﻿ GPT3模型反演GNSS大气可降水量精度评定
 文章快速检索 高级检索
 大地测量与地球动力学  2022, Vol. 42 Issue (5): 483-488  DOI: 10.14075/j.jgg.2022.05.008

### 引用本文

CAI Meng, LIU Lilong, HUANG Liangke, et al. Evaluation of GNSS Precipitable Water Vapor Derived from GPT3 Model[J]. Journal of Geodesy and Geodynamics, 2022, 42(5): 483-488.

### Foundation support

National Natural Science Foundation of China, No.42064002; Natural Science Foundation of Guangxi, No.2017GXNSFDA198016; Guangxi Bagui Scholar Special Fund of Post and Innovation.

### Corresponding author

LIU Lilong, PhD, professor, PhD supervisor, majors in GNSS technology, E-mail: hn_liulilong@163.com.

### About the first author

CAI Meng, postgraduate, majors in GNSS meteorology, E-mail: Caimeng9851@163.com.

### 文章历史

GPT3模型反演GNSS大气可降水量精度评定

1. 桂林理工大学测绘地理信息学院，桂林市雁山街319号，541006;
2. 广西空间信息与测绘重点实验室，桂林市雁山街319号，541006

1 数据来源及计算方法 1.1 数据来源

 图 1 GNSS站与探空站分布 Fig. 1 Distribution of GNSS stations and radiosonde stations
1.2 GNSS PWV反演计算

GPT3模型的表达式为：

 $\begin{gathered} r(t)= \\ A_{0}+A_{1} \cos \left(\frac{\text { doy }}{365.25} 2 \pi\right)+B_{1} \sin \left(\frac{\text { doy }}{365.25} 2 \pi\right)+ \\ A_{2} \cos \left(\frac{\text { doy }}{365.25} 4 \pi\right)+B_{2} \sin \left(\frac{\text { doy }}{365.25} 4 \pi\right) \end{gathered}$ (1)

 $\mathrm{PWV}=K \cdot \mathrm{ZWD}$ (2)
 $K=\frac{10^{6}}{\rho_{\mathrm{w}} R_{v}\left[\left(k_{3} / T_{m}+k_{2}^{\prime}\right)\right]}$ (3)

2 精度评估

 $\operatorname{bias}=\frac{1}{N} \sum\limits_{i=1}^{N}\left(X_{m}^{M_{i}}-X_{m}^{R_{i}}\right)$ (4)
 $\mathrm{RMSE}=\sqrt{\frac{1}{N} \sum\limits_{i=1}^{N}\left(X_{m}^{M_{i}}-X_{m}^{R_{i}}\right)^{2}}$ (5)

2.1 GPT3模型计算气温、气压的精度评估

 图 2 利用探空站检验GPT3模型气压和气温的bias和RMSE分布 Fig. 2 Distribution of bias and RMSE values of air pressure and temperature obtained from GPT3 tested by radiosonde stations

2.2 GPT3模型计算Tm的精度评估

 图 3 利用探空站检验GPT3模型Tm的bias和RMSE分布 Fig. 3 Distribution of bias and RMSE values of Tm obtained from GPT3 model tested by radiosonde stations

 图 4 Tm的bias和RMSE分布直方图 Fig. 4 Histogram of bias and RMSE of Tm

2.3 基于GPT3模型的GNSS反演PWV的精度分析

 图 5 SXTY、HBES、SCGZ和XJKC站PWV时间序列 Fig. 5 PWV time series at SXTY, HBES, SCGZ and XJKC stations

 图 6 SXTY、HBES、SCGZ和XJKC站GPT3模型反演的PWV与探空站反演的PWV的偏差 Fig. 6 Bias of PWV derived from radiosondes with respect to GPT3 model-derived PWV at SXTY, HBES, SCGZ and XJKC stations

 图 7 基于GPT3模型反演出的PWV与探空站推导的PWV之间的RMSE Fig. 7 RMSE of GPT3 model-derived PWV with respect to radiosonde station-derived PWV

3 结语

1) 与49个探空站实测的气压、气温相比，GPT3-1模型的气压和气温bias均值分别为0.73 hPa和1.34 K，RMSE均值分别为4.21 hPa和3.75 K；GPT3-5模型的气压和气温bias均值分别为1.08 hPa和1.31 K，RMSE均值分别为4.35 hPa和3.81 K。总体而言，GPT3-1和GPT3-5模型气温和气压在南方地区的精度优于北方地区和西部地区，并且GPT3-1模型优于GPT3-5模型。

2) GPT3-1模型Tm的精度和稳定性优于GPT3-5模型。相对于南方地区，GPT3模型估算的Tm在北方地区和西部地区表现出较大的误差，可能是受地形起伏大和显著的Tm日周期变化影响。

3) 利用GPT3-1模型提供的气温结合Bevis经验公式得到的Tm反演的PWV和GPT3-1模型提供的Tm反演的PWV与探空站推导的PWV呈现出较好的一致性，且2种方法反演精度相当。总体而言，南方和北方地区站点反演的PWV的RMSE要大于青藏和西北地区，因而GPT3模型在青藏和西北地区反演大气水汽的精度更高，但北方和南方地区多数站点精度也能满足气象学应用要求。

 [1] Bevis M, Businger S, Herring T A, et al. GPS Meteorology: Remote Sensing of Atmospheric Water Vapor Using the Global Positioning System[J]. Journal of Geophysical Research: Atmospheres, 1992, 97(D14): 15 787-15 801 DOI:10.1029/92JD01517 (0) [2] Rocken C, Hove T, Ware R. Near Real-Time GPS Sensing of Atmospheric Water Vapor[J]. Geophysical Research Letters, 1997, 24(24): 3 221-3 224 DOI:10.1029/97GL03312 (0) [3] Huang L K, Jiang W P, Liu L L, et al. A New Global Grid Model for the Determination of Atmospheric Weighted Mean Temperature in GPS Precipitable Water Vapor[J]. Journal of Geodesy, 2019, 93(2): 159-176 DOI:10.1007/s00190-018-1148-9 (0) [4] Huang L K, Mo Z X, Liu L L, et al. Evaluation of Hourly PWV Products Derived from ERA5 and MERRA-2 over the Tibetan Plateau Using Ground-Based GNSS Observations by Two Enhanced Models[J]. Earth and Space Science, 2021, 8(5) (0) [5] Zhang Y L, Cai C S, Chen B Y, et al. Consistency Evaluation of Precipitable Water Vapor Derived from ERA5, ERA-Interim, GNSS, and Radiosondes over China[J]. Radio Science, 2019, 54(7): 561-571 (0) [6] 黄良珂, 彭华, 刘立龙, 等. 顾及垂直递减率函数的中国区域大气加权平均温度模型[J]. 测绘学报, 2020, 49(4): 432-442 (Huang Liangke, Peng Hua, Liu Lilong, et al. An Empirical Atmospheric Weighted Mean Temperature Model Considering the Lapse Rate Function for China[J]. Acta Geodaetica et Cartographica Sinica, 2020, 49(4): 432-442) (0) [7] Landskron D, Böhm J. VMF3/GPT3:Refined Discrete and Empirical Troposphere Mapping Functions[J]. Journal of Geodesy, 2018, 92(4): 349-360 DOI:10.1007/s00190-017-1066-2 (0) [8] 章传银, 郭春喜, 陈俊勇, 等. EGM 2008地球重力场模型在中国大陆适用性分析[J]. 测绘学报, 2009, 38(4): 283-289 (Zhang Chuanyin, Guo Chunxi, Chen Junyong, et al. EGM 2008 and Its Application Analysis in Chinese Mainland[J]. Acta Geodaetica et Cartographica Sinica, 2009, 38(4): 283-289 DOI:10.3321/j.issn:1001-1595.2009.04.001) (0) [9] Saastamoinen J. Contributions to the Theory of Atmospheric Refraction[J]. Bulletin Geéodeésique, 1972, 105: 279-298 (0) [10] 高壮, 何秀凤, 常亮. GPT3模型在中国地区的精度分析[J]. 武汉大学学报: 信息科学版, 2021, 46(4): 538-545 (Gao Zhuang, He Xiufeng, Chang Liang. Accuracy Analysis of GPT3 Model in China[J]. Geomatics and Information Science of Wuhan University, 2021, 46(4): 538-545) (0) [11] 黄良珂, 李琛, 王浩宇, 等. 基于GPT2w模型计算中国地区对流层加权平均温度的精度分析[J]. 大地测量与地球动力学, 2019, 39(5): 496-501 (Huang Liangke, Li Chen, Wang Haoyu, et al. Precision Analysis of the Tropospheric Weighted Mean Temperature Derived from GPT2w Model over China[J]. Journal of Geodesy and Geodynamics, 2019, 39(5): 496-501) (0) [12] Wang H, Wei M, Li G P, et al. Analysis of Precipitable Water Vapor from GPS Measurements in Chengdu Region: Distribution and Evolution Characteristics in Autumn[J]. Advances in Space Research, 2013, 52(4): 656-667 DOI:10.1016/j.asr.2013.04.005 (0)
Evaluation of GNSS Precipitable Water Vapor Derived from GPT3 Model
CAI Meng1     LIU Lilong1,2     HUANG Liangke1     MO Zhixiang1     HUANG Donggui1     LI Haojie1
1. College of Geomatics and Geoinformation, Guilin University of Technology, 319 Yanshan Street, Guilin 541006, China;
2. Guangxi Key Laboratory of Spatial Information and Geomatics, 319 Yanshan Street, Guilin 541006, China
Abstract: Firstly, we evaluate the accuracy of the meteorological parameters estimated by the GPT3 model using the data of 49 radiosonde stations adjacent to GNSS stations in China from 2017 to 2018. Secondly, combining the meteorological parameters estimated by the GPT3 model and 49 GNSS stations, we calculate the daily mean PWV, and evaluate the accuracy by radiosonde stations adjacent to GNSS stations. Finally, the results are obtained through the experiment: 1) In China, the accuracy and stability of the GPT3 model with 1° resolution are better than those with 5° resolution. The air pressure, temperature and Tm annual bias are 0.73 hPa, 1.34 K and -1.67 K, and the annual RMSE are 4.21 hPa, 3.75 K and 4.15 K. 2) The accuracy of PWV based on temperature inversion by GPT3 model combined with Bevis empirical formula is similar to that by GPT3 model Tm, and the PWV obtained by the two methods and PWV obtained by the sounding data show good consistency; furthermore, the accuracy of PWV in Tibet Plateau and northwest China is better than that in the southern and northern regions.
Key words: GPT3 model; GNSS; atmospheric weighted mean temperature; precipitable water vapor