Nonlinear Prediction of Solar Cycle 24

TL;DR

This paper applies nonlinear analysis techniques to forecast the peak of Solar Cycle 24, predicting it will reach maximum activity around December 2012 with about 87 sunspots, aiding space weather prediction.

Contribution

It introduces a novel application of Hurst exponent analysis and Sugihara-May algorithm for decadal-scale prediction of solar activity cycles.

Findings

Hurst exponent indicates strong autocorrelation in sunspot data.

Forecast predicts Solar Cycle 24 maximum in December 2012.

Maximum sunspot number estimated at approximately 87.

Abstract

Sunspot activity is highly variable and challenging to forecast. Yet forecasts are important, since peak activity has profound effects on major geophysical phenomena including space weather (satellite drag, telecommunications outages) and has even been correlated speculatively with changes in global weather patterns. This paper investigates trends in sunspot activity, using new techniques for decadal-scale prediction of the present solar cycle (cycle 24). First, Hurst exponent $H$ analysis is used to investigate the autocorrelation structure of the putative dynamics; then the Sugihara-May algorithm is used to predict the ascension time and the maximum intensity of the current sunspot cycle. Here we report $H$ = 0.86 for the complete sunspot number dataset (1700-2007) and $H$ = 0.88 for the reliable sunspot data set (1848-2007). Using the Sugihara-May algorithm analysis, we forecast that cycle 24 will reach its maximum in December 2012 at approximately 87 sunspots units.