Re-evaluation of Predictive Models in Light of New Data: Sunspot Number Version 2.0

Abstract

The original version of the Zürich sunspot number (Sunspot Number Version 1.0) has been revised by an entirely new series (Sunspot Number Version 2.0). We re-evaluate the performance of our previously proposed models for predicting solar activity in the light of the revised data. We perform new monthly and yearly predictions using the Sunspot Number Version 2.0 as input data and compare them with our original predictions (using the Sunspot Number Version 1.0 series as input data). We show that our previously proposed models are still able to produce quite accurate solar-activity predictions despite the full revision of the Zürich Sunspot Number, indicating that there is no significant degradation in their performance. Extending our new monthly predictions (July 2013 – August 2015) by 50 time-steps (months) ahead in time (from September 2015 to October 2019), we provide evidence that we are heading into a period of dramatically low solar activity. Finally, our new future long-term predictions endorse our previous claim that a prolonged solar activity minimum is expected to occur, lasting up to the year ≈ 2100. © 2016, Springer Science+Business Media Dordrecht.