Gridded population mapping for Germany based on building density, height and type from Earth Observation data using...

Area of Science:

• Geospatial science
• Population geography
• Remote sensing

Background:

• Gridded population data is crucial for global change research and disaster risk assessment.
• Existing methods often lack fine-scale accuracy for human-environmental process studies.

Purpose of the Study:

• To develop and validate a nationwide gridded population dataset for Germany.
• To assess the impact of building density, type, and height on dasymetric mapping accuracy.
• To perform a bottom-up population estimate using integrated datasets.

Main Methods:

• Utilized Copernicus Sentinel-1 and Sentinel-2 data for building type, density, and height.
• Produced a nationwide dataset of predominant building types.
• Examined top-down dasymetric mapping with varying weighting layers.
• Performed a nationwide bottom-up population estimation.

Main Results:

• Integrating building types improved fine-scale mapping by preventing redistribution to non-residential areas.
• Building density enhanced population estimate quality across all scales compared to binary layers.
• Combining building density and height (volume) significantly improved mapping accuracy and reduced regional discrepancies.

Conclusions:

• Building type integration is beneficial for fine-scale population mapping.
• Building density and volume are key factors for accurate gridded population estimates.
• Building density, type, volume, and living floor area per capita enable accurate large-area population estimates.