|
|
Journal of Global Positioning Systems Vol. 22, No. 1, 2026 |
Cover |
|
Journal of Global Positioning Systems
|
JGPS Team Structure, Copyright and Table of Contents |
|
JGPS Team Structure, Copyright
|
1. Advances in Real-Time GNSS Monitoring of Earthquakes and Volcanoes in Indonesia |
|
Muhammad Al Kautsar, Moh. Fifik Syafiudin, Rahmat Triyono, Priatin Hadi Wijaya, Oktadi Prayoga, Sulistiyani, Ayu Nur Safi�i, Thomas Hardy, Ajat Sudrajat,Fanny Zafira Mukti
Indonesia is among the most tectonically and volcanically active regions worldwide, where frequent earthquakes, volcanic eruptions, and long-term surface deformation pose serious risks to population centers and critical infrastructure. Reliable real-time geodetic monitoring is therefore essential for hazard assessment, early warning, and disaster mitigation. This study presents the status and recent advances in GNSS-based real-time geohazard monitoring in Indonesia, with an emphasis on precise point positioning with ambiguity resolution (PPP-AR) implemented through the GSeisRT real-time engine. GSeisRT processes 1 Hz multi-GNSS observations and generates low-latency regional satellite clock and phase bias products, enabling stable and high-precision real-time displacement monitoring. GSeisRT has been deployed within the nationwide Ina-CORS network operated by the BIG and supports continuous monitoring of earthquakes, volcanic deformation, land subsidence, and active faults. Several recent seismic events, including the 2023 Banda Sea Mw 7.1 earthquake and subsequent Mw 5.8-6.6 events, were successfully captured in real time. The derived coseismic displacements and peak ground displacements (PGDs) allow rapid earthquake magnitude determination that agrees well with estimates published by the United States Geological Survey (USGS). Performance comparisons with existing real-time PPP systems demonstrate that GSeisRT provides improved robustness and continuity, effectively suppressing spurious position jumps caused by incorrect ambiguity resolution. Additional developments, including the integration of GNSS with collocated accelerometer data and the availability of global 1 Hz real-time satellite products, further enhance short-period deformation monitoring and seismic response capability. Although challenges remain due to sparse station spacing and communication limitations in remote areas, the results confirm that advanced re-al-time GNSS technologies provide a reliable and scalable foundation for strengthening geohazard early warning and resilience in Indonesia.
|
2. Flood Detection Using CYGNSS GNSS-R Observations - A Case Study of the 2022 Sindh Flood, Pakistan |
|
Afaq Karim, Dongkai Yang1, Omada Friday Ojonugwa
Flood monitoring and mapping are essential
for effective disaster response, particularly in regions
frequently affected by extreme rainfall events.
Conventional flood detection methods are often
constrained by high costs, cloud cover, and limited
temporal resolution. This study investigates the
capability of Global Navigation Satellite System
Reflectometry (GNSS-R) for flood detection using
Level-1 data from the Cyclone Global Navigation
Satellite System (CYGNSS) during the severe August
2022 flood event in Sindh Province, Pakistan. Surface
reflectivity was estimated using a bistatic radar
formula, in which the corrected signal-to-noise ratio
(SNRc) was derived as a proxy for surface water
presence after quality control, incidence-angle
filtering, and outlier removal. To generate continuous
flood reflectivity fields from irregular CYGNSS
observations, a Natural Neighbor interpolation
technique based on Voronoi tessellation was applied.
Flooded areas were identified using an empirically
derived SNRc threshold of approximately 16 dB,
determined from observations over permanent water
bodies. The resulting CYGNSS-derived flood maps
show strong spatial agreement with MODIS-based
flood inundation maps. The overall accuracy achieved
is 0.819, indicating that approximately 82% of the
MODIS and CYGNSS grid cells were correctly
classified.
|
3. Spatio-Temporal Analysis of Soil Moisture Dynamics Using GNSS-R and Ancillary Remote Sensing Datasets in Nigeria |
|
Omada Friday Ojonugwa,Yang Dongkai,Afaq Karim,Mugabo Adolphe Izerimana
This study evaluates CYGNSS Level-3 surface soil moisture (SSM) retrievals across Nigeria using a Random Forest model trained on Level-1 reflectivity and ancillary geophysical data. All performance metrics assess the Random Forest model's ability to spatially disaggregate CYGNSS Level-3 products using ancillary data, not absolute accuracy against ground observations. The model downscales CYGNSS Level-3 SSM from ~36 km to 2 km resolution for soil moisture mapping, with error characterization conducted at 1 km resolution to capture finer-scale uncertainty patterns. Spatial cross validation across four phenological stages demonstrated robust downscaling consistency throughout 2020, yielding R� of 0.918�0.969 and RMSE of 0.014�0.028 m�/m� relative to CYGNSS Level-3 targets. The retrievals accurately captured seasonal moisture gradients from <0.04 m� /m� in the northern Sahel to>
0.40 m�/m� in the Niger Delta, while preserving riparian contrasts. Variable importance shifted seasonally from vegetation dominance to surface reflectivity control. Despite localized uncertainties over rugged terrain, sub-0.03 m�/m� reconstruction precision relative to CYGNSS L3 was maintained without seasonal recalibration. These findings confirm that CYGNSS GNSS-R data provides operationally viable SSM monitoring for agricultural forecasting and drought early warning in data-scarce environments.
|
4. Deep Learning-Driven Synergistic Integration of Satellite-Derived Water Vapor and Meteorological Radar for Enhanced Urban Precipitation Nowcasting: A Spatiotemporal Attention Architecture |
|
Jing Sun,Jiale Wang,Chuang Shi,Yunchang Cao,Hong Liang,Yuyang Zhu,Wu Chen
Intensifying climate variability has escalated the frequency and severity of localized extreme precipitation events, demanding innovative forecasting methodologies to bolster urban resilience and disaster mitigation. Conventional prediction systems struggle with the rapid onset and high-impact nature of short duration intense rainfall. This investigation presents a heterogeneous data fusion architecture that synergistically combines satellite-based observations�specifically GNSS-derived Precipitable Water Vapor (PWV)�with ground-based Radar Reflectivity Composites (RRC) to advance precipitation nowcasting capabilities. GNSS-PWV serves as a sensitive indicator of atmospheric moisture accumulation dynamics, while RRC characterizes the spatial-temporal evolution of precipitation systems. We propose a Spatiotemporal Attention-Enhanced Swin U-Net (STAESwin) architecture that leverages Swin Transformer modules coupled with an integrated spatiotemporal attention mechanism to effectively assimilate these complementary data streams.
|
5. A Gradient-Aware Undersampled Ionospheric Threat Model with Multi Metrics for BDSBAS |
|
Hui Ren,Zhengzhen Jin,Tong Liu,Fangxin Hu,Yiping Jiang
Satellite-based augmentation systems (SBAS) broadcast grid ionospheric vertical delays (GIVDs) together with their integrity bounds, known as grid ionospheric vertical errors (GIVEs), to support accurate and integrity-assured positioning. In regions such as China, characterized by wide latitude coverage and highly dynamic ionospheric behaviour, geometric descriptors of IPP distribution alone are insufficient to capture local ionospheric variability. This limitation prevents conventional undersampled threat models from distinguishing between benign and severe conditions, leading to excessive GIVE inflation and reduced availability. This paper proposes a gradient-aware undersampled ionospheric threat model for BeiDou satellite-based augmentation system (BDSBAS) in China, referred to as the RRG-3D framework. The method extends the conventional geometry-based model by introducing a statistical ionospheric gradient metric, P90G, defined as the 90th percentile of pairwise gradients to characterize local variability.
|
6. LiDAR-Inertial-UWB Integrated Localization with Quality Control and Online NLOS Mitigation for Unmanned Ground Vehicles |
|
Hong Liu,Shuguo Pan,Wang Gao,Jinle Xu,Zhuoxuan Wang,Feixuan Huang
Continuous and precise localization constitutes a fundamental prerequisite for intelligent vehicles and autonomous driving, particularly in challenging environments, such as GNSS signal blockage, laser degeneration and vision illumination variations. In this paper, we propose a system that integrates data from light detection and ranging (LiDAR), inertial sensor, and ultrawideband (UWB) to achieve global positioning through filtering and smoothing techniques in diverse scenes. Firstly, we leverage an Extended Kalman Filter (EKF) to handle ranging measurements, while also online distinguishing and removing None Line of Sight (NLOS) errors. Besides, a quality control module is performed to obtain healthy UWB absolute factor in the backend. Secondly, a precise inertial navigation system (INS) mechanization algorithm is employed to achieve state updates, which also contributes to the LiDAR-inertial partly tight-coupled process. Finally, we propose a lightweight factor graph to optimally fuse marginalization factor, relative LiDAR odometry factor, UWB absolute factor and preintegration factor. We conducted extensive experiments in outdoor and NLOS scenarios using a customized platform. The results show that our integrated positioning approach enhances the accuracy of existing state-of-the-art schemes, while also enabling robust and real-time pose estimation for unmanned ground vehicles.
|
7. Short-term Forecasting and Integrity Overbounding of Low-Latitude Ionospheric Scintillation |
|
Hao Liu, Dongsheng Zhao, Kefei Zhang, Ke Xu, Qichen Zhao
Ionospheric scintillation is a major error source constrains high-accuracy and high-integrity positioning services of the BeiDou Navigation Satellite System (BDS) in low-latitude regions. Its evolution is characterized by strong non-stationarity, abrupt variation, and complex spatiotemporal coupling. In particular, under geomagnetic disturbances or intense solar activity, scintillation can induce rapid carrier-phase fluctuations and increase the probabilities of cycle slips and loss of lock, thereby significantly degrading positioning accuracy, availability, and integrity. Although existing studies have achieved encouraging results in scintillation forecasting, most of them mainly focus on numerical prediction itself, while the continuous evolution of scintillation risk and its integrity-oriented overbounding characterization remain insufficiently investigated. To address this issue, this study proposes a joint framework for multi-step forecasting of ionospheric scintillation risk and construction of an integrity envelope by combining a Cascade long short-term memory (Cascade LSTM) network with a radial basis function (RBF) model.
|
8. Comparative analysis of carrier-phase level FDE for GF and GB models in urban multipath environments |
|
Qi Cheng, Wu Chen
The increasing demand for centimeter-level positioning in autonomous driving and urban air mobility is frequently hindered by multipath (MP) and non-line-of-sight (NLOS) signals in urban canyons. While fault detection and exclusion (FDE) is essential for ambiguity resolution, traditional code-phase level FDE suffers from low sensitivity due to the similar magnitudes of observation noise and multipath-induced faults. In contrast, carrier-phase level FDE offers significantly higher sensitivity, as carrier-phase noise is orders of magnitude smaller than the biases introduced by incorrect integer cycles. This paper provides a rigorous theoretical analysis comparing two primary ambiguity resolution models for carrier-phase level FDE: the geometry-free (GF) model and the geometry-based (GB) model. We demonstrate the specific advantages of the GF model for FDE in urban environments, particularly its ability to isolate incorrectly resolved ambiguities independent of the receiver-satellite geometry. To validate the theoretical findings, vehicular experiments were conducted in the challenging urban environments of Nanjing and Hong Kong. The performance of a GFbased FDE real-time kinematic (RTK) algorithm was compared with the post-processed kinematic (PPK) of an industry-leading commercial software, Leica Infinity. Experimental results indicate that the GFFDE RTK algorithm demonstrates superior robustness, achieving higher ambiguity fix rates and comparable positioning accuracy compared to the commercial benchmark in harsh urban environments. The findings highlight the significant potential of GFbased carrier-phase FDE to enhance the availability and reliability of high-precision navigation engines.
|
9. A subsidence and spatial information monitoring system for mined-out regions: a case study in Jiawang, Xuzhou |
|
Dongsheng Zhao, Hao Liu, Qianxin Wang, Kefei Zhang, Jianfeng Zha, Craig M.Hancock
Building subsidence, atmospheric water vapor, and ionospheric disturbances are essential key information for ecological restoration and high-quality development in mined-out areas. In this study, a subsidence and spatial information monitoring system was established in Jiawang, Jiangsu, which consists of a GNSS geodetic receiver, an ionospheric scina Experimental results show that the geographical environment in the area has been stable over the long term, with no significant relative position tilt observed in buildings. The GNSS-derived PWV time series, with a temporal resolution of 30 s, showed temporal variations corresponding to the evolution of the radar-observed precipitation event. The TEC and ionospheric disturbance indices captured the diurnal variations and disturbance characteristics during the selected space-weather event. The subsidence monitoring, atmospheric water vapor inversion, and ionospheric parameter monitoring conducted in Jiawang are crucial for the restoration of subsidence and the recovery of the mined-out area's productivity.
|
10. Deep Learning-Driven Synergistic Integration of Ground-Based GNSS-Retrieved Precipitable Water Vapor and Meteorological Radar for Enhanced Urban Precipitation Nowcasting: A Spatiotemporal Attention Architecture |
|
Jing Sun, Jiale Wang, Chuang Shi, Yunchang Cao, Hong Liang, Yuyang Zhu, Wu Chen
Intensifying climate variability has escalated the frequency and severity of localized extreme precipitation events, demanding innovative forecasting methodologies to bolster urban resilience and disaster mitigation. Conventional prediction systems struggle with the rapid onset and high-impact nature of short-duration intense rainfall. This investigation presents a heterogeneous data fusion architecture that synergistically combines ground-based GNSS-derived Precipitable Water Vapor (PWV)—a thermodynamic precursor indicator retrieved from Continuously Operating Reference Stations (CORS)—with Radar Reflectivity Composites (RRC) to advance precipitation nowcasting capabilities. GNSS-PWV serves as a sensitive indicator of atmospheric moisture accumulation dynamics, while RRC characterizes the spatial-temporal evolution of precipitation systems. We propose a Spatiotemporal Attention-Enhanced Swin U-Net (STA-Swin) architecture that leverages Swin Transformer modules coupled with an integrated spatiotemporal attention mechanism to effectively assimilate these complementary data streams. The study domain encompasses the Beijing-Tianjin-Hebei metropolitan region, utilizing high-precision PWV retrievals from 250 GNSS Continuously Operating Reference Stations (CORS) via dual-frequency ionosphere-free Precise Point Positioning (PPP) methodology (validation correlation with ERA5 reanalysis: >0.99). Key Findings: Against independent validation datasets from the 2025 flood season, the model demonstrated 68% accuracy in detecting torrential precipitation events. Comparative analysis reveals that the integrated multi-source approach yields an 18.5% improvement in Critical Success Index (CSI) and an 21.5% enhancement in Probability of Detection (POD) relative to single-modality baseline models. Implications: The results substantiate that strategic integration of satellite-derived geophysical parameters with advanced deep learning architectures substantially elevates early-warning precision, offering a scientifically robust and operationally viable solution for regional hazard prevention and climate adaptation strategies.
|
ABSTRACTS OF PHD DISSERTATIONS |
1. RESEARCH ON AIRCRAFT TRAJECTORY PREDICTION METHODS BASED ON DEEP LEARNING |
|
Haoyuan Zhang
Air transportation is essential for global social and economic development, yet a gap persists between air transport services and market needs. To address this, ICAO advocates trajectory-based operations (TBO), highlighting the importance of accurate fourdimensional (4D) trajectory prediction (TP) for improved airspace capacity, efficiency, and safety. Traditional physics-based TP methods face limitations due to complex parameter requirements and observability issues, while machine learning (ML) approaches, though promising, still lack a robust framework and sufficient accuracy for medium- and long-term predictions. Data imbalance, especially under severe weather conditions, further hinders TP performance. Overcoming these challenges is vital for reliable trajectory prediction in air traffic management (ATM).
|
2. ERROR MITIGATION AND QUANTITATIVE ASSESSMENT OF GNSS-BASED STRUCTURAL DEFORMATION MONITORING FOR LONG-SPAN BRIDGES |
|
Liangliang Hu
Long-span bridges are critical components of national transportation systems, and their operational safety is directly related to traffic continuity and socio-economic stability. With the continuous increase in bridge span length and the intensification of external excitations such as wind, temperature variation, and traffic actions, structural dynamic responses and deformation mechanisms exhibit pronounced nonlinearity, multiscale behavior, and multi-source coupling characteristics. Accurate measurement and quantitative analysis of deformation induced by coupled loads across multiple time scales have therefore become key challenges in the long-term reliability assessment of long-span bridges. Owing to its capability for continuous, real-time, and absolute displacement measurement, the Global Navigation Satellite System (GNSS) has become one of the core sensing technologies in structural health monitoring (SHM) systems for long-span bridges.
|
3. Research on deep integration technology of vehicle navigation for GNSS observation quality in harsh environment |
|
Xin FENG
Emerging fields such as autonomous driving and intelligent robotics require continuous and reliable positioning with decimeter-level or even centimeter-level accuracy. However, GNSS signals are highly susceptible to attenuation, obstruction, and reflection in harsh environments such as dense tree canopies and urban canyons, which severely degrades the accuracy and continuity of carrier phase observations and compromises the availability of centimeter-level positioning. Multi-sensor (GNSS, IMU, Camera, etc.) data fusion techniques cannot maintain satisfying long-term positioning accuracy when GNSS carrier phase observations fail. Ultra-tight integration techniques that leverage multi-sensor information can enhance signal tracking performance at the GNSS baseband processing level, radically improving the accuracy and availability of carrier phase observations. However, existing ultra-tight integration techniques based on single low-cost sensors (such as MEMS IMU or camera) cannot provide satisfying assistance accuracy for phase-locked loops (PLLs) when GNSS signals degrade.
|
4. Research on High-Precision Sea Surface Altimetry Methods Based on Land-Sea-AirMulti-Platform GNSS-R |
|
Xinyue Meng
Sea surface height (SSH) is fundamental data in geodesy and marine environmental dynamics, which is widely applied to marine gravity field modeling, tidal and ocean current model establishment and other researches. Nevertheless, existing sea surface altimetry technologies fail to satisfy practical application requirements including low cost, stable observation and flexible networking. Traditional tide gauges impose strict construction conditions and high costs, making large-scale dense deployment difficult. Satellite altimetry is limited in temporal and spatial resolution and performs poorly in coastal areas. Buoy devices are susceptible to ocean currents and difficult to maintain. As an emerging remote sensing technology, Global Navigation Satellite System-Reflectometry (GNSS-R) utilizes abundant free and all-weather GNSS signals for sea surface observation. It features low equipment cost, flexible layout and large-scale networking capability, and can acquire sea surface height data with high temporal and spatial resolution.
|
Back Cover |
|
Journal of Global Positioning Systems
|