Rainfall Modelling with a Transect View in Ghana

Abstract

Rainfall variability is an inherent part of African climate. This variability has important implications for food production and general livelihoods in countries such as Ghana. 53 years of rainfall data for 15 stations were obtained from the Ghana Meteorological Agency and used to describe the variability in the pattern of rainfall in Ghana. The direct method was used to analyse the annual rainfall totals and the total number of rainy days. A Markov chain modelling approach, which involves the fitting of harmonic regression curves to model the probability of rain within the year was also used. The effects of the previous rainy day was obtained throughout the year. The first order Markov model for each station was significant with the probability of rain given dry being higher than the probability of rain given rain in the north, while the probability of rain given rain was higher than the probability of rain given dry in the south.