John Wiley & Sons Ltd
Abstract: This study proposes a structured six-step filter for selecting hydrological models suitable for Kenya?s flood forecasting needs. It evaluates 12 models for performance under data scarcity and diverse catchment conditions, narrowing the list to five candidates including SWAT and HYPE. Key criteria include process realism, calibration feasibility, code accessibility, and operational stability. The methodology incorporates decision trees and Venn diagrams to visualize trade-offs. It offers a replicable modeling framework for national forecasting systems in low-data environments.