基于太阳观测的日冕物质抛射驱动的地磁暴及其预报研究

A geomagnetic storm is a temporary disturbance of the Earth’s magnetosphere caused by a solar wind shock wave and/or magnetic cloud that interacts with the Earth’s magnetic field. A violent geomagnetic storm may bring about the global communication outage and also poses great threats to the spacecraft. What’s more, the strong fluctuation of the geomagnetic field may result in serious damage on the electric transmission line and the pipe laying, especially at the high latitude area. It has been known that most big geomagnetic storms are generated by the coronal mass ejection (CME), but not all the CMEs can produce geomagnetic stroms and the mechanism of the storm is still not fully understood. For the forecasting the geomagnetic storm more accurately by using the corona data, only the velocity, the location, and the number of the CMEs seems not enough. In this project, we plan to use the corona data directly to investigate and forecast the geomagnetic storm which can save 1-2 days by comparing with the traditional forecast by using the solar wind data at the L1 point. By using the high resolution solar activities (especially CME) data from Chinese observatory, SOHO, STEREO A and B on different solar cycle and different phases in one solar cycle, we will use the modern statistical method (like machine learning and SVM) to investigate the solar influences on the geomagnetic activities especially at the middle and low latitude, the relationship between the violent geomagnetic storm and the type of CMEs, and discuss the generation mechanism and solar source of the geomagnetic storm. We will build models directly from the solar activities to forecast the solar wind parameters at L1 and the Dst indices.

磁暴是一种典型的灾害性空间天气现象。磁暴期间，电离层、中高层大气和高能粒子环境都会有强烈的响应，形成非常恶劣的空间环境，导致近地轨道上高层大气阻力迅速增加、充电效应、电离层通信中断以及高纬地区的输电线路和输油管线损坏等。研究表明，绝大多数大磁暴由CME驱动，但并不是所有CME都能产生大磁暴，目前使用的CME速度位置等参数预报磁暴Dst指数的方法预报效果不理想，太阳风扰动是否影响中低纬的地磁活动存在很大争议。针对这些问题，本项目拟结合现代统计方法，利用可靠的特征选择方法选择最有效的参数，依据太阳观测，确定出最优的模型输入参数，提出利用日冕信息预报L1点太阳风参数以及磁暴Dst指数的模型，研究地磁暴的产生与太阳活动的关系，揭示太阳活动对中低纬地磁活动的影响规律以及回答大磁暴的产生由哪类CME引起等科学问题。

利用太阳活动的天文观测数据，结合支持向量机、人工神经网络技术、广义相加模型等统计方法，通过知识选择学习（LC）和共线性等特征选择方法选择最有效的参数确定出最优的模型输入参数，提出了利用日冕信息预报L1点太阳风参数以及磁暴Dst指数的模型，并研究地磁暴的产生与太阳活动的关系，揭示出太阳活动对中低纬地磁活动的影响规律等科学问题。