Modeling Urban Surface Runoff in Tropical Watersheds Using the SCS-CN Method on Google Earth Engine: Case of the Gourou Basin

Abstract

The Gourou watershed, located in the metropolitan region of Abidjan, is highly vulnerable to flooding due to rapid urbanization and the increasing frequency of extreme rainfall events. To address the challenge of runoff estimation in this ungauged tropical urban basin, this study develops and applies an innovative, spatially explicit methodology that quantifies surface runoff through the Soil Conservation Service Curve Number (SCS-CN) method, integrating pedological characteristics (soil textures) and dynamic land use patterns. Our approach leverages an automated processing chain implemented in Google Earth Engine, enabling high-resolution land use classification (overall accuracy: 99.58%; Kappa index: 0.97), spatial interpolation of soil data (based on 19 sampling points), computation of Curve Numbers (CN) through matrix cross-referencing of soil texture and land use classes, and refinement of CN values through slope adjustments to enhance hydrological accuracy. Hydrological modeling was performed using the SCS equation. The analysis reveals alarming hydrological indicators: a significant reduction in effective infiltration, an average runoff rate of 42.6% (runoff coefficient = 0.42), a total runoff volume of 827,113 m³ from cumulative rainfall of 1,941,220 m³, and widespread imperviousness — 77.8% of the watershed consists of built-up areas or bare soils (CN > 78). These conditions drive rapid surface flow concentration, intensifying flood peaks and highlighting the critical need for spatially targeted mitigation strategies in data-scarce urban watersheds.