In this paper,we propose a long short-term memory(LSTM)deep learning model to deal with the smoothed monthly sunspot number(SSN),aiming to address the problem whereby the prediction results of the existing sunspot pre...In this paper,we propose a long short-term memory(LSTM)deep learning model to deal with the smoothed monthly sunspot number(SSN),aiming to address the problem whereby the prediction results of the existing sunspot prediction methods are not uniform and have large deviations.Our method optimizes the number of hidden nodes and batch sizes of the LSTM network structures to 19 and 20,respectively.The best length of time series and the value of the timesteps were then determined for the network training,and one-step and multi-step predictions for Cycle 22 to Cycle 24 were made using the well-established network.The results showed that the maximum root-mean-square error(RMSE)of the one-step prediction model was6.12 and the minimum was only 2.45.The maximum amplitude prediction error of the multi-step prediction was 17.2%and the minimum was only 3.0%.Finally,the next solar cycles(Cycle 25)peak amplitude was predicted to occur around 2023,with a peak value of about 114.3.The accuracy of this prediction method is better than that of the other commonly used methods,and the method has high applicability.展开更多
Historical earthquakes registered in Chile (from 1900 up to 2015) with epicenters located between 17?30'S and 56?0'S latitude and yearly mean total sunspot number have been considered in order to evaluate a si...Historical earthquakes registered in Chile (from 1900 up to 2015) with epicenters located between 17?30'S and 56?0'S latitude and yearly mean total sunspot number have been considered in order to evaluate a significant linkage between them. The occurrence of strong earthquakes along Chile and the sunspots activity are analyzed to inspect possible influence of solar cycles on earthquakes. The cross wavelet transform and wavelet coherence analysis were applied for sequences of sunspots and earthquakes activity. An 8 - 12 years modulation of earthquakes activity has been identified.展开更多
Solar activity refers to any natural phenomenon occurring on the sun such as sunspots, solar flare and coronal mass ejection etc. Such phenomena have their roots deep inside the sun, where the dynamo mechanism operate...Solar activity refers to any natural phenomenon occurring on the sun such as sunspots, solar flare and coronal mass ejection etc. Such phenomena have their roots deep inside the sun, where the dynamo mechanism operates and fluid motions occur in a turbulent way. It is mainly driven by the variability of the sun’s magnetic field. The present paper studies the relation between various solar features during January 2009 to December 2011. A good correlation between various parameters indicates similar origin.展开更多
The prediction for the smoothed monthly mean sunspot numbers (hereafter SMSNs) of solar cycle 23, which was given with a similar cycle method proposed by us at the beginning time of cycle 23, is analyzed and verified ...The prediction for the smoothed monthly mean sunspot numbers (hereafter SMSNs) of solar cycle 23, which was given with a similar cycle method proposed by us at the beginning time of cycle 23, is analyzed and verified in this paper. Using our predicted maximum SMSN and the ascending length for solar cycle 24, and as- suming their relative errors to be respectively 20% and ± 7 months, solar cycles 2, 4, 8, 11, 17, 20 and 23 are selected to be the similar cycles to cycle 24. The selected solar cycles are divided into two groups. The first group consists of all the selected cycles; while the second group consists of only cycles 11, 17, 20 and 23. Two SMSN time profiles then may be obtained, respectively, for the two similar cycle groups. No significant difference is found between the two predicted time profiles. Consid- ering the latest observed sunspot number so far available for cycle 23 and the pre- dictions for the minimum SMSN of cycle 24, a date calibration is done for the ob- tained time profiles, and thus, SMSNs for 127 months of cycle 24, from October 2007 to April 2018, are predicted.展开更多
It is a significant task to predict the solar activity for space weather and solar physics. All kinds of approaches have been used to forecast solar activities, and they have been applied to many areas such as the sol...It is a significant task to predict the solar activity for space weather and solar physics. All kinds of approaches have been used to forecast solar activities, and they have been applied to many areas such as the solar dynamo of simulation and space mission planning. In this paper, we employ the long-shortterm memory(LSTM) and neural network autoregression(NNAR) deep learning methods to predict the upcoming 25 th solar cycle using the sunspot area(SSA) data during the period of May 1874 to December2020. Our results show that the 25 th solar cycle will be 55% stronger than Solar Cycle 24 with a maximum sunspot area of 3115±401 and the cycle reaching its peak in October 2022 by using the LSTM method. It also shows that deep learning algorithms perform better than the other commonly used methods and have high application value.展开更多
There is increasing interest in the relation between the solar activity and climate change. Regarding the solar activity, the fractal property of the sunspot number (SSN) has been studied by many previous works. In ge...There is increasing interest in the relation between the solar activity and climate change. Regarding the solar activity, the fractal property of the sunspot number (SSN) has been studied by many previous works. In general, fractal properties have been observed in the time series of the dynamics of complex systems. The purpose of this research is to investigate the relationship between the solar activity, total ozone, and the North Atlantic Oscillation (NAO) from a viewpoint of multi-fractality. To detect the changes of multifractality, we performed the wavelets analysis, and plotted the τ-function derived from the wavelets of these time series. We showed that the solar activity relate to the NAO, by observing the matching in monofractality or multifractality of these indices. When the SSN increased and the solar activity was stable, the NAO also became stable. When the SSN became maximum, the fractality of the SSN, F10.7 flux, geomagnetic aa, and NAO indices changed from multifractality to monofractality and those states became stable for most of the solar cycles. When the SSN became maximum, the fluctuations became large and multifractality became strong, and a change from multifractal to monofractal behavior was observed in the SSN, F10.7 flux, geomagnetic aa, and NAO indices. The strong interactions of the solar flux, geomagnetic activity, total ozone, and NAO occur in the SSN maximum. The strong interactions were inferred from the similarity of fractality changes and the wavelet coherence. The influence of the solar activity on the NAO was shown from a viewpoint of multi-fractality. These findings will contribute to the research on the effects of the solar activity on climate change.展开更多
We study the relation between monthly average counting rates of the cosmic ray intensity (CRI) observed at Moscow Neutron Monitoring Station, solar flare index (SFI) and coronal index during the solar cycles 22 and 23...We study the relation between monthly average counting rates of the cosmic ray intensity (CRI) observed at Moscow Neutron Monitoring Station, solar flare index (SFI) and coronal index during the solar cycles 22 and 23, for the period 1986-2008. The long-term behaviour of various solar activity parameters: sunspot numbers (SSN), solar flare index (Hα flare index), coronal index (CI) in relation to the duration of solar cycles 22 and 23 is examined. We find that the correlation coefficient of CRI with the coronal index as well as Hα flare index is relatively large anti-correlation during solar cycle 22. However, the monthly mean values of sunspot number, Hα flare index, and coronal index are well positively correlated with each other. We have analyzed the statistical analysis of the above parameters using of linear model and second order polynomial fits model.展开更多
Sunspot number, sunspot area and sunspot unit area are usually used to show sunspot activity. In this paper, periodicity of sunspot activity of modern solar cycles has been investigated through analyzing the monthly m...Sunspot number, sunspot area and sunspot unit area are usually used to show sunspot activity. In this paper, periodicity of sunspot activity of modern solar cycles has been investigated through analyzing the monthly mean val- ues of the three indices in the time interval of May 1874 to May 2004 by use of the wavelet transform. Their global power spectra and local power spectra are given while the statistical tests of these spectra are taken into account. The main results are (1) the local wavelet power spectrum of the sunspot number seems like that of the sunspot area, indicat- ing that the periodicity of the both indices is similar. The local power spectrum of the sunspot unit area resembles the local power spectra of the previous two indices, but looks more complicated. (2) the possible periods in sunspot activity are about 10.6 (or 10.9 years for the sunspot unit area), 31, and 42 years, and the period of about 10.6 years is statisti- cally significant in the considered time. For the periods of about 31 and 42 years, their power peaks are under the 95% confidence level line but over the mean red-noise spectral line, and for the other rest periods, their power peaks are even under the mean red-noise spectral line, which are sta- tistically insignificant. (3) the local power of the three periods is higher in the late stage than in the early stage of the con- sidered time. (4) the period characteristics of the three indi- ces, shown in the global power spectra and the local power spectra, are similar but there is difference in detail.展开更多
基金the National Natural Science Foundation of China(Grant No.U1531128)。
文摘In this paper,we propose a long short-term memory(LSTM)deep learning model to deal with the smoothed monthly sunspot number(SSN),aiming to address the problem whereby the prediction results of the existing sunspot prediction methods are not uniform and have large deviations.Our method optimizes the number of hidden nodes and batch sizes of the LSTM network structures to 19 and 20,respectively.The best length of time series and the value of the timesteps were then determined for the network training,and one-step and multi-step predictions for Cycle 22 to Cycle 24 were made using the well-established network.The results showed that the maximum root-mean-square error(RMSE)of the one-step prediction model was6.12 and the minimum was only 2.45.The maximum amplitude prediction error of the multi-step prediction was 17.2%and the minimum was only 3.0%.Finally,the next solar cycles(Cycle 25)peak amplitude was predicted to occur around 2023,with a peak value of about 114.3.The accuracy of this prediction method is better than that of the other commonly used methods,and the method has high applicability.
文摘Historical earthquakes registered in Chile (from 1900 up to 2015) with epicenters located between 17?30'S and 56?0'S latitude and yearly mean total sunspot number have been considered in order to evaluate a significant linkage between them. The occurrence of strong earthquakes along Chile and the sunspots activity are analyzed to inspect possible influence of solar cycles on earthquakes. The cross wavelet transform and wavelet coherence analysis were applied for sequences of sunspots and earthquakes activity. An 8 - 12 years modulation of earthquakes activity has been identified.
文摘Solar activity refers to any natural phenomenon occurring on the sun such as sunspots, solar flare and coronal mass ejection etc. Such phenomena have their roots deep inside the sun, where the dynamo mechanism operates and fluid motions occur in a turbulent way. It is mainly driven by the variability of the sun’s magnetic field. The present paper studies the relation between various solar features during January 2009 to December 2011. A good correlation between various parameters indicates similar origin.
基金the National Natural Science Foundation of China (Grant Nos. 10673017 and 10733020) the National Basic Research Program of China (Grant No. 2006CB806307)
文摘The prediction for the smoothed monthly mean sunspot numbers (hereafter SMSNs) of solar cycle 23, which was given with a similar cycle method proposed by us at the beginning time of cycle 23, is analyzed and verified in this paper. Using our predicted maximum SMSN and the ascending length for solar cycle 24, and as- suming their relative errors to be respectively 20% and ± 7 months, solar cycles 2, 4, 8, 11, 17, 20 and 23 are selected to be the similar cycles to cycle 24. The selected solar cycles are divided into two groups. The first group consists of all the selected cycles; while the second group consists of only cycles 11, 17, 20 and 23. Two SMSN time profiles then may be obtained, respectively, for the two similar cycle groups. No significant difference is found between the two predicted time profiles. Consid- ering the latest observed sunspot number so far available for cycle 23 and the pre- dictions for the minimum SMSN of cycle 24, a date calibration is done for the ob- tained time profiles, and thus, SMSNs for 127 months of cycle 24, from October 2007 to April 2018, are predicted.
基金supported by the National Natural Science Foundation of China under Grant numbers U2031202,U1731124 and U1531247the special foundation work of the Ministry of Science and Technology of the People’s Republic of China under Grant number 2014FY120300the 13th Five-year Informatization Plan of Chinese Academy of Sciences under Grant number XXH13505-04。
文摘It is a significant task to predict the solar activity for space weather and solar physics. All kinds of approaches have been used to forecast solar activities, and they have been applied to many areas such as the solar dynamo of simulation and space mission planning. In this paper, we employ the long-shortterm memory(LSTM) and neural network autoregression(NNAR) deep learning methods to predict the upcoming 25 th solar cycle using the sunspot area(SSA) data during the period of May 1874 to December2020. Our results show that the 25 th solar cycle will be 55% stronger than Solar Cycle 24 with a maximum sunspot area of 3115±401 and the cycle reaching its peak in October 2022 by using the LSTM method. It also shows that deep learning algorithms perform better than the other commonly used methods and have high application value.
文摘There is increasing interest in the relation between the solar activity and climate change. Regarding the solar activity, the fractal property of the sunspot number (SSN) has been studied by many previous works. In general, fractal properties have been observed in the time series of the dynamics of complex systems. The purpose of this research is to investigate the relationship between the solar activity, total ozone, and the North Atlantic Oscillation (NAO) from a viewpoint of multi-fractality. To detect the changes of multifractality, we performed the wavelets analysis, and plotted the τ-function derived from the wavelets of these time series. We showed that the solar activity relate to the NAO, by observing the matching in monofractality or multifractality of these indices. When the SSN increased and the solar activity was stable, the NAO also became stable. When the SSN became maximum, the fractality of the SSN, F10.7 flux, geomagnetic aa, and NAO indices changed from multifractality to monofractality and those states became stable for most of the solar cycles. When the SSN became maximum, the fluctuations became large and multifractality became strong, and a change from multifractal to monofractal behavior was observed in the SSN, F10.7 flux, geomagnetic aa, and NAO indices. The strong interactions of the solar flux, geomagnetic activity, total ozone, and NAO occur in the SSN maximum. The strong interactions were inferred from the similarity of fractality changes and the wavelet coherence. The influence of the solar activity on the NAO was shown from a viewpoint of multi-fractality. These findings will contribute to the research on the effects of the solar activity on climate change.
文摘We study the relation between monthly average counting rates of the cosmic ray intensity (CRI) observed at Moscow Neutron Monitoring Station, solar flare index (SFI) and coronal index during the solar cycles 22 and 23, for the period 1986-2008. The long-term behaviour of various solar activity parameters: sunspot numbers (SSN), solar flare index (Hα flare index), coronal index (CI) in relation to the duration of solar cycles 22 and 23 is examined. We find that the correlation coefficient of CRI with the coronal index as well as Hα flare index is relatively large anti-correlation during solar cycle 22. However, the monthly mean values of sunspot number, Hα flare index, and coronal index are well positively correlated with each other. We have analyzed the statistical analysis of the above parameters using of linear model and second order polynomial fits model.
文摘Sunspot number, sunspot area and sunspot unit area are usually used to show sunspot activity. In this paper, periodicity of sunspot activity of modern solar cycles has been investigated through analyzing the monthly mean val- ues of the three indices in the time interval of May 1874 to May 2004 by use of the wavelet transform. Their global power spectra and local power spectra are given while the statistical tests of these spectra are taken into account. The main results are (1) the local wavelet power spectrum of the sunspot number seems like that of the sunspot area, indicat- ing that the periodicity of the both indices is similar. The local power spectrum of the sunspot unit area resembles the local power spectra of the previous two indices, but looks more complicated. (2) the possible periods in sunspot activity are about 10.6 (or 10.9 years for the sunspot unit area), 31, and 42 years, and the period of about 10.6 years is statisti- cally significant in the considered time. For the periods of about 31 and 42 years, their power peaks are under the 95% confidence level line but over the mean red-noise spectral line, and for the other rest periods, their power peaks are even under the mean red-noise spectral line, which are sta- tistically insignificant. (3) the local power of the three periods is higher in the late stage than in the early stage of the con- sidered time. (4) the period characteristics of the three indi- ces, shown in the global power spectra and the local power spectra, are similar but there is difference in detail.
