DISCRETIZATION OF CLIMATIC CHARACTERISTICS AND MARKOV CHAIN MODELLING OF CROP GROWTH

Abstract

Climate change is impacted by multiple variables, and modeling the joint impact of climatic
variables is of paramount interest; hence, this study presents a unique method that uses the logical
operator to map bivariate data series to the univariate sequence. Each of the bivariate random
variables can take only categorical values. The logical “AND” and “OR” were used for mapping
these sequences and, subsequently, the Markov chain analysis. The method was applied to climatic
variables (Rainfall and Temperature) to obtain favourable and unfavourable climate conditions for
the growth of the yam crop. The Markov chain analysis indicates that the sequence of state for the
yam crop is ergodic and thus, the favourable and unfavourable climatic conditions has a stable
distribution. The logical “AND” has a low probability of favorability of 0.36 compared to the
logical “OR”, 0/75. The climate change impact (CCI) revealed that climate change adversely
affects the growth of the yam crop. The mean recurrent time for favourable climate gave an insight
into how to adapt to avoid losses. The study recommends that farmers invest more in the crop in
question, considering climate change adaptation (CCA). This is because there is high climatic
favorability during this period.