Optical and SuperDARN radar observations of duskside shock aurora over Zhongshan Station

We present observations of a duskside shock aurora occurred on 21 April 2001 by the SuperDARN radar at Syowa Station and the all-sky camera at Zhongshan Station （ZHS） in Antarctica when the radar was operated in fast-scan mode covering the ZHS region. With the two independent data sets, we examine ionospheric plasma convection and aurora arising from a sudden impulse （SI） event associated with an interplanetary shock. During the transient shock compression, the aurora was quiescent without any optical emission at the preliminary impulse of the SI. About 7 min later, a new thin auroral arc with brighter emissions and a lifetime of -14 rain expanded westward from the region above ZHS during the main impulse of the SI. SuperDARN radar line-of-sight measurements showed periodical oscillation in the flow direction with ultra-low-frequency waves having a period of -8 min during the shock compression. We suggest that downward field-aligned current during the preliminary impulse stage of the SI was the main driver of the first plasma flow reversal, and the subsequent new discrete auroral arc may be associated with field-aligned acceleration in the region of the main impulse related upward field-aligned currents. The ground magnetometer observations suggest that the oscillation of the ionospheric convection on the duskside was associated with field line resonance activity.

深度聚类算法在SuperDARN雷达目标回波分类中的应用

SuperDARN雷达目标回波中通常包含多种类型散射的回波,例如电离层不规则体回波、地面/海面散射回波、极区中层夏季回波以及流星余迹回波等.利用SuperDARN采集的电离层回波制作的电离层对流图对于空间天气研究具有重要意义.SuperDARN接收到的电离层回波通常会与地面海面的散射回波混淆,从而造成绘制的电离层对流图不准确,因此对于SuperDARN目标回波进行聚类分析具有重要意义.本文首次将基于自动编码器网络的图嵌入深度聚类算法应用于SuperDARN目标回波数据,有效地对SuperDARN回波数据进行了分类.此外,还将该模型与传统算法和机器学习聚类算法进行了比较.该模型在样本数据中的应用表明,深度聚类算法能够捕捉到回波数据的深层结构特征,提高了回波聚类的准确性.

Statistical characteristics of ionospheric backscatter observed by SuperDARN Zhongshan radar in Antarctica

Zhongshan HF radar, as one component of SuperDARN, has been established and in operation since April, 2010. Using data from the first two years of its operation, this paper investigates the radar＇s performance, the diurnal and seasonal variations o1 ionospheric echoes, and their dependence on geomagnetic activity. Statistical studies show that the occurrence of echoes in different beams varies at different frequencies, which arises from the direction of the beam and the area over which the beam can achieve the orthogonality condition between the wave vector and the Earth＇s magnetic field. The diurnal variation is obvious with double peak structures both in the occurrence rate and average power at 04-08 UT and 16-17 UT. The line-of-sight velocities are mainly positive on the dayside and negative on the nightside for Beam 0, which is the opposite of the trend for Beam 15. The spec- tral widths on the dayside are often higher than those on the nightside owing to the high energy particle precipitation in the cusp region. The seasonal variations are more obvious for those beams with larger numbers. The occurrence, the average power, the line-of-sight velocity, and the spectral widths are generally larger in the winter months than in the summer months. The influence of geomagnetic activity on radar echoes is significant. The peak echo occurrence appears on the dayside during geomagnetically quiet times, and shifts toward the nightside and exhibits an obvious decrease with increasing Kp. With increasing geomagnetic activity, the line-of-sight velocities increase, whereas the spectral widths decrease. The frequency dependence is investigated and it is found that in the operating frequency bands in 2010, 9-10 MHz is the most appropriate band for the SuperDARN Zhongshan radar.

Northward-propagating nighttime medium-scale traveling ionospheric disturbances observed with SuperDARN Hokkaido HF radar and GEONET

We report on the characteristics of nighttime medium-scale traveling ionospheric disturbances （MSTIDs） propagating northward observed with the SuperDARN Hokkaido HF radar, which has a field of view to the north of Japan, and occasionally with the GNSS Earth Observation NETwork （GEONET）, which provides total electron content （TEC） data over Japan. From statis- tical analysis of MSTIDs observed with the Hokkaido radar during nighttime （1700-0700 LT） from January 2007 to July 2009, we find that these MSTIDs traveling northward, although rare in comparison with those traveling southwestward, have a relatively high occurrence rate after sunset and around midnight in May and August, which is partly consistent with the occurrence rate of MSTIDs over Japan observed with GEONET in 2002, when the MSTID event database is available. We also use the data from simultaneous observation of nightside MSTIDs by the Hokkaido radar and GEONET to find that when the HF radar observed northward-propagating MSTIDs, GEONET did not always observe such MSTIDs with the same propagation direction. Judging from this result and considering the HF radar field of view located to the north of the GEONET coverage area, we speculate that some physical parameters of the ionosphere/thermosphere over Japan differ from those to the north of Japan, which may result in the inconsistency of MSTID propagation direction. The present results provide new knowledge of MSTIDs propagating northward using the Hokkaido radar, whose field of view was not covered by GEONET.

Study of ionospheric disturbances during solar flare events using the SuperDARN Hokkaido radar

It is well known that many types of ionospheric disturbances occur during solar flare events. The sudden increase in total electron content （SITEC） has been studied for several decades, but total electron content （TEC） data do not provide information on the altitudinal distribution of electron density changes. Previous studies used HF Doppler system data to investigate the contributions of the D-region and F-region ionospheric electron density changes by examining the HF radio wave frequency dependence on the Doppler shift values. In this study we examined the dependence of the elevation angle of the Doppler shift of ground scatter echoes using the SuperDARN Hokkaido radar. We analyzed solar flare events from Dec 2006 to Mar 2012. A sudden fade-out of echoes was observed in almost all the events we analyzed, which was the result of the radio absorption associated with a significant increase in electron density within the D-region ionosphere. In addition, we discovered positive Doppler shifts just before the sudden fade-out of echoes. The Doppler shift is negatively correlated with the elevation angle of received radar waves. It indicates that variation of electron density in the D-region ionosphere is dominant during solar flare events. This result is consistent with a previous study. We also compared the irradiation by X-ray and extreme ultra violet rays observed by the GOES-14 and GOES-15 satellites, which generated Doppler shifts. A positive Doppler shift is consistent with a change of X-ray

基于深度学习的SuperDARN雷达极区电离层跨极盖电场模型构建

通过超级双子极光雷达网(SuperDARN)获得的跨极盖电势计算了极区电离层对流电场。利用2014年的极区电离层对流电场数据为基础,引入对流电场的历史数据,分别基于多元线性回归算法和后向传播神经网络算法构建电离层电场模型。利用独立的数据集,验证了两种模型的准确性和稳定性。结果表明,模型值与测量值的均方根误差在2.0~3.5 mV·m^(-1)之间,平均绝对误差范围为1.5~3.0 mV·m^(-1),线性相关系数均大于0.6,最高可达0.9。引入前20分钟的历史数据作为模型的输入,后向传播神经网络模型比多元线性回归模型具有更好的预测性能。

极区电离层闪烁指数和对流速度关系的建模研究

电离层对流速度会显著影响电离层闪烁的发生,尤其是在具有较大对流速度的高纬极区,且其对相位闪烁指数和幅度闪烁指数影响各不相同.本文基于弱散射条件下电离层闪烁理论,初步构建了极区电离层闪烁指数与对流速度的显式关系模型,模型结果表明当相位谱指数取定值3时,相位闪烁指数和幅度闪烁指数的比值与对流速度存在线性关系;进而,利用加拿大Hall Beach台站的观测数据,验证了由电离层闪烁指数比值推断电离层对流速度这一方法在极区的适用性.这不仅提升我们对电离层闪烁机理的认识,也提供了一种获取极区电离层对流速度的新途径,将有助于改善和提高我国北斗全球导航系统的服务精度.