First, do no harm - Missing data treatment to support lake ecological condition assessment

Abstract

Indicators of ecological condition status of water bodies associated with field measurements are often subject to data gaps. This obstacle can often lead to abandonment of assessment. Furthermore, it can lead to the use of methods, based merely on their availability. In response to these challenges, a systematic approach for expert-analyst interaction for missing data treatment is proposed. A combination of algorithms with hierarchical clustering of results is used. A particular emphasis is put on the preparation and interpretation of input data and the role of an expert in the workflow. The proposed approach enhances the decision-making process by improving communication and transparency throughout interactions between experts, analysts and decision makers. Future research should focus on assessing the scale of the ecological data drift phenomenon, which, based on the observed climate change, anthropological pressure and biodiversity loss, may impact the broad concept of indicator construction for lake water ecological assessment