中国区域基于PPP与双差网解法的GNSS水汽反演精度分析 

展开
  • 1. 云南省大气探测技术保障中心,云南 昆明 650034
    2. 中国气象局气象探测中心,北京 100081
    3. 北京航空航天大学 电子信息工程学院,北京 100191
    4. 广西壮族自治区气象技术装备中心,广西 南宁 530022

网络出版日期: 2025-06-26

基金资助

中国气象局创新发展专项(CXFZ2024J061);中国气象局气象探测中心观测试验项目(GCSYJH24-02SY2023004);国家自然科学基金面上项目(42174027);云南省科技厅重点研发计划项目(202403AC100012);中国气象局大气探测重点开放实验室开放课题(2023KLAS12M

Accuracy Analysis of GNSS Water Vapor Inversion Based on PPP and Double-Difference Network in China Region

Expand
  • 1. Yunnan Atmospheric Observation Technology Support CentreKunming 650034YunnanChina
    2. CMA Meteorological Observation CentreBeijing 100081China
    3. School o f Electronic and In formation EngineeringBeihang UniversityBeijing 100191China
    4. Guangxi Meteorological Technical Equipment CenterNanning 530022GuangxiChina

Online published: 2025-06-26

摘要

利用中国区域2023年夏季945个地基全球导航卫星系统(GNSS)测站的观测数据,分别采用双差网解法与精密单点定位法(Precise Point PositioningPPP)对大气可降水量(Precipitable Water VaporPWV)进行了反演,以同址探空站和ERA5再分析资料的PWV为参考值,研究分析了两种方法在中国不同气候区域反演 PWV 的精度及稳定性特征。结果表明:与 PPP 解相比,双差解与探空和 ERA5资料的PWV的相关性更强,偏差(Bias)频率分布更集中,峰值区概率更高,偏差范围更小。以探空资料获取的RS-PWV为参考值时,双差解与PPP解的平均Bias分别为-0. 1 mm1. 1 mm,平均均方根误差(RMSE)分别为 2. 4 mm 3. 1 mm,以 ERA5-PWV 为参考值时,双差解与 PPP 解的平均 Bias 分别为-0. 2 mm 0. 1 mm,平均 RMSE分别为 2. 7 mm 3. 2 mm,双差解的平均 RMSE均小于 3 mm,这表明双差网解法反演的PWV具有更高的精度和稳定性。GNSS探测水汽的精度总体表现为西部非季风区优于东部季风区,双差解在各气候区域的RMSE都更集中于中位数附近,而PPP解在不同测站多表现出不同的精度水平,在水汽充足且探测精度偏低的温带和亚热带季风气候区域精度离散程度较大,具有较强的不稳定性。

本文引用格式

罗 晋, 曹云昌, 徐八林, 梁 宏, 周凌昊, 王乙竹, 梁静舒 . 中国区域基于PPP与双差网解法的GNSS水汽反演精度分析 [J]. 高原气象, 0 : 1 . DOI: 10.7522/j.issn.1000-0534.2025.00011

Abstract

Using observation data from 945 ground-based Global Navigation Satellite SystemGNSSstations in the summer of 2023 in Chinathe double-differenceDDnetwork algorithm and Precise Point Positioning PPPalgorithm were used to invert the Precipitable Water VaporPWVof the atmosphere. The accuracy and stability characteristics of the two methods for inverting PWV in different climate regions of China were studied and analyzed using PWV from radiosonde stations at the same location and ERA5 reanalysis data as reference values. The results show that compared with the PPP solutionthe DD solution has a stronger correlation with the PWV of radiosonde and ERA5 dataa more concentrated bias frequency distributiona higher probability of peak areaand a smaller range of bias. Compared with the Radio-PWVthe average biases for the DD-PWV and PPP-PWV are -0. 1 mm and 1. 1 mmrespectivelyand the corresponding root mean square errorRMSEare 2. 4 mm and 3. 1 mmrespectively. While compared with the ERA5-PWVthe average Bias of the DD-PWV and PPP-PWV are -0. 2 mm and 0. 1 mmrespectivelyand the corresponding RMSE are 2. 7 mm and 3. 2 mmrespectively. The average RMSE of the DD solution is less than 3 mmindicating that the PWV inverted by the DD network method has higher accuracy and stability. The accuracy of GNSS detection of water vapor is generally better in the western non monsoon region than in the eastern monsoon region. The RMSE of the DD solution is more concentrated around the median in various climate regionswhile the PPP solution shows different levels of accuracy at different GNSS stations. The PPP solution has a greater degree of dispersion in accuracy and strong instability in temperate and subtropical monsoon climate regions with sufficient water vapor and low detection accuracy.

文章导航

/