Solar cycle full-shape predictions: a global error evaluation for cycle 24

TL;DR

This paper evaluates the accuracy of various solar cycle 24 predictions using a global error metric and introduces a percent cycle similarity degree for better shape comparison.

Contribution

It provides a quantitative, a posteriori assessment of solar cycle prediction methods, including a new metric for shape similarity evaluation.

Findings

Non-linear dynamics method showed promising prediction accuracy.

Global error metric effectively measures prediction performance.

Percent cycle similarity degree enhances shape comparison of solar cycle predictions.

Abstract

There are many proposed prediction methods for solar cycles behavior. In a previous paper we updated the full-shape curve prediction of the current solar cycle 24 using a non-linear dynamics method and we compared the results with the predictions collected by the NOAA/SEC prediction panel, using observed data up to October 2010. The aim of the present paper is to give a quantitative evaluation, a posteriori, of the performances of these prediction methods using a specific global error, updated on a monthly basis, which is a measure of the global performance on the predicted shape (both amplitude and phase) of the solar cycle. We suggest also the use of a percent cycle similarity degree, to better evaluate the predicted shape of the solar cycle curve.