GWMT Global Atmospheric Weighted Mean Temperature Models: Development and Refinement

摘要

The atmospheric weighted mean temperature T_m plays a pivotal role in remote sensing water vapor with GNSS technique. However, real-time T_m cannot be obtained at given place due to limitation of modern technology. In order to remove this barrier, a global and season-specific T_m model (GWMT-Ⅰ) based on spherical harmonics analysis was built using land radiosonde data. Combining Global Pressure and Temperature (GPT) model with the Bevis formula to provide virtual T_m observation at sea, a modified T_m model (GWMT-Ⅱ) was presented aimed to improve the accuracy of GWMT-Ⅰ in the ocean areas. Thereafter, the third generation T_m (GWMT-Ⅲ) model, derived from GGOS Atmosphere surface global grid T_m data, came into existence. But on the one hand, these three models fail to accurately describe vertical T_m variation. On the other hand, they are expanded into spherical harmonics up to degree and order nine which may impose restriction to their precision. This paper reconstructs the latest model (GWMT-Ⅳ) with 4D global NCEP/DOE 2 Reanalysis data from 2005 to 2009. Comparison of result between 2010 NCEP2 data and the GWMT-IV model shows that it achieves high accuracy all over the world and its modeling method better conforms to reality.