نمذجة و تحليل البقع الشمسية الشهرية باستخدام نماذج المكونات غير المشاهدة

Abstract EN

Predicting the number of sunspots which is likely to occur during any given cycle is important for scientists who understanding the terrestrial consequences of solar activity.

Many methods have been devised and combined and refined in an attempt to predict the activity of sunspots.

In this paper, Unobserved Components Models (UCM) were highlighted and used for modeling and analyzing the monthly sunspot data for the period from January 1900 to December 2018 (1428 months), which is unstable and has a clear cycle.

The analysis showed the superiority of unobserved components model to fit the sunspot data by passing the diagnostic tests as well as its high predicting.

The analysis revealed that the level and seasonal components are stochastic models, and the auto regression, cycles and slope are deterministic models.

The number of sunspots was predicted for a period of 100 months from January 2019 to April 2026.

The sunspot number maximum in cycle 25 is predicted to occur in April 2023 with a sunspot number 146 and a solar cycle length of 125.38331 months or 10.45 years.

SAS 9.4 was used in the analysis.