GPS PPP-derived precipitable water vapor retrieval based on Tm/Ps from multiple sources of meteorological data sets in China

Abstract

Weighted mean temperature (Tm) and pressure (Ps) are two parameters of great relevance to precipitable water vapor (PWV) retrieval from global positioning system (GPS) data. However, information about the Tm and Ps cannot be available for those GPS stations that are not colocated with meteorological sensors. To investigate the optimal GPS-PWV retrieval method for China, two enhanced Tm models, GM-Tm (temperature dependent) and GH-Tm (temperature independent), are developed. Additionally, the potentials of the Ps data from the two reanalysis data sets, the National Centers for Environmental Prediction (NCEP)-Department of Energy (DOE) Reanalysis II (NCEP II) and ERA-Interim, and from the empirical model GPT2w for GPS-PWV retrieval are investigated over China. To evaluate the performances of multisources Tm and Ps data for GPS-PWV retrieval, GPS data (2011-2013) collected from 22 stations of the Crustal Movement Observation Network of China (CMONOC) were processed by using the precise point positioning (PPP) technique, estimating the zenith tropospheric delay (ZTD) so as to be subsequently converted to GPS-PWV. The retrieved GPS-PWVs are compared with their counterparts derived from NCEP II and radiosonde data over China. The results show that (1) the GM-Tm model consistently shows the highest accuracy (with root mean square error of 2.3 K), and the GH-Tm model should be selected when temperature observations are not available, and that (2) the performances of Ps from NCEP II and ERA-Interim differ marginally for GPS-PWV retrieval, and significant seasonal variations are found in the agreement between the GPS-PWVs and the PWVs derived from NCEP II and radiosonde data over China.