High-precision Global Navigation Satellite System Research Articles

High-resolution reconstruction and correction of FY-4A precipitable water vapor in China using back propagation neural network

Precipitable water vapor (PWV) with high spatial and temporal resolution is essential for a deeper understanding of the study area's hydrological cycle and climate change. This study uses the European Center for Medium-Range Weather Forecasts Reanalysis 5 (ERA5) PWV as auxiliary verification data and proposes a reconstruction method based on back propagation neural network (BPNN) downscaling. Combining with high temporal resolution meteorological elements, digital elevation model (DEM), location, and time, reconstructed the Fengyun-4A (FY-4A) PWV data set with high spatial and temporal resolution in China was obtained. Subsequently, using the characteristics of FY-4A PWV with DEM, location, and time, a BPNN-based correction method is proposed to correct the reconstructed FY-4A PWV using the high-precision global navigation satellite system (GNSS) PWV. The results showed that the root mean square error (RMSE), Bias, and Pearson correlation coefficient (R) of the reconstructed model with respect to the FY-4A PWV are 1.32 mm, 0 mm, and 0.99, respectively. Compared with the auxiliary verification data ERA5 PWV, the RMSE, Bias, and R of the reconstructed FY-4A PWV are 0.85 mm, 0.18 mm, and 0.99, respectively. Compared with the GNSS PWV, the RMSE of the corrected FY-4A PWV is 2.44 mm, which is 28.86% lower than that of the reconstructed FY-4A PWV, the Bias is improved from −1.46 mm to 0 mm, and the R is enhanced from 0.98 to 0.99. The reconstruction and correction method of FY-4A PWV with high spatial and temporal resolution proposed in this study dramatically improves the completeness and accuracy of FY-4A PWV in China. It lays the foundation for studying rapid changes in China's hydrological cycle and climate.

Improving GNSS PPP-RTK through global forecast system zenith wet delay augmentation

The precise point positioning real-time kinematic (PPP-RTK) is a high-precision global navigation satellite system (GNSS) positioning technique that combines the advantages of wide-area coverage in precise point positioning (PPP) and of rapid convergence in real-time kinematic (RTK). However, the PPP-RTK convergence is still limited by the precision of slant ionospheric delays and tropospheric zenith wet delay (ZWD), which affects the PPP-RTK network parameters estimation and user positioning performance. The present study aims to construct a PPP-RTK model augmented with a priori ZWD values derived from the global forecast system (GFS) product (a global numerical weather prediction (NWP) model) to improve the PPP-RTK performance. This study gives a priori ZWD values and conversion based on the GFS products, and the full-rank GFS-augmented undifferenced and uncombined (UDUC) PPP-RTK network model is derived. To verify the performance of GFS-augmented UDUC PPP-RTK, a comprehensive evaluation using 10-day GNSS observation data from three different GNSS station networks in the United States (US), Australia, and Europe is conducted. The results show that with the GFS ZWD a priori information, PPP-RTK performance significantly improves at the initial filtering stage, but this advantage gradually decays over time. Based on 10-day positioning results for all user stations, the GFS ZWD-augmented PPP-RTK approach reduces the average convergence time by 46% from 10.0 to 5.4 min, the three-dimensional root-mean-square (3D-RMS) error by 5.7% from 3.5 to 3.3 cm, and the time to first fix (TTFF) value by 35.8% from 6.7 to 4.3 min, all when compared to the traditional PPP-RTK without GFS ZWD constraints.

ConvGRU-MHM: a CNN GRU-enhanced MHM for mitigating GNSS multipath

In high-precision global navigation satellite system (GNSS) short-baseline positioning, multipath is the main source of errors. If the station environment is quasi-static, repeat periods of satellites can be utilized to generate time- or space-dependent multipath models to mitigate multipaths. However, two general problems are associated with the multipath models constructed based on satellite mechanics: (1) an accuracy decrease occurs when the above models are applied to multipath mitigation over a long time-span; (2) when constructing the spatial and temporal grids of the satellite-based spatially dependent multipath model, it is challenging to balance computational efficiency and spatial resolution. We propose a convolutional neural network-gated recurrent unit enhanced multipath hemispherical map (ConvGRU-MHM) in the observational domain to address these problems. The proposed method directly mines the deep features of elevation, azimuth angle, and multipath and the mapping relationship between these to establish a real-time prediction model. The predicted multipath is obtained and returned to the observation equation for multipath mitigation when the real-time position of the satellite is placed in the pre-trained model. We compared the multipath mitigation performance of sidereal filtering and a MHM with that of the ConvGRU-MHM to demonstrate the advantages of the proposed method. The experimental results are as follows: (1) in the short time-span (first 20 d), the mean accuracy improvements of the ConvGRU-MHM in the E/N/U direction performed better than those of the SF and MHM; and (2) in the long-term time (after 50 d), the mean accuracy improvements of the ConvGRU-MHM in the E/N/U direction are higher than that of the SF and MHM by 10%–20%. As a lightweight model, the ConvGRU-MHM can effectively improve the measurement accuracy of GNSS real-time monitoring in fields, such as deformation monitoring and seismic research.

A global grid model for the estimation of zenith tropospheric delay considering the variations at different altitudes

Abstract. The accuracy of tropospheric delay correction heavily depends on the quality of the tropospheric model, and the zenith tropospheric delay (ZTD) is an important factor affecting the tropospheric delay. Therefore, it is essential to establish a precise ZTD empirical model. The existing ZTD models are constrained by a single fitting function, lack consideration for daily cycle variations, and rely solely on data with one resolution for modeling. To address these limitations, we proposed a global piecewise ZTD empirical grid (GGZTD-P) model. This model considers the daily cycle variation and latitude factor of ZTD, using the sliding window algorithm based on fifth-generation European Centre for Medium-Range Weather Forecasts atmospheric reanalysis data (ERA5). The ZTD data from 545 radiosonde stations and the second Modern-Era Retrospective analysis for Research and Applications (MERRA-2) atmospheric reanalysis data are used to validate the accuracy of the GGZTD-P model. The results indicate that the GGZTD-P model outperforms the global pressure and temperature 3 (GPT3) model, exhibiting 26 % and 53 % lower bias and rms, respectively, when using radiosonde stations as reference values. Furthermore, when evaluated using MERRA-2 atmospheric reanalysis data, the GGZTD-P model consistently exhibits superior performance across various latitude regions. It is expected that the application of this new model will provide improved services for high-precision global navigation satellite system (GNSS) positioning and GNSS meteorology.

Multi-GNSS ultra-rapid orbit determination through epoch-parallel processing

High-precision Global Navigation Satellite Systems (GNSS) orbits are critical for real-time clock estimation and precise positioning service; however, the prediction error grows gradually with the increasing prediction session. In this study, we present a new efficient precise orbit determination (POD) strategy referred to as the epoch-parallel processing to reduce the orbit update latency, in which a 24-h processing job is split into several sub-sessions that are processed in parallel and then stacked to solve and recover parameters subsequently. With a delicate handling of parameters crossing different sub-sessions, such as ambiguities, the method is rigorously equivalent to the one-session batch solution, but is much more efficient, halving the time-consuming roughly. Together with paralleling other procedures such as orbit integration and using open multi-processing (openMP), the multi-GNSS POD of 120 satellites using 90 stations can be fulfilled within 30 min. The lower update latency enables users to access orbits closer to the estimation part, that is, 30–60-min prediction with a 30-min update latency, which significantly improves the orbit quality. Compared to the hourly updated orbit, the averaged 1D RMS values of predicted orbit in terms of overlap for GPS, GLONASS, Galileo, and BDS MEO are improved by 39%, 35%, 41%, and 37%, respectively, and that of BDS GEO and IGSO satellites is improved by 47%. We also demonstrate that the boundary discontinuities of half-hourly orbit are within 2 cm for the GPS, GLONASS, and Galileo satellites, and for BDS the values are 2.6, 15.5, and 9.8 cm for MEO, GEO, and IGSO satellites, respectively. This method can also be implemented for any batch-based GNSS processing to improve the efficiency.

Machine learning-based multipath modeling in spatial domain applied to GNSS short baseline processing

Multipath is the main unmodeled error source hindering high-precision Global Navigation Satellite System data processing. Conventional multipath mitigation methods, such as sidereal filtering (SF) and multipath hemispherical map (MHM), have certain disadvantages: They are either not easy to use or not effective enough for multipath mitigation. In this study, we propose a machine learning (ML)-based multipath mitigation method. Multipath modeling was formulated as a regression task, and the multipath errors were fitted with respect to azimuth and elevation in the spatial domain. We collected 30 days of 1 Hz GPS data to validate the proposed method. In total, five short baselines were formed and multipath errors were extracted from the postfit residuals. ML-based multipath models, as well as observation-domain SF and MHM models, were constructed using 5 days of residuals before the target day and later applied for multipath correction. It was found that the XGBoost (XGB) method outperformed SF and MHM. It achieved the highest residual reduction rates, which were 24.9%, 36.2%, 25.5% and 20.4% for GPS P1, P2, L1 and L2 observations, respectively. After applying the XGB-based multipath corrections, kinematic positioning precisions of 1.6 mm, 1.9 mm and 4.5 mm could be achieved in east, north and up components, respectively, corresponding to 20.0%, 17.4% and 16.7% improvements compared to the original solutions. The effectiveness of the ML-based multipath model was further validated using 30 s sampling data and data from a low-cost device. We conclude that the ML-based multipath mitigation method is effective, easy to use, and can be easily extended by adding auxiliary input features, such as signal-to-noise ratio, during model training.

GNSS Receiver Antenna Absolute Field Calibration System Development: Testing and Preliminary Results

For high-precision Global Navigation Satellite Systems (GNSS) positioning based on carrier-phase measurements, knowledge of the GNSS receiver antenna electrical signal reception characteristics, i.e., phase center, is crucial. Numerous studies have led to the understanding of the influence of GNSS receiver antenna phase center corrections (PCCs) on GNSS positioning accuracy and other estimated parameters (e.g., receiver clock estimates, ambiguities, etc.). With the goal of determining the PCC model of GNSS receiver antennas, only a few antenna calibration systems/facilities are in operation or under development worldwide. The International GNSS Service (IGS) publishes type-mean PCC models for almost all geodetic-grade GNSS antennas. However, the type-mean models are not perfect and do not fully reflect the signal reception properties of individual GNSS receiver antennas. Relevant published scientific research has shown that the application of individual PCC models significantly improves the accuracy of GNSS positioning and other estimated parameters. In this article, the new automated GNSS antenna calibration system, recently developed at the Laboratory for Measurements and Measuring Technique (LMMT) of the Faculty of Geodesy of the University of Zagreb in Croatia, is presented. The developed system is an absolute field calibration system based on the utilization of a Mitsubishi MELFA 6-axis industrial robot. During calibration, the robot tilts and rotates the GNSS antenna under test (AUT) around a fixed point within the antenna. The antenna PCC modelling is based on time-differenced double-difference carrier-phase observations. Our preliminary results for the Global Positioning System (GPS) L1 (G01) frequency show a submillimeter repeatability of the estimated PCC model and a submillimeter agreement with the Geo++ GmbH calibration results.

A Relative Field Antenna Calibration Method Designed for Low-Cost GNSS Antennas by Exploiting Triple-Differenced Measurements

Performing the high-precision Global Navigation Satellite System (GNSS) applications with low-cost antennas is an up-and-coming research field. However, the antenna-induced phase biases, i.e., phase center corrections (PCCs), of the low-cost antennas can be up to centimeters and need to be calibrated in advance. The relative field antenna calibration method is easy to conduct, but the classical procedure entails integer ambiguity resolution, which may face the problem of low success rate under the centimeter-level PCCs. In this contribution, we designed a relative field calibration method suitable for the low-cost GNSS antennas. The triple-differencing operations were utilized to eliminate the carrier-phase ambiguities and then construct PCC measurements; the time-differencing interval was set to a relatively long time span, such as one hour, and the reference satellite was selected according to the angular distance it passed over during a time-differencing interval. To reduce the effect of significant triple-differencing noise, a weight setting method based on the area of a spherical quadrilateral was proposed for the spherical harmonics fitting process. The duration of the data collection with respect to GPS and BDS was discussed. The performance of the proposed method was assessed with real GPS and BDS observations and a variety of simulated phase patterns, showing that calibration results could be obtained with millimeter-level accuracy. The impact of cycle slip and elevation mask angle on the calibration results was also analyzed.

Detecting Maritime GPS Spoofing Attacks Based on NMEA Sentence Integrity Monitoring

Today’s maritime transportation relies on global navigation satellite systems (GNSSs) for accurate navigation. The high-precision GNSS receivers on board modern vessels are often considered trustworthy. However, due to technological advances and malicious activities, this assumption is no longer always true. Numerous incidents of tampered GNSS signals have been reported. Furthermore, researchers have demonstrated that manipulations can be carried out even with inexpensive hardware and little expert knowledge, lowering the barrier for malicious attacks with far-reaching consequences. Hence, exclusive trust in GNSS is misplaced, and methods for reliable detection are urgently needed. However, many of the proposed solutions require expensive replacement of existing hardware. In this paper, therefore, we present MAritime Nmea-based Anomaly detection (MANA), a novel low-cost framework for GPS spoofing detection. MANA monitors NMEA-0183 data and advantageously combines several software-based methods. Using simulations supported by real-world experiments that generate an extensive dataset, we investigate our approach and finally evaluate its effectiveness.

A STAP anti-interference technology with zero phase bias in wireless IoT systems based on high-precision positioning

Fog computing has been applied to the data processing for the Internet of Things (IoT) based on distributed high-precision Global Navigation Satellite Systems (GNSS). However, the space-time adaptive processing (STAP) interference suppression technology in the system will cause fog computing data deviation that includes carrier phase bias and pseudocode offset. An unbiased STAP technique is proposed to eliminate these deviations. First, it is analyzed that the carrier phase bias and pseudocode offset are caused by the non-linear phase response of the STAP equivalent filter. Then, a coefficient-constrained method based on practical engineering processing is proposed, which can eliminate these deviations by restricting the tap coefficients to be symmetrically equal around the center-tap. Moreover, by analyzing the coherent integral function of the pseudocode after filtering, the tap structure of STAP is modified to eliminate the group offset of the pseudocode without increasing the computational complexity and hardware resources. Finally, the unbiased performance and anti-interference performance of the system are verified by numerical and real data simulations.

A Data Quality Assessment Approach for High-Precision GNSS Continuously Operating Reference Stations (CORS) with Case Studies in Hong Kong and Canada/USA

Centimeter-level or better positioning accuracy is needed in engineering surveying applications. When employing the Global Navigation Satellite System (GNSS) in engineering surveying, the high-precision real-time kinematic (RTK) positioning method must be used to achieve such positioning accuracy. Currently, precise point positioning (PPP) cannot reliably achieve the positioning accuracy needed for engineering surveying in real time. The high-precision RTK positioning method needs carrier-phase measurements and a reference station/network. Surveyors may not need a GNSS receiver in their organizations/companies to act as the reference station. Continuously Operating Reference Stations (CORS), run by international/national organizations/agencies or private companies such as GNSS receiver manufacturers, let users freely access the raw GNSS measurements or corrections for real-time and post-processing applications. The positioning accuracy of the GNSS rover is affected by the data quality of the reference stations, including virtual reference stations (VRS). The International GNSS Service (IGS) currently provides the number of cycle slips and the L1 and L2 average pseudorange multipath errors per station daily. The US National Geodetic Survey (NGS) provides daily station coordinate residuals. Carrier-phase data quality of the CORS stations is not provided by their organizations/agencies. Nowadays, many CORS stations track multi-GNSS satellites. This paper proposes a multi-GNSS and multi-frequency data quality assessment approach for CORS stations with a focus on carrier-phase data quality. The proposed approach is demonstrated with case studies on IGS/CORS networks in Hong Kong and Canada/USA. In other words, a strategy to obtain non-linear combined carrier-phase multipath errors and noise is proposed in this work. The data quality of a CORS station depends on the site environment, monument type and height, and GNSS receiver/antenna. An account of the data quality at some selected stations is given; the main focus of the paper is on the proposed data quality assessment approach.

A GNSS/LiDAR/IMU Pose Estimation System Based on Collaborative Fusion of Factor Map and Filtering

One of the core issues of mobile measurement is the pose estimation of the carrier. The classic Global Navigation Satellite System/Inertial Measurement Unit (GNSS/IMU) integrated navigation scheme has high positioning accuracy but is vulnerable to Global Navigation Satellite System (GNSS) signal occlusion and multipath effect. Simultaneous Localization and Mapping (SLAM) is not affected by signal occlusion, but there are problems such as error accumulation and scene degradation. The multi-sensor fusion scheme combining the two technologies can effectively expand the scene coverage and improve the positioning accuracy and system robustness. However, the current scheme has some limitations. On the one hand, GNSS plays a less important role in most SLAM systems, only for initialization or as a closed-loop factor and other auxiliary work. On the other hand, in the fusion method, most of the current systems only use filtering or graph optimization, without taking into account the advantages of both. Aiming at pose estimation for mobile carriers, this paper combines the advantages of the global optimization of the factor graph and the local optimization of filtering and proposes a GNSS-IMU-LiDAR Constraint Kalman Filter (abbreviated as GIL-CKF), which has the characteristics of full coverage and effectively improving absolute accuracy and high output frequency. The scheme proposed in this paper consists of three parts. Firstly, an extensible factor map is used to fuse the positioning information from different sources, including GNSS, IMU, LiDAR, and a closed-loop map, to maintain a high-precision SLAM system, and the output is used as Multi-Sensor-Fusion-Odometry (MSFO). Then, the scene is divided according to the GNSS quality factor, and a Scene Optimizer (SO) is designed to filter GNSS pose and MSFO. Finally, the results are input into the Extended Kalman Filter (EKF) together with the original IMU data for further optimization and smoothing. The experimental results show that the integration of high-precision GNSS positioning information with IMU, LiDAR, a closed-loop map, and other information through the factor map can effectively suppress error accumulation and improve the overall accuracy of the SLAM system. The SO based on GNSS indicators can fully retain high-precision GNSS positioning information, effectively play their respective advantages of filtering and graph optimization, and alleviate the conflict between global and local optimization. SO with EKF filtering furthers optimization, can improve the output frequency, and smooth the trajectory. GIL-CKF can effectively improve the accuracy and robustness of pose estimation and has obvious advantages in multi-sensor scene complementarity, partial road section accuracy improvement, and high input frequency.

Improving GNSS positioning performance of Android smart devices by a novel pseudorange correction method

To achieve high-precision global navigation satellite system (GNSS) positioning of smart devices, it is necessary to correct the pseudorange measurements, especially to reduce time-related errors. Traditional pseudorange smoothing methods fail to cope with measurement anomalies and do not consider the time correlation. Therefore, a time-correlated pseudorange correction (TCPC) method is proposed. The Mann-Kendall trend test and breakpoint detection are used to address data anomalies and pseudorange errors are estimated with an adaptive smoothing method. Using the proposed method, the GNSS receivers can improve the pseudorange accuracy and availability for better single-point positioning performance. The positioning errors are reduced by 10%–45% for different GNSS receivers. Moreover, the circular antenna motion mechanism is used to reduce the measurement correlation and enhance observability. Using the TCPC method in the smartphone antenna motion case, the improvement rates of positioning accuracies are 43%–76% and the horizontal positioning accuracy is improved from meter level to decimeter level.

Quantifying seaweed and seagrass beach deposits using high-resolution UAV imagery

Macroalgae and seagrass wash ashore by tidal waters and episodic events and create an ocean-to-land transport of carbon and nutrients. On land, these deposits (beach wrack) are consumed by macrofauna, remineralized by microorganisms, or washed back to the sea, during which recycling of carbon and nitrogen affect the biochemical cycles in coastal zones. Manual quantification of beach wracks is time-consuming and often difficult due to complex topography and remote locations. Here, we present a novel method using Unoccupied Aerial Vehicle (UAV) photogrammetry combined with in situ measurements of carbon and nitrogen contents of wrack to quantify marine carbon and nutrient deposits in beach zones. The UAV method was tested against placed cubes ranging from 125 to 88,218 cm3 and demonstrated a high accuracy (R2 > 0.99) for volume acquisition when compared to manual measurements. Also, the UAV-based assessments of the cross-sectional area of beach deposits demonstrated a high accuracy when compared to manual and high-precision GNSS (Global Navigation Satellite System) measurements without significant differences between the methods. This demonstrated that UAVs can provide detailed spatial maps, three-dimensional (3D) surface models, and accurate volumetric assessments of beach wrack deposits. In three case studies, combined with carbon and nitrogen measures, total organic carbon and nitrogen deposits in beach wracks were quantified ranging from 4.3 to 9.7 and from 0.3 to 0.5 kg per meter coastline, respectively. In conclusion, this UAV method demonstrated an effective tool to quantify ecosystem carbon and nitrogen deposits relevant to ecosystem assessments and quantification of blue carbon stocks. The method is optimal when the terrain below beach wrack deposits is known, as in the case with before-and-after or repeated surveys. Further, UAVs display strong time- and cost-effective advantages over manual methods which is amplified with increasing project scale. We propose it as a valuable method for multiple scientific and commercial applications related to environmental monitoring and management, including marine resource exploration and exploitation.

A trajectory similarity-based method to evaluate GNSS kinematic precise positioning performance with a case study

There is a lack of effective testing methods to evaluate high-precision global navigation satellite system (GNSS) kinematic positioning solutions, such as GNSS real-time kinematic (RTK) or post-processed kinematic (PPK), with centimeter-level accuracy. Current methods either process static GNSS data in kinematic mode to perform a pseudo-kinematic test or use a precise motion table to make a real-kinematic test but within a very limited travel distance. This study proposes a trajectory similarity method by moving a track trolley platform along a railway track, which can match the GNSS positioning trajectory and the pre-surveyed reference track. The GNSS trajectory offsets from the reference track along the cross-track and vertical directions are regarded as GNSS kinematic positioning errors. Lever-arm compensation is applied to achieve millimeter-level accuracy for this evaluation method. A case study was conducted to evaluate the positioning performance of the Global Positioning System/BeiDou Navigation Satellite System (GPS/BDS) PPK using the proposed method. The results indicate that the proposed method can provide a reference trajectory on the order of a few millimeters, which is sufficiently accurate even for PPK positioning performance evaluation and error source tracing in wide regions. In this case, cycle slips as small as 10 cm in the carrier phase measurements can be detected and studied based on the proposed method.

A Wideband Antenna for GNSS Applications with Improved Orthogonal Dipole

This letter presents a wideband antenna for high-precision Global Navigation Satellite System (GNSS) applications. The antenna consists of a pair of improved orthogonal dipoles that are fed by a feed network with Composite Right/Left-Handed Transmission Lines (CRLH-TLs), a polygonal patch, and a reflector. By using improved orthogonal dipole, the beamwidth of the antenna is broadened. The CRLH-TLs is adopted in feed network to obtain outputs with stable phase difference. Experiment results indicate that the antenna has a less than 10-dB return-loss bandwidth in the range of 1.1 to 1.7 GHz, a 3-dB axial-ratio bandwidth from 1.2 to 1.6 GHz, and a larger than 6.37 dBi gain in the whole operating band. The measured results show that the proposed antenna has stable performance in the whole operating band, which means that it is a suitable antenna used for GNSS applications.

Continuous and Precise Positioning in Urban Environments by Tightly Coupled Integration of GNSS, INS and Vision

Accurate, continuous and seamless state estimation is the fundamental module for intelligent navigation applications, such as self-driving cars and autonomous robots. However, it is often difficult for a standalone sensor to fulfill the demanding requirements of precise navigation in complex scenarios. To fill this gap, this letter proposes to exploit the complementariness of the global navigation satellite system (GNSS), inertial measurement unit (IMU) and vision via a tightly coupled integration method, aiming to achieve continuous and accurate navigation in urban environments. Specifically, the raw GNSS carrier phase and pseudorange measurements, IMU data, and visual features are directly fused at the observation level through a centralized Extended Kalman Filter (EKF) to make full use of the multi-sensor information and reject potential outlier measurements. Furthermore, the widely used high-precision GNSS models including precise point positioning (PPP) and real-time kinematic (RTK) are unified in the proposed integrated system to increase usability and flexibility. We validate the performance of the proposed method on several challenging datasets collected in urban canyons and compare against the loosely coupled and state-of-the-art methods.

Experimental Validation of a High Precision GNSS System for Monitoring of Civil Infrastructures

In recent years, Global Navigation Satellite System (GNSS) technologies, which take full advantage of both real-time kinematic (RTK) and precise point positioning (PPP), managed to reach centimeter-level positioning accuracy with ambiguity resolution (AR) quick convergence techniques. One great advantage over traditional structural health monitoring (SHM) systems is that GNSS technologies will be functional in disaster management situations, when terrestrial communication links become unavailable. In this study, a multi-GNSS system, based on GPS and Galileo constellations and exploiting advanced RTK and PPP-AR technologies with update rate of 100 Hz is tested on two benchmark structures as an SHM system. The first case study served as a baseline to outline the methodology: first, a setup phase of the instrumentation, then a signal processing phase and last, the validation of the results. The methodology was then applied to a real-case scenario, in which the GNSS was tested on a road bridge. A comparative analysis with the results acquired by a set of accelerometers showed that the GNSS was able to identify the crossing of heavy vehicles. The work is paving the way for the development of an affordable and efficient multi-GNSS-based tool for the monitoring of civil infrastructures.

Transformer-Based Global Zenith Tropospheric Delay Forecasting Model

Zenith tropospheric delay (ZTD) plays an important role in high-precision global navigation satellite system (GNSS) positioning and meteorology. At present, commonly used ZTD forecasting models comprise empirical, meteorological parameter, and neural network models. The empirical model can only fit approximate periodic variations, and its accuracy is relatively low. The accuracy of the meteorological parameter model depends heavily on the accuracy of the meteorological parameters. The recurrent neural network (RNN) is suitable for short-term series data prediction, but for long-term series, the ZTD prediction accuracy is clearly reduced. Long short-term memory (LSTM) has superior forecasting accuracy for long-term ZTD series; however, the LSTM model is complex, cannot be parallelized, and is time-consuming. In this study, we propose a novel ZTD time-series forecasting utilizing transformer-based machine-learning methods that are popular in natural language processing (NLP) and forecasting global ZTD, the training parameters provided by the global geodetic observing system (GGOS). The proposed transformer model leverages self-attention mechanisms by encoder and decoder modules to learn complex patterns and dynamics from long ZTD time series. The numeric results showed that the root mean square error (RMSE) of the forecasting ZTD results were 1.8 cm and mean bias, STD, MAE, and R 0.0, 1.7, 1.3, and 0.95, respectively, which is superior to that of the LSTM, RNN, convolutional neural network (CNN), and GPT3 series models. We investigated the global distribution of these accuracy indicators, and the results demonstrated that the accuracy in continents was superior to maritime space transformer ZTD forecasting model accuracy at high latitudes superior to that at low latitude. In addition to the overall accuracy improvement, the proposed transformer ZTD forecast model also mitigates the accuracy variations in space and time, thereby guaranteeing high accuracy globally. This study provides a novel method to estimate the ZTD, which could potentially contribute to precise GNSS positioning and meteorology.

High-precision GNSS PWV retrieval using dense GNSS sites and in-situ meteorological observations for the evaluation of MERRA-2 and ERA5 reanalysis products over China

Precipitable water vapor (PWV) product with high accuracy and high spatiotemporal resolution is important for climate change research. Hourly PWV products with a high spatiotemporal resolution can be provided by global navigation satellite system (GNSS) and global reanalysis data. The National Aeronautics and Space Administration (NASA) and the European Center for Medium-Range Weather Forecasts (ECMWF) have recently released their global reanalysis of the Modern-Era Retrospective Analysis for Research and Applications, Version 2 (MERRA-2) and the fifth-generation ECMWF Reanalysis (ERA5), respectively. A comprehensive evaluation of these two PWV reanalysis products over China has yet to be fully investigated. In this study, the PWV products of MERRA-2 and ERA5 reanalysis over China have been systematically evaluated based on 339 GNSS sites and in-situ ground meteorological observations from year 2016 to 2018. To mitigate the uncertainties of this evaluation, three enhanced vertical correction models for temperature, pressure and PWV that we developed previously are adopted. Four weighted mean temperature (Tm) models are established for the four regions of China (North China, South China, Tibet Plateau, and Northwest China), and they exhibit better accuracy than the other classical Tm models in China. The evaluation results of the MERRA-2- and ERA5-derived PWV show that the correlation coefficient, biases, and root mean square error (RMSE) in China are 0.98/0.99, −0.51/0.38 mm, and 2.50/1.99 mm, respectively. They both exhibit relatively large difference when applied to South China. For the same grid spatial resolution, ERA5 PWV product exhibits similar performance with MERRA-2. The monthly-mean-reanalysis-derived PWV exhibits obvious seasonality and has the largest difference in summer. ERA5 PWV can provide more details of the spatial pattern compared to MERRA-2 PWV. The diurnal anomaly variation for ERA5 is more consistent with GNSS data, and some differences at certain moments during autumn and winter have been observed. These results show that ERA5 PWV and MERRA-2 PWV exhibits excellent applicability in China and can help us to efficiently use PWV products for climate research in China.