Scientific and methodological challenges make it hard to evaluate links between climate change and human conflict. We find the criticisms made in a review of these challenges (see C. Adams et al. Nature Clim. Change 8, 200–203; 2018), as outlined in your Editorial (Nature 554, 275–276; 2018), to be a flawed oversimplification.
One criticism is that the field suffers from a potential sampling bias because research focuses only on regions of violent conflict to draw conclusions about the social and political effects of climate change. In fact, researchers use such cases to identify the variety of factors that produced these conflicts, just as epidemiologists study disease outbreaks to understand the factors that produce epidemics.
Researchers are also accused of ignoring instances of cooperation in favour of conflict. That is not why studying cooperation is important. Comparing triggers of cooperation and conflict helps to understand human responses to environmental stresses.
We agree that research results should not be overgeneralized or inappropriately used to justify causality between climate and conflict. Both practices are rare in our experience, contrary to your implication. Most researchers take pains to describe climate and weather merely as ‘contributing factors’ to some conflicts (see, for example, P. H. Gleick Weather Clim. Soc. 6, 331–340; 2014 and C. P. Kelley et al. Proc. Natl Acad. Sci. USA 112, 3241–3246; 2015).
Nature 555, 587 (2018)