Supporting Quality Assurance for Citizen Science Observations of Plants and Animals

For citizen science observations of plants and animals, quality assurance is a vital issue. In many cases, experts check reports, judging the plausibility of observations based on their own expert knowledge, but also using... [ view full abstract ]

For citizen science observations of plants and animals, quality assurance is a vital issue. In many cases, experts check reports, judging the plausibility of observations based on their own expert knowledge, but also using spatial and temporal context information. In fact, earlier observations that were reported to the same project or to other projects are a valuable source of context. Was the species observed at or near the same place before? What other species were reported at that place? Are there reports from the same time of year? Questions like these can be answered not only by humans, but also by computers that have the necessary data and analysis methods available. We followed exactly this path to build a decision support system designed to help experts make sound decisions about acceptance or rejection of species observations reported by volunteers to online portals. When triggered by an expert for a certain new observation, the application analyzes existing, certified reports, using as parameters the new observation’s properties (species, date, place and reporter). The application then produces a dossier of information for that observation, including an estimation of the observer’s degree of expertise, a visualization of the observation’s date, and results of spatial analyses (looking at the species reported close to the observation). This talk will present important lessons learned from the use of this tool for checking species observations in a German citizen science portal, which are subsequently employed for administration and planning purposes. An outlook is given on advanced methods for (semi)automatic estimation of plausibility for citizen science species observations that are currently being developed and examined in depth in a research project at Heidelberg University (GIScience Group, Department of Geography).