Analyzing the Hydrological Disconnect Between Extreme Precipitation and River Discharge in Douala, Cameroon

This study investigates the critical temporal disconnect between extreme rainfall and subsequent flood discharge in Douala, Cameroon, a rapidly urbanizing coastal city increasingly vulnerable to flooding. Using a 13-year dataset (2010–2023) that integrates CHIRPS precipitation data Funk et al. (2015) and ERA5-Land discharge data (Muñoz-Sabater et al. 2021), we applied Extreme Value Theory (EVT) through a Peak-Over-Threshold (POT) framework to quantify and compare return levels and return periods for both meteorological and hydrological extremes. Our analysis reveals a marked hydrological transformation: while extreme rainfall exhibits heavy-tailed behavior (shape parameter ξ = 0.19), discharge extremes show even heavier tails (ξ = 0.44), suggesting that urban watershed processes amplify flood risks beyond what rainfall statistics alone would predict. Cross-correlation and event-based analyses indicate a systematic temporal lag of 1–2 days between precipitation peaks and discharge responses, with weak coupling (r ≈ 0.2–0.4) between daily rainfall and runoff intensity. Moreover, a 100-days return period rainfall event (98.08 mm) generates only a 37.93 mm·day⁻¹ discharge peak, reflecting over 60% attenuation between precipitation input and hydrological output. Diagnostic plots, threshold stability checks, and goodness-of-fit tests confirmed the robustness of the Generalized Pareto Distribution (GPD) models. The inclusion of return date analysis, following the methodology proposed by Padji et al. (2024), demonstrated the approach’s predictive capability for estimating the timing of flood events rather than their magnitude alone. These findings reveal that Douala’s flood hazard is not solely determined by rainfall extremes but is amplified by rapid urbanization, reduced infiltration, and altered drainage dynamics. The combination of weak rainfall–discharge coupling, temporal lag, and hydrological attenuation underscores the urgent need for integrated flood forecasting and water management strategies that explicitly incorporate urban hydrological dynamics rather than relying exclusively on rainfall forecasts.