Atmospheric precipitable water vapor (PWV) plays a key role in the study of global climate change and numerical weather prediction. However, the vertical correction of PWV mainly relies on a simple empirical correction model at present, which limite the high precision fusion of multi-source water vapor products and the comparisons between different water vapor products in a way. In this paper, for the characteristics of highly undulating terrain and diverse climate in China, the PWV vertical

more »

... rection models which considering the time-varying lapse rate according to the whole areas in Mainland China and geographical divisions, named as C-PWVC1 model and C-PWVC2 model respectively, are established using ERA5 reanalysis data provided by the European Center for Medium Range Weather Forecast (ECMWF) from 2012 to 2017. Taking the profiles of 86 radiosonde stations in China in 2017 as reference values, the ERA5 and MERRA-2 reanalysis gridded data are interpolated into radiosonde stations to evaluate the performance of the newly established models, respectively. The results show that the performance of C-PWVC1 and C-PWVC2 models in PWV vertical correction is comparable. Compared with the condition without vertical correction, the performances of the C-PWVC model for ERA5 and MERRA-2 are improved by 16% and 8%, respectively. Compared with the commonly used PWV vertical correction model, C-PWVC model has little improvement for ERA5, but 12% for MERRA-2. C-PWVC model shows significant advantages when the height difference between two kinds of PWVs is large, and the correction capability for MERRA-2 is greater than ERA5. Moreover, for the ERA5 with different spatial resolutions, C-PWVC model has better interpolation precision and stability when compared with the commonly used PWV vertical correction model, especially in southern and western China. Therefore, the C-PWVC model has a good performance in PWV vertical correction for Mainland China, which can provide important application for the comparison and fusion of multi-s [...]