Abstract

Accurate catchment estimation is essential for water management, flood forecasting, and environmental planning. Traditional grid-based and contour-based methods are commonly used as baselines but often show limited accuracy, especially in large terrains with long or incomplete contours. This study presents an enhanced, physically based algorithm for catchment size estimation. The algorithm relies on elevation and gradient information obtained from a BEM-based numerical solution of Laplace’s equation, and is designed to work effectively with shortened, contour-based inputs. The algorithm is integrated into an adaptive framework that automates contour creation, applies robust preprocessing, and delineates surface water paths (SWPs) from real-world digital elevation models (DEMs). This integration enables accurate catchment size estimation while reducing computational complexity. Validation on both synthetic and real-world terrains confirms the framework’s effectiveness. On synthetic surfaces, the algorithm achieved estimation errors as low as 1% and improved with higher DEM resolution, unlike grid-based methods, which showed over 20% error with minimal improvement. Real-world tests showed that the framework handles noisy data, reduces processing time, and captures full upslope catchment areas—even in diverging SWP conditions. It also benefits from increased DEM resolution, where GIS-based methods often plateau. The outcome is a practical algorithm that improves catchment estimation accuracy with lower computational demands. It supports applications in water flow prediction, flood, and drought management, and landslide risk assessment. Future work will focus on scalability through graphics processing unit (GPU) and cloud computing, hydrological model validation, testing on complex terrains, and geographic information system (GIS) tool integration.