Abstract The precipitable water vapor(PWV) retrieved by MODIS is continuous planar data, but the accuracy is not high. The PWV accuracy of GNSS water vapor detection technology is high, but it is discrete point data. Therefore, we combine the advantages of both, propose a MODIS water vapor correction method that combines inverse distance weighting and Fourier transform, and use GNSS PWV to correct MODIS PWV products. The GNSS data and MODIS water vapor products in Hong Kong are used for experimental verification. Firstly, the spatial interpolation problem is considered to make GNSS PWV and MODIS PWV consistent at spatial points. By comparing the accuracy of six interpolation algorithms with different MODIS pixel points, it is found that the algorithm with 5 interpolation parameters has the best effect, and the inverse distance weighting method has the highest accuracy, with an average deviation of -0.99 mm. Secondly, considering the water vapor correction problem, three models are constructed to correct MODIS PWV. It is shown that the root mean square error improvement rate of Fourier transform model can reach more than 70%. The MODIS water vapor correction method of the new model can weaken the influence of local data, and more accurately reflect the water vapor distribution in the region.
YAN Xiangrong,YANG Weifang,DING Nan et al. MODIS PWV Correction Method Based on Inverse Distance Weighting and Fourier Transform Combined Model[J]. jgg, 2024, 44(6): 584-590.
YAN Xiangrong,YANG Weifang,DING Nan et al. MODIS PWV Correction Method Based on Inverse Distance Weighting and Fourier Transform Combined Model[J]. jgg, 2024, 44(6): 584-590.