Two years ago, the Defense Department made headlines when it deemed climate change a "threat multiplier" for the nation's armed forces -- an X factor that could exacerbate existing tensions and threats in unpredictable ways.

That unpredictability has been challenging for physical and political scientists trying to understand how shifting weather patterns, rising seas and more extreme weather will affect conflict in coming decades, experts said yesterday at a conference sponsored by the National Center for Science Education.

What appears to be an intuitive link between climate change, resource scarcity and conflict may be illusory, said Kaitlin Shilling, who recently completed a Ph.D. at Stanford University, where she studied the relationship between climate, agriculture and conflict in sub-Saharan Africa.

"We're already committed to a certain amount of climate change in coming decades," she said. "But we need to move beyond the idea that conflict is the inevitable result of climate-driven resource scarcity" to investigate the mechanisms by which shifting climatic conditions may influence human behavior.

One recent study, published last year in the journal Nature, concluded that El Niño, a global climate pattern that warms the central Pacific Ocean, doubles the risk of civil conflicts in tropical nations.

But the study's authors, researchers from Princeton and Columbia universities, said they were not sure how, exactly, El Niño contributes to outbreaks of conflict in 93 tropical countries singled out by the research.

Peering into the 'black box' of future reactions

The question for researchers, Shilling said yesterday, is how to peer into that "black box," where the complex climate system meets the equally complex web of emotions, instinct and logic that governs human behavior.

"The qualitative intuition on climate and conflict is clear," said Justin Mankin, a doctoral student at Stanford. "But the clear, quantitative evidence is mixed."

Mankin is studying the influence of climate on agriculture in Afghanistan. There, farmers must decide whether to reduce their plantings of wheat -- a staple crop in a country that has suffered wheat shortages for the past three decades -- to plant more opium poppies, which can be sold to the Taliban for income -- turning the act of planting poppies into a de facto loan on future profits.

Mankin is trying to tease out the relative influence of climate, economic need and local politics on Afghanistan's farmers. "Guessing how humans will actually respond in a given situation is the most difficult question," he said.

Some of the strongest evidence linking climate and conflict concerns so-called communal conflicts, tensions on a smaller scale than civil wars or wars between nations, said Cullen Hendrix, a professor of international relations at the College of William and Mary.

That category includes studies that have concluded climatic conditions are "highly influential" on the frequency of cattle raids in Africa's Sahel and played a role in post-election rioting in Kenya in 2007 and 2008. Other research suggests that, in some agriculture-dependent areas, "better" climatic conditions can actually increase violence.

"Climate variability can have impacts that are not in any obvious way resource wars," Hendrix said.