# Forecasting the number of sunspots for solar cycle 25 utilizing the facebook prophet model

## TL;DR

This paper uses the Facebook Prophet model to forecast sunspot numbers for Solar Cycle 25 and part of Cycle 26, showing high accuracy compared to existing predictions.

## Contribution

The novel use of the Facebook Prophet model for long-term sunspot forecasting with high accuracy.

## Key findings

- The model predicts Solar Cycle 25 will peak in early 2025 with a smoothed sunspot number of around 118.
- The model forecasts Solar Cycle 26 to peak in mid-2034 with a similar sunspot number.
- The model's coefficient of determination (R²) is 89.23%, indicating strong predictive accuracy.

## Abstract

The solar cycle, also referred to as the solar magnetic activity cycle, represents a nearly periodic change in solar activity occurring approximately every 11 years, as evidenced by the observation of sunspot numbers. The terms solar maximum and minimum denote the phases of peak and trough sunspot activity, respectively. Solar Cycle 25 commenced in December 2019, starting with a minimum smooth sunspot number of 1.8, and is projected to persist until the conclusion of December 2030. In this study, we employed the FB Prophet Prediction Model, utilizing sunspot data collected from January 1749 to March 2025 (spanning over 276.25 years), to forecast sunspot numbers for the latter half of Solar Cycle 25 (69 months). We forecast sunspot numbers for the remainder of Solar Cycle 25 and the entirety of Solar Cycle 26 (through 2036). This study employed the FB Prophet model on 276 years of sunspot data (January 1749–March 2025) to generate two forecasts: one for the remainder of Solar Cycle 25, and a second for the early portion of Solar Cycle 26, extending through 2036. Our model predicts that Solar Cycle 25 will peak in early 2025 and that Solar Cycle 26 will peak in mid-2034, both with a smoothed sunspot number of approximately 118. A comparison between our predicted outcomes and the NOAA published forecast data demonstrates the effectiveness and suitability of the FB Prophet Prediction model for predicting sunspot activity during Cycle 25. The coefficient of determination, commonly referred to as (\documentclass[12pt]{minimal}
\usepackage{amsmath}
\usepackage{wasysym}
\usepackage{amsfonts}
\usepackage{amssymb}
\usepackage{amsbsy}
\usepackage{mathrsfs}
\usepackage{upgreek}
\setlength{\oddsidemargin}{-69pt}
\begin{document}$$R^{2}$$\end{document}), assesses the extent to which the model reflects the observed outcomes and signifies the percentage of variance in the dependent variable that can be forecasted based on the independent variables within the model. Its value is 89.23%, which demonstrates the model’s high level of predictive accuracy. This demonstrates the good agreement results and also confirms the effectiveness and suitability of the FB Prophet Prediction model for predicting sunspot activity during Cycle 25.