Prediction of the Length of Upcoming Solar Cycles

Abstract

The forecast of solar cycle (SC) characteristics is crucial particularly for several space-based missions. In the present study, we propose a new model for predicting the length of the SC. The model uses the information of the width of an autocorrelation function that is derived from the daily sunspot data for each SC. We tested the model on Versions 1 and 2 of the daily international sunspot number data for SCs 10 – 24. We found that the autocorrelation width $A_{\mathrm{w}} ^{n}$ of SC $n$ during the second half of its ascending phase correlates well with the modified length that is defined as $T_{\mathrm{cy}}^{n+2} - T_{\mathrm{a}}^{n}$ . Here $T_{\mathrm{cy}}^{n+2}$ and $T_{ \mathrm{a}}^{n}$ are the length and ascent time of SCs $n+2$ and $n$ , respectively. The estimated correlation coefficient between the model parameters is 0.93 (0.91) for Version 1 (Version 2) sunspot series. The standard errors in the observed and predicted lengths of the SCs for Version 1 and Version 2 data are 0.38 and 0.44 years, respectively. The advantage of the proposed model is that the predictions of the length of the upcoming two SCs (i.e., $n+1$ , $n+2$ ) are readily available at the time of the peak of SC $n$ . The present model gives a forecast of 11.01, 10.52, and 11.91 years (11.01, 12.20, and 11.68 years) for the length of SCs 24, 25, and 26, respectively, for Version 1 (Version 2).