Background: A number of global conflicts have been linked to serious drought preceding the emergence of the conflict (e.g.: https://www.nber.org/papers/w16461 ), with one example being the drought in Syria prior to the civil war. We would like to build a model that predicts destabilization of regions on a global scale, utilizing atmospheric data like average rainfall. This is not only interesting from a Political Science point of view, but provides a risk assessment tool for foreign investors in emerging markets. An additional business value is from a humanitarian aid perspective: mitigation strategies for atmospheric events might preclude destabilization.

Data collection: We wish to join a number of different data sources, but at the heart of the project is the global conflicts databases here: https://icr.ethz.ch/data/
Other data we want to utilize include: global rainfall averages over time, development indices, existing economical and demographic data, etc.

Analysis: We intend to include geospatial measurements in our feature set to be able to capture more information due to geography and politics of a region.