Optimization of the neural-network geomagnetic model for forecasting large-amplitude substorm events

Artificial neural networks (NN) have been used to model solar wind-driven auroral electrojet dynamics, with emphasis on the reliable real-time forecasting of auroral electrojet activity (the AE index) from solar wind input. Practical limitations of the NN-based models used earlier are clarified. These include the inability to accurately predict large-amplitude substorm events, which is the most important feature for many applications. A novel technique for improving predictions is suggested based on application-specific threshold mapping and symbolic encoding of the AE index. This approach allows us to disregard relatively unimportant details of small-amplitude perturbations and effectively improve forecasting of large-amplitude events. Results from our new model imply that application-oriented optimization of the real-time substorm forecasting system can be an important factor in the overall improvement of the prediction accuracy.

[1]  L. Lanzerotti.  Geomagnetic induction effects in ground-based systems , 1983 .

[2]  Tang,et al.  Symbol sequence statistics in noisy chaotic signal reconstruction. , 1995, Physical review. E, Statistical physics, plasmas, fluids, and related interdisciplinary topics.

[3]  W. Horton,et al.  A low‐dimensional energy‐conserving state space model for substorm dynamics , 1996 .

[4]  Wendell Horton,et al.  A Low-Dimensional Dynamical Model for the Solar Wind Driven Geotail-Ionosphere System , 1998 .

[5]  R. Schunk,et al.  Global ionosphere‐polar wind system during changing magnetic activity , 1997 .

[6]  Konstantinos Papadopoulos,et al.  Reconstruction of low‐dimensional magnetospheric dynamics by singular spectrum analysis , 1993 .

[7]  P. Wintoft,et al.  Prediction of geomagnetic storms from solar wind data with the use of a neural network , 1994 .

[8]  Juan Alejandro Valdivia,et al.  Spatiotemporal activity of magnetic storms , 1999 .

[9]  Daniel N. Baker,et al.  Magnetospheric Impulse Response for Many Levels of Geomagnetic Activity , 1985 .

[10]  P. Palmadesso,et al.  Plasma transport in the auroral return current region , 1987 .

[11]  Daniel N. Baker,et al.  A description of the solar wind-magnetosphere coupling based on nonlinear filters , 1995 .

[12]  J. Takalo,et al.  Correlation dimension and affinity of AE data and bicolored noise , 1993 .

[13]  J. Timonen,et al.  Properties of AE data and bicolored noise , 1994 .

[14]  D. Prichard,et al.  The non‐linear response of the magnetosphere: 30 October 1978 , 1993, comp-gas/9305003.

[15]  J. Joselyn.  Geomagnetic activity forecasting: The state of the art , 1995 .

[16]  Patricia H. Reiff,et al.  Effects of the March 1989 solar activity , 1989 .

[17]  Geoffrey E. Hinton,et al.  Learning representations by back-propagating errors , 1986, Nature.

[18]  G. Ganguli,et al.  Dispersive properties of a magnetized plasma with a field‐aligned drift and inhomogeneous transverse flow , 1996 .

[19]  Toshiki Tajima,et al.  Forecasting auroral electrojet activity from solar wind input with neural networks , 1999 .

[20]  G. Ganguli,et al.  Electrostatic oscillations due to filamentary structures in the magnetic-field-aligned flow: The ion-acoustic branch , 1999 .

[21]  G. Ganguli,et al.  Coupling of microprocesses and macroprocesses due to velocity shear: An application to the low-altitude ionosphere , 1994 .

[22]  Charles F. Kennel,et al.  Topside current instabilities , 1971 .

[23]  R. A. Smith,et al.  Prediction of geomagnetic activity , 1993 .

[24]  J. Lyon,et al.  Global numerical simulation of the growth phase and the expansion onset for a substorm observed by Viking , 1995 .

[25]  V. Gavrishchaka,et al.  Three‐dimensional simulations of the ionospheric plasma transport in the presence of the structured field‐aligned flows , 1999 .

[26]  Daniel N. Baker,et al.  Substorm recurrence during steady and variable solar wind driving: Evidence for a normal mode in the unloading dynamics of the magnetosphere , 1994 .

[27]  H. Lundstedt,et al.  Prediction of geomagnetic storms from solar wind data using Elman Recurrent Neural Networks , 1996 .

[28]  W. Krueger,et al.  Cross‐field transport due to low‐frequency oscillations in the auroral region: A three‐dimensional simulation , 1999 .

[29]  Vadim M. Uritsky,et al.  Low frequency 1/f-like fluctuations of the AE-index as a possible manifestation of self-organized criticality in the magnetosphere , 1998 .

[30]  D. Baker,et al.  Dst index prediction using data‐derived analogues of the magnetospheric dynamics , 1998 .

[31]  José Carlos Príncipe,et al.  The gamma model--A new neural model for temporal processing , 1992, Neural Networks.

[32]  R. Schunk,et al.  A three‐dimensional time‐dependent model of the polar wind , 1989 .

[33]  Daniel N. Baker,et al.  Indications of low dimensionality in magnetospheric dynamics , 1991 .

[34]  G. Ganguli,et al.  Kinetic theory for electrostatic waves due to transverse velocity shears , 1988 .

[35]  H. Gleisner,et al.  Response of the auroral electrojets to the solar wind modeled with neural networks , 1997 .

[36]  C. Mobarry,et al.  Topological structure of the magnetotail as a function of interplanetary magnetic field direction , 1995 .

[37]  Tom Chang,et al.  Self-organized criticality, multi-fractal spectra, sporadic localized reconnections and intermittent turbulence in the magnetotail , 1999 .

[38]  Daniel N. Baker,et al.  A nonlinear dynamical analogue model of geomagnetic activity , 1992 .

[39]  P. Wintoft,et al.  Neural network models predicting the magnetospheric response to the 1997 January Halo‐CME event , 1998 .

[40]  K. Papadopoulos.  A review of anomalous resistivity for the ionosphere , 1976 .

[41]  J. Timonen,et al.  Characteristic time scale of auroral electrojet data , 1994 .

[42]  Konstantinos Papadopoulos,et al.  Low-dimensional chaos in magnetospheric activity from AE time series , 1990 .

[43]  H. Gleisner,et al.  Predicting geomagnetic storms from solar-wind data using time-delay neural networks , 1996 .

[44]  Daniel N. Baker,et al.  The organized nonlinear dynamics of the magnetosphere , 1996 .

[45]  Daniel N. Baker,et al.  Satellite anomalies linked to electron increase in the magnetosphere , 1994 .

[46]  Juan Alejandro Valdivia,et al.  Prediction of magnetic storms by nonlinear models , 1996 .

[47]  Tang,et al.  Self-Organized Criticality: An Explanation of 1/f Noise , 2011 .

[48]  Toshiki Tajima,et al.  Neural net forecasting for geomagnetic activity , 1993 .

[49]  C. Goertz,et al.  Chaotic appearance of the AE index , 1991 .