Arima Model for Forecasting of Monthly Rainfall And Temperature in the Lake Victoria Basin

Abstract

Economic activities in Lake Victoria Basin such as agriculture, sh- ing, mining and transportation depend's heavily on the climatic conditions of the Lake and its Basin. Global climatic change caused by Greenhouse Gas emission (GHG) has resulted in a disruptive and erratic weather pattern for these economic activities. This unpre- dictable weather variations is responsible for loss of life and destruc- tion of property. The primary cause of this negative impact is lack of a reliable information addressing climatic variation within the region. The main objective of this project was to identify a suit- able time series model that can be used in predicting, forecasting and analyzing weather variations.The Box jenskin methodology is used to build ARIMA model for rainfall and temperature. Data ob- tained from the Kenya Meteorological Department's Kisumu, Lwak, and Migori for the years 2007 to 2014 produced An ARIMA (2,0,1) model for rainfall is using R package. Data for the years 2011 to 2013 were estimated using values from the years 2008 to 2010 and the relationship showed a strong positive relationship indicating a high accuracy level on predictability by the model.