磁层顶穿越事件自动识别算法

磁层顶是太阳风与磁层进行质量、动量、能量交换的关键区域.磁层顶穿越事件（MCEs）可通过对卫星探测到的粒子能谱和磁场数据图进行人工分析的方式来识别.因太阳风动压和行星际磁场的易变性，位于磁层顶附近的卫星经过长期观测可能会经历成千上万次的磁层顶穿越.人工分析的方法工作量巨大，而且识别速度慢.本文发展了一种新的日下点附近MCEs自动识别算法.此算法综合分析卫星探测到的粒子和磁场数据，能有效地减少误判的发生.为了验证算法的有效性，采用单CPU计算机对THEMIS卫星在2007—2018年靠近日下点附近观测到的数据进行MCEs自动识别，最终在约6h共识别出16758个MCEs.这些自动识别出来的MCEs样本可用于统计研究磁层顶相关的诸多物理问题，如凹陷磁层顶、太阳风与磁层相互作用，磁层顶磁场重联等.同时还分析了算法的精确性和局限性. 

Automatic Identification of Magnetopause Crossing Events

The magnetopause is the key area of mass, momentum and energy coupling between the solar wind and the magnetosphere. The Magnetopause Crossing Events (MCEs) are usually identified by visual inspection of the plots, which is labor intensive and inefficient. Here we develop a novel procedure that is able to rapidly identify MCEs around the subsolar point and accurately define the transition layer between the magnetosphere and the magnetosheath. Here we synthetically consider the characteristics of the variations of the magnetic field and the particle flux for the identification criteria so that false identifications are avoided to the greatest possible. To demonstrate the efficiency of this procedure, it is applied to the THEMIS observations from 2007 to 2018 when THEMIS's apogee is near the subsolar point. The procedure successfully determines 16758MCEs in about 6 hours on a common PC. The huge number of identified samples of MCEs would benefit the investigations on the magnetopause-related scientific problems, such as indentation of the magnetopause, solar wind-magnetosphere interaction, magnetic reconnection process, and so on. The accuracy and limitation of this algorithm are also analyzed in this paper.