On the scaling of climate impact indicators with global mean temperature increase: a case study of terrestrial ecosystems and water resources

We assessed whether the impacts of various increases in global mean temperature from preindustrial levels (∆GMT) on terrestrial ecosystems and water resources could be approximated by linear scaling of the impacts of ∆GMT = 2 °C at global and large regional scales. Impacts on net primary production, CO2 emissions from biomass burning, soil erosion, and surface runoff calculated by impact model simulations driven by multiple climate scenarios were assessed for a ∆GMT range of 1.5–4 °C. The results showed that the linear scaling was tolerable for net primary production, biomass burning, and surface runoff for a global average. However, for regional averages, the linear scaling was unacceptable for net primary production and biomass burning as well as for soil erosion at around 3 °C in numerous regions around the world. The linear scaling was judged to be tolerable for surface runoff in most regions where the impacts of 2 °C were statistically significant, but there were large uncertainties in future changes in surface runoff in many regions. Exploring the applicability of linear scaling could help simplify and streamline climate-change impact assessments at various ∆GMTs. Our approach leaves room for refinement, and further investigation will be worthwhile.