Key points are not available for this paper at this time.
In the upcoming decades, climate change impacts will increasingly emerge, requiring regions worldwide to obtain actionable climate information. Global Climate Models (GCMs) are often used to explore future conditions, but the variability of projections among GCMs complicates regional climate risk assessments. Often, multi-model means of climate responses to various emission scenarios are used to reduce analysis resources and uncertainties. Unfortunately, emission scenarios can only explain a small fraction of variance in the mean climate responses of local precipitation patterns and are dominated by model uncertainty and internal variability. Emission scenarios model means, therefore, lead to similar mean responses across emission scenarios. This results in inefficient use of analysis resources and a narrow view of potential climate risks in the region. Since precipitation is a key driver for many hazards like flood, drought and wildfire, local assessment of these risks using emission-based multi-model means is probably not optimal. This study proposes a method to select more impact-relevant scenarios by determining regionally relevant climatic impact drivers and grouping GCMs on their projected changes in these drivers. We quantify the effectiveness of our approach by comparing future impacts covered by emission-based scenarios with our approach, expressed as an “exploratory amplification” factor. We illustrate the method for flood risk in the Latvian Lielupe basin and find the novel method has an exploratory amplification up to a factor of eight for the mid-century. We conclude that our method significantly improves regional exploration of future impacts, enabling more informed adaptation decisions.
Buskop et al. (Wed,) studied this question.