Improving Interpolating Accuracy of Weighted Mean Temperature by Using a Novel Lapse Rate Model in Compact VMF1 Product

Abstract

In GNSS (Global Navigation Satellite Systems) meteorology, the accuracy of precipitable water vapor (PWV) retrieved from the tropospheric delay of GNSS signals is affected by the conversion factor. Compact VMF1 product (known as GGOS Atmosphere data) provides high-accuracy global grid-wise weighted mean temperature (T_m) values, which can be utilized to calculate the conversion factor. However, the T_m provided in the compact VMF1 data are solely ground surface values. To enhance the performance of compact VMF1 product, a new T_m lapse rate model for each grid point was developed for the purpose of reducing its surface T_m to the elevation of the GNSS site. Then the reduced T_m values over the neighboring grid points together with horizontal interpolation were used to obtain the interpolated T_m for the GNSS station. The sample data for the development of the new model were the T_m profiles obtained from ERA5 monthly averaged data spanning 2009–2018. To assess the model's performance, global radiosonde data at 504 radiosonde stations spanning 2019–2021 were employed. Results demonstrated that implementing the T_m lapse rate model significantly enhanced the accuracy of interpolating T_m values for GNSS stations with substantial height disparities from adjacent grid points, thereby improving PWV conversion accuracy. This indicates that employing the new T_m lapse rate model to adjust surface T_m data in the compact VMF1 product holds promise for enhancing its utility in GNSS meteorology.