针对NPP VIIRS数据的云检测方法研究

为了获得新型红外成像辐射仪套件(Visible Infrared Imager Radiometer Suite, VIIRS)传感器准确的云掩码数据,克服当前VIIRS云检测算法在中国区域存在的部分缺陷.本研究通过分析当前较为成熟的中分辨率光谱成像仪(Moderate Resolution Imaging Spectroradiometer,MODIS)云掩码算法,结合VIIRS传感器的波段特性,提出了适合中国区域的云检测算法.针对1.38μm高(卷)云检测算法在高海拔区域存在的限制,本研究使用亮温进行辅助检测,降低因低水汽含量造成的误报;针对当前VIIRS M12-M13云检测阈值在我国存在误报的问题,对M12-M13差值云检测在我国适用范围和阈值进行了分析讨论,并使用亮温辅助M12-M13进一步克服地表二项性反射造成的干扰.使用中国区域的两景数据进行应用分析表明:亮温辅助1.38um高(卷)云检测能够较好的抑制地表污染,亮温辅助M12-M13差值云检测比当前VIIRS M12-M13云检测能更好的抑制误报.通过人工解译,将检测结果和解译结果做了对比分析,实例数据检测精度均高于85%,能够满足当前云检测精度要求.

Cloud detection application on NPP VIIRS

Cloud is an important factor that affects global climate change, so the research of cloud monitoring is very important. In order to identify pixel as cloudy or clear from the Visible Infrared Imager Radiometer Suite (VIIRS) sensor and overcome the shortage of current VIIRS cloud mask algorithm in China, this paper made an introduction for the basic theory of current cloud detection algorithm, and then the cloud detection method was presented.The brightness temperature of was used to detect high cloud to overcome the limit of using 1.38 μm in elevated surface, and the appropriatethreshold of M12-M13 cloud detection was determined by some analysis. Application analysis showed that the method presents in this paper could be used to accurately detect the cloud, and the brightness temperature auxiliary M12-M13and 1.38 μm could inhibition the surface reflection. With the artificial interpret of the remote sensing image, the results showed the estimation accuracy was higher than 85% which can satisfy the requirements of the cloud detection accuracy for the National Polar-orbiting Operational Environmental Satellite System (NPOESS) program.