一种基于紫外极光图像的亚暴膨胀期起始时刻的自动检测方法  

A method for automatic identification of substorm expansion phase onset from UVI images

作　　者：杨秋菊[1] 梁继民[1] 刘俊明[2] 胡泽骏[2] 胡红桥[2] 

机构地区：[1]西安电子科技大学电子工程学院,西安710071 [2]中国极地研究中心,上海200136

出　　处：《地球物理学报》2013年第5期1435-1447,共13页Chinese Journal of Geophysics

基　　金：国家自然科学基金(60872154;41031064;40904041;41274164;41104090;41274148);海洋公益性行业科研专项(201005017);国家重点基础研究发展计划(973计划)(2010CB950503-06);海洋局极地考察办公室对外合作支持项目(IC201205)资助

摘　　要：准确界定亚暴起始时刻是理解亚暴相关问题的关键.已有研究主要集中在两方面:一是从极光图像中人工挑选亚暴事件进行案例分析或统计分析来研究亚暴发生机制及亚暴期间的地磁环境;二是基于一些空间物理参数,如AE指数、SME(SuperMAG electrojet)指数、Pi2、正弯扰等,采用人眼判断或是模式识别的方法从中找出亚暴起始时刻.本文尝试采用模式识别的方法从紫外极光图像中自动地检测出亚暴膨胀期起始时刻.首先,将紫外极光图像通过网格化处理转换到磁地方时-地磁纬度(MLT-MLAT)直角坐标下,然后通过模糊c均值聚类方法提取亮斑,再考察亮斑强度是否增强、面积是否极向膨胀来判断是不是亚暴事件.本文方法在1996年12月—1997年2月这三个月的Polar卫星紫外极光图像上进行了实验验证.我们将检测到的亚暴起始时刻与Liou(J.Geophys.Res.,2010,115:A12219)的人工标记进行了对比,并详细分析了与标记不一致的多检和漏检事件.本文提出的自动检测方法可以快速地从海量紫外极光图像中完成亚暴事件的初步筛选,方便研究人员进一步深入研究极光亚暴.Substorm research largely depends on the precise definition and the timing accuracy of the substorm onset used in various observations.At present,the substorms are studied mainly in two ways.One is the case study or statistical study by selecting substorm events from auroral images via visual inspection;and the other is examining,manually or automatically,some physical indices,such as the auroral electrojet(AE) index,SuperMAG electrojet(SME) index,Pi2,and positive bays.In this paper,we propose to automatically identify auroral substorm onset time from global Ultraviolet Imager(UVI) images.We first transformed the original UVI images into the MLT-MLAT rectangular coordinate system,and the bright bulge was determined with the spatial fuzzy c-means method.The proposed technique was tested using Polar UVI observations acquired from three months of the winter season in 1996—1997.The identified onset results were compared with the available manual statistical report made by Liou(J.Geophys.Res.,2010,115: A12219),and the missed substorms and the extra identified events were analyzed in detail.The proposed method can easily provide the substorm candidates from massive UVI images for researchers to further explore the nature of auroral substorm.

关 键 词：亚暴 紫外极光图像 MLT-MLAT网格图 模糊C均值聚类 

分 类 号：P353[天文地球—空间物理学]