Abstract
Two neural network algorithms are applied to the short-term,1 to 3 days, prediction of theAp geomagnetic index. A multi-layer, back-propagation (MBP) network is used to implement a self-prediction filter forAp and this provides a forecast of the numerical value of the index. A probabilistic neural network (PNN) is used to estimate the probability distribution of theAp index, in six activity classes, and to provide a forecast of the single most likely activity class for each day. BothAp and an index of solar activity, based on the daily reports issued by the Space Environment Services Centre (Boulder), are input to the probabilistic net. It is found that the numerical forecasts of the MBP filter are most accurate at low, non-storm, levels of activity. This non-linear method provides quantitatively better estimates of activity than are produced by an existing linear prediction filter, particularly with increasing forward forecasting lag. At high levels of the solar activity index the PNN is found to anticipate storm classAp with around 60% accuracy in 1992 and 1993. Some details of the algorithms and implementation issues are described. It is concluded that interplanetary field and solar wind data will be significant components of any of the possible future developments which are discussed.
Similar content being viewed by others
References
Beale, R., andJackson, T.,Neural Computing: An Introduction (Adam Hilger, Bristol 1990).
Caudhill, M., andButler, C.,Understanding Neural Networks: Computer Explorations, Volume 2, Advanced Networks (MIT Press, Cambridge, Massachusetts 1992).
Joselyn, J.-A., Lundtedt, H., andTrolinger, J.,Proceedings of an International Workshop on Artificial Intelligence Applications in Solar-Terrestrial Physics at Lund, Sweden (NOAA, Boulder, USA 1993).
Lundstedt, H. (1992),Neural Networks and Predictions of Solar-Terrestrial Effects, Planet. Space Sci.40, 457–464.
SECS,SESC Forecasters' Handbook (NOAA, Boulder, USA 1990).
SESC,SESC Glossary of Solar-terrestrial Terms (NOAA, Boulder, USA 1992).
Thomson, A. W. P.,Neural Networks and Non-linear Prediction Filters for the Ap Geomagnetic Index (Brit. Geol. Surv. Tech. Rept.WM/92/33 1992).
Thomson, A. W. P.,Neural Networks and Non-linear Prediction of Geomagnetic Activity. InProceedings of an International Workshop on Artificial Intelligence Applications in Solar-terrestrial Physics at Lund, Sweden (eds. Joselyn, J-A., Lundstedt, H., and Trolinger, J.) (NOAA, Boulder, USA 1993a) pp. 133–138.
Thomson, A. W. P.,The Use of Solar Activity Data in the Short-term Prediction of the Ap Geomagnetic Index (Brit. Geol. Surv. Tech. Rept.WM/93/19R 1993b).
Thomson, A. W. P., Clark, T. D. G., andKerridge, D. J.,Computer Algorithms and FORTRAN Programs for Forecasting Solar and Geomagnetic Activity in the Short Term (Brit. Geol. Surv. Tech. Rept.WM/92/19C 1992).
Thomson, A. W. P., Clark, T. D. G., andKerridge, D. J.,Forecasting Ap in the Short-term using Time Series Analysis, InSolar Terrestrial Predictions Volume 3: Proceedings of a Workshop at Ottawa, Canada (eds. Hruska, J., Shea, M. A., Smart, D. F., and Heckman, G.) (World Data Center A for Solar-Terrestrial Physics, NOAA, Boulder, USA 1993) pp. 269–283.
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Thomson, A.W.P. Non-linear predictions ofAp by activity class and numerical value. PAGEOPH 146, 163–193 (1996). https://doi.org/10.1007/BF00876675
Received:
Accepted:
Issue Date:
DOI: https://doi.org/10.1007/BF00876675