Prediction of Sunspot Cycles by Data Assimilation Method

TL;DR

This paper applies the Ensemble Kalman Filter to a dynamo model to forecast sunspot cycles, successfully predicting a weaker upcoming cycle consistent with observed trends.

Contribution

It introduces a novel application of data assimilation with EnKF to solar dynamo modeling for improved sunspot cycle prediction.

Findings

Forecasted the next sunspot cycle to be ~30% weaker.

Predictions show reasonable agreement with past cycles.

Model indicates a continuing trend of low solar activity.

Abstract

Despite the known general properties of the solar cycles, a reliable forecast of the 11-year sunspot number variations is still a problem. The difficulties are caused by the apparent chaotic behavior of the sunspot numbers from cycle to cycle and by the influence of various turbulent dynamo processes, which are far from understanding. For predicting the solar cycle properties we make an initial attempt to use the Ensemble Kalman Filter (EnKF), a data assimilation method, which takes into account uncertainties of a dynamo model and measurements, and allows to estimate future observational data. We present the results of forecasting of the solar cycles obtained by the EnKF method in application to a low-mode nonlinear dynamical system modeling the solar $\alpha\Omega$-dynamo process with variable magnetic helicity. Calculations of the predictions for the previous sunspot cycles show a reasonable agreement with the actual data. This forecast model predicts that the next sunspot cycle will be significantly weaker (by $\sim 30%$) than the previous cycle, continuing the trend of low solar activity.