Abstract: Resampling interpolation method is one of the problems in the field of meteorological information processing research. This paper introduces a direct and indirect resampling interpolation methods which are based on map projection coordinate conversion for the meteorological remote sensing images, and introduces bilinear interpolation method and Bessel interpolation method for the meteorological grid point field. Taking meteorological satellite（FY2E and FY2D） remote sensing images with different resolutions and ECMWF（European Centre for Medium-Range Weather Forecasts） precipitation forecast field for example, the paper analyzes different resampling interpolation methods. The results show that the calculation amount of the nearest neighbor algorithm is less than that of the weighted nearest neighbor algorithm based on indirect resampling method, while the weighted nearest neighbor algorithm can get better results than the nearest neighbor algorithm. With the improvement of resolution, the comparative advantage of calculation amount of the nearest neighbor algorithm is more obvious. The weighted nearest neighbor algorithm is more suitable for high-resolution meteorological remote sensing images. The results also show that Bessel interpolation algorithm is better than bilinear interpolation algorithm for the meteorological grid point field.

Key words: meteorological remote sensing images, meteorological grid point field, resampling, interpolation method

摘要： 重采样插值方法是气象信息处理领域研究的问题之一。针对气象遥感图像，介绍了基于地图投影坐标转换的直接重采样插值方法和间接重采样插值方法；针对气象格点场，介绍了双线性插值方法和贝塞尔插值方法。以气象业务中不同分辨率的气象卫星（FY2E和FY2D）遥感图像以及欧洲中期天气预报中心（ECMWF）降水预报场为例，分别对不同重采样插值方法进行了分析比较。结果表明：基于间接重采样的气象遥感图像最近邻点插值法的计算量小于邻点权重插值方法，而邻点权重插值方法的效果优于最近邻点插值方法；随着图像的分辨率提高，最近邻点插值法与邻点权重插值方法相比，计算量小的优势更加明显；对于高分辨率的气象遥感图像建议采用基于间接重采样的最近邻点法；对于气象格点场，贝塞尔插值方法的插值效果优于双线性插值方法。

关键词: 气象遥感图像, 气象格点场, 重采样, 插值方法