Thesis (M. Sc. (Statistics)) -- University of Limpopo, 2021 / The effects of ozone depletion on climate change has, in recent years, become a reality, impacting on changes in rainfall patterns and severity of extreme floods or extreme droughts. The majority of people across the entire African continent live in semi-arid and drought-prone areas. Extreme droughts are prevalent in Somalia and eastern Africa, while life-threatening floods are common in Mozambique and some parts of other SADC (Southern African Development Community) countries. Research has cautioned that climate change in South Africa might lead to increased temperatures and reduced amounts of rainfall, thereby altering their timing and putting more pressure on the country’s scarce water resources, with implications for agriculture, employment and food security. The average annual rainfall for South Africa is about 464mm, falling far below the average annual global rainfall of 860mm. The Limpopo Province, which is one of the nine provinces in South Africa, and of interest to this study, is predominantly agrarian, basically relying on availability of water, with rainfall being the major source for water supply. It is, therefore, pertinent that the rainfall pattern in the province be monitored effectively to ascertain the rainy period for farming activities and other uses. Modelling and forecasting rainfall have been studied for a long time worldwide. However, from time to time, researchers are always looking for new models that can predict rainfall more accurately in the midst of climate change and capture the underlying dynamics such as seasonality and the trend, attributed to rainfall. This study employed Exponetial Smoothing (ETS) State Space and Seasonal Autoregressive Integrated Moving Average (SARIMA) models and compared their forecasting ability using root mean square error (RMSE). Both models were used to capture the sporadic behaviour of rainfall. These two models have been widely applied to climatic data by many scholars and adjudged to perform creditably well. In an attempt to find a suitable prediction model for monthly rainfall patterns in Limpopo Province, data ranging from January 1900 to December 2015, for seven weather stations: Macuville Agriculture, Mara Agriculture, Marnits, Groendraal, Letaba, Pietersburg Hospital and Nebo, were analysed. The results showed that the two models were adequate in predicting rainfall patterns for the different stations in the Limpopo Province. / National Research Foundation (NRF)