Urban Flood-Risk Mapping Using Explainable Machine Learning and Remote-Sensing Indicators: A Simulated Spatial Analytics Study

This article presents a simulated spatial machine-learning study integrating rainfall intensity, elevation, slope, imperviousness, drainage density, vegetation index, and proximity-to-water indicators to classify neighborhood flood susceptibility. Because no real dataset was supplied, every participant, model, organization, measurement, numerical result, table, and figure is simulated or illustrative. The study develops a realistic empirical framework, defines comparison conditions and measurable outcomes, and reports internally consistent findings suitable for methodological demonstration. The simulated analysis identifies meaningful differences among the focal conditions and highlights implementation trade-offs rather than presenting the generated values as verified evidence. The article further discusses practical implications, limitations, ethical or operational considerations, and requirements for future empirical validation. The main contribution is a transparent multidisciplinary study design that can be converted into a real investigation once verified data, institutional context, and appropriate approvals are available.