Correlation between solar flare productivity and photospheric vector magnetic fields

Studying the statistical correlation between the solar flare productivity and photospheric magnetic fields is very important and necessary. It is helpful to set up a practical flare forecast model based on magnetic properties and improve the physical understanding of solar flare eruptions. In the previous study ([Cui, Y.M., Li, R., Zhang, L.Y., He, Y.L., Wang, H.N. Correlation between solar flare productivity and photospheric magnetic field properties 1. Maximum horizontal gradient, length of neutral line, number of singular points. Sol. Phys. 237, 45–59, 2006]; from now on we refer to this paper as ‘Paper I’), three measures of the maximum horizontal gradient, the length of the neutral line, and the number of singular points are computed from 23990 SOHO/MDI longitudinal magnetograms. The statistical relationship between the solar flare productivity and these three measures is well fitted with sigmoid functions. In the current work, the three measures of the length of strong-shear neutral line, total unsigned current, and total unsigned current helicity are computed from 1353 vector magnetograms observed at Huairou Solar Observing Station. The relationship between the solar flare productivity and the current three measures can also be well fitted with sigmoid functions. These results are expected to be beneficial to future operational flare forecasting models.     

Sunspot cycle prediction using Warped Gaussian process regression

Solar cycle prediction is a key activity in space weather research. Several techniques have been employed in recent decades in order to try to forecast the next sunspot-cycle maxima and time. In this work, the Gaussian process, a machine-learning technique, is used to make a prediction for the solar cycle 25 based on the annual sunspot number 2.0 data from 1700 to 2018. A variation known as Warped Gaussian process is employed in order to deal with the non-negativity constraint and asymmetrical data distribution. Tests using holdout data yielded a root mean square error of 10.0 within 5 years and 25.0–35.0 within 10 years. Simulations using the predictive distribution were performed to account for the uncertainty in the prediction. Cycle 25 is expected to last from 2019 to 2029, with a peak sunspot number about 117 (110 by the median) occurring most likely in 2024. Thus our method predicts that solar Cycle 25 will be weaker than previous ones, implying a continuing trend of declining solar activity as observed in the past two cycles.     

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.     

Asymmetric precipitation in a coronal loop as explanation of a singular observed spectrum

Almost 10 years of solar submillimeter observations have shown new aspects of solar activity, such as the presence of rapid solar spikes associated with the launch of coronal mass ejections and an increasing submillimeter spectral component in flares. We analyse the singular microwave–submillimeter spectrum of an M class solar flare on 20 December, 2002. Flux density observations measured by Sun patrol telescopes and the Solar Submillimeter Telescope are used to build the radio spectrum, which is fitted using Ramaty’s code. At submillimeter frequencies the spectrum shows a component different from the microwave classical burst. The fitting is achieved proposing two homogeneous sources of emission. This theoretical fitting is in agreement with differential precipitation through a magnetically asymmetric loop or set of loops. From a coronal magnetic field model we infer an asymmetric magnetic structure at the flare location. The model proposed to quantify the differential precipitation rates due to the asymmetry results in a total precipitation ratio Q₂/Q₁≈10⁴–10⁵, where Q₁(Q₂) represents the total precipitation in the loop foot with the high (low) magnetic field intensity. This ratio agrees with the electron total number ratio of the two sources proposed to fit the radio spectrum.     

神经网络方法在太阳质子事件短期预报中的应用

在大量统计结果的基础上,深入研究了太阳质子事件预报机理．总结了质子事件爆发与太阳活动区面积、位置、McIntosh结构、磁结构以及前两天活动区爆发耀斑事件数目之间的关系．然后,在神经网络的基础上建立了太阳质子事件短期预报模型,并对2000年以后12个未参加训练的样本进行测试,结果对事件预报的准确率为83％．此外,我们还利用该模型对2002年1-4月发生的几次质子事件进行了预报试验,结果发现,这期间发生的6次事件都被预报．其中3次质子事件系统预报提前了3天,两次事件预报提前了2天,一次事件提前1天预报．     

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。      

利用质子强度变化对太阳质子事件进行短期预报方法初步研究

采用GOSE-10卫星4～9 MeV(P2),9～15 MeV(P3),15～40 MeV(P4),40～80 MeV(P5)能段上的质子通量数据,结合质子能谱,对太阳质子事件发生前各能谱参数的变化特征进行分析,详细介绍利用能谱参数的变化特征及能量E>10 MeV的质子通量数据对太阳质子事件进行预报的新方法,并运用这种方法对2002-2006年期间太阳质子事件进行了预报.预报结果显示,预报提前量最多达到100 h以上,对质子事件的报准率达97.5%,预报方法具备一定的有效性和实用性.      

Solar cycles: A tutorial

The Sun is the nearest stellar and astrophysical laboratory, available for detailed studies in several fields of physics and astronomy. It is a sphere of hot gas with a complex and highly variable magnetic field which plays a very important role. The Sun shows an unprecedented wealth of phenomena that can be studied extensively and to the greatest detail, in a way we will never be in a position to study in other stars. Humans have studied the Sun for millennia and after the discovery of the telescope they realized that the Sun varies with time, i.e., solar activity is highly variable, in tune scales of millennia to seconds. The study of these variabilities helps us to understand how the Sun works and how it affects the interplanetary medium, Earth and the other planets. Solar power varies substantially and greatly affects the Earth and humans. Solar activity has several important periodicities, and quasi-periodicities. Knowledge of these periodicities helps us to forecast, to an extent, solar events that affect our planet. The most prominent periodicity of solar activity is the one of 11 years. The actual period is in fact 22 years because the magnetic field polarity of the Sun has to be taken into account. The Sun can be considered as a non-linear RLC electric circuit with a period of 22 years. The RLC equivalent circuit of the Sun is a van der Pol oscillator and such a model can explain many solar phenomena, including the variability of solar energy with time. Other quasi-periodicities such as the ones of 154 days, the 1.3, 1.7 to 2 years, etc., some of which might be harmonics of the 22 year cycle are also present in solar activity, and their study is very interesting and important since they affect the Earth and human activities. The period of 27 days related to solar rotation plays also a very important role in geophysical phenomena. It is noticeable that almost all periodicities are highly variable with time as wavelet analysis reveals. It is very important for humans to be in a position to forecast solar activity during the next hour, day, year, decade and century, because solar phenomena affect life on Earth and such predictions will help politicians and policy makers to better serve their countries and our planet.     

太阳光球磁场特征物理量在质子事件短期预报中的应用

利用SOHO/MDI全日面纵向磁图, 计算了三个描述太阳活动区磁场复杂性和非势性的特征物理量, 即纵向磁场最大水平梯度Bz, 强梯度中性线长度L, 孤立奇点数目η. 为检验太阳光球磁场特征在质子事件短期预报中是否有效, 采用BP神经网络方法, 建立了基于这三个磁场特征物理量简单的太阳质子事件短期(24h)预报模型. 模 型在对2002年和2003年连续两年的样本检测中, 有很高的准确率(2002年和2003年 分别为90 %, 87.54 %)和较高的 质子事件报准率(2002年和2003年分别为60 %, 75 %),从而为光球磁场特征物理量作为质子事件预报的有效因子提供了依据.      

Imbalance of magnetic field observed in vector magnetograms

The systematic investigation of the three components of the magnetic field is made on 6629 vector magnetograms obtained with the Solar Magnetic Field Telescope at Huairou Solar Observing Station over 18 years 1988–2005. The sign distribution of these values has been analyzed over the solar hemispheres and the solar activity cycle as follows:     

LSTM神经网络在太阳F10.7射电流量中期预报中的应用

F_10.7指数作为大气密度经验模型的重要输入参量，其预报精度直接影响航天器轨道预报精度.研究发现，太阳活动表现出长时间尺度上平均11年和中短时间尺度平均27天的周期性变化特征.依据这一观测事实，基于长短期记忆单元（Long Short-term Memory，LSTM）递归神经网络方法进行F_10.7指数未来27天的中期预报.利用一个连续长时段F_10.7数据作为训练数据，构建LSTM神经网络训练和预测模型，分别预测太阳活动高低年未来27天的F_10.7指数.结果表明，太阳活动高年的第27天F_10.7指数预报平均相对误差最优可达10%以内，低年最优可达2%以内.      

基于SDO/HMI全日面实时磁图的活动区自动识别

太阳活动区是太阳活动的主要发生源区，活动区的形态、结构、特征是预报太阳爆发的主要依据.因此，活动区的识别是实现太阳爆发预报的前提.SDO/HMI能够提供连续、高时空精度的全日面光球观测图像.参照文献[1]SOHO/MDI综合磁图中活动区的自动识别方法，利用实时可得的HMI全日面磁图，通过阈值法、数学形态法和区域增长法相结合的方式，开展了活动区的快速自动识别研究.将2010—2018年的自动识别结果与NOAA/SWPC每日发布的活动区结果进行比较发现:通过磁图自动识别的活动区数目与SWPC活动区数目整体变化趋势基本一致，两者的相关系数为0.87；从活动区整体标识的数目上来看，通过磁图自动识别的活动区数目少于SWPC发布的结果.未被自动标识的活动区主要为面积小、磁场弱且磁类型简单的活动区，引发太阳爆发的可能性极小，因此不会对太阳爆发的实际预报产生影响.本文的研究方法和结果能够为太阳活动预报提供实时的活动区数据，加速太阳爆发预报模型的实际应用.      

SMY和第21太阳周期间地磁及太阳活动特点的分析

本文比较第17—21太阳周黑子数、地磁A_p指数、各周极大年≥2级耀斑数、磁暴数及第一、二、三大磁暴情况;分析了≥2级耀斑数及磁暴的分布。21周3级耀斑对应磁暴比例低于19、20周,Ⅳ型及米波射电爆发是产生磁暴的重要条件。进一步分析了21周最大磁暴、最大射电爆发引起的磁暴,最严重的电离层短波通讯干扰及有明亮物质抛射的大耀斑、双带大耀斑引起的磁暴等典型例子。最后对SMY期间22个无黑子耀斑作了分析,它们可能引起中小幅度的磁暴。      

A simple model to predict solar radiation under clear sky conditions

Solar radiation is one of the major factors that dominate the thermal behaviors of aerostats in the daytime and the primary energy source of high altitude long endurance aerostats. Therefore, it is necessary to propose an accurate model to predict the solar irradiances. A comprehensive review of the well-known solar radiation models is conducted to help develop the new model. Based on the analysis of the existing models and the available radiation data, the extensive computer tests of the regression and optimization are conducted, from which the new solar radiation model for direct and diffuse irradiances under clear sky conditions is proposed. The new model has excellent prediction accuracy. The coefficient of determination for direct radiation is 0.992, with the root mean square error (RMSE) of 16.9 W/m² and the mean absolute error (MAE) of 2.2%. The coefficient of determination for diffuse radiation is 0.86, with RMSE = 8.7 W/m² and MAE = 9.9%. Comparisons with the well-known existing models show that the new model is much more accurate than the best existing ones.     

联合太阳和行星际物理参数预测行星际激波能否到达地球

选取第23太阳活动周（1997—2006年）期间542例由太阳爆发活动驱动的行星际激波事件,分析确定了太阳源头和行星际空间中影响行星际激波能否到达地球轨道的关键物理参数;在此基础上,建立了预测行星际激波能否到达地球的新预报模型（EdEaSPM）. 回溯预报结果表明,EdEaSPM模型的预报成功率约为66%,略高于国际一流预报模型的预报成功率;EdEaSPM模型的虚报率未超过50%,改善了当前国际主流模型虚报率较大的情况;对于偏度指标,虽然当前所有模型的偏度值均大于1,但EdEaSPM模型的偏度值最接近于1且明显小于其他模型的偏度值;EdEaSPM模型的其他评价指标也都高于国际主流模型的相应指标. 此外,选取2012年期间的激波事件对EdEaSPM模型进行了预报检验,预测结果与实际情况吻合. EdEaSPM模型不仅能够提前约1～3天进行预报,而且预报效果与国际一流模型具有可比性,尤其是在提高预报成功率及降低虚报率方面具有一定优势.      

Comparison of solar soft X-ray irradiance from broadband photometers to a high spectral resolution rocket observation

The solar soft X-ray (XUV; 1–30 nm) radiation is highly variable on all time scales and strongly affects the ionosphere and upper atmosphere of Earth, Mars, as well as the atmospheres and surfaces of other planets and moons in the solar system; consequently, the solar XUV irradiance is important for atmospheric studies and for space weather applications. While there have been several recent measurements of the solar XUV irradiance, detailed understanding of the solar XUV irradiance, especially its variability during flares, has been hampered by the lack of high spectral resolution measurements in this wavelength range. The conversion of the XUV photometer signal into irradiance requires the use of a solar spectral model, but there has not been direct validation of these spectral models for the XUV range. For example, the irradiance algorithm for the XUV Photometer System (XPS) measurements uses multiple CHIANTI spectral models, but validation has been limited to other solar broadband measurements or with comparisons of the atmospheric response to solar variations. A new rocket observation of the solar XUV irradiance with 0.1 nm resolution above 6 nm was obtained on 14 April 2008, and these new results provide a first direct validation of the spectral models used in the XPS data processing. The rocket observation indicates very large differences for the spectral model for many individual emission features, but the differences are significantly smaller at lower resolution, as expected since the spectral models are scaled to match the broadband measurements. While this rocket measurement can help improve a spectral model for quiet Sun conditions, many additional measurements over a wide range of solar activity are needed to fully address the spectral model variations. Such measurements are planned with a similar instrument included on NASA’s Solar Dynamics Observatory (SDO), whose launch is expected in 2009.     

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.      

Interplanetary shocks and sudden impulses during solar maximum (2000) and solar minimum (1995–1996)

In this work a study is performed on the correlation between fast forward interplanetary shock parameters at 1 Astronomical Unit and sudden impulse (SI) amplitudes in the H-component of the geomagnetic field, for periods of solar activity maximum (year 2000) and minimum (year 1995–1996). Solar wind temperature, density and speed, and total magnetic field, were taken to calculate the static pressures (thermal and magnetic) both in the upstream and downstream sides of the shocks. The variations of the solar wind parameters and pressures were then correlated with SI amplitudes. The solar wind speed variations presented good correlations with sudden impulses, with correlation coefficients larger than 0.70 both in solar maximum and solar minimum, whereas the solar wind density presented very low correlation. The parameter better correlated with SI was the square root dynamic pressure variation, showing a larger correlation during solar maximum (r = 0.82) than during solar minimum (r = 0.77). The correlations of SI with square root thermal and magnetic pressure were smaller than with the dynamic pressure, but they also present a good correlation, with r > 0.70 during both solar maximum and minimum. Multiple linear correlation analysis of SI in terms of the three pressure terms have shown that 78% and 85% of the variance in SI during solar maximum and minimum, respectively, are explained by the three pressure variations. Average sudden impulse amplitude was 25 nT during solar maximum and 21 nT during solar minimum, while average square root dynamic pressure variation is 1.20 and 0.86 nPa^1/2 during solar maximum and minimum, respectively. Thus on average, fast forward interplanetary shocks are 33% stronger during solar maximum than during solar minimum, and the magnetospheric SI response has amplitude 20% higher during solar maximum than during solar minimum. A comparison with theoretical predictions (Tsyganenko’s model corrected by Earth’s induced currents) of the coefficient of sudden impulse change with solar wind dynamic pressure variation showed excellent agreement, with values around 17 nT/nPa^1/2.     

基于机器学习相似度算法的Kp指数预报

基于机器学习中的相似度算法,建立了在历史太阳风数据中寻找与当前太阳风特征相近事例的推荐模型,用来预报地磁Kp指数.使用1998-2019年间随机选择的120个太阳风事例作为测试数据集,该模型能够推荐得到历史上与输入太阳风造成相似地磁影响的太阳风事例,最优事例的Kp指数与实际值的均方根误差为0.79,相关系数为0.93....