Evaluation of the weighted mean temperature over China using multiple reanalysis data and radiosonde

Global Navigation Satellite Systems (GNSS) meteorology has received extensive attention for weather research in recent years. As a crucial conversion parameter, weighted mean temperature (Tm) is used to calculate precipitable water vapor (PWV) by the signal path wet delay in GNSS meteorology. Comparing with Tm calculated by radiosonde, Tm from reanalysis data is more convenient due to higher spatiotemporal resolution, such as the second Modern-Era Retrospective Analysis for Research and Applications (MERRA-2), the fifth-generation ECMWF Reanalysis (ERA5), Japanese 55-year Reanalysis Project (JRA-55), the global atmospheric reanalysis of China Meteorological Administration (CRA40). To investigate their different performances, a comprehensive assessment of multiple Tm datasets from 2011 to 2019 was conducted using the radiosonde data over China as references, in which the root mean square error (RMSE) and the mean absolute error (MAE) were selected as statistics. The results showed that the Tm derived from multiple reanalysis data exhibited excellent applicability in China, especially in the southeast region. The RMSE and MAE were 1.69/1.36, 1.70/1.36, 1.79/1.45, 1.99/1.61 K for ERA5, JRA-55, CRA40, MERRA-2, respectively. By comparing Tm from ERA5 with three different resolutions, it's found that the accuracy of Tm was getting better with the increase of the spatial resolution. Additionally, the accuracy of these Tm was of great relevance to season, moment, latitude, and elevation. By exploring the intercomparison of Tm from reanalysis data, their correlation coefficients were all above 0.98. This study indicated that the Tm derived from various reanalysis data can be utilized for climate research in China, and the ERA5 data with high resolution is recommended first.