Global and regional climate modes modulate armed conflict risk.

Because of their impacts on droughts, famines, and floods, modes of climate variability can shape patterns of social instability. Yet the mechanisms linking climate variability to armed conflict remain contested, especially relative to the myriad sociopolitical and economic determinants of conflict. A key challenge is that most studies rely on coarse, state-level data and treat climatic teleconnections as invariant. As such, it is unknown whether there are distinct climate hazards that select for conflict risk; whether conflict scales with climate hazard exposure; and whether such associations exist for more regional forms of climate variability. Here we leverage empirical modeling using a high-resolution gridded dataset of armed conflicts and the natural experiment afforded by two major climate modes-the El Niño-Southern Oscillation (ENSO) and the Indian Ocean Dipole (IOD)-to clarify how systematic hydroclimatic anomalies influence conflict emergence. Our results reveal the following: First, conflict risk heightens during El Niño, but ENSO-associated risk does not scale linearly with teleconnection strength; and, evidence for threshold behavior varies with spatial aggregation. Second, El Niño-related increases in conflict risk arise through its dry teleconnections, with limited evidence for wet teleconnections. Finally, the more regionally confined IOD also influences conflict, with both positive and negative phases elevating risk in strongly teleconnected regions, namely the Horn of Africa and Southeast Asia. These results reveal that modes of climate variability can differentially shape conflict risk, offering insight into societies' vulnerabilities to natural climate fluctuations and, by extension, anthropogenic climate change.