They also point to a change in relative risk since 1950, after which changes in temperature have emerged as a higher risk factor than changes in precipitation, and the risk of group violence is greater than the risk of person-to-person conflict.
The researchers caution that climate clearly is not the only factor at work. Not all climate events affect all conflicts. Nor does a changing climate alone determine whether conflicts will occur.
Still, "for a long time, people have been arguing: Is there an effect or not, yes or no?" says Solomon Hsiang, a researcher at the University of California, Berkeley who focuses on sustainable development and the effects of climate on societies.
The answer is yes, says Dr. Hsiang, who along with two colleagues from Princeton University and Berkeley conducted the analysis. The results are set to appear in Friday's issue of the journal Science.
The new questions, he continues, are why, and how heavily does climate influence conflict compared with other influences, such as personal or national wealth, political stability within a region or country, or how governments respond to migration within or across their national borders.
Other researchers exploring the links between climate, social, economic, and political conditions and conflict say the new study overstates climate's effect on the risk of conflict.
John O'Loughlin, a geographer at the University of Colorado at Boulder who explores the interplay of environmental conditions and conflict, notes that in his research, as in the new study, climate conditions play a role. But his own work suggests that warming temperatures or extreme precipitation are poor predictors of risk compared with the social, economic, and political drivers for conflict.