Rainfall-Runoff Modelling for Sustainable Water Resource Management in the Louisiana State

Abstract

Abstract. Urban watersheds in Louisiana face increasing pressure as urbanization and environmental challenges mount in the 21st century. This study leverages Google Earth Engine (GEE) to model and predict runoff patterns in Louisiana from 2018 to 2023, aiming to improve water resource management. An iterative methodology integrates key data such as soil texture, land use, and daily precipitation, focusing on the Curve Number (CN) approach to represent runoff potential based on land characteristics. The analysis uses the Soil Conservation Service Curve Number (SCS CN) method, factoring in precipitation and antecedent moisture, and visualizes spatial distributions of CN, precipitation, soil, and runoff. Results show an accurate model with an R2 of 0.964, validated against the European Centre for Medium-Range Weather Forecasts (ECMWF) Reanalysis 5th Generation (ERA5) Land Monthly data. Model performance was further assessed with the receiver operating characteristic (ROC) curve, achieving a high area under the curve (AUC) value of 0.937, underscoring the model’s precision in predicting daily runoff. This highlights the role of comprehensive, data-driven approaches in understanding Louisiana's hydrological conditions, emphasizing the importance of soil, land use, and rainfall data in managing water resources and predicting surface runoff within urban watersheds.