Scholarlink

<p>Precipitable water vapor (PWV) reflects the vertically integrated amount of water vapor in the atmosphere, but it only represents the two-dimensional distribution of water vapor. With the development of GNSS satellite constellation and station network, GNSS water vapor tomography has emerged as a prominent technique for obtaining the three-dimensional water vapor distribution and become a hotspot in GNSS meteorology. We propose an optimal method for GNSS water vapor tomography based on parameter adaptation to solve the ill-posed problem caused by the discretization parameter redundancy in the tomographic model. This method introduces the minimum probability factor of voxels being penetrated by signals, determines the number and position of the redundant parameters in the observation equation by iterating the coefficient matrix, establishes the optimal tomographic model by eliminating the zero elements of the coefficient matrix and the redundant parameters, and calculates the water vapor density of all voxels by constructing a Laplace smoothing equation. The tomography experiments conducted in Hong Kong show that the average root mean square error (RMSE) and mean absolute error (MAE) for the proposed and traditional method are 1.25/0.70 and 1.83/0.92&#xa0;mm in the internal accurate testing, respectively, compared to the GAMIT-estimated slant water vapor (SWV) as a reference. The average RMSE and normalized RMSE are 10.9&#xa0;mm and 8.2% for the proposed method, while these values are 13.2&#xa0;mm and 10% for the traditional method. Using the radiosonde data as references, the proposed method improves average RMSE and MAE by 20.4% and 25.3% compared with the traditional method.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

An optimal method for GNSS water vapor tomography based on parameter adaptation

  • 期刊论文

  • 发表日期：

    2025年12月10日

查看全文
  • Fei Yang,
  • Ruixian Hao,
  • Haoyu Wang,
  • Lei Wang,
  • Zhicai Li,
  • Junxi Zheng,
  • Yingying Wang,
  • Ran Chen

摘要

Precipitable water vapor (PWV) reflects the vertically integrated amount of water vapor in the atmosphere, but it only represents the two-dimensional distribution of water vapor. With the development of GNSS satellite constellation and station network, GNSS water vapor tomography has emerged as a prominent technique for obtaining the three-dimensional water vapor distribution and become a hotspot in GNSS meteorology. We propose an optimal method for GNSS water vapor tomography based on parameter adaptation to solve the ill-posed problem caused by the discretization parameter redundancy in the tomographic model. This method introduces the minimum probability factor of voxels being penetrated by signals, determines the number and position of the redundant parameters in the observation equation by iterating the coefficient matrix, establishes the optimal tomographic model by eliminating the zero elements of the coefficient matrix and the redundant parameters, and calculates the water vapor density of all voxels by constructing a Laplace smoothing equation. The tomography experiments conducted in Hong Kong show that the average root mean square error (RMSE) and mean absolute error (MAE) for the proposed and traditional method are 1.25/0.70 and 1.83/0.92 mm in the internal accurate testing, respectively, compared to the GAMIT-estimated slant water vapor (SWV) as a reference. The average RMSE and normalized RMSE are 10.9 mm and 8.2% for the proposed method, while these values are 13.2 mm and 10% for the traditional method. Using the radiosonde data as references, the proposed method improves average RMSE and MAE by 20.4% and 25.3% compared with the traditional method.