Prediction of Weekly Rainfall Both in Discrete and Continuous Time Using Markov Model

In this chapter, a Markov model is presented for studying weekly rainfall in both discrete and continuous time. Using eleven years of rainfall data, the model predicts and analyses the weekly rainfall pattern of Makurdi, Nigeria (2005-2015). The discrete time Markov model stabilises to equilibrium probabilities after a few effective iterations, revealing that in the long run, 22 percent of the weeks during the rainy season in Makurdi will have no rainfall, 50 percent will have low rainfall, 25 percent will have moderate rainfall, and 2 percent will have high rainfall. It was discovered that if the continuous time Markov model is in the No rainfall state in a given week, it would take at most 49 percent, 27 percent, and 16 percent of the time to shift to Low rainfall, Moderate rainfall, and High rainfall in the far future, respectively. Thus, given a week’s rainfall, the likelihood of finding weekly rainfall in other states the following week and in the long run can be measured quantitatively. A week of High rainfall cannot be preceded by another week of High rainfall, nor can a week of High rainfall be followed by a week of No rainfall, nor can a week of Moderate rainfall precede a week of High rainfall, according to the model. The rainfall pattern of the study region is better understood thanks to the combined results of the discrete and continuous time Markov models. These findings are crucial for Markudi residents and environmental management scientists to prepare effectively and grow viable crops.

Author (s) Details

Dr. Lawal Adamu
Department of Mathematics, Federal University of Technology, Minna, Nigeria.

Prof. U. Y. Abubakar
Department of Mathematics, Federal University of Technology, Minna, Nigeria.

Prof. Danladi Hakimi
Department of Mathematics, Federal University of Technology, Minna, Nigeria.

Prof. Andrew Saba Gana
Department of Crop Production, Federal University of Technology, Minna, Nigeria.

View Book :- https://stm.bookpi.org/IREGES-V8/issue/view/37

Leave a Reply

Your email address will not be published. Required fields are marked *