BirdNET App data: 2020 case studies

Datasets for four case studies used to test whether BirdNET App data could be used to replicate known patterns in avian ecology and thus serve as a reliable research resource. The case studies were selected based on the availability of results we could attempt to reproduce (e.g., song dialect mapping has been somewhat limited at broad scales because of data limitations) and to encompass a range of phenomena (e.g., simple vs complex songs and the migration of small vs large populations). Researchers interested in collaborations using app data are also encouraged to email .

Please note that BirdNET confidence scores are not equivalent to probabilities and are unique to each species. That is, a confidence score of 0.85 may have a very different probability of being correct for species A and species B.