Satellite data analysis in conflict and famine-affected areas

More than half of Somalia’s population is composed of pastoralists, many of whom practice nomadic pastoralism. While their mobility enables adaptability, it also results in higher vulnerability to external factors including slow-onset environmental events, which have increased in severity and frequency over the years. The collection of longitudinal data on nomadic populations is challenging, especially because they are often situated in remote regions and change locations frequently. This project therefore implemented and evaluated additional technological methods to address the gaps in current data collection practices.

The Governance Lab (GovLab) at New York University, Signal Program at the Harvard Humanitarian Initiative (HHI), and UNICEF considered satellite imagery analysis to develop a standardized methodology and assist in the identification and monitoring of patterns of population movements in Somalia. The project transitioned from satellite imagery analysis to agent-based modeling as the primary means of analysis. There is no single methodology that addresses all the components of the work completed, but the tested methods add unique value to ongoing humanitarian work. The next steps for the completed work included refining methodologies and development of case studies that were shared to inform ground operations.

This project addressed the following SDGs: SDG 3 - Good health & well-being, SDG 4 - Quality education, SDG 10 - Reduced inequalities, and SDG 13 - Climate action. The project resulted in the generation of two major products: 1) imagery interpretation guide and 2) an agent-based model (ABM). The Java-based ABM code is available via GitHub and the Imagery Interpretation Guide, titled Satellite Imagery Interpretation Guide of Landscape Features in Somaliland, were posted on HHI's website. The videos elaborating on the project background, the satellite imagery analysis, and ABM development are hosted by Harvard Humanitarian on YouTube. 

Imagery Interpretation Guide

Lessons Learned


Disclaimer: As many of the links are external, some links may not work.