The magnetic field of the umbrae is sometimes found to be saturated in the magnetograms taken by the Michelson Doppler Imager （MDI） onboard the Solar and Heliospheric Observatory （SOHO）. It is suggested that the c...The magnetic field of the umbrae is sometimes found to be saturated in the magnetograms taken by the Michelson Doppler Imager （MDI） onboard the Solar and Heliospheric Observatory （SOHO）. It is suggested that the combination of the low intensity of sunspot umbrae and the limitation of the 15-bit onboard numerical data acquisition leads to this saturation. In this paper, we propose to use the MDI＇s intensity data to correct this saturation. This method is based on the well-established relationship between the continuum intensity and the magnetic field （the so-called I-B relationship）. A comparison between the corrected magnetic field and the data taken by the Stokes-Polarimeter of the Solar Optical Telescope （SOT/SP） onboard Hinode shows a reasonable agreement, suggesting that this correction is effective.展开更多
为了全面了解太阳活动规律,需要连续观测时长覆盖多个太阳活动周的、高质量数据.太阳磁图是研究太阳活动的重要数据,连续、长时间和高空间分辨率的太阳磁图能够提供更精细的太阳磁场演化信息,有助于更准确地预报太阳活动和空间天气事件...为了全面了解太阳活动规律,需要连续观测时长覆盖多个太阳活动周的、高质量数据.太阳磁图是研究太阳活动的重要数据,连续、长时间和高空间分辨率的太阳磁图能够提供更精细的太阳磁场演化信息,有助于更准确地预报太阳活动和空间天气事件.因此,提出一种基于深度学习的超分率算法,对MDI(Michelson Doppler Imager)磁图进行超分,取得与日震和磁成像仪(Helioseismic and Magnetic Imager,HMI)磁图一致的分辨率,从而能够获得持续时长将近两个太阳活动周的高质量太阳磁图数据库.为了引导网络学习磁图中有效的特征信息,将注意力机制引入到网络中,学习注意力权重图.此外,采用了不确定性损失作为模型训练的损失函数,该方法能够对于带有磁场变化的纹理和边缘分配更大的权重,同时不增加网络参数和计算量.实验证明,提出的算法显著提高了超分太阳磁图的质量,在峰值信噪比(Peak Signal-to-Noise Ratio,PSNR:33.3168)、结构相似性(Structure Similarity Index Measure,SSIM:0.8754)、相关性(Correlation Coefficient,CC:0.9323)和均方根误差(Root Mean Square Error,RMSE:21.8544)等指标上取得了最优的结果.展开更多
Reviewing briefly the recent progress in a joint program of specifying the polar ionosphere primarily on the basis of ground magnetometer data, this paper em-phasizes the importance of processing data from around the ...Reviewing briefly the recent progress in a joint program of specifying the polar ionosphere primarily on the basis of ground magnetometer data, this paper em-phasizes the importance of processing data from around the world in real time for space weather predictions. The output parameters from the program include ionospheric electric fields and currents and field-aligned currents. These real-time records are essential for running computer simulations under realistic boundary conditions and thus for making numerical predictions of space weather efficient as reliable as possible. Data from individual ground magnetometers as well as from the solar wind are collected and are used as input for the KRM and AMIE mag-netogram-inversion algorithms, through which the two-dimensional distribution of the ionospheric parameters is calculated. One of the goals of the program is to specify the solar-terrestrial environment in terms of ionospheric processes and to provide the scientific community with more than what geomagnetic activity indices and statistical models indicate.展开更多
文摘The magnetic field of the umbrae is sometimes found to be saturated in the magnetograms taken by the Michelson Doppler Imager （MDI） onboard the Solar and Heliospheric Observatory （SOHO）. It is suggested that the combination of the low intensity of sunspot umbrae and the limitation of the 15-bit onboard numerical data acquisition leads to this saturation. In this paper, we propose to use the MDI＇s intensity data to correct this saturation. This method is based on the well-established relationship between the continuum intensity and the magnetic field （the so-called I-B relationship）. A comparison between the corrected magnetic field and the data taken by the Stokes-Polarimeter of the Solar Optical Telescope （SOT/SP） onboard Hinode shows a reasonable agreement, suggesting that this correction is effective.
文摘为了全面了解太阳活动规律,需要连续观测时长覆盖多个太阳活动周的、高质量数据.太阳磁图是研究太阳活动的重要数据,连续、长时间和高空间分辨率的太阳磁图能够提供更精细的太阳磁场演化信息,有助于更准确地预报太阳活动和空间天气事件.因此,提出一种基于深度学习的超分率算法,对MDI(Michelson Doppler Imager)磁图进行超分,取得与日震和磁成像仪(Helioseismic and Magnetic Imager,HMI)磁图一致的分辨率,从而能够获得持续时长将近两个太阳活动周的高质量太阳磁图数据库.为了引导网络学习磁图中有效的特征信息,将注意力机制引入到网络中,学习注意力权重图.此外,采用了不确定性损失作为模型训练的损失函数,该方法能够对于带有磁场变化的纹理和边缘分配更大的权重,同时不增加网络参数和计算量.实验证明,提出的算法显著提高了超分太阳磁图的质量,在峰值信噪比(Peak Signal-to-Noise Ratio,PSNR:33.3168)、结构相似性(Structure Similarity Index Measure,SSIM:0.8754)、相关性(Correlation Coefficient,CC:0.9323)和均方根误差(Root Mean Square Error,RMSE:21.8544)等指标上取得了最优的结果.
文摘Reviewing briefly the recent progress in a joint program of specifying the polar ionosphere primarily on the basis of ground magnetometer data, this paper em-phasizes the importance of processing data from around the world in real time for space weather predictions. The output parameters from the program include ionospheric electric fields and currents and field-aligned currents. These real-time records are essential for running computer simulations under realistic boundary conditions and thus for making numerical predictions of space weather efficient as reliable as possible. Data from individual ground magnetometers as well as from the solar wind are collected and are used as input for the KRM and AMIE mag-netogram-inversion algorithms, through which the two-dimensional distribution of the ionospheric parameters is calculated. One of the goals of the program is to specify the solar-terrestrial environment in terms of ionospheric processes and to provide the scientific community with more than what geomagnetic activity indices and statistical models indicate.
