Abstract

This study examines rainfall-runoff interactions in Nigeria’s flood-prone Bida watershed from 2010 to 2023, comparing traditional statistical methods with the Runge-Kutta Fourth Order (RK4) simulation for runoff prediction. Daily rainfall and runoff data were sourced from the Nigerian Meteorological Agency (NiMet), processed using Excel and Python to compute monthly averages. Pearson and Spearman correlation coefficients assessed linear and monotonic relationships, respectively. The RK4 method simulated runoff based on rainfall and watershed parameters. Results indicate peak rainfall in August and September (300-350 mm/month), with significant events in 2012 and 2018. Corresponding peak runoff occurred in September, reaching up to 160 m³/s, showing a one-month lag due to catchment storage. Strong correlations were observed (Pearson r = 0.89; Spearman’s ρ = 0.83), confirming the robustness of the rainfall-runoff relationship. The RK4 simulation demonstrated high predictive accuracy (NSE = 0.88; RMSE = 5.2 m³/s). Anomalies in 2012, 2018, and 2022 aligned with major flood events, highlighting increasing hydro-climatic variability. These findings validate RK4 as an effective tool for runoff forecasting and underscore the importance of integrated modeling approaches in enhancing flood prediction and water management strategies under changing climate conditions.

Keywords: Rainfall-runoff modeling, Runge-Kutta simulation (RK4), Spearman correlation, Hydro-climatic variability, Watershed