基于深度学习的太阳黑子Wilson山磁类型识别方法

太阳黑子是太阳光球层中带有较强磁场的区域,通常是太阳爆发活动的源区。Wilson山磁分类是当前最为主流的太阳黑子分类方法之一,对研究太阳爆发有重要意义。利用2010－2017年间SDO/HMI成像仪观测到的720s_SHARP磁图和白光图数据,研究使用深度学习对太阳黑子群Wilson山磁分类的方法。实验结果表明,Xception网络在识别太阳黑子Wilson山磁类型上能取得最优的效果,其中对α类型黑子的F₁得分为96.50%,β类为93.20%,其他类型的黑子为84.65%。      

Flux emergence,flux imbalance,magnetic free energy and solar flares

Emergence of complex magnetic flux in the solar active regions lead to several observational effects such as a change in sunspot area and flux embalance in photospheric magnetograms. The flux emergence also results in twisted magnetic field lines that add to free energy content. The magnetic field configuration of these active regions relax to near potential-field configuration after energy release through solar flares and coronal mass ejections. In this paper, we study the relation of flare productivity of active regions with their evolution of magnetic flux emergence, flux imbalance and free energy content. We use the sunspot area and number for flux emergence study as they contain most of the concentrated magnetic flux in the active region. The magnetic flux imbalance and the free energy are estimated using the HMI/SDO magnetograms and Virial theorem method. We find that the active regions that undergo large changes in sunspot area are most flare productive. The active regions become flary when the free energy content exceeds 50% of the total energy. Although, the flary active regions show magnetic flux imbalance, it is hard to predict flare activity based on this parameter alone.     

太阳黑子自动识别与特征参量自动提取

日面上黑子数目反映了太阳活动水平的高低.黑子形态的复杂性和磁场的非势性与太阳活动爆发密切相关.随着高时空精度的太阳观测数据量的急剧增长，快速准确地自动识别日面上的黑子以及对黑子群特征自动提取已成为太阳活动预报的现实需求.本文针对SDO/HMI的活动区白光数据，利用数学形态法开展黑子自动识别研究，并在黑子识别基础上对黑子群的面积和黑子数进行了计算.通过对利用2011-2017年HMI活动区数据计算得到的黑子群面积和黑子数与NOAA/SWPC发布的活动区相应参量进行比较，发现本文计算结果与SWPC发布数据的变化趋势基本一致，相关性较好.其中黑子群面积的相关系数为0.77，黑子数的相关系数为0.79.研究结果表明，利用本文方法对SDO/HMI数据进行处理，能够得到高时间分辨率的黑子群特征参量，可为太阳活动预报提供及时准确的输入.      

太阳活动区EUV图像的生成式模型耀斑分级与预报

近年来,不断发射的空基观测台持续传送回海量日面图像及日地间气象数据,为采用人工智能技术对太阳活动进行预报预警提供了数据基础。但是,极端天气爆发少,样本量较少;中等程度爆发稍多,样本量较多;常规无爆发天气常见,样本较为集中,样本不均衡状况严重影响机器学习方法在空间天气领域的广泛应用。本文面向多源多通道多尺度日面图像信息,构建了来自SOHO和SDO的1996-2015年日面活动区图像数据集;针对数据分布的不平衡,对太阳活动区图像作耀斑分级与预报。在对比分析元学习算法的基础上,设计了结合分类头设计和卷积核初始化的生成式模型;在使网络轻量化的基础上,能够将M和X级耀斑预报的检测率指标相较于普通的深度学习模型和无监督度量式模型分别提升10%和7%。     

基于长短期记忆神经网络的太阳耀斑短期预报

提出了一个基于长短期记忆神经网络的耀斑预报模型,利用过去24 h太阳活动区的磁场变化时序构建样本,通过长短期记忆神经网络对磁场特征时序演化进行分析,预报未来48 h内是否发生≥M级别耀斑事件。使用的数据集为2010年5月到2017年5月所有活动区样本,选取了SDO/HMI SHARP的10个磁场特征参量。在建模过程中通过XGBoost方法选取权重、增益率和覆盖率均较高的6个特征参量作为输入参数。通过测试对比,模型的虚报率和准确率与传统机器学习模型相近,报准率和临界成功指数分别为0.7483和0.7402,优于传统机器学习模型。模型总体效果优于传统机器学习模型。      

Sunspot motions and magnetic shears as precusors of flares

Using full-disc white light photoheliograms, we have studied umbrae motion and variations in sunspot areas in a large activity complex over 4 solar rotations. On the basis of the observational data we illustrate with typical examples to what extent rapid spot motions are associated with flare occurrences.     

1996-2002年太阳耀斑的统计分析

分析了1996-2002年南北半球的太阳黑子相对数和南北半球太阳X射线耀斑级别(简称Imp)≥M1．0的太阳X射线耀斑的特征和不对称性．分析结果表明，南北半球的太阳耀斑活动的程度交替上升，在2001年7月以前北半球的太阳耀斑活动强于南半球，2001年7月开始耀斑活动逐渐以南半球为主．本文还逐月分析了1996—2001年南北半球的耀斑指数．2000年7月为第23周太阳指数最大的一个月，与第23周太阳黑子相对数最大月均值吻合．     

2011年8月4日螺旋状日冕物质抛射爆发事件

利用多卫星多波段的综合观测数据,通过追踪光球表面等离子体速度分析计算了耀斑爆发前后磁螺度的变化,发现耀斑爆发前活动区中光球表面存在强的水平剪切运动,活动区磁螺度的注入主要由这种剪切运动所产生;使用CESE-MHD-NLFFF重建了耀斑爆发前后活动区的磁场位形,推测出耀斑过程中存在磁绳结构的抛射.基于这些分析,给出了这一螺旋状抛射结构的形成机制:爆发前暗条西侧足点的持续剪切运动驱动磁通量绳增加扭转,高度扭缠的通量绳与东侧足点附近的开放磁力线重联并与东侧足点断开,进而向外抛出并伴随解螺旋运动.另外,利用1AU处WIND卫星的观测数据在对应的行星际日冕物质抛射中找到典型磁云的观测特征.这表明除了传统上双足点均在太阳表面的磁云模型,这种单足点固定于太阳表面的磁通量绳爆发图景同样可能在行星系际空间形成磁云结构.研究结果对进一步认识磁云结构具有重要意义.      

太阳活动对电离层TEC变化影响分析ormalsize

为研究太阳活动对电离层TEC变化的影响,从整体到局部分析了2000—2016年的太阳黑子数、太阳射电流量F_10.7指数日均值与电离层TEC的关系,并重点分析了2017年9月6日太阳爆发X9.3级特大耀斑前后15天太阳活动与电离层TEC变化的相关性.结果表明:由2000—2016年的数据整体看来,太阳黑子数、太阳F_10.7指数、TEC两两之间具有很强的整体相关性,但局部相关性强弱不均;此次耀斑爆发前后太阳黑子数、太阳F_10.7指数和TEC具有很强的正相关特性,太阳活动对TEC的影响时延约为2天;太阳活动对全球电离层TEC的影响不同步,从高纬至低纬约有1天的延迟,且对低纬度的影响远大于中高纬度.太阳活动是影响电离层TEC变化的主要原因,但局部也可能存在其他重要影响因素.      

基于PCA方法对太阳活动区主要参量的分析

太阳耀斑与太阳质子事件的发生通常与太阳活动区存在非常密切的关系, 对这种关系的深入分析有助于太阳耀斑和太阳质子事件预报模型的建立. 本文利用主成分分析(Principal Component Analysis, PCA)方法对1997-2010年太阳质子事件所在活动区的主要参量进行分析, 选取的参量包括黑子磁分类、 McIntosh分类、太阳黑子群面积、10.7 cm射电流量、耀斑指数、质子耀斑位置和软X射线耀斑强度. 结果得到81个太阳活动主成分得分值排序(得分值代表每个事件的强弱), 与太阳质子事件峰值流量、太阳黑子年均值以及10.7 cm射电流量年均值的对比显示相似度非常高, 表明主成分得分值一定程度上可以反映太阳活动的强弱规律.      

Application of support vector machine combined with K-nearest neighbors in solar flare and solar proton events forecasting

The support vector machine (SVM) combined with K-nearest neighbors (KNN), called the SVM-KNN method, is new classing algorithm that take the advantages of the SVM and KNN. This method is applied to the forecasting models for solar flares and proton events. For the solar flare forecasting model, the sunspot area, the sunspot magnetic class, and the McIntosh class of sunspot group and 10 cm solar radio flux are chosen as inputs; for the solar proton event forecasting model, the inputs include the longitude of active regions, the flux of soft X-ray, and those for the solar flare forecasting model. Detailed tests are implemented for both of the proposed forecasting models, in which the SVM-KNN and the SVM methods are compared. The testing results demonstrate that the SVM-KNN method provide a higher forecasting accuracy in contrast to the SVM. It also gives an increased rate of ‘Low’ prediction at the same time. The ‘Low’ prediction means occurrence of solar flares or proton events with predictions of non-occurrence. This method show promise for forecasting models of solar flare and proton events.     

Solar flare forecasting model supported with artificial neural network techniques

Nowadays operational models for solar activity forecasting are still based on the statistical relationship between solar activity and solar magnetic field evolution. In order to set up this relationship, many parameters have been proposed to be the measures. Conventional measures are based on the sunspot group classification which provides limited information from sunspots. For this reason, new measures based on solar magnetic field observations are proposed and a solar flare forecasting model supported with an artificial neural network is introduced. This model is equivalent to a person with a long period of solar flare forecasting experience.     

Detailed analysis of dynamic evolution of three Active Regions at the photospheric level before flare and CME occurrence

We present a combined analysis of the applications of the weighted horizontal magnetic gradient (denoted as ${\mathit{WG}}_M$ in Korsós et al. (2015)) method and the magnetic helicity tool (Berger and Field, 1984) employed for three active regions (ARs), namely NOAA AR 11261, AR 11283 and AR 11429. We analysed the time series of photospheric data from the Solar Dynamics Observatory taken between August 2011 and March 2012. During this period the three ARs produced a series of flares (eight M- and six X-class) and coronal mass ejections (CMEs). AR 11261 had four M-class flares and one of them was accompanied by a fast CME. AR 11283 had similar activities with two M- and two X-class flares, but only with a slow CME. Finally, AR 11429 was the most powerful of the three ARs as it hosted five compact and large solar flare and CME eruptions. For applying the ${\mathit{WG}}_M$ method we employed the Debrecen sunspot data catalogue, and, for estimating the magnetic helicity at photospheric level we used the Space-weather HMI Active Region Patches (SHARP’s) vector magnetograms from SDO/HMI (Solar Dynamics Observatory/Helioseismic and Magnetic Imager). We followed the evolution of the components of the ${\mathit{WG}}_M$ and the magnetic helicity before the flare and CME occurrences. We found a unique and mutually shared behaviour, called the U-shaped pattern, of the weighted distance component of ${\mathit{WG}}_M$ and of the shearing component of the helicity flux before the flare and CME eruptions. This common pattern is associated with the decreasing-receding phases yet reported only known to be a necessary feature prior to solar flare eruption(s) but found now at the same time in the evolution of the shearing helicity flux. This result leads to the conclusions that (i) the shearing motion of photospheric magnetic field may be a key driver for solar eruption in addition to the flux emerging process, and that (ii) the found decreasing-approaching pattern in the evolution of shearing helicity flux may be another precursor indicator for improving the forecasting of solar eruptions.     

Some statistical properties of transequatorial loops

Yohkoh soft X-ray telescope brought plenty of high quality images, it provides a good chance to research coronal loops, especially for transequatorial loops (TLs). In this paper, we focus on the statistical results of TLs including static properties and dynamic properties.

There are two types of classification about TLs: according to configuration and according to magnetic polarities of footpoints, respectively. The footpoints of TLs never root in sunspot, in a general way, they exist in moderately strong field. The mean separation value of TLs is close to 30° and the separation value varies with solar cycle. The helicity patterns of active regions connected by TLs are discussed, the mean twist value of TLs is close to zero. The formation of TLs is generally thought to be caused by magnetic reconnection, the relationship of TLs eruption with flare and CME is introduced.     

Ionospheric heating due to solar flares as measured by SROSS-C2 satellite

The data on thermal fluctuations of the topside ionosphere have been measured by Retarding Potential Analyser (RPA) payload aboard the SROSS-C2 satellite over the Indian region for half of the solar cycle (1995–2000). The data on solar flare has been obtained from National Geophysical Data Center (NGDC) Boulder, Colorado (USA) and other solar indices (solar radio flux and sunspot number) were download from NGDC website. The ionospheric electron and ion temperatures show a consistent enhancement during the solar flares. The enhancement in the electron temperature is 28–92% and for ion temperature it is 18–39% compared to the normal day’s average temperature. The enhancement of ionospheric temperatures due to solar flares is correlated with the variation of sunspot and solar radio flux (F10.7cm). All the events studied in the present paper fall in the category of subflare with almost same intensity. The ionospheric electron and ion temperatures enhancement have been compared with the IRI model values.     

Prediction and evaluation of solar particle events based on precursor information.

Protection from the radiation effects of solar particle events for deep space mission crews requires a warning system to observe solar flares and predict subsequent charged particle fluxes. Such a system relates precursor information observed in each flare to the intensity, delay, and duration of the subsequent Solar Particle Event (SPE) at other locations in the solar system. A warning system of this type is now in operation at the NOAA Space Environment Services Center in Boulder, Colorado for support of space missions. It has been used to predict flare particle fluxes at the earth for flares of Solar Cycle 22. The flare parameters used and the effectiveness of the current warning system, based on Solar Cycle 22 experience, are presented, with an examination of the shortcomings. Needed improvements to the system include more complete observations of solar activity, especially information on the occurrences of solar mass ejections; and consideration of the effects of propagation conditions in the solar corona and interplanetary medium. Requirements for solar observations and forecasting systems on board the spacecraft are discussed.     

太阳活动区磁场扭绞与耀斑产率

本文以1972年10月的太阳活动区McMath 12094为范例, 研究了活动区磁场扭绞与耀斑产率的关系.先在常α无力场模型假定下, 以观测到的活动区光球磁场为边值, 对活动区在日面中心附近4天(10月28—31日), 推算出代表活动区磁场平均扭绞程度的无力因子α, 从而外推出活动区在这4天的三维磁力线形态.然后以这些资料为基础, 进一步讨论了活动区磁场演化特征, 磁场扭绞与耀斑产率的关系, 并且近似用单极场模型估算了通过活动区前导大黑子A的电流、电流密度以及因大黑子逆时针旋转造成磁场扭绞所贮存的能量.本文主要结论为:(1)活动区McMath 12094从10月27日起保持较强扭绞, 10月30日达到极大, 10月31日后扭绞减弱.活动区磁场扭绞的主要原因是光球中的磁流体力学作用所导致的前导大黑子A的逆时针旋转。(2)代表活动区磁场平均扭绞程度的无力因子α与活动区耀斑产率同步变化, 表明活动区磁场扭绞与耀斑产率成正相关.(3)通过活动区前导大黑子A的本影电流为4.3—6.6×1012A, 因扭绞产生的自由能贮存为0.44—1.11×1032erg.活动区中的电流密度达到0.96—1.47×10A·m-2.这样高的电流密度可能是该活动区高耀斑产率的重要原因.      

1986年2月太阳的高活动I活动区4711的演化和特征

本文使用太阳黑子、磁场、Hα色球、10.7cm射电及软X射线流量等观测资料,对太阳活动谷期的高活动区4711(SESC编号)从光球、色球和日冕三个方面做了综述.指出该活动区演化过程的特征是:(1)黑子群在主要发展阶段呈一个紧密的结构复杂的强磁区;(2)两次大的太阳爆发均发生在黑子群面积衰减阶段的初期;(3)黑子群的转动可能是活动区日冕加热和耀斑活动的主要供能机制;(4)色球暗条的频繁活动是爆发的先兆;(5) 10.7cm射电辐射和软X射线辐射的逐日流量有彼此不重合的双峰.