Development Of Global Refined Model For Tropospheric Key Parameters Based On The Sliding Window Algorithm

The tropospheric zenith delay(ZTD)and the weighted average temperature(Tm)of the atmosphere are both key parameters of the troposphere.As GNSS navigation and positioning and GNSS water vapor detection requiring high precision and high temporal and spatial resolution for key tropospheric parameters increasingly,high-precision modeling of key tropospheric parameters has gained widespread attention in recent years.Aiming at the problem of too many model parameters in existing tropospheric key parameter models or using single grid data for modeling,a global high-precision ZTD and Tm model are developed by using the latest MERRA-2 atmospheric reanalysis data provided by NASA.The main research content and work of this paper are as follows:1.The ZTD/Tm calculated by the integrated atmosphere reanalysis data is evaluated by using the ZTD data of the IGS site and the Tm data of the radiosonde to MERRA-2.The mean annual ZTD/Tm RMSE calculated by the integrated atmosphere reanalysis data of MERRA-2 in the global region is 1.31 cm and 1.23 K,respectively,indicating that the ZTD and Tm data obtained by the integration of MERRA-2 reanalysis data have high accuracy and stability,and can be used as a data source for the construction of refined model of key parameters of the global troposphere.2.Analyzed the spatial and temporal characteristics of ZTD height scale factors,a global ZTD model named as GGZTD-H model is developed at the average elevation of each window.The accuracy of the GGZTD-H model was verified with the ZTD data from 305 IGS stations and 319 radiosonde data in 2017,and the accuracy was compared with the UNB3 m and GPT2 w models.The results show that:(1)Using MERRA-2 surface ZTD data as a reference value,the bias and RMSE of the GGZTD-H model both show smaller values,which are 0.08 cm and 3.42 cm,respectively.Compared with other models,the GGZTD-H model shows the best accuracy and stability in the world.(2)Taking the IGS station ZTD as the reference value,the GGZTD-H model has the smallest mean annual RMSE value,which is3.29 cm,compared with the UNB3 m model,the GGZTD-H model is increased by1.66cm(36%).Compared with the GPT2w-1 and GPT2w-5 models,the GGZTD-H model is increased by 0.1cm and 0.16 cm,respectively.(3)Taking the radiosonde data as the reference value,the GGZTD-H model has the smallest annual mean RMSE value.Compared with the UNB3 m model,the GGZTD-H model is increased by1.18 cm.The GGZTD-H model also show improvement compared with GPT2w-5 and GPT2w-1 models.3.A new method of constructing a global Tm model based on the sliding window algorithm was proposed,and the temporal and spatial characteristics of the global Tm vertical decline rate model were explored,and a global Tm vertical decline rate model was established.On this basis,a new global Tm grid model was constructed(GGTm-H model).The accuracy of the GGTm-H model is verified by combining the MERRA-2 surface Tm data and 319 radiosonde data.The results show that:(1)Using MERRA-2 surface Tm data as a reference value,the mean annual RMSE of the GGTm-H model is 2.72 K.Compared with the Bevis,GPT2w-5 and GPT2w-1 models,it shows the improvement of 1.5K,0.33 K,0.21 K,respectively.(2)Taking radiosonde data as a reference value,the mean bias and RMSE of the GGTm-H model are-0.41 K and 3.82 K,respectively.Compared with other models,the GGTm-H model has the smallest mean annual bias and RMSE.The new model does not require any meteorological parameters when calculating Tm,so it has important applications in real-time high-precision GNSS water vapor monitoring.