GENERATION OF RAINFALL DATA SERIES USING SECOND ORDER MARKOV CHAIN MODEL: A CASE STUDY OF PORT HARCOURT, RIVERS STATE, NIGERIA.

This study investigates the generation of rainfall data series using the Second Order Markov Chain Model. The research develop a robust methodology for simulating realistic rainfall patterns, leveraging on historical rainfall data spanning over 20 years period (1996-2015). The historic observed rainfall data sets were obtained from the Nigerian Meteorological Agency (NiMet). The study explores the temporal dependencies between consecutive days' rainfall states, categorized into four distinct states: rainy days followed by rainy days (RR), rainy days followed by dry days (RD), dry days followed by rainy days (DR), and dry days followed by dry days (DD). The Markov Chain Model effectively captured these dependencies, allowing for the prediction of rainfall trends and the generation of synthetic rainfall sequences. The analysis revealed significant variability in rainfall patterns, with notable implications for flood management, urban planning, and environmental sustainability in Port Harcourt. The results indicate that while the Second Order Markov Chain Model is suitable for short-term rainfall forecasting, its accuracy decreases over longer periods due to the complexity of the rainfall processes. Recommendations include enhancing data collection efforts, integrating more advanced models, and considering the impacts of climate change on future rainfall patterns. This research contributes valuable insights into the application of stochastic models in hydrology, providing a framework for improving water resource management and mitigating the risks associated with extreme weather events in coastal regions like Port Harcourt.