Evaluation of indicators for agricultural vulnerability to climate change: The case of Swedish agriculture

Agriculture is often described as one of the sectors most vulnerable to future climate change, and its vulnerability is commonly assessed through quantitative indices. However, such indices differ significantly depending on their selected indicators, weighting mechanisms, and summarizing methods, often leading to divergent assessments of vulnerability for the same geographic area. The use of generic indicators might also lead to a loss of information about contextual risks and vulnerabilities. This may reduce the perceived usefulness of indices among stakeholders. This study analyses the role of indicators in assessing agricultural vulnerability to climate change. It analyses how indices are understood and used through three separate focus group sessions, involving agricultural experts professionally active in south-eastern Sweden. The paper presents how agricultural practitioners perceive a set of common vulnerability indicators, presented through a visualization tool, and their relevance, logic, and applicability to assess and address vulnerability to climate change. The results of this study contribute with perspectives on (i) the relevance and applicability of the commonly used generic indicators for agricultural vulnerability (ii) the assumed correlation of indicators with climate vulnerability and (iii) the identification of missing vulnerability indicators. The study finds that commonly used vulnerability indicators are perceived hard to apply in practice, as definitions and thresholds are often depending on the geographical and temporal scale, as well as the regional context. Additional exposure factors that were identified included extreme events, such as heavy precipitation and external factors such as global food demand and trade-patterns. Further, participants expressed that it is important to include indices that combine effects of multiple climatic changes and in-direct factors, such as policies, regulations and measures. Inherent complexities, context dependencies, and multiple factors should further be included, but entail difficulties in developing suitable indicators. These factors must be addressed by a broader set of qualitative and quantitative indicators, and greater flexibility in the assessment methodology. The interactive vulnerability assessments presented in this paper indicate a need for an integration of quantitative and qualitative aspects and how such indicators could be developed and applied.