Addition to the Gnevyshev-Ohl rule and prediction of solar cycle 25

TL;DR

This paper extends the Gnevyshev-Ohl rule by identifying new sunspot cycle precursors, enabling more accurate predictions of solar cycle 25's amplitude and timing using sunspot number relations.

Contribution

It introduces novel precursor methods based on sunspot numbers for predicting the amplitude and timing of odd and even solar cycles, improving cycle prediction accuracy.

Findings

Cycle 25 predicted to reach 154 sunspot units in 2023

Strong correlation (up to 0.94) between precursor sunspot numbers and cycle maxima

New precursors enable cycle predictions with high correlation coefficients

Abstract

In addition to the Gnevyshev-Ohl rule (GOR), the relation of the odd cycle with the subsequent even one in the 22-year Hale solar cycle was found. It is shown that 3 years before the 11-year minimum $m$, the value of the relative sunspot number SN in an odd cycle is closely related to the value of the maximum in the next even cycle (correlation coefficient $\rho=0.94$), and the same relation of an odd cycle with the previous even one is weaker. Like GOR, cycles are linked in pairs, but opposite to the Rule. Based on this result, we propose to use SN$_{m-3}$ on the descending phase of the previous odd cycle as a precursor of the subsequent EVEN cycle (Figure 3a) -- a precursor called MI3E. For the prediction of an odd cycle or a prediction without consideration of parity (as in the article by Braj\v{s}a et al., 2022), this method gives less reliable results. To predict the amplitude of an ODD cycle, we propose to use the precursor of the seventh year to its maximum $M$ MA7O -- SN$_{M-7}$ on the descending phase of the previous even cycle (Figure 3b). It turned out that in this case, we can also predict the years near the maximum with a high correlation coefficient ($\rho=0.90{-}0.94$). Thus, the proposed approaches allow us to predict cycles of different parity. According to our prediction, the current solar Cycle 25 in 2023 will reach a maximum of 154 units with a prediction error of $\pm25$ (68% confidence) and $\pm53$ (95% confidence). In 2024, SN will be almost as high as in 2023 -- 147 units, so with smaller time averaging scales, the maximum will fall at the end of 2023.