated.

Pub.: 24 Dec '16, Pinned: 07 Aug '17