基于CEEMD-SSWT对太阳南北半球Rieger-type周期的不对称性分析

CEEMD(Complete Ensemble Empirical Mode Decomposition)是近年提出的一种自适应滤波算法,SSWT(Synchrosqueezed Wavelet Transform)是目前的一种高分辨率时频分析算法.采用CEEMD对太阳黑子日面积数据进行预滤波处理,再利用SSWT进行太阳南北半球Rieger-type周期不对称性分析,实现对太阳周期信号更精确的提取.研究结果表明:(1)虽然整个北半球的活动强度要高于南半球,但其整个13个活动周的Rieger-type周期信号的平均周期尺度却是极其相近的;(2)南北半球的太阳活动强度明显不对称,但在大部分活动周中,在太阳活动的极大期也是南北半球太阳活动不对称性最弱的时期;(3)太阳活动的强度与太阳的Rieger-type周期信号平均周期没有明显的相关关系,并不是在每个活动周中太阳活动强度越高的半球,所对应的半球Rieger-type周期信号的平均周期就一定越小,活动强度越低所对应的半球Rieger-type周期信号的平均周期就一定越大;(4)Rieger-type周期通常出现在太阳活动的极大期附近,但在某些活动周中Rieger-type周期会呈现多次波动,也会出现在太阳活动的上升期或下降期.

基于深度学习的太阳F_(10.7)辐射通量的短期预报研究

F_(10.7)太阳辐射通量作为输入参数被广泛运用于大气经验模型、电离层模型等空间环境模型,其预报精度直接影响航天器轨道预报精度.采用时间序列法统计了太阳辐射通量F_(10.7)指数和太阳黑子数(SSN)的关系,给出了两者之间的线性关系,在此基础上提出了一种基于长短时记忆神经网络(Long and Short Term Memory,LSTM)的预报方法,方法结合了54 d太阳辐射通量指数和SSN历史数据来对F_(10.7)进行未来7 d短期预报,并与其他预报方法的预报结果进行了比较,结果表明:(1)所建短期预报7 d方法模型的性能优于美国空间天气预报中心(Space Weather Prediction Center,SWPC)的方法,预测值和观测值的相关系数(CC)达到0.96,同时其均方根误差约为11.62个太阳辐射通量单位(sfu),预报结果的均方根误差(RMSE)低于SWPC,下降约11%;(2)对预测的23、24周太阳活动年结果统计表明,太阳活动高年的第7 d F_(10.7)指数预报平均绝对百分比误差(MAPE)最优可达12.9%以内,低年最优可达2.5%以内;(3)联合SSN的LSTM结果和仅使用单变量F_(10.7)的LSTM结果对比显示,新引入的SSN在改进LSTM预测方面是有效的,并且这两个模型的RMSE较SWPC分别低约11%和5%.

太阳总辐照的演化特征分析

太阳总辐照在23和24太阳活动周的显著周期分别为35 d和26 d,进而推断太阳的准旋转周期在23和24太阳活动周也分别为35 d和26 d太阳总辐照在24周极小期的值可能与蒙德极小期的值相近.在一个太阳旋转周到几个月的时间尺度上,太阳黑子是引起太阳总辐照变化的主要原因,但不是唯一的原因;在几天到一个太阳旋转周的时间尺度上,太阳总辐照的变化与MgⅡ特征指数是不相关的.

Relationship between air temperature oscillations and solar variability on short and medium time scales

The contemporary science of climate change is increasingly focusing on the temporal and spatial characteristics of temperature oscillations and determining possible underlying causes.In particular,the effect of variations in solar irradiance on the variability of the climate remains a hot topic of debate.Most studies focus on the effects of solar variation on the Earth's climate on long time scales.This study presents the responses of regional climates to solar variations on shorter time scales using two datasets:one for the air temperature in Nanjing and the Greenwich sunspot number,and the other for the air temperature in Shijiazhuang and the United States sunspot number.Employing empirical mode decomposition,both the 11-year quasi-period of the sunspot number and similar periods including approximately 5.5-and 10.5-year cycles of the air temperature in Nanjing and Shijiazhuang are obtained.However,correlation analysis of similar periodic components for the sunspot number and air temperature indicates that changes in the air temperature on short and medium time scales are not linked to solar variations.This is further confirmed by a test of whether a mode component is a stochastic noise signal.Many shorter periods are also found at the 95% confidence level;in particular,the 3.1-year period of the Nanjing air temperature coincides with a previously obtained empirical result.Moreover,no temperature variations on shorter time scales correlate with solar variability.

Phase asynchrony between flare index and sunspot activity

In this paper, the relative phase relationship between flare index and sunspot activity （sunspot numbers and sunspot areas） is investigated. It is found that （i） the flare index and sunspot activity are asynchronous in phase space at all period scales, and the former lags behind the latter, which implies our results are supported for the integral response model; （ii） their different definitions and physical meanings may be a major reason for their phase asynchrony between them, and the solar flare activity favor to be related to the magnetic complex rather than magnetic strength.

Cases and statistical study on Hot Flow Anomalies with Cluster spacecraft data

Hot Flow Anomalies (HFAs) are phenomena that frequently appear in the vicinity of the Earth's bow shock. We have identified 765 HFA events with Cluster spacecraft data from 2003 to 2009. We study the plasma and magnetic field variations during typical HFAs. Then we study the average structure of HFAs using the superposed epoch method during a 200 s time interval, with the HFA onset time as the epoch time. The results show that HFAs can be classified into four classes based on variations of the dynamic pressure over time, namely "-+" (down-up), "+-" (up-down), "M" (up-down-up) and "W" (up-down-up-down-up), where the letters represent similar shapes with the variation trends of the dynamic pressure. Trends of other parameters are highly related to those of the dynamic pressure with obvious characteristics of the classification. Moreover, statistical results suggest that the number of HFA events varies in years. Compared with the speed of solar wind and sunspot number, the number of HFA events in each year has positive correlation with the former, while it has little relation with the latter. The result of this paper will provide data base for further studies on the mechanisms of the formation, the structural evolution and other relative questions of HFAs.

Is a higher correlation necessary for a more accurate prediction?

A higher correlation tends to yield a more accurate prediction,so that a correlation as high as possible has been searched for and employed in the prediction of solar activity.Instead of using geomagnetic activity during the descending phase of the solar cycle,the minimum annual aa index (aa min) is used as an indicator for the ensuing maximum amplitude (R m) of the sunspot cycle.A four-cycle periodicity is roughly shown in the correlation between R m and aa min.The widely accepted Ohl's precursor prediction method often fails due to the prediction error relative to its estimated uncertainty.An accurate prediction depends on the positive variation of the correlation rather than a higher correlation.Previous experiences by using this method indicate that a prediction for the next cycle,R m (24)=80 ± 17,is likely to fail,implying that the sunspot maximum of Cycle 24 may be either smaller than 63 or greater than 97.

Starspots, stellar cycles and stellar flares： Lessons from solar dynamo models

In this review, we discuss whether the present solar dynamo models can be extrapolated to explain various aspects of stellar activity. We begin with a summary of the following kinds of data for solar-like stars:(i) data pertaining to stellar cycles from Ca H/K emission over many years;(ii) X-ray data indicating hot coronal activity;(iii) starspot data(especially about giant polar spots); and(iv) data pertaining to stellar superflares. Then we describe the current status of solar dynamo modelling—giving an introduction to the flux transport dynamo model, the currently favoured model for the solar cycle. While an extrapolation of this model to solar-like stars can explain some aspects of observational data, some other aspects of the data still remain to be theoretically explained. It is not clear right now whether we need a different kind of dynamo mechanism for stars having giant starspots or producing very strong superflares.