The availability of many high-degree Global Geopotential Models(GGMs), namely EGM2008, EIGEN-6C4,GECO, SGG-UGM-1, SGG-UGM-2, XGM2019e_2159, and GGMPlus, challenges users regarding which model is best for Vietnam. This...The availability of many high-degree Global Geopotential Models(GGMs), namely EGM2008, EIGEN-6C4,GECO, SGG-UGM-1, SGG-UGM-2, XGM2019e_2159, and GGMPlus, challenges users regarding which model is best for Vietnam. This study, therefore, evaluates their performance by comparing them with GNSS/leveling data over Vietnam. Results show that their absolute and relative performances are largely independent of topographic conditions and geographical location and can be ranked into three classes:(1)XGM2019e_2159 has the highest accuracy,(2) the models EIGEN-6C4, GECO, SGG-UGM-1, SGG-UGM-2, and GGMPlus, have a very similar level of medium accuracy, while(3) EGM2008 is found to be the least accurate. In an absolute sense, the differences between GNSS/leveling and EGM2008-based height anomalies have a standard deviation(STD) of 0.290 ± 0.010 m, whereas, for XGM2019e_2159, this is 0.156 ± 0.006 m.All other models have STDs of(0.18-0.19) ± 0.007 m. Regarding relative performance without fitting, all GGMs have comparable accuracies for baseline length of 5-20 km, while for baselines longer than 20 km,the STD of XGM2019e_2159 is 1.5 ppm-0.5 ppm(approximately 19%-40%) lower compared with EGM2008, and 0.5 ppm-0.25 ppm(approximately 7%-36%) lower compared with EIGEN6C4, GECO,SGG-UGM-1, SGG-UGM-2, and GGMPlus. In addition, the STDs decrease significantly from 20 to 12 ppm in the range of 5-10 km, slightly from 12 to 6 ppm for 10-35 km, very slightly from 6 to 2.5 ppm for35-200 km, and then remain almost unchanged for longer baselines. After fitting, the relative accuracies of all GGMs are at the same level with negligible STD/RMSE values. Furthermore, only EGM2008 experiences significant regional differences, while other GGMs show more homogeneous spatial variation of absolute accuracy over Vietnam. These findings can contribute to the development of local quasigeoid models in Vietnam and may be helpful with the improvement of GGMs in the future.展开更多
This research presents the results for analyses done to five geopotential global models(GGM),comparing them with ground data from GNSS and leveling in heterogenic zones from the geodetic perspective,in Chile and Spain...This research presents the results for analyses done to five geopotential global models(GGM),comparing them with ground data from GNSS and leveling in heterogenic zones from the geodetic perspective,in Chile and Spain.While the official and complete implementation of the International Height Reference Frame(IHRF)has not yet been established,the vertical geodetic system of many countries is not calculated on a global scale;instead,it is calculated by the variation of relative heights between one or more local tide gauges,such as in the case of Spain and Chile.This aspect creates regional and specific altimetry data,which disables the use of GGM to directly obtain the orthometric height of the vertical reference system(VRS)from the GNSS heights.Global models currently reach centimetric precision due to their high resolution but are directly incompatible for a local level.To solve this,we expose in this article the contrast between geometric geoidal undulation(ellipsoidal heights and orthometric data from the leveling networks in Spain and Chile)with the geoidal undulation in more recent models and higher resolution:Earth Gravitational Model 2008(EGM08),European Improved Gravity model of the Earth by New techniques(EIGEN6 C4),Gravity Observation Combination(GOCO05 C),Experimental Gravity Field Model(XGM2016),and Ultra-High Resolution Global Geopotential Model(SGG-UGM),adjusting the residual between both referential heights by different parametric models and polynomials of determined order.Once evaluated,their geoidal undulations are combined with GNSS/leveling data from the corresponding VRS to generate a correcting surface,which is also known as a hybrid geoid,resulting in a model of optimal adjustment for the combination of five parameters of the EIGEN-6 C4 with orthometric heights and ellipsoids of both Chile and Spain.The results show 2-3 cm precisions,which were statistically analyzed to determine the suitability for use.The final products are three grids of independent hybrid geoids,one for northern Spain and two for Chile(central and north),which allow continuous access to the VRS of each country using the GNSS’s full potential until the IHRF is available and ready for use.展开更多
【目的】随着智慧城市建设中信息技术的深度应用,GNSS轨迹数据呈爆炸式增长,但其轨迹生成过程易受信号干扰与传感器故障影响而产生噪声。本文旨在设计新型噪声识别与修复算法,以提升原始GNSS轨迹数据的处理精度与质量。【方法】针对轨...【目的】随着智慧城市建设中信息技术的深度应用,GNSS轨迹数据呈爆炸式增长,但其轨迹生成过程易受信号干扰与传感器故障影响而产生噪声。本文旨在设计新型噪声识别与修复算法,以提升原始GNSS轨迹数据的处理精度与质量。【方法】针对轨迹噪声识别问题,本文提出基于密度矩阵的自适应DBSCAN算法,其具有超参数无关特性,可敏感捕获低幅值噪声点,同时避免连续转向点的误判。针对噪声修复问题,提出基于轨迹分段的函数构造式修复算法:首先采用道格拉斯-普克(Douglas-Peucker,DP)算法压缩轨迹数据实现分段;其次定位含噪声轨迹段,基于段内有效点构造拟合函数;最终依据相邻点时空属性修复噪声数据。相较于主流插值算法(如拉格朗日、牛顿、埃尔米特、线性、三次样条及最近邻插值),本方法通过规避全局特征依赖,显著保留了噪声点蕴含的局部信息特征。【结果】基于长春市1500名志愿者2024年8月19日—9月1日的原始GNSS轨迹数据,设计2组对比实验。第1组将新型识别算法与原始DBSCAN及其主流衍生算法(KANN-DBSCAN、BDT-ADBSCAN)进行对比。实验表明:新算法在轮廓系数(SC)、Calinski-Harabasz指数(CHI)、Da‐vies-Bouldin指数(DBI)3项指标均取得最优值,优化幅度分别为40.17%~381.80%、20.03%~235.18%、23.42%~79.53%。第2组实验对比新型修复算法与6类经典插值方法(拉格朗日、牛顿、埃尔米特、线性、三次样条、最近邻),结果显示:新算法在轨迹相似性度量指标(Dynamic Time Warping,DTW)上全面优于对比方法,整体优化幅度达43.18%~80.43%。【结论】本文提出的噪声识别与修复算法显著提升了原始GNSS轨迹的质量精度,可高效支撑大规模轨迹数据预处理任务,为时空轨迹挖掘研究提供高质量数据基础。展开更多
现有GNSS水汽层析研究主要聚焦于如何提升卫星观测数据利用率,但在卫星信号数据优选方面研究较少,导致穿过同一组网格集的层析观测方程线性近似且方程系数矩阵列向量元素多数为零,水汽层析模型病态严重。针对该现状,本文提出一种GNSS卫...现有GNSS水汽层析研究主要聚焦于如何提升卫星观测数据利用率,但在卫星信号数据优选方面研究较少,导致穿过同一组网格集的层析观测方程线性近似且方程系数矩阵列向量元素多数为零,水汽层析模型病态严重。针对该现状,本文提出一种GNSS卫星信号自适应优选的水汽层析方法,解决层析模型设计矩阵零元素较多和层析模型病态的难题。该方法基于网格覆盖率最大原则确定层析区域水平网格划分,并发展联合卫星高度角与方位角阈值的卫星信号自适应优选方法,克服水汽层析模型观测方程线性近似的难题。本文选取香港地区2013年5月2日—2013年5月7日共6 d 12个GNSS测站及1个无线电探空站数据为例进行试验。与现有方法相比,本文方法能在降低卫星信号利用率的同时保证网格覆盖率,克服相似卫星信号造成层析模型设计矩阵病态的现状。以无线电探空数据为真值,发现本文方法反演水汽密度廓线的平均RMS、MAE和Bias分别为1.03、0.80和0.13 g/m^(3),优于传统方法的1.25、0.97和0.10 g/m^(3),其RMS改善率为20.78%;此外,本文方法在模型解算效率方面也优于传统方法,其模型计算效率平均提升9.51%。展开更多
文摘The availability of many high-degree Global Geopotential Models(GGMs), namely EGM2008, EIGEN-6C4,GECO, SGG-UGM-1, SGG-UGM-2, XGM2019e_2159, and GGMPlus, challenges users regarding which model is best for Vietnam. This study, therefore, evaluates their performance by comparing them with GNSS/leveling data over Vietnam. Results show that their absolute and relative performances are largely independent of topographic conditions and geographical location and can be ranked into three classes:(1)XGM2019e_2159 has the highest accuracy,(2) the models EIGEN-6C4, GECO, SGG-UGM-1, SGG-UGM-2, and GGMPlus, have a very similar level of medium accuracy, while(3) EGM2008 is found to be the least accurate. In an absolute sense, the differences between GNSS/leveling and EGM2008-based height anomalies have a standard deviation(STD) of 0.290 ± 0.010 m, whereas, for XGM2019e_2159, this is 0.156 ± 0.006 m.All other models have STDs of(0.18-0.19) ± 0.007 m. Regarding relative performance without fitting, all GGMs have comparable accuracies for baseline length of 5-20 km, while for baselines longer than 20 km,the STD of XGM2019e_2159 is 1.5 ppm-0.5 ppm(approximately 19%-40%) lower compared with EGM2008, and 0.5 ppm-0.25 ppm(approximately 7%-36%) lower compared with EIGEN6C4, GECO,SGG-UGM-1, SGG-UGM-2, and GGMPlus. In addition, the STDs decrease significantly from 20 to 12 ppm in the range of 5-10 km, slightly from 12 to 6 ppm for 10-35 km, very slightly from 6 to 2.5 ppm for35-200 km, and then remain almost unchanged for longer baselines. After fitting, the relative accuracies of all GGMs are at the same level with negligible STD/RMSE values. Furthermore, only EGM2008 experiences significant regional differences, while other GGMs show more homogeneous spatial variation of absolute accuracy over Vietnam. These findings can contribute to the development of local quasigeoid models in Vietnam and may be helpful with the improvement of GGMs in the future.
基金financial support for the Chilean part of the project from the Scientific and Technological Research Department of USACH(DICYT in Spanish)through the project DICYT-Regular 091612TM。
文摘This research presents the results for analyses done to five geopotential global models(GGM),comparing them with ground data from GNSS and leveling in heterogenic zones from the geodetic perspective,in Chile and Spain.While the official and complete implementation of the International Height Reference Frame(IHRF)has not yet been established,the vertical geodetic system of many countries is not calculated on a global scale;instead,it is calculated by the variation of relative heights between one or more local tide gauges,such as in the case of Spain and Chile.This aspect creates regional and specific altimetry data,which disables the use of GGM to directly obtain the orthometric height of the vertical reference system(VRS)from the GNSS heights.Global models currently reach centimetric precision due to their high resolution but are directly incompatible for a local level.To solve this,we expose in this article the contrast between geometric geoidal undulation(ellipsoidal heights and orthometric data from the leveling networks in Spain and Chile)with the geoidal undulation in more recent models and higher resolution:Earth Gravitational Model 2008(EGM08),European Improved Gravity model of the Earth by New techniques(EIGEN6 C4),Gravity Observation Combination(GOCO05 C),Experimental Gravity Field Model(XGM2016),and Ultra-High Resolution Global Geopotential Model(SGG-UGM),adjusting the residual between both referential heights by different parametric models and polynomials of determined order.Once evaluated,their geoidal undulations are combined with GNSS/leveling data from the corresponding VRS to generate a correcting surface,which is also known as a hybrid geoid,resulting in a model of optimal adjustment for the combination of five parameters of the EIGEN-6 C4 with orthometric heights and ellipsoids of both Chile and Spain.The results show 2-3 cm precisions,which were statistically analyzed to determine the suitability for use.The final products are three grids of independent hybrid geoids,one for northern Spain and two for Chile(central and north),which allow continuous access to the VRS of each country using the GNSS’s full potential until the IHRF is available and ready for use.
文摘【目的】随着智慧城市建设中信息技术的深度应用,GNSS轨迹数据呈爆炸式增长,但其轨迹生成过程易受信号干扰与传感器故障影响而产生噪声。本文旨在设计新型噪声识别与修复算法,以提升原始GNSS轨迹数据的处理精度与质量。【方法】针对轨迹噪声识别问题,本文提出基于密度矩阵的自适应DBSCAN算法,其具有超参数无关特性,可敏感捕获低幅值噪声点,同时避免连续转向点的误判。针对噪声修复问题,提出基于轨迹分段的函数构造式修复算法:首先采用道格拉斯-普克(Douglas-Peucker,DP)算法压缩轨迹数据实现分段;其次定位含噪声轨迹段,基于段内有效点构造拟合函数;最终依据相邻点时空属性修复噪声数据。相较于主流插值算法(如拉格朗日、牛顿、埃尔米特、线性、三次样条及最近邻插值),本方法通过规避全局特征依赖,显著保留了噪声点蕴含的局部信息特征。【结果】基于长春市1500名志愿者2024年8月19日—9月1日的原始GNSS轨迹数据,设计2组对比实验。第1组将新型识别算法与原始DBSCAN及其主流衍生算法(KANN-DBSCAN、BDT-ADBSCAN)进行对比。实验表明:新算法在轮廓系数(SC)、Calinski-Harabasz指数(CHI)、Da‐vies-Bouldin指数(DBI)3项指标均取得最优值,优化幅度分别为40.17%~381.80%、20.03%~235.18%、23.42%~79.53%。第2组实验对比新型修复算法与6类经典插值方法(拉格朗日、牛顿、埃尔米特、线性、三次样条、最近邻),结果显示:新算法在轨迹相似性度量指标(Dynamic Time Warping,DTW)上全面优于对比方法,整体优化幅度达43.18%~80.43%。【结论】本文提出的噪声识别与修复算法显著提升了原始GNSS轨迹的质量精度,可高效支撑大规模轨迹数据预处理任务,为时空轨迹挖掘研究提供高质量数据基础。
文摘现有GNSS水汽层析研究主要聚焦于如何提升卫星观测数据利用率,但在卫星信号数据优选方面研究较少,导致穿过同一组网格集的层析观测方程线性近似且方程系数矩阵列向量元素多数为零,水汽层析模型病态严重。针对该现状,本文提出一种GNSS卫星信号自适应优选的水汽层析方法,解决层析模型设计矩阵零元素较多和层析模型病态的难题。该方法基于网格覆盖率最大原则确定层析区域水平网格划分,并发展联合卫星高度角与方位角阈值的卫星信号自适应优选方法,克服水汽层析模型观测方程线性近似的难题。本文选取香港地区2013年5月2日—2013年5月7日共6 d 12个GNSS测站及1个无线电探空站数据为例进行试验。与现有方法相比,本文方法能在降低卫星信号利用率的同时保证网格覆盖率,克服相似卫星信号造成层析模型设计矩阵病态的现状。以无线电探空数据为真值,发现本文方法反演水汽密度廓线的平均RMS、MAE和Bias分别为1.03、0.80和0.13 g/m^(3),优于传统方法的1.25、0.97和0.10 g/m^(3),其RMS改善率为20.78%;此外,本文方法在模型解算效率方面也优于传统方法,其模型计算效率平均提升9.51%。
