부산대학교 도서관

Climate has changed over the past century due to anthropogenic greenhouse gas emissions. In parallel, societies and their environment have evolved rapidly. To identify the impacts of historical climate change on human or natural systems, it is therefore necessary to separate the effect of different drivers. By definition this is done by comparing the observed situation to a counterfactual one in which climate change is absent and other drivers change according to observations. As such a counterfactual baseline cannot be observed it has to be estimated by process-based or empirical models. We here present ATTRICI (ATTRIbuting Climate Impacts), an approach to remove the signal of global warming from observational climate data to generate forcing data for the simulation of a counterfactual baseline of impact indicators. Our method identifies the interannual and annual cycle shifts that are correlated to global mean temperature change. We use quantile mapping to a baseline distribution that removes the global mean temperature related shifts to find counterfactual values for the observed daily climate data. Applied to each variable of two climate datasets, we produce two counterfactual datasets that are made available through the Inter-Sectoral Impact Model Intercomparison Project (ISIMIP) along with the original datasets. Our method preserves the internal variability of the observed data in the sense that observed (factual) and counterfactual data for a given day remain in the same quantile in their respective statistical distribution. That makes it possible to compare observed impact events and counterfactual impact events. Our approach adjusts for the long-term trends associated with global warming but does not address the attribution of climate change to anthropogenic greenhouse gas emissions. [ABSTRACT FROM AUTHOR]