基于聚类分析的不同形态极光的产生条件研究

项目名称： 基于聚类分析的不同形态极光的产生条件研究

项目编号： No.41504115

项目类型： 青年科学基金项目

立项/批准年度： 2016

项目学科： 天文学、地球科学

项目作者： 王倩

作者单位： 西安邮电大学

项目金额： 22万元

中文摘要： 极光是极区日地物理过程的最集中的表现形式，是研究太阳风暴的最佳窗口。不同的极光形态可能是各种不同动力学过程独立或交互作用产生的结果。为了研究各种极光形态的产生条件，首先应该对极光现象依据形态进行分类，然而极光形态和动态演变过程极为复杂，至今尚无明确的形态特征描述和合理的分类方法。本项目拟基于高分辨率的多波段全天空极光观测图像，提取极光的形态和瞬时运动信息，并且联合其他多种同步极光观测数据验证聚类有效性，实现极光的自动聚类，对现有的分类机制进行补充和矫正。然后，通过极光类型标记与各种同步观测数据的回归分析试验，定量分析极光类型标记与各种磁层动力学过程的相关物理参数的相关性，估计各种极光形态的产生条件，可望在极光类型的量化描述上取得新进展。

中文关键词： 极光；形态；聚类分析；产生条件；回归分析

英文摘要： Changeable auroral morphology maps solar-terrestrial physics along the Earth’s magnetic field lines projected onto the polar ionosphere. The aurora phenomena provide a best window of solar storms from the outer magnetosphere. The auroral appearances are ascribed to cumulative effects of the solar wind–magnetosphere interaction and the physics of the magnetosphere–ionosphere interaction. In order to study the occurrence conditions of the various auroral morphologies, the first step should be classifying the auroral phenomena based on the morphology. However, the available classification schemes are derived by observing with naked eyes and subjectively judgment. And the automatic classification of the massive amounts of auroral data is based on manually labeling. There is no generally accepted set of auroral types for auroral classification studies. In order to achieve the classification of the auroral data by clustering method, this proposal will extract the morphological and motion information based on the all-sky image, and obtain the clustering validities by combining with multiple auroral observation data. Then, by regressive analysis of clustering labels and other synchronous observation data of various magnetic layer dynamic process, their dependency and correlation can be obtained. Thereby, the occurrence conditions of auroral morphologies can be concluded, which will make quantitative representation of auroral types more fruitful.

英文关键词： Aurora;Morphology;Clustering analysis;Occurrence conditions;Regression analysis