A rainfall simulation model for agricultural development in Bangladesh

Abstract

A rainfall simulation model based on a first-order Markov chain has been developed to simulate the annual variation in rainfall amount that is observed in Bangladesh. The model has been tested in the Barind Tract of Bangladesh. Few significant differences were found between the actual and simulated seasonal, annual and average monthly. The distribution of number of success is asymptotic normal distribution. When actual and simulated daily rainfall data were used to drive a crop simulation model, there was no significant difference of rice yield response. The results suggest that the rainfall simulation model perform adequately for many applications.

Keywords

• Simulation modelling
• Markov chain
• Probability distribution
• Latin Hypercube Sampling
• Agro-technology.