Objectives  Using moderate-resolution imaging spectroradiometer (MODIS) can detect high spatial resolution but low precision precipitable water vapor (PWV), while global navigation satellite system (GNSS) can retrieve high precision but low spatial resolution PWV, in order to get high spatial resolution and precision PWV, it is necessary to integrate two kinds of data.

  Methods  Taking Chinese mainland as an example, MODIS PWV was done the correction based on GNSS observation and meteorological data provided by crustal movement observation network of China(CMONOC). Firstly, considering the geographical environment, altitude, climate type and other factors, Chinese mainland was divided into 16 regions, and the correlation analysis of regional MODIS PWV and GNSS PWV was carried out. And then, it was constructed for the correction-model of MODIS PWV by different regions and seasons based on GNSS PWV. After that, the regional model result, the single site model result and the measured GNSS PWV were compared to test the reliability of the model. Finally, MODIS PWV distribution in Chinese mainland can be obtained by MODIS PWV correction and image superposition.

  Results  The results show that the accuracy of the regional MODIS PWV correction model is similar to that of the single-site model, which can replace the single-site model for MODIS PWV correction.

  Conclusions  The MODIS PWV correction model of different regions can be used to effectively improve the precision of MODIS PWV. It can be used for providing reference for short-term weather prediction and interferometric synthetic aperture radar atmospheric correction.