Dataset for "A probabilistic social vulnerability index to assess spatio-temporal vulnerability dynamics using Bayesian machine learning"

This dataset and accompanying Python codes support the published article “A Probabilistic Framework for Spatio-Temporal Social Vulnerability Assessment to Flood Hazards in Canada using Bayesian Machine Learning” by Chakraborty et al. (2026), published in the Journal of Geovisualization and Spatial Analysis. The repository includes geospatial data layers (shapefiles), socioeconomic indicators, and racial/ethnic variables used to construct a probabilistic social vulnerability index through Bayesian machine learning and Monte Carlo simulation. Data preprocessing, spatial analysis, and modeling were conducted using GIS and Python. The Python codes are provided as Jupyter Notebook source files and are specifically designed to support risk analysis at the census dissemination area (DA) geographic level. This dataset is intended to support reproducible research and further applications in disaster risk reduction, dynamic social vulnerability assessment, probabilistic risk analysis, and environmental justice research. All datasets and codes are associated with and derived from the published Springer Nature article and are subject to copyright under Springer Nature. Users are required to provide appropriate reference, and citation to the original publication for any reuse, redistribution, or derivative work. Required citation (APA style):Chakraborty, L., Taherimashhadi, H., Thistlethwaite, J., et al. (2026). A probabilistic framework for spatio-temporal social vulnerability assessment to flood hazards in Canada using Bayesian machine learning. Journal of Geovisualization and Spatial Analysis, 10(3). https://doi.org/10.1007/s41651-025-00247-y