Abstract: In Nigeria, rainfall has a significant impact on land use patterns, agricultural output, and ecological systems. For agriculture, water resource management, and other climate-sensitive industries to make wise decisions, reliable rainfall forecasting is essential. The Nigerian Meteorological Agency (NiMet) provided monthly rainfall data (measured in millimeters) for southern Nigeria from Jan. 2009 to Dec. 2024. This study uses time series techniques to evaluate the data. The Kwiatkowski–Phillips–Schmidt–Shin (KPSS) and Augmented Dickey–Fuller (ADF) tests were used to evaluate the series' stationarity. The SARIMA(5,0,5)(0,0,1)[12] model offered the greatest fit, according to model identification and selection based on the Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC). Diagnostic analyses confirmed the adequacy of the model, as the residuals were approximately normally distributed and exhibited no significant autocorrelation. The selected model was subsequently used to forecast for the monthly rainfall from Jan. 2025 to Dec. 2026, with 80% and 90% confidence intervals. The results suggest a relatively stable rainfall pattern characterized by pronounced seasonal variations, highlighting the reliability of the model’s predictive performance. These findings offer valuable insights for agricultural stakeholders, supporting informed crop scheduling, irrigation planning, and disaster risk management. Overall, the study demonstrates the effectiveness of SARIMA-based forecasting as a tool for climate-responsive planning and resilience enhancement in southern Nigeria.”

Keywords: Rainfall Forecasting, Time Series Analysis, SARIMA Model, Southern Nigeria, Agricultural Planning

Pages: 94-103

WSEAS Transactions on