The Use of Markov Model in Continuous Time for Prediction of Rainfall for Crop Production

Abstract

In a first order Markov process, if the state is known for any specific values of the time parameter, that information is sufficient to predict the next behavior of the process beyond that point. This principle was used to formulate a four-state model in continuous time to study the annual rainfall data with respect to the annual rainfall distribution for crop production in minna. It was observed that if it is low rainfall in a given year, it would take at most 25%,33% and 27% of the time to make a transition to moderate rainfall also well spread, high rainfall, and moderate rainfall but not well spread respectively in the far future. Thus given the rainfall in a year, it is possible to determine quantitatively the probability of finding rainfall in other states in the following year and in the long run. This is an important information that could assist the farmers to plan strategies for high crop production in the region.