Global high-resolution estimates of the UN Human Development Index using satellite imagery and machine learning

Area of Science:

• Socioeconomic indicators
• Geospatial analysis
• Machine learning applications

Background:

• The United Nations Human Development Index (HDI) is a key metric for national development, but its coarse resolution limits granular analysis.
• Existing HDI data (N=191 countries) fails to capture sub-national variations, hindering targeted policy interventions and accurate development assessments.

Purpose of the Study:

• To develop and distribute novel, high-resolution global Human Development Index (HDI) estimates at municipal and grid levels.
• To address the limitations of country-level HDI data by providing spatially detailed insights.
• To demonstrate the impact of aggregation bias on development assessments.

Main Methods:

• Utilized machine learning and advanced satellite imagery to create downscaled HDI estimates.
• Developed a generalizable downscaling technique applicable to various administrative data shapes and sizes.
• Generated global HDI estimates for 61,530 municipalities and 819,309 grid cells (0.1° × 0.1°).

Main Results:

• Produced unprecedented global HDI estimates at municipal and grid resolutions.
• Identified that over 50% of the global population was previously misclassified into incorrect HDI quintiles due to aggregation bias.
• Quantified the significant impact of spatial resolution on development metrics.

Conclusions:

• High-resolution HDI data derived from satellite imagery significantly improves development assessment accuracy.
• The developed downscaling methodology and satellite features can enhance spatial resolution for various administrative datasets.
• Addressing aggregation bias is crucial for precise global development monitoring and policy formulation.