Uncertainty assessment in ecosystem services analyses: Seven challenges and practical responses

TLDR

Ecosystem services analyses are increasingly used to address societal challenges, yet they often lack uncertainty assessment, limiting the validity of their findings and undermining science‑based decisions; the multidisciplinary nature of ES science provides existing knowledge to quantify uncertainties, and insights from integrated assessment can improve such analyses. This paper summarizes and analyzes seven commonly perceived challenges to conducting uncertainty assessment in ES studies to explain why it is frequently treated superficially. The authors link these challenges to solutions drawn from relevant scientific literature and guidance documents. They identify issues that hinder the adoption of uncertainty assessment as standard practice and present a synthesis that serves as a starting point for analysts and decision‑makers to set realistic expectations for confidence in ES assessments.

Abstract

Ecosystem services (ES) analyses are increasingly used to address societal challenges, but too often are not accompanied by uncertainty assessment. This omission limits the validity of their findings and may undermine the 'science-based' decisions they inform. We summarize and analyze seven commonly perceived challenges to conducting uncertainty assessment that help explain why it often receives superficial treatment in ES studies. We connect these challenges to solutions in relevant scientific literature and guidance documents. Since ES science is based on a multiplicity of disciplines (e.g. ecology, hydrology, economics, environmental modeling, policy sciences), substantial knowledge already exists to identify, quantify, and communicate uncertainties. The integration of these disciplines for solution-oriented modeling has been the focus of the integrated assessment community for many years, and we argue that many insights and best practices from this field can be directly used to improve ES assessments. We also recognize a number of issues that hinder the adoption of uncertainty assessment as part of standard practice. Our synthesis provides a starting point for ES analysts and other applied modelers looking for further guidance on uncertainty assessment and helps scientists and decision-makers to set reasonable expectations for characterizing the level of confidence associated with an ES assessment.