Downscaling and disaggregating NAO-conflict nexus in pre-industrial Europe

Recently, the desiccation effect of the North Atlantic Oscillation (NAO) is found to be positively correlated with violent conflict in pre-industrial Europe, with agricultural shrinkage and its subsequent economic shocks to be their causal link. However, it remains unexplored whether the correlation persists if the study period is extended backward in time, a different definition of violent conflict is applied, or the relationship is examined at lower geographic levels. In this study, we based on 835 internal disturbance incidents in Europe during 1049–1800 to conduct long-temporal and multi-scalar examination on the NAO-conflict nexus. ′Time-series′ and ′panel data′ disaggregation approaches, together with Granger Causality, Multiple Regression, and Survival Analyses were applied to verify the nexus quantitatively. Results show that the positive NAO-conflict correlation was significant at the continent and physiographic zone levels. During the positive NAO phases, the annual probability of internal disturbance outbreak increased by 70.0% in the southern Europe and the Mediterranean, a zone most affected by the NAO-induced desiccation effect. Yet, the NAO-conflict correlation was rather inconsistent when it was downscaled to the sub-regional level. Moreover, the NAO-conflict correlation was inflated under the time-series approach, while the panel data approach demonstrated the region-specific nature of the NAO forcing more clearly. The associated implications in examining climate-conflict nexus are discussed. Our findings may be crucial in examining violent conflict in the northwestern Africa, a highly agricultural region affected by the NAO.