地表信息对静止卫星云检测算法的影响
|
摘要
针对静止卫星资料的云检测存在通道较少、不适用多通道云检测算法的问题,引入地表背景信息,优化了云检测算法。利用FY2C卫星资料以及其他辅助资料,分析了地表背景信息(包括最高红外亮温底图、地表纹理、地表高程信息等)对静止卫星云检测的影响。在综合考虑这些地表背景信息的基础上,提高了静止卫星云检测精度。经过实际研究和业务化应用,引入地表信息将有效提高云检测效率。具体的研究结论包括:(1)优化的动态亮温底图进一步提高了中高纬地区的云检测精度,降低了中低纬地区云检测的误判率;(2)引入地表植被指数的空间纹理法,进一步消除了由于地表差异带来的误差;(3)引入地表高程数据,对于可见光通道云检测精度的提高非常有效,对于红外通道的动态阈值确定也非常有效。
Integrating kinds of methods of cloud detection can obtain more accurate results and is suitable for operational applications.In the present paper,the surface information is introduced,including brightness temperature background,surface texture,elevation,etc.,when kinds of methods of cloud detection are combined,based on FY2 Cdata and some other assistant data.Nearly four years' applications at Remote Sensing Center,Institute of Meteorology,PLA University of Science and Technology indicate that the accuracy comes to 93%.The main results are as follows:(1)optimal maximum brightness temperature plays the major role in cloud detection,and it relieves the difficulty of finding threshold due to variable land covers,especially at high latitude areas;(2)spatial texture method further eliminates the impacts from surface variations;(3)by introducing the elevation information,the cloud detection achieves high accuracy for infrared channels,and dynamic thresholds can also be found effectively for visible channels.
Integrating kinds of methods of cloud detection can obtain more accurate results and is suitable for operational applications.In the present paper,the surface information is introduced,including brightness temperature background,surface texture,elevation,etc.,when kinds of methods of cloud detection are combined,based on FY2 Cdata and some other assistant data.Nearly four years' applications at Remote Sensing Center,Institute of Meteorology,PLA University of Science and Technology indicate that the accuracy comes to 93%.The main results are as follows:(1)optimal maximum brightness temperature plays the major role in cloud detection,and it relieves the difficulty of finding threshold due to variable land covers,especially at high latitude areas;(2)spatial texture method further eliminates the impacts from surface variations;(3)by introducing the elevation information,the cloud detection achieves high accuracy for infrared channels,and dynamic thresholds can also be found effectively for visible channels.
引文
[1]王毅.国际新一代对地观测系统的发展及其主要应用[M],北京:气象出版社,2006.
[2]马芳,张强,郭铌,等.多通道卫星云图云检测方法的研究[J].大气科学,2007,31(1):119-128.
[3]WILLIAM B R,LEONID C G.Cloud detection using satellite measurements of infrared and visible radiances for ISCCP[J].Journal of Climate 1993,6(12):2341-2369.
[4]PEAK J E,TAG P M.Segmentation of satellite imagery using hierarchical thresholding and neural networks[J].Journal of Applied Meteorology,1994,33(1):605-616.
[5]师春香,瞿建华.用神经网络方法对NOAA-AVHRR资料进行云客观分类[J].气象学报,2002,60(2):123-128.
[6]李艳兵,李元祥,翟景秋.卫星云图形态特征提取和表示的一种方法[J].南京气象学院学报,2006,29(5):74-81.
[7]吴咏明,张韧,蒋国荣,等.多光谱卫星图像的一种模糊聚类方法[J].热带气象学报,2004,6(3/4):74-81.
[8]洪梅,张韧,孙照渤.多光谱卫星云图的高维特征聚类与降水天气判别[J].遥感学报,2006,10(2):42-48.
[9]郑君杰,黄峰,张韧,等.基于纹理与分形理论的气象卫星云图目标物识别[J].气象科学,2005,25(3):244-248.
[10]郭洪涛,王毅,刘向培,等.卫星云图云检测的一种综合优化方法[J].解放军理工大气学报(自然科学版),2010,11(2):221-227.
[2]马芳,张强,郭铌,等.多通道卫星云图云检测方法的研究[J].大气科学,2007,31(1):119-128.
[3]WILLIAM B R,LEONID C G.Cloud detection using satellite measurements of infrared and visible radiances for ISCCP[J].Journal of Climate 1993,6(12):2341-2369.
[4]PEAK J E,TAG P M.Segmentation of satellite imagery using hierarchical thresholding and neural networks[J].Journal of Applied Meteorology,1994,33(1):605-616.
[5]师春香,瞿建华.用神经网络方法对NOAA-AVHRR资料进行云客观分类[J].气象学报,2002,60(2):123-128.
[6]李艳兵,李元祥,翟景秋.卫星云图形态特征提取和表示的一种方法[J].南京气象学院学报,2006,29(5):74-81.
[7]吴咏明,张韧,蒋国荣,等.多光谱卫星图像的一种模糊聚类方法[J].热带气象学报,2004,6(3/4):74-81.
[8]洪梅,张韧,孙照渤.多光谱卫星云图的高维特征聚类与降水天气判别[J].遥感学报,2006,10(2):42-48.
[9]郑君杰,黄峰,张韧,等.基于纹理与分形理论的气象卫星云图目标物识别[J].气象科学,2005,25(3):244-248.
[10]郭洪涛,王毅,刘向培,等.卫星云图云检测的一种综合优化方法[J].解放军理工大气学报(自然科学版),2010,11(2):221-227.