Abstract The positioning accuracy of global navigation satellite systems (GNSS) is important for conducting effective fieldwork, designing geodetic networks, and achieving desired deformation rates. Over the years, the positioning accuracy of the global positioning system (GPS) has been extensively studied, and mathematical models for the north, east, and up coordinate components have been developed. Today researchers have turned their attention to studying the accuracy of multiGNSS (MGEX) positioning. NASA’s GNSS software GIPSY-X provides a framework capable of producing positioning information for GPS and European satellite navigation system (GALILEO) data. A positioning-accuracy model is also available for GIPSY-X solutions; however, it only uses GPS data, and is outdated today. In order to determine the accuracy of GIPSY-X GNSS positioning, we use the processing products of NASA’s Jet Propulsion Laboratory (JPL) in which the data are also available for Global Navigation Satellite System of Russia (GLONASS), and this multiGNSS product set was available only for a short period of time in 2019. We assess the accuracy performance of the GPS-only solution (G) against each of the various combinations of these three constellations (GRE, GR, GE) for continuous data. Using the publicly available data from the NASA 2019 experiment, the accuracy of positioning from short observation sessions (i.e., considered to be 4- and 8-h GNSS campaigns) is also assessed against 24-h continuous measurements. The PPP module of GIPSY-X is employed to produce the position solutions. Furthermore, we derive accuracy prediction models based on the observation session duration for each of the various GNSS constellation combinations. As expected, the best accuracy is obtained by the combination of all three constellations (GRE solution), and the GRE solution, on average, is superior to the GPS-only solution by about 9%, 16%, and 14% for north, east, and up coordinates. The impact is slightly greater on the GNSS campaign solutions and hence is promising for the studies conducting episodic GNSS measurements to monitor natural hazards and engineering structures.
Abstract Digital elevation models (DEMs) are essential for infrastructure design and flood modeling, yet publicly available DEMs typically exhibit vertical accuracies of one to three meters, insufficient for these applications. This study develops and systematically evaluates a constraint-based DEM enhancement approach that integrates sparse, high-accuracy light detection and ranging (LiDAR) reference data collected at bridge locations to improve regional DEM quality. The methodology employs nearest-neighbor spatial interpolation to generate elevation correction surfaces from constraint points (points of known accurate elevations), followed by Gaussian smoothing to preserve topographic continuity. The approach was evaluated using LiDAR data from 15 bridge locations in the Austin metropolitan area in Central Texas. Four spatial interpolation methods (nearest neighbor, inverse distance weighting, natural neighbor, and linear interpolation) were compared, with nearest neighbor achieving optimal performance with a 28.79% improvement in mean Root Mean Square Error (RMSE) within the study area. Gaussian filtering with an optimized smoothing parameter ( σ = 0.3 m ) further enhanced accuracy, achieving a mean RMSE to 0.105 m. Spatial configuration analysis across six-, nine-, and 15-bridge configurations revealed critical dependencies: distributed constraint arrangements consistently achieved optimal accuracy with six to eight constraint locations, while clustered configurations (constraint points clustered inside the area of interest) with peripheral constraints exhibited performance degradation despite increasing constraint count. This performance resulted from nearest-neighbor interpolation’s reliance on geometric proximity without terrain similarity consideration. The findings provide practical guidance for transportation agencies and flood management programs seeking to leverage existing infrastructure-derived LiDAR surveys for regional DEM enhancement, demonstrating that strategic constraint placement throughout the area of interest is essential for maximizing performance.
As the demand for marine development along the South China Coast continues to grow, the absence of a unified land-sea vertical reference datum system in this region has become an important problem to be solved. The establishment of a seamless chart datum and its separation/conversion model with other vertical data is crucial for unifying vertical reference measurements across land and sea. In this paper, a comprehensive study is conducted, involving the collection of 78 tide gauges, 846 satellite altimetry points, and leveling connections from the coastal Guangdong Province (20 degrees-25 degrees N, 105 degrees-120 degrees E) over a period of 19 years (2003-2021). Through the application of a national tidal datum epoch and an algorithmic approach, a seamless chart datum model, a topography of sea surface model, a mean sea level model, and a separation/conversion model were established. Model accuracy assessment and validation results indicate that the sea surface topography model exhibits a root mean square error (RMSE) of 3.2 cm. The mean difference between the mean sea level height and observed values is found to be 4.45 cm. Further, the separation/conversion model achieves an accuracy of 6.5 cm. When tide gauge and satellite altimetry data are incorporated into the accuracy statistics, the overall RMSE for the seamless chart datum model is 7.93 cm. However, when only the data from the tide gauge station are used to evaluate, the accuracy of the unified land-sea vertical datum and conversion is determined to be better than 7 cm.
The automatic astrogeodetic survey method based on total stations has emerged as the second near-Earth astrogeodetic technology after digital zenith cameras, offering high automation, precision, portability, and cost-effectiveness. The method employs an automatic scheduling algorithm for optimal observation sequencing and integrates Global Navigation Satellite System (GNSS) receiver synchronization with a precision oscillator to ensure time accuracy. This enables high-precision surveys with flexible field deployment, effectively replacing manual observations. Using the improved Image Total Station Astrogeodetic Positioning and Orientation System (ITSAPOS) as the measurement platform, we examined its operating principle, focusing on accuracy and reliability. Between 2022 and 2024, 76 experiments were conducted at 24 observation sites across plains, plateaus, hills, and mountains given varying seasonal conditions. ITSAPOS, the entire system weighing no more than 15 kg, consistently outperformed the Astrogeodetic Surveying System Based on Electronic Theodolites (ASSET), achieving standard deviations approximately half those of ASSET, thereby doubling precision. Longitude standard deviations ranged from 0.002 to 0.009 s and latitude standard deviations from 0.003 '' to 0.110 '' based on multinight observations at the same points. ITSAPOS also achieved stable closure errors for the meridian triangle within shorter times, reaching 0.01 s accuracy. Longitudinal external point accuracy was high, with a maximum difference of 0.012 s, average absolute difference of 0.007 s, and root mean square error (RMSE) of 0.009 s. For latitude, the maximum difference was 0.481 '', with average absolute difference of 0.253 '' and RMSE of 0.301 ''. In comparison, ASSET showed 0.053 s and 0.532 '' for maximum differences, 0.018 s and 0.300 '' for average differences, and 0.025 s and 0.334 '' for RMSE. Overall, the automatic astrogeodetic survey method based on total station images provides autonomous measurement, portability, cost-effectiveness, and stable high precision. The method allows rapid adjustment of observation points and cycles, offering transformative advantages for remote deployments and continuous monitoring.
Mobile laser scanning (MLS) is central to modern corridor surveying, but moving objects introduce systematic geometric distortions that compromise point cloud reliability. These distortions reduce the geometric accuracy needed for surveying and asset management. Although prior studies have noted these effects, no systematic analytical framework currently exists to model their nature and magnitude. This study introduces the first closed-form deterministic equations for motion-induced distortions in MLS under constant-velocity motion on locally planar terrain. Two principal distortion types are identified: length distortion (causing elongation, contraction, or layover depending on relative motion) and parallelepiped distortion (where cuboid objects skew into tilted parallelepipeds). The derived equations explicitly relate distortion parameters to scanner geometry and relative kinematics. Validation using a light detection and ranging (LiDAR) simulation engine across representative motion angles (0 degrees, 40 degrees, 90 degrees, 130 degrees, 180 degrees) and velocity regimes confirmed the equations, with root mean square error (RMSE) values of 0.17 m for length distortion and 0.57 degrees, 0.49 degrees, 0.29 degrees for the three face angles of parallelepiped distortion. Field data sets qualitatively illustrate characteristic distortion patterns in real traffic scenes. Together, the analytical framework and validation provide a foundation for interpreting, quantifying, and correcting motion-induced distortions in MLS point clouds used for dynamic surveying.
Although traditional forest inventory methods are comprehensive, they continue to be labor-intensive and time-consuming. Light detection and ranging (lidar) data collected from various platforms can address these shortcomings. This paper presents a case study of multimodal laser scanning for forest inventory at the Millhopper VARIETIES II Forest experimental site, a pine plantation owned by the University of Florida. The study explores the integration of mobile laser scanning (MLS), terrestrial laser scanning (TLS), and uncrewed aerial laser scanning (ULS) for enhanced forest metric extraction. It employed varying scanner mounting angles and laser pulse repetition rates (PRR) in MLS data collection to optimize point cloud density and reduce occlusions. Additionally, it combined MLS and ULS to improve the accuracy of individual tree detection (ITD) and tree height estimation compared to single-platform approaches. A least squares technique was utilized to fit spherical targets geometrically, ensuring precise alignment and fusion of datasets across all modalities. Results indicated that the integrated MLS+TLS+ULS dataset achieved superior ITD accuracy with a precision of approximately 90%, recall of about 99%, F1-score of roughly 94%, and a height root mean square error (RMSE) of 0.36 m. Notably, the MLS+ULS dataset reached a recall of approximately 95%, a precision of around 90%, an F1-score of about 93%, and an RMSE of 0.31 m. The integrated approach provided more accurate and comprehensive forest data than individual laser scanner modality datasets. These findings emphasize the importance of precise lidar data fusion and tuning lidar sensor parameters for scalable and accurate forest inventory, enabling improved tree count estimation and enhanced forest health assessment for sustainable forest management.
Coastal vegetation plays a critical role in shoreline stabilization, habitat preservation, and maintaining biodiversity. Accurate mapping of these communities is essential for informed ecosystem management, particularly in ecologically sensitive coastal areas. This study investigates the application of machine learning for mapping vegetation communities using high-resolution multispectral imagery acquired by an unmanned aerial system (UAS)-mounted MicaSense RedEdge-MX sensor at the Jupiter Inlet Lighthouse Outstanding Natural Area (ONA), FL. Four dominant vegetation communities, oak scrub, palm, sand pine, and mangrove, were classified using both pixel-based and object-based approaches with support vector machine and random forest algorithms. To enhance classification performance in spectrally similar communities such as mangrove and sand pine, a high-resolution digital elevation model derived from airborne LiDAR (USGS 3DEP) was incorporated as an additional input feature. Results show that integrating elevation data improved classification accuracy by up to 15%, achieving a maximum overall accuracy of 90%. The inclusion of elevation was particularly effective in distinguishing ecologically distinct yet spectrally similar vegetation communities. These findings highlight the value of combining spectral and elevation data to improve coastal vegetation mapping and offer practical guidance for land managers and remote sensing practitioners in similar environments.
The Survey of India has recently established the Indian Continuously Operating Reference Stations (CORS) network to advance geodetic activities and create a precise geodetic infrastructure nationwide. Numerous online and offline software tools are available for high precision Global Navigation Satellite System (GNSS) data processing using precise point positioning (PPP) and relative positioning techniques to facilitate such infrastructure. This article offers an in-depth comparative analysis of various popular offline and web-based software, including Bernese, GNSS at MIT (GAMIT)/Global Kalman filter (GLOBK), GipsyX, AUSPOS, and Online Positioning Users Services (OPUS)-Static, using the newly established Indian CORS subnetwork. The study examines the models employed by these software programs, which are crucial for achieving precise coordinates. The findings show that among offline scientific software, Bernese and GAMIT/GLOBK agree with repeatability in X, Y, and Z at the millimeter level, supporting their use as the superior GNSS data processing software for monitoring stations despite the complex strategies involved in processing the datasets. Although GipsyX provides PPP solutions, it delivers precise results at the subcentimeter level except for a few outliers. AUSPOS, based on Bernese v5.2, with subtle changes, shows good results, but complexities may arise when handling large datasets. OPUS-Static uses a combination of three international GNSS services and/or the National Geodetic Survey CORS as reference stations to compute a solution, resulting in extremely long baselines, which in turn compromises accuracy with repeatability values reaching up to 4 cm in Y coordinates for a few stations. Some statistical tests, such as Bartlett's test, the Welch analysis of variance test, and the Games-Howell test, were also applied to compare the population mean and variances of X, Y, and Z coordinates. These tests show that the population means of Bernese and GAMIT/GLOBK were statistically similar for more than 80% of stations, while GipsyX failed the Games-Howell test for pairwise comparison of population means for the majority of the stations when compared with the results of two datasets obtained from Bernese and GAMIT/GLOBK software. This rigorous study aids in understanding the strengths and drawbacks of each software, leading to reliable decision-making strategies for various precise GNSS applications.
This work assesses the performance of uncalibrated phase delay (UPD) products at different scales in multi-Global Navigation Satellite System (GNSS) precise point positioning ambiguity resolution (PPP-AR). Two sets of wide-lane (WL) and narrow-lane (NL) UPD products were estimated and generated using the Multi-GNSS Experiment global station network and the Heilongjiang regional station network. A systematic evaluation assessed the stability, precision, and impact of these products on positioning performance. The results show that regional UPD products surpass global ones in accuracy and stability, mainly due to denser measurement stations and reduced unmodeled errors, with a notable advantage in NL UPD stability. Conversely, global UPD products demonstrate broader observation coverage and greater applicability, effectively mitigating regional errors and exhibiting superior WL UPD stability. The positioning experiments confirm that both global and regional UPD products enable high-precision PPP-AR positioning within the user domain. However, for regional users, PPP-AR based on regional UPD products delivers superior performance. In single-system mode, the BeiDou Navigation Satellite System (BDS) with regional UPD products achieves an average convergence time of 13.6 min, a 23.6% improvement over the global solution. BDS positioning accuracy is comparable to GPS, with marginally better vertical performance. Furthermore, the integration of multiple systems significantly enhances positioning performance. The GPS, Galileo, and BDS (GEC) combination with regional UPD products reduces the convergence time to 7.3 min, a 24.7% improvement over the global solution. After 15 min of convergence, the horizontal and vertical positioning accuracies reach 3.2 and 3.6 cm, outperforming global UPD accuracies of 4.3 and 4.6 cm. These findings offer critical insights into optimizing high-precision regional positioning services and emphasize the benefits of regional UPD products and multi-GNSS integration.
The quality of elevation data plays a pivotal role in various land transport activities, such as accessibility analyses, route optimization, and estimates of energy demand and gas emissions. This study evaluates the accuracy of altimetric profiles obtained during a car trip. The compared methods include (1) barometric altimetry using a low-cost pressure module; (2) single-point satellite positioning using a low-cost receiver; (3) postprocessed relative kinematic (PPK) satellite positioning using a geodetic receiver; (4) the Shuttle Radar Topography Mission digital elevation model (DEM); and (5) the digital terrain model provided by the GeoSampa portal for S & atilde;o Paulo City, used as the reference dataset. Moreover, the paper explores corrections for barometric altimetry to address the well-known issues of systematic errors and high-frequency noise in low-cost modules. A high correlation among the altimetric profiles was identified, but a notably higher accuracy was observed in the profiles and road grades obtained by PPK satellite positioning and barometric altimetry after the correction for systematic errors and filtering through moving average for noise reduction. The results highlight the importance of a careful choice of altimetric bases and processing techniques to prevent inaccuracies in road projects and analytical models.
This case study investigates the integrated application of photogrammetry, geographic information systems, light detection and ranging (LiDAR), unmanned aerial vehicles (UAVs), global navigation satellite systems, and three-dimensional (3D) laser scanning to mitigate risks in NEOM's Oxagon port, a floating industrial hub within a 26,500 km2 megaproject in northwest Saudi Arabia. Covering preconstruction (foundation alignment), construction (structural monitoring), and postconstruction (as-built verification and maintenance) phases from 2022-2024, these technologies addressed environmental risks (coastal ecosystems with 2,500 marine species), logistical challenges (desert terrain, sandstorms), financial constraints ($50 billion budget), and technical complexities (floating structures). Photogrammetry achieved 1.5 mm accuracy in steel framework monitoring, validated by 3D scanning, while LiDAR and UAVs reduced survey times by 50% compared to conventional total station methods. A three-phase framework yielded 15%-20% cost savings ($7.5-$10 billion), 25% error reduction, 15% reduced environmental impact, and 10% maintenance cost savings, validated by ecological surveys and project reports. Compared to conventional methods, integrated technologies reduced survey times by 40% and rework costs by 30%. Despite challenges such as high LiDAR costs and sandstorm disruptions, this study provides a rigorously validated, scalable framework for megaproject risk management in extreme environments, offering actionable insights for global infrastructure development.
Enhancing the performance of real-time precise point positioning (RT-PPP) holds significant application value, with pseudorange accuracy being one of the key factors constraining its performance. To address this, this paper proposes a novel algorithm called B2b+ID-PPP, which integrates the PPP-B2b service with integrated Doppler observations. Experimental results demonstrate that the satellite-to-ground distance root mean square error (RMSE) characterized by integrated Doppler reaches the subcentimeter level (0.55 cm), enabling high-precision approximation of distance variations. The residual of the pseudorange positioning equation incorporating this observation significantly decreases, with the pseudorange residual RMSE improving from 76.1 to 6.53 cm-an accuracy enhancement of 91.4%. In the single BDS-3 system, B2b+ID-PPP achieves average positioning accuracies of 14.3, 8.7, and 21.4 cm in the east (E), north (N), and up (U) directions, respectively, representing improvements of 5.3%, 16.4%, and 4.5% over B2b-PPP. The convergence time for seven consecutive days of solutions was consistently under 15 min, and the average was 9.2 min. Moreover, given mild ionospheric scintillation conditions, the proposed method further accelerated the convergence speed. Additionally, in simulated severe multipath environments, B2b+ID-PPP effectively enhanced the up-direction accuracy from meter- to decimeter-level. These results confirm that this method effectively enhances pseudorange accuracy through subcentimeter-level integrated Doppler observations, thereby significantly improving the overall performance of PPP-B2b-based RT-PPP in China and its surrounding regions.
This paper proposes a new method [the improved spatiotemporal random effect model (ISTRE)] based on the spatiotemporal Kalman filter model for fusing Global Navigation Satellite System (GNSS) and Interferometric Synthetic Aperture Radar (InSAR) deformation to obtain high spatiotemporal resolution deformation. This method realizes the fusion of GNSS vertical and InSAR LOS deformation based on the spatiotemporal random effect model (STRE) combining robust estimation theory and correlation analysis. During modeling, the model parameters of an improved STRE model (ISTRE) including the spatial basis, state transfer matrix, observation noise, and state noise were determined using the small baseline subset (SBAS)-InSAR deformation. Then the vertical deformation value of small baseline subset interferometric synthetic aperture radar (SBAS-InSAR) was calculated by the forward ISTRE with the reweighted observation noise due to the deformation replacement of SBAS-InSAR by the nearest GNSS site. And the final deformation fusion in the vertical direction was carried out using the corrected SBAS-InSAR and GNSS. The study applies SBAS-InSAR and GNSS deformation data from a tailings pond in Fushun City to validate the vertical correction and fusion quantity of ISTRE, explores the impact of the root mean square error (RMSE) accuracy improvement rate and cross-correlation ratio between GNSS and SBAS-INSAR on the accuracy of vertical correction, and analyzes the superiority of ISTRE by comparing with the improved kriging Kalman filter (IKKF) and spatial-temporal Kalman filter (ISTKF). The results indicate a strong correlation between the vertical correction accuracy of deformation velocity and the cross-correlation coefficient. The deformation velocity accuracy of GNSS point P02 has the best cross-correlation coefficient quality with an average of 0.9053 and can be improved by approximately 70% specifically. The accuracy of cumulative deformation shows a strong correlation with both the cross-correlation ratio and GNSS cross-correlation; the deformation accuracy of GNSS point P05 with an average GNSS cross-correlation of 0.8950 can be enhanced by up to 240%. The RMSEs of the ISTRE fusion results with four of the five sets of GNSS data are smaller than those of IKKF and ISTKF, and the RMSEs of the SBAS-InSAR vertical correction results are similarly generally smaller. ISTRE shows superiority over IKKF and ISTKF in terms of stability, vertical correction, and fusion accuracy by comparative analysis.
This article presents a texture mapping procedure for transferring information from historical images, potentially dating back several decades, even to the late 19th century, onto a three-dimensional (3D) model representing the current condition of the studied system. The method is implemented as a MATLAB toolbox, initially developed for 3D infrared thermography and subsequently modified and adapted to allow fast texture mapping using a variety of historical image types, including photographs, postcards, and technical maps (COMAP3 toolbox, freely available under a General Public License). This approach ensures accurate alignment and reliable integration of the images onto the 3D model. Besides the point cloud onto which data from one or more historical images are projected, the procedure includes quality analysis of the results in terms of residuals, defined as differences between moving points recognized manually or automatically via a feature-based approach, and transformations of the corresponding fixed points. It also evaluates the actual ground pixel size with respect to the ground sampling distance for each pixel. The paper demonstrates the method's effectiveness through a dedicated test and applies it to the parish church of San Giorgio in Argenta, Ferrara, Italy, mapping flood levels from 1917, 1955, and 2024 onto a photogrammetric point cloud.
Outlier detection is a crucial aspect of model fitting, particularly in geodetic applications where data reliability is paramount. The traditional least squares estimation method, while optimal under ideal conditions, is highly sensitive to deviations caused by outliers. This study proposes a robust M-estimation method by incorporating a redundancy design to enhance the reliability of outlier detection. An iterative weight adjustment method is developed by changing the partial redundancy values of the observations, thereby improving the sensitivity of residuals to outliers. The proposed method was evaluated through extensive Monte Carlo simulations using a simple regression model and leveling network for a small outlier. The results demonstrate that the proposed method outperforms conventional techniques such as the Baarda's and Pope's tests, particularly when multiple outliers are present. While the method slightly reduces detection power for observations initially exhibiting high redundancy, it significantly enhances the detection rate for low-redundancy observations. In addition, the method was applied to uncorrelated observations. The overall findings highlight the importance of redundancy management in robust estimation processes and suggest that the proposed method offers a viable alternative for outlier detection in adjustment models.
Surveys collected with a 10-m-long lighter amphibious resupply cargo (LARC) vessel and a 3.3-m-long jetski are evaluated by comparison with surveys using the 11-m-tall coastal research amphibious buggy (CRAB) in calm and rough sea states representing the lower and upper operational bounds when considering safety and data quality. The centimeter-level accuracy of CRAB surveys, performed since 1981, is well established. The field test consisted of repetitively surveying two cross-shore profile lines from near the shoreline to 6-m water depth, 700 m offshore. The survey lines were repeated four times by the CRAB and nine to ten times by the LARC and jetski. The CRAB data were averaged to define a reference cross-shore elevation and profile shape to serve as ground truth to determine the accuracy of the LARC and jetski surveys. The two systems compare best with the CRAB surveys on the mildly sloping shoreface, seaward of the nearshore sandbar, where alongshore currents were weakest, depth-limited breaking waves were infrequent, and small-scale morphological features that may not be resolved by the CRAB were minimal. The root mean square (RMS) elevation error between the LARC measurements and the CRAB mean profile was 0.03 m under calm conditions and 0.05 m under rougher conditions, whereas the jetski RMS was 0.09 and 0.10 m, respectively.
To address the challenges associated with the large data volume, extensive distribution, and inherent complexity of railway data, as well as the difficulty in removing objects adjacent to the railway tracks, we propose a progressive railway track extraction method. This method integrates a fused attention mechanism and residual architecture to enhance the accuracy and efficiency of track extraction. Our method is divided into two distinct modules: the ballast bed extraction module and the railway track extraction module. The approach begins with preprocessing the raw point cloud data to facilitate the initial extraction of the ballast bed region. Following this, vegetation points adjacent to the ballast bed are isolated to enhance the accuracy of extraction process. For the railway track extraction, we utilize a sampling and grouping strategy to capture local features from diverse regions using PointNet. The feature learning phase incorporates both residual and attention mechanism to effectively fuse spatial and channel-wise information, thereby enhancing feature representation. Finally, the extracted features are passed through a fully connected layer for the final railway track extraction. Experimental results demonstrate that the proposed method delivers robust performance in accurately extracting railway tracks across various railway environments, showcasing its effectiveness and adaptability in complex scenarios. The proposed method achieves an overall accuracy of 95.84%, with a mean intersection over union of 89.33% for railway track extraction. Additionally, the F1-score for ballast bed extraction exceeds 94%, highlighting the effectiveness and precision of the approach in both track and ballast bed extraction tasks. These results underscore the robustness and reliability of the method in handling complex railway environments.
