Prediction of even and odd hemispheric sunspot cycles

Abstract

Sunspots form when the Sun's poloidal magnetic field gets twisted in a toroidal magnetic field, which form tight flux tubes. Due to convective motion of the solar dynamo, they rise through the solar surface and form sunspots. This means that they are an indirect indicator of the Sun's magnetism, which is the cause of many solar phenomena, such as solar eruptions. In the mid-1800s, the 11-year sunspot cycle was discovered, which made it possible to quantify the solar activity through the number of sunspots.

Predicting the sunspots has been an age-old quest, and over the last century or so, countless of methods have been proposed and implemented. These vary from statistical methods to complex physical simulations and neural networks, but majority of them are physically motivated forecasting methods. Meaning they model some aspects of the Sun, such as the solar dynamo or the movement of the magnetic field on the solar surface. One very successful method is the use of some precursors, which means using some physical aspect as an indirect indicator for the next cycle. It has been shown that, for example, the geomagnetic aa index near solar solar minimum and sunspot number some years before the solar minimum are very good precursor for the next cycles sunspot number.

This thesis studies the hemispheric sunspot cycles, while taking into account the 22-year magnetic Hale cycle. It is taken into account by separating the even- and odd-numbered cycles. A 4-parameter asymmetric Gaussian function is used to model the sunspot cycles, after which connections between the four unknown parameters are sought. In addition, precursors are also used to look relationships between them and the parameters. In the southern hemisphere, the best precursor for the even cycles was found to be the southern hemisphere sunspot number 45 months before the southern hemisphere sunspot minimum, and for the odd cycles, was the maximum geomagnetic aa index close to the autumnal equinox within the a 3-year window preceding the southern hemisphere sunspot minimum. In the northern hemisphere, the best precursor for the even cycles was found to be the northern hemisphere sunspot number 26 months before the northern hemisphere sunspot minimum, and for the odd cycles, was the maximum geomagnetic aa index close to the autumnal equinox within the a 3-year window preceding the northern hemisphere sunspot minimum. Using cross-validation methods, each individual past cycle is predicted. For the future cycle 25, the southern hemisphere is predicted to have an amplitude of 92.4 ± 22.5, which would occur in September 2023 ± 0.96 years, and for northern hemisphere, an amplitude of 88.7 ± 16.6, which would occur in October 2023 ± 0.56 years.