基于TC方法青藏高原卫星与ERA5水汽误差分析及融合

Analysis and Fusion of Water Vapor Error between Satellite and ERA5 over the Tibetan Plateau based on TC Method

为分析中国的静止卫星FY-4A搭载的多通道光谱成像仪(AGRI)、国外的极轨卫星Terra搭载的中分辨率光谱成像仪(MODIS)和ERA5再分析数据的水汽产品在青藏高原地区的差异,采用Triple-Collocation方法,在该地区0.25°×0.25°空间尺度上对AGRI、MODIS和ERA5来源的大气可降水量产品进行随机误差分析,并开展基于随机误差融合算法的研究。结果表明:不同来源的大气可降水量产品的随机误差存在空间上的差异,随机误差从小到大依次为ERA5、AGRI和MODIS。根据该随机误差,计算3种产品的融合权重系数,随机误差越大则融合权重系数越小,反之越大。基于该融合系数进行数据融合,得到的融合产品在空间完整性上较卫星数据有较大改善,并且对于MODIS的大气可降水量产品在青藏高原地区的应用有了较大的改善。

In order to analyze the Advanced Geostationary Radiation Imager(AGRI) carried by the Chinese geostationary satellite FY-4A,the Moderate Resolution Imaging Spectroradiometer(MODIS) carried by the foreign polar-orbiting satellite Terra,and the water vapor products of the ERA5 reanalyzed data,the water vapor products in the Qinghai-Tibet Plateau difference,the Triple-Collocation method is used to analyze the random error of the precipitable water vapor products from AGRI,MODIS and ERA5 on the spatial scale of the region,and to carry out the research based on the random error fusion algorithm.The results show that there are spatial differences in the random errors of the precipitable water vapor products from different sources,and the random errors are ERA5,AGRI and MODIS in descending order.According to the random error,the fusion weight coefficient of the three products is calculated.The larger the random error,the smaller the fusion weight coefficient,and vice versa.Based on the fusion coefficient for data fusion,the spatial integrity of the obtained fusion product has been greatly improved compared with the satellite data,and the precipitable water vapor product of MODIS has been greatly improved in the Qinghai-Tibet Plateau.