Modelling armed conflict risk under climate change with machine learning and time-series data

Using machine learning, the authors reveal that stable background conditions explain most variation in armed conflict risk worldwide. Positive temperature deviations and precipitation extremes also increase the risk of conflict onset and incidence.