Bird populations are known to be showing rapid responses to environmental change. For example, due to climate change, migratory events have been shifting to an earlier time in conjunction with warmer spring temperatures. This has consequences to bird population dynamics and functioning, which are being examined with a new mobile application Muuttolintujen kevät (Spring of Migratory Birds) that engages citizens to participate in collecting data.
Muuttolintujen kevät, which is designed for audio-based bird monitoring, has been launched this spring together with the University of Jyväskylä, Finland and the Finnish broadcasting company Yle. The mobile application is connected to a machine learning model which is housed in CSC’s cPouta cloud service and is used for real-time bird species identification. The collected data is stored for research use in CSC’s Allas data storage service.
Easy-to-use mobile application
The application allows anyone to record bird sounds in nature with their own mobile device and identify different bird species. The collected data will be used as a basis for research. The recordings of bird sounds enable researchers to obtain information on when exactly the migratory events are happening and in which locations different bird species can be found.
The results appear in the mobile application itself as well as on a webpage in the form of a map which shows the bird species that have been observed to vocalise during a selected time period. The www-page has been created, hosted and maintained by Yle and can be accessed here. At the moment, the application is able to identify 150 different bird species, but the model is continuously being improved.
The user can see from the application how the recording that has been made appears on the map and how it updates current predictions as well as where the recordings of other users have taken place.
The functionalities of the “Muuttolintujen kevät” mobile application
In addition to quick recordings, the application can be utilised to record for long periods at a time, for example overnight. This type of data is most valuable because not only does it provide systematic information about where and when birds have been observed, but also information on where and when they have not been observed.
High interest among the Finnish population
Within three weeks since the launch, the mobile application has been downloaded more than 30 000 times. Approximately 1 million recordings have already been submitted, with especially Southern Finland being well covered. More than 100 000 classifications have reached at least the 90% probability threshold of correct species identification.
Large attention and participation by citizens interested in birds leads to more accurate data on migration than has previously been collected.
Digital twin on bird distribution modelling
In the BioDT project, digital twins are being developed to mimic behaviour observed in nature, with the purpose of developing an improved understanding of biodiversity dynamics in response to diverse pressures, including climate change. To meet the specific requirements of each use case, the BioDT project will develop multiple digital twins. To process and analyse the large volumes of data involved, the digital twins benefit from access to the LUMI supercomputer.
One of the aims of BioDT is to develop new ways to combine digital twins and high-performance computing with citizen science. To achieve this, data gathered via the Muuttolintujen kevät mobile application will be used as input data for a purpose-designed digital twin for bird distribution modelling. Data on the identities and locations of bird species will be transferred to the LUMI supercomputer, which is also used by the digital twins developed in BioDT. Citizens will therefore be able to directly contribute data to research that makes use of the fastest supercomputer in Europe.
Collected data and wider impact
The data recorded allows not only to analyse when the birds are arriving to Finland and where do they reside but also investigate how changing environmental conditions influence bird populations. When observations about birds are combined with information about climatic and habitat conditions such as forest structure, researchers are able to analyse and identify which factors affect the migration and breeding periods of birds. Ultimately, the data collected facilitates understanding of the connection between environmental change, bird populations and ecosystem functioning.
This article was originally published on CSC’s website.
Authors: Anna-Liisa Allas and Jesse Harrison, CSC – IT Center for Science, Finland
Image: Adobe Stock