scholarly journals Forecasting auroral electrojet activity from solar wind input with neural networks

1999 ◽  
Vol 26 (10) ◽  
pp. 1353-1356 ◽  
Author(s):  
R. S. Weigel ◽  
W. Horton ◽  
T. Tajima ◽  
T. Detman
Keyword(s):  
Neural Networks ◽  
Solar Wind ◽  
Auroral Electrojet ◽  
Wind Input

Prediction of the Dst Geomagnetic Index Using Adaptive Methods

2021 ◽  
Vol 3 ◽  
pp. 38-46
Author(s):  
I. N. Myagkova ◽  
◽  
V. R. Shirokii ◽  
R. D. Vladimirov ◽  
O. G. Barinov ◽  
...  
Keyword(s):  
Magnetic Field ◽  
Neural Networks ◽  
Time Series ◽  
Solar Wind ◽  
Artificial Neural Networks ◽  
Adaptive Methods ◽  
Gradient Boosting ◽  
Multilayer Perceptrons ◽  
Lagrange Point ◽  
Geomagnetic Index

The potential is investigated of predicting the time series of the Dst geomagnetic index using various adaptive methods: artificial neural networks (classical multilayer perceptrons), decision trees (random forest), gradient boosting. The prediction is based on the parameters of the solar wind and interplanetary magnetic field measured at the Lagrange point L1 in the ACE spacecraft experiment. It is shown that the best prediction skill of the three adaptive methods is demonstrated by gradient boosting.

scholarly journals Cross-correlation and cross-wavelet analyses of the solar wind IMF <i>B</i><sub><i>z</i></sub> and auroral electrojet index AE coupling during HILDCAAs

2018 ◽  
Vol 36 (1) ◽  
pp. 205-211 ◽  
Author(s):  
Adriane Marques de Souza ◽  
Ezequiel Echer ◽  
Mauricio José Alves Bolzan ◽  
Rajkumar Hajra
Keyword(s):  
Solar Wind ◽  
High Speed ◽  
Cross Correlation ◽  
Auroral Electrojet ◽  
Long Duration ◽  
Ae Index ◽  
Speed Solar Wind ◽  
Wavelet Analyses ◽  
Cross Wavelet ◽  
High Speed Solar Wind

Abstract. Solar-wind–geomagnetic activity coupling during high-intensity long-duration continuous AE (auroral electrojet) activities (HILDCAAs) is investigated in this work. The 1 min AE index and the interplanetary magnetic field (IMF) Bz component in the geocentric solar magnetospheric (GSM) coordinate system were used in this study. We have considered HILDCAA events occurring between 1995 and 2011. Cross-wavelet and cross-correlation analyses results show that the coupling between the solar wind and the magnetosphere during HILDCAAs occurs mainly in the period ≤ 8 h. These periods are similar to the periods observed in the interplanetary Alfvén waves embedded in the high-speed solar wind streams (HSSs). This result is consistent with the fact that most of the HILDCAA events under present study are related to HSSs. Furthermore, the classical correlation analysis indicates that the correlation between IMF Bz and AE may be classified as moderate (0.4–0.7) and that more than 80 % of the HILDCAAs exhibit a lag of 20–30 min between IMF Bz and AE. This result corroborates with Tsurutani et al. (1990) where the lag was found to be close to 20–25 min. These results enable us to conclude that the main mechanism for solar-wind–magnetosphere coupling during HILDCAAs is the magnetic reconnection between the fluctuating, negative component of IMF Bz and Earth's magnetopause fields at periods lower than 8 h and with a lag of about 20–30 min. Keywords. Magnetospheric physics (solar-wind–magnetosphere interactions)

scholarly journals Comparative study of dynamical critical scaling in the auroral electrojet index versus solar wind fluctuations

2001 ◽  
Vol 28 (19) ◽  
pp. 3809-3812 ◽  
Author(s):  
Vadim M. Uritsky ◽  
Alex J. Klimas ◽  
Dimitris Vassiliadis
Keyword(s):  
Solar Wind ◽  
Comparative Study ◽  
Auroral Electrojet ◽  
Critical Scaling

scholarly journals Effects of solar wind density on auroral electrojets and brightness under influence of substorms

2009 ◽  
Vol 27 (1) ◽  
pp. 113-119 ◽  
Author(s):  
J.-H. Shue ◽  
Y. Kamide ◽  
J. W. Gjerloev
Keyword(s):  
Solar Wind ◽  
Auroral Oval ◽  
Auroral Electrojet ◽  
The Other ◽  
Auroral Ionosphere ◽  
Stored Energy ◽  
Solar Wind Density ◽  
The Earth ◽  
Density Enhancement ◽  
Solar Wind Parameters

Abstract. Using the auroral electrojet indices and Polar Ultraviolet Imager auroral images, we examined two fortuitous events during which the solar wind density had clear enhancements while the other solar wind parameters were relatively constant. Two electrojet enhancements were found in each event. The first electrojet enhancement was likely to be related to a substorm in which an auroral bulge appeared at premidnight. The second electrojet enhancement was driven by the density enhancement in the solar wind. The auroral oval became wider in latitude and the auroral distribution became dispersed after the density enhancement arrived at the Earth. The total auroral power integrated over the entire nightside region from 50 to 80° MLAT, however, did not increase significantly in response to the density enhancement. Our interpretation is that the substorm that occurred prior to the solar wind density enhancement had drained out a significant portion of the stored energy in the magnetotail; therefore, less precipitation energy was deposited into the auroral ionosphere by the density enhancement.

scholarly journals Predicting geomagnetic storms from solar-wind data using time-delay neural networks

1996 ◽  
Vol 14 (7) ◽  
pp. 679-686 ◽  
Author(s):  
H. Gleisner ◽  
H. Lundstedt ◽  
P. Wintoft
Keyword(s):  
Neural Networks ◽  
Solar Wind ◽  
Time Delay ◽  
Input Data ◽  
Ring Current ◽  
Geomagnetic Storms ◽  
Time History ◽  
Recovery Phase ◽  
Wind Data ◽  
Data Sequence

Abstract. We have used time-delay feed-forward neural networks to compute the geomagnetic-activity index Dst one hour ahead from a temporal sequence of solar-wind data. The input data include solar-wind density n, velocity V and the southward component Bz of the interplanetary magnetic field. Dst is not included in the input data. The networks implement an explicit functional relationship between the solar wind and the geomagnetic disturbance, including both direct and time-delayed non-linear relations. In this study we especially consider the influence of varying the temporal size of the input-data sequence. The networks are trained on data covering 6600 h, and tested on data covering 2100 h. It is found that the initial and main phases of geomagnetic storms are well predicted, almost independent of the length of the input-data sequence. However, to predict the recovery phase, we have to use up to 20 h of solar-wind input data. The recovery phase is mainly governed by the ring-current loss processes, and is very much dependent on the ring-current history, and thus also the solar-wind history. With due consideration of the time history when optimizing the networks, we can reproduce 84% of the Dst variance.

scholarly journals An hybrid neuro-wavelet approach for long-term prediction of solar wind

2010 ◽  
Vol 6 (S274) ◽  
pp. 153-155 ◽  
Author(s):  
Christian Napoli ◽  
Francesco Bonanno ◽  
Giacomo Capizzi
Keyword(s):  
Neural Networks ◽  
Solar Wind ◽  
Wavelet Decomposition ◽  
Weather Forecasting ◽  
Learning Algorithm ◽  
Correct Prediction ◽  
Data Series ◽  
Intrinsic Structure ◽  
Data Prediction ◽  
Van Allen Belts

AbstractNowadays the interest for space weather and solar wind forecasting is increasing to become a main relevance problem especially for telecommunication industry, military, and for scientific research. At present the goal for weather forecasting reach the ultimate high ground of the cosmos where the environment can affect the technological instrumentation. Some interests then rise about the correct prediction of space events, like ionized turbulence in the ionosphere or impacts from the energetic particles in the Van Allen belts, then of the intensity and features of the solar wind and magnetospheric response. The problem of data prediction can be faced using hybrid computation methods so as wavelet decomposition and recurrent neural networks (RNNs). Wavelet analysis was used in order to reduce the data redundancies so obtaining representation which can express their intrinsic structure. The main advantage of the wavelet use is the ability to pack the energy of a signal, and in turn the relevant carried informations, in few significant uncoupled coefficients. Neural networks (NNs) are a promising technique to exploit the complexity of non-linear data correlation. To obtain a correct prediction of solar wind an RNN was designed starting on the data series. As reported in literature, because of the temporal memory of the data an Adaptative Amplitude Real Time Recurrent Learning algorithm was used for a full connected RNN with temporal delays. The inputs for the RNN were given by the set of coefficients coming from the biorthogonal wavelet decomposition of the solar wind velocity time series. The experimental data were collected during the NASA mission WIND. It is a spin stabilized spacecraft launched in 1994 in a halo orbit around the L1 point. The data are provided by the SWE, a subsystem of the main craft designed to measure the flux of thermal protons and positive ions.

AL index dependence on the solar wind input revisited

2005 ◽  
Vol 36 (12) ◽  
pp. 2440-2444 ◽  
Author(s):  
A.A. Petrukovich ◽  
A.A. Rusanov
Keyword(s):  
Solar Wind ◽  
Wind Input
Sign in / Sign up

Export Citation Format

Share Document