So2Sat POP - A Curated Benchmark Data Set for Population Estimation from Space on a Continental Scale

Area of Science:

• Geospatial Science
• Urban Planning
• Environmental Science

Background:

• Accurate population distribution is vital for urban planning, disaster management, and resource allocation.
• Traditional census methods for population data collection are time-consuming and costly.
• Existing statistical and machine learning methods often rely on small-scale or unavailable datasets, hindering research.

Purpose of the Study:

• To address the need for accessible, large-scale population data for research.
• To provide a comprehensive dataset for developing and evaluating population estimation methods.
• To facilitate advancements in dynamic population distribution modeling.

Main Methods:

• Compilation of a multi-source dataset covering 98 European cities.
• Integration of diverse geospatial data: digital elevation models, local climate zones, land use, and nighttime lights.
• Inclusion of multi-spectral Sentinel-2 imagery and OpenStreetMap data.

Main Results:

• A novel, comprehensive dataset for population estimation across 98 European cities has been created.
• The dataset integrates multiple data sources, offering a rich resource for geospatial analysis.
• This resource aims to overcome limitations of previous small-scale or inaccessible datasets.

Conclusions:

• The developed dataset is a valuable resource for the research community.
• It supports the development of sophisticated approaches for population estimation.
• Facilitates improved decision-making in areas like urban planning and disaster management.