Beijing Block-Level Population Dataset (2020)

By linking the census data with street and road network vector data, it achieves a refined spatial distribution of block-level population in Beijing. The dataset is of significant importance for research in urban planning, resource allocation, and emergency management.

The study utilizes OSM road network data to generate natural neighborhood boundaries for Beijing, integrating Gaode building footprints and urban rooftop vector data. Combining attributes like building heights, a linear regression model was developed to obtain a refined block-level population dataset. Cross-validation with WorldPop data shows that the model has an R² of 0.91, a Pearson correlation coefficient of 0.953, and errors within reasonable bounds, verifying the high accuracy and applicability of the dataset.