January 2024 – December 2024
Intended to support the ICRC to precisely map local populations to improve effective planning of humanitarian action, the results of this Humanitarian Action Challenges (HAC) research project were published in Nature in November 2022. Machine learning-based methodologies were used to estimate fine-grained population maps using open-source geodata in developing countries – the outcome of the model was gridded population estimates at a spatial resolution of 100 x 100m for Tanzania and Zambia.
The implementation and scale-up project is progressing well, with a consultant working on it until the end of 2024. The focus here is to implement and promote the use of gridded population datasets, particularly the Popcorn datasets, within the institution. These datasets are being refined and prepared for external cloud computing to allow ICRC to generate population estimates using open-source sentinel images.
ICRC focal point: Thao Ton That Whelan