基于MODIS云产品的AIRS云相态的识别和云量产品的检验

AIRS CLOUD PHASE IDENTIFICATION AND CLOUD FRACTION PRODUCTS EXAMINATION BASED ON MODIS CLOUD PRODUCTS

选用中国中东部（95～130°E，20—60°N）午后同一时刻相对应的AIRS与MODIS各15景样本数据，将MODIS在红外波段上四种基于亮温的云识别算法对AIRS晴空像元和云像元中不同的云相态进行区分，分别统计出AIRS不同类型像元中这四种云识别算法的亮温差异，根据AIRS在晴空、水云和冰云三种不同相态下亮温的差异，确定出其阈值，实现AIRS晴空和云相态识别。在此基础上，对AIRS不同云相态下反演的有效云量进行检验和改进。结果表明：运用亮温阈值法对AIRS进行云相态识别能够较好地反映AIRS的晴空和云相态特征，云相态分布与MODIS云相态产品对应效果较好，特别在云边缘区域。对AIRS不同云相态下反演的有效云量进行对比和误差分析后发现：当水云有效云量较高、冰云有效云量较低时，AIRS反演的有效云量误差较大。在误差分析的基础上，提出了AIRS有效云量的偏差订正算法，对AIRS反演的有效云量具有一定改进，为AIRS云产品的应用提供参考依据。

Selecting fifteen samples of AIRS and MODIS data of the same post-midday moment of time each in Central and East China （95-130 °E, 20-60 °N）, four cloud recognition algorithms based on brightness temperature in the infrared wavelength of MODIS are used to identify clear sky pixels and different cloud phases in the cloud pixels of AIRS, and determine the brightness temperature differences between different AIRS types of pixels in the four cloud recognition algorithms. Based on the brightness temperature differences of AIRS in the three different phases for clear sky, water cloud and ice cloud, respective thresholds are determined to identify clear sky and cloud phase for AIRS. On the basis of the above work, the AIRS effective cloud fraction is tested and improved for retrieval in the different phase. The results are shown as follows. Applying the method of brightness temperature threshold to AIRS cloud phase identification can better reflect the characteristics of clear sky and cloud phase for AIRS, and cloud phase distribution corresponds better with MODIS cloud phase product, especially in the edge of the cloud. Having analyzing the comparison and deviation for AIRS effective cloud fraction in different cloud phases, the following conclusions can be drawn. When effective cloud fraction is low with water clouds but high with ice clouds, the effective cloud fraction as retrieved by AIRS is large. On the basis of the error analysis, an effective cloud fraction deviation correction algorithm is presented for AIRS. It can increase to a certain extent the AIRS effective cloud fraction and provide reference for application of AIRS cloud products.