A regional atmospheric weighted mean temperature model considering diurnal variation and nonlinear vertical correction
Atmospheric weighted mean temperature(Tm)is a key factor of Global Navigation Satellite System(GNSS)to retrieve atmospheric water vapor.In response to the problems of existing regional models not being able to simultaneously consider nonlinear changes in the vertical direction and fine diurnal variations,as well as using only a single grid point data when constructing the model,this paper takes China as the research area and proposes a sliding window based regional modeling method.The Tm grid model(CNTm model)that considering diurnal variations and nonlinear vertical corrections is developed using ERA5 data from the European Centre for Medium Range Weather Forecasts(ECMWF)in 2012 to 2017.Both ERA5 data and radiosonde data that were not involved in the modeling in 2018 are treated as reference values to assess the accuracy and applicability of CNTm model,GPT3 model and IGPT2w model are also used to compared with CNTm model.The results show that based on ERA5 data and radiosonde data in 2018,the root mean square error(RMS)of CNTm model are 3.31 K and 3.21 K,respectively.Compared with GPT3 model,the accuracy of CNTm model is improved by 16％and 23％,and compared with IGPT2w model,the accuracy is improved by 7％and 11％,respectively;The estimated value of CNTm model that considering nonlinear vertical correction is close to the vertical trend of Tm,and CNTm model can provide the diurnal variation of Tm.Because CNTm model shows excellent performance in the study area,it has an important application in real-time high-precision and high-resolution GNSS water vapor monitoring in the study area.
CNTm modelAtmospheric weighted mean temperatureGNSSNonlinear vertical correctionDiurnal variation
NETL
NSTL
万方数据
