Solar activity forecast: Spectral analysis and neurofuzzy prediction

• Published in: Journal of atmospheric and solar-terrestrial physics 2005-04, Vol.67 (6), p.595-603
• Main Authors: Gholipour, Ali, Lucas, Caro, Araabi, Babak N., Shafiee, Masoud
• Format: Article
• Language: English
• Subjects: Earth, ocean, space; Exact sciences and technology; External geophysics; Forecasting; Long-term prediction; Neurofuzzy modeling; Physics of the high neutral atmosphere; Physics of the ionosphere; Physics of the magnetosphere; Singular spectrum analysis; Solar activity; Sunspot numbers
• ISSN: 1364-6826; 1879-1824
• DOI: 10.1016/j.jastp.2004.12.001

Active research in the last two decades indicates that the physical precursor and solar dynamo techniques are preferred as practical tools for long-term prediction of solar activity. But why should we omit more than 23 cycles of solar activity history, and just use empirical methods or simple autoregressive methods on the basis of observations for the latest eight cycles? In this article, a method based on spectral analysis and neurofuzzy modeling is proposed that is capable of issuing very accurate long-term prediction of sunspot number time series. A locally linear neurofuzzy model is optimized for each of the principal components obtained from singular spectrum analysis, and the multi-step predicted values are recombined to make the sunspot number time series. The proposed method is used for solar cycles 22 and 23 and the results are remarkably good in comparison to the predictions made by solar dynamo and precursor methods. An early prediction of the maximum smoothed international sunspot number for cycle 24 is 145 in 2011–2012.